Improving Prevalence Estimates of Hepatitis C in Key Populations: A Simulated Data-Based Comparison of Missing Data Techniques

Improving Prevalence Estimates of Hepatitis C in Key Populations: A Simulated Data-Based Comparison of Missing Data Techniques | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Improving Prevalence Estimates of Hepatitis C in Key Populations: A Simulated Data-Based Comparison of Missing Data Techniques Adewunmi Akingbola, Olajumoke Adewole, Abiodun Adegbesan, Joel Chuku This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6994675/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 11 You are reading this latest preprint version Abstract Background: Hepatitis C virus (HCV) remains a major public health challenge, particularly among People Who Inject Drugs (PWIDs). Missing data in surveillance systems can bias prevalence estimates, affecting decision-making. This study compares Complete-Case Analysis (CCA) and Multiple Imputation (MI) for handling missing data in the estimation of HCV prevalence using a simulated dataset derived from the UK’s Unlinked Anonymous Monitoring (UAM) Survey. Methodology: We conducted a cross-sectional analysis using a simulated version of the UAM dataset, focusing on key demographic and behavioural variables. HCV prevalence was estimated using both CCA and MI approaches. MI was performed using chained equations with five imputations. The effect of missing data handling on prevalence estimates and associated confidence intervals was compared between methods. Results: HCV prevalence estimates obtained via MI were consistently higher than those from CCA, with narrower confidence intervals. The CCA approach excluded a substantial proportion of cases due to missing data, introducing potential bias. MI preserved sample size and yielded more robust estimates, particularly among subgroups with higher missingness. Conclusion: Multiple Imputation outperformed Complete-Case Analysis in estimating HCV prevalence from the simulated UAM data. These findings highlight the importance of appropriate missing data methods in epidemiological surveillance and public health research. Biological sciences/Computational biology and bioinformatics Health sciences/Diseases Health sciences/Gastroenterology Health sciences/Health care Health sciences/Medical research Health sciences/Risk factors Hepatitis C People Who Inject Drugs Missing Data Multiple Imputation Complete-Case Analysis Surveillance Data Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 INTRODUCTION Hepatitis C, an infectious disease and a type of viral Hepatitis caused by the Hepatitis C virus accounts for an estimated 15–20% of acute hepatitis cases worldwide. About 16% of patients develop cirrhosis within 20 years and about 25% of Hepatocellular carcinoma cases have been attributed to this disease ( 1 ). It is a major public health concern with an estimated 58 million people currently infected and commonly spreads through sharing needles or syringes, amongst other routes, thus people who inject drugs (PWID) account for an estimated 11 million cases globally ( 2 ). In the United Kingdom, an estimated 118,000 people lived with chronic hepatitis C infection in 2019( 3 ), with an estimated 80% of them contracting the disease due to drug use through injections ( 4 ). Through an Unlinked Anonymous Monitoring Survey conducted in the United Kingdom between 1990 and 2019, the median age of people who inject psychoactive substances which was previously at 27 years, was shown to have increased to 33 years in 2008 and 40 years in 2019. This indicates that before the 1990s, drug injection was prevalent amongst young people and a shift to middle-aged years has been identified, due to fewer new and a lot of existing old drug users. 74.9% of the 66,545 respondents of the survey were Males, with the North-west region and London having the highest percentage of older respondents in England, Wales, and Northern Ireland, while Glasgow and Clyde had the highest percentage of older people in Scotland. This means that the Male Gender has a higher inclination to use drugs, thus a higher potential of contracting the Hepatitis C virus, and the age-by-location distribution offers insights into the availability of drugs and the effectiveness of public health policies in curbing this trend in those locations ( 5 ). Further extensive analysis of the 2019 Unlinked Anonymous Monitoring Survey for England, Wales, and Northern Ireland, 51.9% of PWIDs respondents alluded to having engaged in sexualized drug use in the last year, 70% of them have been in prison and 50.4% have been homeless. These provide valuable insights into some social and behavioral events associated with PWIDs and can as well, predispose them to contracting the Hepatitis C disease. ( 6 ) However, routinely collected epidemiological data like the Unlinked Anonymous Monitoring Survey has a high propensity for missing values which can occur because of loss to follow-ups of patients, non-response and/or errors during data entry. This results in a biased result with questionable validity and reliability, alongside reducing the statistical power of the data such that the probability of rejecting the null hypothesis when it is false is reduced, if not appropriately handled. ( 7 , 8 ). Traditionally, missing values in a dataset are handled using the list-wise deletion, also known as complete-case analysis which can cause bias in the estimates and thus, also impact the results and inferences from it. Several alternative statistical methods exist to handle missing data including Multiple Imputation. This method predicts likely values with similar variability and uncertainty with the right values, from other values in the dataset, thus safeguarding the validity of the results obtained from the imputed data ( 9 ). In this research project, complete-case analysis and multiple imputation techniques will be employed in the analysis of the data and a comparison of both methods will be made, insights and differences will be identified, and their implications assessed. METHODOLOGY This study employed a cross-sectional, observational design using a simulated version of the Unlinked Anonymous Monitoring (UAM) Survey dataset, which is routinely used to monitor the prevalence of Hepatitis C (HCV) among people who inject drugs (PWID) in the UK ( 10 , 11 ). The simulated dataset replicates the structure and content of the original UAM survey, allowing for methodological exploration without involving real participants. The original UAM survey gathers anonymous data from PWID recruited through convenience sampling at drug treatment centers, needle exchange programs, and related services, enabling access to a typically hard-to-reach population ( 12 ). Data collection is conducted anonymously to encourage truthful responses and minimize stigma, thereby improving data reliability ( 13 , 14 ). While the original design offers advantages in studying epidemiological trends and associations efficiently and cost-effectively, it carries inherent limitations such as selection bias, recall bias, and issues related to data completeness ( 10 ). These challenges persist in the simulated dataset, making it an appropriate tool for evaluating missing data strategies and analytical approaches in public health research. Figure 1 below summarizes the steps in analysis. Variables of Interest : In this study, the variables of interest were classified into dependent and independent variables ( 15 , 16 ) to analyze the prevalence and risk factors associated with Hepatitis C (HCV) among people who inject drugs (PWID). Analysis of Missing Data Mechanisms This study conducted a comprehensive analysis to understand the mechanisms underlying the missing data in the dataset. Using Little's MCAR test, which evaluates whether missing data is Missing Completely at Random (MCAR) by comparing the observed data patterns to what would be expected if the data were truly random, using a chi-square statistic to determine significance ( 17 ) with results in Appendix A below, I assessed whether the missingness in each variable was completely at random. To further investigate, we performed a logistic regression analysis on the missingness indicators for the 'HCV RNA' outcome variable (see Appendix B) but did not assess the plausibility of an MNAR scenario. Complete Case Analysis: - To handle missing data in the key variables, I performed a complete case analysis function in R, which involves retaining only those observations with no missing values in the specified variables. The process was as follows: First, we identified the key variables of interest, and I generated a logical vector indicating which rows in the dataset contained complete data across the selected variables in Table 2 above. We then subset the dataset to include only these complete cases. Multiple Imputation Process : Multiple imputation was performed using the Multiple Imputation by Chained Equations (MICE) package in R (version 4.4.0). This method provides a robust way to impute missing values, ensuring that the relationships between variables are preserved and the imputed data accurately reflects the underlying patterns ( 18 ). The multiple imputation procedure involved the following steps: Imputation Method Selection: The imputation was conducted using the MICE function, which employs a chained equations approach. The method chosen for imputation was predictive mean matching (PMM), suitable for handling both continuous and categorical data types. PMM matches missing values to observed values that are similar in the predictive model, thus preserving the original distribution of the data ( 19 ). Execution of Multiple Imputation: The MICE function generated multiple imputed datasets. The imputation was configured as follows: data_for_imputation represented the dataset with missing values. m = 5 specified the creation of five imputed datasets, balancing computational efficiency with the reliability of the imputation results. maxit = 50 indicated the maximum number of iterations for the imputation algorithm to converge. Seed = 500 ensured reproducibility by setting a seed for the random number generator. Method = imputation_methods applied specific imputation methods tailored to each variable, particularly Predictive Mean Matching (PMM). 3. Analysis and Pooling of Imputed Data: Each of the five imputed datasets was analyzed separately to obtain estimates for the statistical analyses. The results from these analyses were then combined using Rubin’s rules. This pooling process accounted for the variability between the imputations and provided final estimates, standard errors, and confidence intervals ( 20 , 21 ). This comprehensive approach helped integrate the imputed datasets, thus reflecting the uncertainty due to missing data and yielding more robust and reliable statistical inferences ( 22 ). Statistical Analysis The statistical analysis was employed to summarize and infer key socio-demographic variables and factors related to Hepatitis C Virus (HCV) prevalence within the study population. Descriptive statistics provided a comprehensive overview of the critical variables, while inferential statistics facilitated the identification of significant associations and trends. The analysis included data processing, summary measures, and advanced statistical modelling to draw meaningful conclusions. Descriptive and Inferential analysis included Chi-square tests ( 23 ), Wilcoxon Rank Sum tests ( 24 , 25 ), T-test ( 26 ), Logistic regression models ( 27 ). All analyses were conducted using R software version 4.4.0. RESULTS DESCRIPTIVE ANALYSIS The complete-case analysis and multiple imputation age distribution in Fig. 2 below are bell-shaped curves indicating that they are typically distributed, and their highest frequencies are observed between ages 35 and 45 years with their peak ages approximately 40 years old. However, the frequency of the complete-case analysis is very close to 1200, and that of multiple imputation is close to 4000, which is the number of individuals within the most populous age group. The density plots for injection duration in Fig. 3 illustrate the distribution among participants, comparing Complete-Case Analysis and Multiple Imputation methods. Both plots exhibit a similar shape, peaking around 10 to 15 years of injection duration. The Multiple Imputation plot shows a more continuous distribution and fluctuation, while the Complete-Case Analysis plot has slightly less pronounced peaks and troughs, reflecting differences in data handling between the two methods. The analysis included several variables, each with varying percentages of missing data, as listed in Table 1 below. The age variable had a missing percentage of 1.35%, while gender had 0.40%. Condom use (Condom1) was the variable with the highest missing percentage at 45.05%, followed by sharing needles(share) at 27.05%. Previous imprisonment (epris1) had a missing percentage of 2.14%, homeless status (homeless1) was missing for 1.94% of cases, and the exchange of needle (exch) had a 0.92% missing rate. Prison before injection (prisbeforeinj) had 10.00% missing data, and the HCV RNA status (hcvrna) variable had 21.70% missing data. These percentages highlight the need for robust statistical methods to handle incomplete datasets due to the large proportion of data that would be omitted in Complete Case (C.C), which may be used in Multiple Imputation (M.I) methods. Table 1 Variables and Missing Percentage. Variable Missing Percentage age 1.35% Gender(gen) 0.40% Use of Condom (condom1) 45.05% Ever been in Prison (epris1) 2.14% Ever Homeless (homeless1) 1.94% Needle Exchange (exch) 0.92% Prison before Injection (prisbeforeinj) 10.00% Sharing of Needle (Share) 27.05% HCV Prevalence (HCV RNA) 21.70% Injection Duration (Injdur) 4.54% The prevalence of HCV RNA over the years was tracked to understand trends and patterns. Figure 4 and Table 2 below shows the Complete Case (C.C) image, a line graph marked by notable volatility. From 2011 to 2013, the graph shows relatively stable rates, indicating consistent monitoring and recording of very similar prevalence during these years. However, between 2013 and 2014, there was a dramatic spike, suggesting data incongruencies as all negative results were observed to be missing between those years as RNA Testing was done retrospectively. This sharp rise is visually prominent and indicates a heightened concern that continued into 2015 when it peaked. Following this peak, the graph depicts a steep decline starting in 2015 and continuing into the following years, suggesting effective intervention measures were implemented. By the end of the period, from 2018 to 2021, the trend line stabilizes, showing a gradual decrease. In contrast, the Multiple Imputation (M.I.) image portrays a smoother and more consistent trend over the same period. The line graph starts with a similar stability from 2011 to 2013. However, unlike the C.C. method, there was no dramatic spike in 2013 or 2015. Instead, the M.I. method shows a gradual increase during these years, reflecting a more moderated rise in prevalence. After 2015, the M.I. graph shows a steady decline starting from 2018, like the C.C method, but with a more gradual slope. The trend continues downward to 2021, presenting a stable decrease and indicating sustained effectiveness of intervention measures. In the unadjusted results, as seen in Table 5 below, the year variables show variations between complete case analysis (C.C) and multiple imputation (M.I). For instance, C.C. indicates higher ORs for 2014 (2.92, 95% CI: 2.19–3.88) and 2015 (4.02, 95% CI: 2.95–5.49), which are statistically significant, while the M.I results for these years are closer to 1 and non-significant. Both methods show that years like 2018 (C.C OR: 1.39, 95% CI: 1.06–1.82, p-value: 0.0162; M.I OR: 0.84, 95% CI: 0.72–0.98, p-value: 0.0296) and 2021 (C.C OR: 0.47, 95% CI: 0.31–0.73, p-value: 0.000824; M.I OR: 1.90, 95% CI: 1.46–2.47, p-value: <0.0001) are statistically significant, with M.I showing consistently lower p-values for these years. In the adjusted results from the multivariate analysis in Table 9 below, C.C shows significant ORs for 2014 (2.49, 95% CI: 1.85–3.34) and 2015 (3.37, 95% CI: 2.45–4.63), while these years are non-significant in M.I (close to 1). The years 2019, 2020, and 2021 are significant in both methods, with C.C showing ORs of 0.71 (95% CI: 0.54–0.94), 0.51 (95% CI: 0.31–0.84), and 0.37 (95% CI: 0.24–0.58), and M.I showing ORs of 0.77 (95% CI: 0.65–0.91), 0.47 (95% CI: 0.34–0.64), and 0.40 (95% CI: 0.31–0.53), respectively, all with strong statistical significance. Table 2 Prevalence per Year Table: The prevalence of HCV RNA, as calculated by Complete Case Analysis and Multiple Imputation, shows trends in prevalence rates from 2011 to 2021. Year Complete-Case Analysis Prevalence per Year (%) Multiple Imputation Prevalence per Year (%) 2011 29.7 30.7 2012 29.8 30.5 2013 29.9 30.8 2014 55.2 30.8 2015 63.0 32.5 2016 30.4 33.8 2017 28.8 30.5 2018 37.0 34.4 2019 27.0 29.2 2020 23.1 21.1 2021 16.7 18.9 Variables and HCV RNA The association between various variables and HCV RNA status was analyzed using different statistical tests as shown in Table 3 below. The Injection Duration and HCV RNA variables showed significant associations in both C.C (Wilcoxon, p < 0.0001) and M.I (Wilcoxon, p < 0.0001) analyses. This suggests that the duration of injection use is related to the presence or level of HCV RNA, and the consistency of this result across both the C.C. and M.I. analyses further reinforce the robustness of this finding. The Ever-Used Needle Exchange Program (exch) variable also exhibited significant associations (C.C: Chi-square, p = 0.0122; M.I: Chi-square, p = 0.0012). The Share variable for Sharing of Needle did not show significant associations in either method (C.C: Chi-square, p = 0.3207; M.I: Chi-square, p = 1.0000). Ever Homeless status (Homeless1) had significant associations in both methods (C.C: Chi-square, p < 0.0001; M.I: Chi-square, p < 0.0001), meaning that current or previous homelessness is related to the presence of HCV RNA/Hepatitis C infection. Prison before Injection Commencement variable was significant in the C.C method (Chi-square, p < 0.0001), meaning that having been in prison before starting injections is strongly associated with having Hepatitis C, but not in the M.I method (Chi-square, p = 0.2434). Gender showed significant associations in both methods (C.C: Chi-square, p < 0.0001; M.I: Chi-square, p < 0.0001). Condom Use was significant in the C.C method (Chi-square, p = 0.0170) but not in the M.I. method (Chi-square, p = 0.3547). Finally, Ever in Prison or Young Offenders Institute was significant in both methods (Chi-square, p < 0.0001). The age variable also showed significant associations (C.C: t-test, p < 0.0001; M.I: t-test, p < 0.0001). Table 3 Association Between Variables and HCV RNA: Table summarizing the association between various variables and HCV RNA status, using different statistical tests, and comparing Complete Case (C.C) and Multiple Imputation (M.I) methods. Variable Test Complete Case Analysis (Statistic, p-value) Interpretation (C.C) Multiple Imputation (Statistic, p-value) Interpretation (M.I) Injection Duration (injdur) Wilcoxon W = 2057993, p < 0.0001 Significant W = 14000, p < 0.0001 Significant Needle Exchange (exch) Chi-square X² = 6.28, p = 0.0122 Significant X² = 22.87, p = 0.0012 Significant Sharing of Needle (Share) Chi-square X² = 0.99, p = 0.3207 Not Significant X² = 0.60, p = 1.0000 Not Significant Ever Homeless (homeless1) Chi-square X² = 35.73, p < 0.0001 Significant X² = 104.49, p < 0.0001 Significant Prison before Injection commencement (prisbeforeinj) Chi-square X² = 16.80, p < 0.0001 Significant X² = 5.78, p = 0.2434 Not Significant Gender (gen) Chi-square X² = 16.956, p < 0.0001 Significant X² = 83.71, p < 0.0001 Significant Condom Use (condom1) Chi-square X² = 8.16, p = 0.0170 Significant X² = 9.61, p = 0.3547 Not Significant Ever been in Prison (epris1) Chi-square X² = 101.68, p < 0.0001 Significant X² = 101.68, p < 0.0001 Significant Age T-test t = -7.33, p < 0.0001 Significant t = 13.86, p < 0.0001 Significant As shown in Table 4 , the Pearson correlation results provide insights into the strength and direction of relationships between variables. For injection duration, the C.C. method showed a very weak positive correlation (0.0805), while the M.I. method indicated a moderate negative correlation (-0.5253). The exchange variable had a moderate positive correlation in the C.C. method (0.5284) and a strong positive correlation in the M.I. method (0.8045). The sharing of injection showed a moderate negative correlation in the C.C. method (-0.5896) and a strong negative correlation in the M.I. method (-0.7759). Condom use had a very weak positive correlation in the C.C. method (0.1484) and a weak positive correlation in the M.I. method (0.2768). Homeless Status showed weak positive correlations in both methods (C.C: 0.3287; M.I: 0.3681). Previous imprisonment before Injection had very weak positive correlations in both methods (C.C: 0.1544; M.I: 0.1728). The Ever in Prison or Young Offenders Institute had a strong positive correlation in the C.C. method (0.6776) and a moderate positive correlation in the M.I. method (0.5117). Table 4 Pearson Correlation Table: Pearson correlation coefficients and interpretations for Complete Case (C.C) and Multiple Imputation (M.I) methods, illustrating the strength and direction of relationships between variables and HCV RNA status. Variable Pearson Corr (C.C) Interpretation (C.C) Pearson Corr (M.I) Interpretation (M.I) Injection Duration (injdur) 0.0805 Very weak positive -0.5253 Moderate negative Needle Exchange Programs (exch) 0.5284 Moderate positive 0.8045 Strong positive Sharing of Needle (Share) -0.5896 Moderate negative -0.7759 Strong negative Use of Condom (condom1) 0.1484 Very weak positive 0.2768 Weak positive Ever Homeless (homeless1) 0.3287 Weak positive 0.3681 Weak positive Prison before Injection Commencement (prisbeforeinj)(prisbeforeinj) 0.1544 Very weak positive 0.1728 Very weak positive Ever in Prison (epris1) 0.6776 Strong positive 0.5117 Moderate positive Epris1 (“Ever in Prison or Young Offenders Institute”) In the Complete Case (C.C) part of Fig. 5, Ever in Prison/Young Offenders Institute, represented by the blue line, shows significant variability. From 2011 to 2021, the trend starts high and remains relatively high throughout the period, with sharp fluctuation from 2014 and 2016 through 2021. A notable peak occurred around 2013, indicating increased reported cases of previous imprisonment. Following this peak, there is a gradual decline, although the values remain elevated compared to other variables. The sharp rise and fall suggest data inconsistencies or heightened reporting during these years. The Multiple Imputation (M.I) image shows a more stable trend for epris1, represented by the blue line. It started at a high level in 2011 and peaked in 2013, but the trend remains relatively constant with fewer fluctuations. There has been a slight decline over the years, but the overall trend is smoother and more stable. Homeless Status The homeless Status variable in the C.C. image of Fig. 5, depicted by the orange line, shows moderate fluctuations. From 2011, the trend starts at a lower level compared to epris1for “Ever in Prison or Young Offenders Institute” but shows an increase around 2014–2015, drops around 2016–2017, and followed by stabilization in subsequent years. In the M.I. image of Fig. 5, the homeless status variable, shown by the orange line, presents a similar trend. The prevalence starts higher than in the C.C. method and remains relatively constant. It also records the decline around 2016–2017, similarly seen in the Complete Case Analysis method. Prison Before Injection Commencement The prison before injection variable in the C.C. image of Fig. 5, represented by the green line, exhibits a relatively stable trend with minor fluctuations. From 2011, there was a sharp increase around 2013 and a corresponding decrease in 2014, but overall, the trend does not show significant changes. This suggests that this variable's data might be more consistently reported or less affected by missing data than the other variables. In the M.I. image of Fig. 5, the prison before injection variable, depicted by the brown line, also shows a stable trend but a generally slight decline over the years, with its nadir in 2014. The trend is smoother than the C.C method, indicating that the imputation method effectively fills in missing data, leading to a more reliable representation of the variable over time. The smoother trend in the M.I. image suggests that imputation helps create a more accurate and less volatile depiction of the data. Needle Exchange Programs In the Complete Case (C.C) image of Fig. 6, the needle exchange prevalence trend (purple line) is consistently high from 2011 to 2021. The values remain close to the maximum (100%), indicating a high prevalence of needle exchange programs use amongst participants or consistent data collection practices. The trend line shows minor fluctuations but maintains a steady trajectory, suggesting reliable reporting and a low impact of missing data on this variable. The multiple imputation (M.I.) image in Fig. 6 shows the needle exchange prevalence trend (purple line) as equally high. The trend line in the M.I. image is almost like that of the Complete Case Analysis method. The prevalence remains high across all years, identical to the C.C. method, which implies that the data for this variable is consistently robust in both methods. Needle Sharing Prevalence The needle-sharing prevalence trend (Blue line) in the C.C. image of Fig. 6 starts at a moderate level in 2011 and shows noticeable variability over the years. Initially, the trend is relatively low, but it increases gradually but irregularly, peaking significantly around 2019–2021. This increase suggests variability in data collection or an actual rise in needle-sharing practices during these years. In the M.I. image, the needle-sharing prevalence trend (Blue line) is more stable and smoother than the C.C method. The trend started at a similar moderate level in 2011 but exhibited a more gradual increase over the years. Injection Duration The injection duration trend (red line) in the C.C. image of Fig. 6 shows moderate stability with some fluctuations. The trend starts at a low level in 2011 and gradually increases, peaking around 2014–2015. After this peak, the trend stabilized somewhat but still showed minor fluctuations in the later years, especially a decline between 2017 and 2019. This pattern suggests variability in data reporting or an actual change in injection practices over time, with the peak around 2014–2015 potentially reflecting changes in behavior or reporting accuracy. In the M.I. image of Fig. 6, the injection duration trend (red line) shows a more consistent and smoother pattern. Starting at a low level in 2011, the trend gradually increased with less pronounced fluctuations than the C.C method. The trend line remains relatively stable, with a slight upward trajectory, indicating that the imputation method provides a more precise and reliable depiction of injection duration trends over time, addressing the missing data. 3.2.7 Condom Use Prevalence In the Complete Case (C.C) image of Fig. 7 below, the condom use prevalence trend (blue line) started at a moderate level around 2011 and maintained a relatively stable pattern with slight fluctuations throughout the period. The multiple Imputation (M.I) method in Fig. 7 shows the condom use prevalence trend (blue line) to have a similar trend and pattern as the Complete case method image. The trend started at a similar moderate level in 2011 and continued with fluctuations throughout the period and a few sharp peaks and significant increases, like the C.C. method. 3.3 Comparative Analysis of Univariate Regression: Complete Case (C.C) vs. Multiple Imputation (M.I) The univariate regression analysis, as shown in Table 5 , examines the relationships between the independent and dependent variables. This section compares the results obtained from the Complete Case (C.C) and the Multiple Imputation (M.I) methods, highlighting the differences in odds ratios, confidence intervals, and p-values. For injection duration (injdur), the odds ratio (OR) in the C.C method is 1.03 (95% CI: 1.02–1.04) with a highly significant p-value of < 0.0001. In the M.I. method, the OR is slightly higher at 1.03 (95% CI: 1.03–1.04) with an equally significant p-value of < 0.0001. Both methods indicate a significant positive association between injury duration and the dependent variable, HCVRNA. The C.C. method’s injection exchange programs (exch) variable shows an OR of 1.86 (95% CI: 1.18–3.07), with a significant p-value of 0.0101. The M.I. method provides an OR of 1.55 (95% CI: 1.29–1.87) with a p-value of < 0.0001. Both methods demonstrate a significant positive association, with similar ORs and overlapping confidence intervals, indicating consistent findings and less impact of missing values. For sharing of injection (share), the C.C. method shows an OR of 1.08 (95% CI: 0.93–1.26) with a non-significant p-value of 0.3019. The M.I. method reports an OR of 1.02 (95% CI: 0.91–1.14) with a non-significant p-value of 0.738. Both methods indicate a non-significant association, with close ORs and overlapping confidence intervals, confirming the lack of a significant effect. The homeless status variable in the C.C. method has an OR of 1.38 (95% CI: 1.15–1.65) with a highly significant p-value of 0.0005. In the M.I. method, the OR is higher at 1.47 (95% CI: 1.31–1.65) with a p-value of < 0.0001. Both methods show a significant positive association, with the M.I. method showing a marginally higher Odds Ratio. Ever in Prison, or Young Offenders Institute variable (prisbeforeinj) reveals an OR of 1.31 (95% CI: 1.15–1.49) in the C.C method, with a significant p-value of < 0.0001. The M.I. method shows a significantly higher OR of 2.03 (95% CI: 1.86–2.22) with a highly significant p-value of < 0.0001. This suggests that imputing missing data uncovers a stronger relationship between previous imprisonment and the dependent variable. For age, the C.C. method reports an OR of 1.03 (95% CI: 1.02–1.04) with a highly significant p-value of < 0.0001. The M.I. method provides an OR of 1.03 (95% CI: 1.02–1.03) with a p-value of < 0.0001. Both methods indicate a significant positive association, with very close ORs and confidence intervals, confirming the robustness of the association. The gender variable in the C.C method shows an OR of 0.75 (95% CI: 0.65–0.86) with a significant p-value of < 0.0001. The M.I method has an OR of 0.69 (95% CI: 0.64–0.75) with a p-value of < 0.0001. Both methods demonstrate a significant negative association, with the M.I. method indicating a slightly weaker effect but consistent findings due to overlapping confidence intervals. For condom use, the C.C. method shows an OR of 1.05 (95% CI: 0.92–1.21) with a non-significant p-value of 0.4519. The M.I. method reports an OR of 1.08 (95% CI: 0.97–1.20) with a non-significant p-value of 0.1721. This discrepancy highlights the impact of handling missing data, where both methods suggest the relationship is insignificant. Table 5 Univariate Table: Univariate analysis showing odds ratios, confidence intervals, and p-values for Complete Case (C.C) and Multiple Imputation (M.I) methods, identifying key variables associated with HCV RNA status. Variable C.C OR (CI) C.C P-Value Interpretation (C.C) M.I OR (CI) M.I P-Value Interpretation (M.I) Injection Duration (injdur) 1.03 (1.02–1.04) < 0.0001 Statistically significant 1.03 (1.03–1.04) < 0.0001 Statistically significant Needle Exchange Programs (exch) 1.86 (1.16-3.00) 0.0101 Statistically significant 1.55 (1.29–1.87) < 0.0001 Statistically significant Sharing of Needle (Share) 1.08 (0.93–1.26) 0.302 Non-statistically significant 1.02 (0.91–1.14) 0.738 Non-statistically significant Prison before injection Commencement (prisbeforeinj1) 1.31 (1.15–1.49) < 0.0001 Statistically significant 2.03 (1.86–2.22) < 0.0001 Statistically significant Age 1.03 (1.02–1.04) < 0.0001 Statistically significant 1.03 (1.02–1.03) < 0.0001 Statistically significant Gender (gen1) 0.75 (0.65–0.86) < 0.0001 Statistically significant 0.69 (0.64–0.76) < 0.0001 Statistically significant Year − 2012 1.00 (0.79–1.27) 0.974 Non-statistically significant 1.01 (0.87–1.18) 0.9015 Non-statistically significant Year − 2013 1.01 (0.79–1.29) 0.934 Non-statistically significant 0.99 (0.85–1.16) 0.9343 Non-statistically significant Year - 2014 2.92 (2.19–3.88) < 0.0001 Non-statistically significant 1.00 (0.85–1.17) 0.9503 Non-statistically significant Year − 2015 4.02 (2.95–5.49) < 0.0001 Non-statistically significant 0.92 (0.78–1.08) 0.2993 Non-statistically significant Year − 2016 1.03 (0.79–1.35) 0.814 Non-statistically significant 0.87 (0.74–1.02) 0.0777 Non-statistically significant Year − 2017 0.96 (0.73–1.26) 0.754 Non-statistically significant 1.01 (0.85–1.19) 0.9519 Non-statistically significant Year − 2018 1.39 (1.06–1.82) 0.0162 Statistically significant 0.84 (0.72–0.98) 0.0296 Statistically significant Year − 2019 0.88 (0.67–1.15) 0.339 Non-statistically significant 1.07 (0.91–1.26) 0.3918 Non-statistically significant Year − 2020 0.71 (0.44–1.16) 0.169 Non-statistically significant 1.66 (1.23–2.23) 0.00281 Statistically significant Year − 2021 0.47 (0.31–0.73) 0.000824 Statistically significant 1.90 (1.46–2.47) < 0.0001 Statistically significant Ever Homeless (homeless1) 2.30 (1.62–3.27) 0.000469 Statistically significant 1.47 (1.31–1.65) < 0.0001 Statistically significant Condom Use (condom1) 1.36 (0.99–1.87) 0.452 Non-statistically significant 1.08 (0.97–1.20) 0.1721 Non-statistically significant Comparative Analysis of Multivariate Logistic Regression: Complete Case (C.C) vs. Multiple Imputation (M.I) The multivariate logistic regression analysis results, as shown in Table 6 , provide insights into the relationships between multiple independent variables and the dependent variable simultaneously. This analysis compares the results obtained from the Complete Case (C.C) method and the Multiple Imputation (M.I) method, using forest plots to visually represent the odds ratios and confidence intervals. In the Complete Case (C.C) method, the intercept shows an odds ratio (OR) of 0.10 (95% CI: 0.05–0.19), with a highly significant p-value of < 0.0001. This suggests a strong baseline effect when all other variables are at zero. In contrast, the Multiple Imputation (M.I) method provides an OR of 0.12 (95% CI: 0.09–0.16), with a significant p-value of < 0.0001, indicating a robust baseline effect with imputed data. For injection duration (injdur), the C.C. method reports an OR of 1.02 (95% CI: 1.01–1.04) with a significant p-value < 0.0001, indicating a positive association. The M.I. method shows a slightly lower OR of 1.03 (95% CI: 1.02–1.04) with a p-value < 0.0001, suggesting a consistently significant association but slightly reduced effect size. The Injection Exchange Programs (exch1) variable in the C.C method has an OR of 1.51 (95% CI: 0.92–2.46) and a non-significant p-value of 0.102. The M.I. method reports a lower OR of 1.30 (95% CI: 1.07–1.56) with a significant p-value of 0.0069, indicating a positive association with imputed data. Sharing of injection (share1) in the C.C method has an OR of 1.24 (95% CI: 1.06–1.46) with a p-value of 0.0081, suggesting a significant association. The M.I. method provides an OR of 1.13 (95% CI: 1.01–1.27) with a significant p-value of 0.035, indicating that the association is also significant when missing data is accounted for. For people experiencing homelessness but not last year (homeless11), the C.C. method shows an OR of 1.18 (95% CI: 0.98–1.43) with a p-value of 0.0811, indicating a non-significant association. The M.I. method reports a higher OR of 1.35 (95% CI: 1.20–1.52) with a highly significant p-value of < 0.0001, suggesting a significant association with imputed data. The second category for Homeless last year (homeless12) in the C.C method shows an OR of 1.59 (95% CI: 1.30–1.86) with a highly significant p-value of < 0.0001. The M.I. method has a similar OR of 1.61 (95% CI: 1.44–1.81) with a significant p-value < 0.0001. Both methods indicate a strong positive association, with the M.I. method providing a slightly narrower confidence interval. For condom use (condom11), the C.C. method shows an OR of 1.10 (95% CI: 0.95–1.28) with a non-significant p-value of 0.1880. The M.I. method has an OR of 1.11 (95% CI: 0.99–1.23) with a p-value of 0.0929, which is also non-significant. Both methods indicate that the association between condom use and HCVRNA is not significant. In the second category of condom use (condom12), the C.C. method shows an OR of 1.26 (95% CI: 1.04–1.51) with a significant p-value of 0.0155. The M.I. method reports an OR of 1.14 (95% CI: 0.97–1.34) with a non-significant p-value of 0.1522. This indicates that the initial significant association observed in the C.C. method is not supported when missing data is imputed. For previous imprisonment before injection commencement (prisbeforeinj1), the C.C method shows an OR of 1.30 (95% CI: 1.13–1.51) with a significant p-value of 0.0004. The M.I. method reports an OR of 1.01 (95% CI: 0.91–1.13) with a non-significant p-value of 0.8133, suggesting that the significant association in the C.C. method is not present when missing data is accounted for. Age in the C.C. method has an OR of 1.01 (95% CI: 1.00–1.02) with a non-significant p-value of 0.0596. The M.I. method shows an OR of 1.01 (95% CI: 1.01–1.02) with a significant p-value of 0.0001. This indicates that age becomes a significant factor after imputation. Finally, for gender (gen1), the C.C. method shows an OR of 0.89 (95% CI: 0.77–1.04) with a non-significant p-value of 0.147. The M.I. method reports a lower OR of 0.77 (95% CI: 0.71–0.84) with a highly significant p-value of < 0.0001, suggesting a stronger and more significant association when missing data is accounted for. Table 6 Multivariate Regression Table: This multivariate regression analysis compares odds ratios, confidence intervals, and p-values for Complete Case (C.C) and Multiple Imputation (M.I) methods, highlighting significant associations with HCV RNA status. Variable OddsRatio (95% CI) C.C p_value C.C Interpretation C.C OddsRatio (95% CI) M.M p_value M.M Interpretation M.M (Intercept) 0.10 (0.05–0.19) < 0.0001 Statistically significant 0.12 (0.09–0.16) < 0.0001 Statistically significant Injection Duration (injdur) 1.02 (1.01–1.04) < 0.0001 Statistically significant 1.03 (1.02–1.04) < 0.0001 Statistically significant Needle exchange programs (exch1) 1.51 (0.92–2.46) 0.1020 Not statistically significant 1.30 (1.07–1.56) 0.0069 Statistically significant Sharing of Needle (Share) 1.24 (1.06–1.46) 0.0081 Statistically significant 1.13 (1.01–1.27) 0.0359 Statistically significant Homeless, but not in the last year (homeless11) 1.18 (0.98–1.43) 0.0811 Not statistically significant 1.35 (1.20–1.52) < 0.0001 Statistically significant Homeless in the last year (homeless12) 1.56 (1.30–1.86) < 0.0001 Statistically significant 1.61 (1.44–1.81) < 0.0001 Statistically significant Prison before injection commencement (prisbeforeinj1) 1.30 (1.13–1.51) 0.0004 Statistically significant 1.01 (0.91–1.13) 0.8133 Not statistically significant Sometimes Used Condoms (condom11) 1.10 (0.95–1.28) 0.1880 Not statistically significant 1.11 (0.99–1.23) 0.0929 Not statistically significant Always Used Condoms (condom12) 1.26 (1.04–1.51) 0.0155 Statistically significant 1.14 (0.97–1.34) 0.1522 Not statistically significant Age 1.01 (1.00–1.02) 0.0596 Borderline Non-statistical significance 1.01 (1.01–1.02) 0.0001 Statistically significant Gender (gen1) 0.89 (0.77–1.04) 0.1420 Not statistically significant 0.77 (0.71–0.84) < 0.0001 Statistically significant Year − 2012 1.00 (0.78–1.27) 0.9830 Not statistically significant 1.00 (0.86–1.17) 0.9992 Not statistically significant Year − 2013 0.95 (0.74–1.22) 0.6730 Not statistically significant 0.95 (0.81–1.10) 0.4812 Not statistically significant Year − 2014 2.49 (1.85–3.34) < 0.0001 Statistically significant 0.92 (0.78–1.09) 0.3233 Not statistically significant Year − 2015 3.37 (2.45–4.63) < 0.0001 Statistically significant 0.98 (0.83–1.16) 0.8209 Not statistically significant Year − 2016 0.88 (0.67–1.16) 0.3560 Not statistically significant 1.01 (0.86–1.19) 0.8739 Not statistically significant Year − 2017 0.81 (0.61–1.07) 0.1360 Not statistically significant 0.85 (0.71–1.01) 0.0717 Not statistically significant Year − 2018 1.20 (0.91–1.58) 0.1980 Not statistically significant 1.00 (0.85–1.17) 0.9862 Not statistically significant Year − 2019 0.71 (0.54–0.94) 0.0172 Statistically significant 0.77 (0.65–0.91) 0.0023 Statistically significant Year − 2020 0.51 (0.31–0.84) 0.0081 Statistically significant 0.47 (0.34–0.64) < 0.0001 Statistically significant Year − 2021 0.37 (0.24–0.58) < 0.0001 Statistically significant 0.40 (0.31–0.53) < 0.0001 Statistically significant Forest Plots The forest plots in Fig. 8 below provide a visual representation of the odds ratios and confidence intervals for the variables in both the C.C. and M.I. methods. The plots clearly illustrate the differences in the magnitude and significance of associations between the two methods. In general, the M.I. method shows tighter confidence intervals and more significant associations for several variables. DISCUSSION The discussion section interprets the results of our study, comparing the implications of using complete case analysis versus multiple imputation methods for estimating HCV prevalence and identifying key risk factors. We also explore the broader impact of our findings on public health strategies and policies aimed at controlling and preventing HCV among high-risk populations. The Complete Case analysis results of this study revealed a bell-shaped age distribution curve with the highest frequencies between ages 35 and 45, a median age of 35.5 years, and a mean age of 35.86 years. In comparison, the multiple imputation analysis results also had the highest frequencies between ages 35 and 45, with a slightly higher median age of 37 and a mean age of 37.68 years, reflecting the inclusion of missing data. These findings align with trends observed in the literature and data from the Office for Health Improvements and Disparities, where the age of people in treatment for substance use are the older age groups. More than half of the individuals in treatment were over 40 years old, with significant proportions in the 40 to 44, 45 to 49, and 50 to 54 age brackets. This trend is attributed to many opiate users having initiated heroin use during the epidemics of the 1980s and 1990s, resulting in a current treatment population that is predominantly older. The median age for those in treatment for opiates was 43, reinforcing the pattern of an ageing cohort within this group. This comparison highlights the importance of considering age distribution trends when analyzing and interpreting data related to substance use treatment populations ( 29 ) and brings to the fore their susceptibility to developing liver cirrhosis ( 30 ) Our study's findings on the mean duration of injection practices using the C.C. and M.I. methods align with observations from the Needle Exchange Surveillance Initiative (NESI) report. In our study, the mean injection duration was 13.78 years for the C.C. method and 14.83 years for the M.I. method. These figures reflect long-term injection behaviors, which is a common trend among people who inject drugs (PWID). The NESI report does not explicitly provide an average duration of drug injection in years. However, it does highlight the ageing cohort of PWID, indicating long-term injection practices. The report mentions an increasing average age of participants, rising from 33 years old in 2008–09 to 41 years old in 2019–20, suggesting prolonged engagement in injection drug use. The NESI report also discusses the static nature of the age at first injection, reinforcing the idea of long-term injection behaviors ( 31 ). This consistency aligns with our findings, as it implies sustained injection practices among older PWID rather than new injectors starting at an older age and brings to the fore the need for ongoing support services to mitigate health risks, such as blood-borne viruses, that increase with prolonged injection use. HCV Prevalence The C.C method showed a volatile trend with a spike between 2013 and 2014 and a peak in 2015 which may have occurred because HCV RNA data was missing for all antibody-negative results in 2013/2014, while the M.I method indicated a smoother trend with a gradual increase from 2011 to 2016 (30.7% -33.8%), a sharp drop in 2017 (30.5%) then another increase in 2018 (34.4%) which was followed subsequently by a steady decline from 2017 to 2021. For the years 2017–2021, the unadjusted results, C.C. indicates higher ORs for 2014 (2.92) and 2015 (4.02), which are statistically significant, while M.I results for these years are closer to 1 and non-significant. Both methods show significance for 2018 and 2021, but M.I. shows consistently lower p-values, indicating higher reliability and higher statistical power. In the adjusted results, C.C. shows significant ORs for 2014 (2.49) and 2015 (3.37), which are non-significant in M.I. Both methods indicate significance for 2019, 2020, and 2021, but the ORs are closer in M.I (e.g., 2021: C.C OR 0.37, M.I OR 0.40), suggesting more consistent and reliable estimates. Association of variables with HCV RNA Injection Duration and HCV RNA Injection duration (Injdur) was significantly associated with HCV RNA status in both methods in the chi-square analysis, adjusted and unadjusted regression analysis, which is consistent with findings from Hope et al. (2020) ( 32 ), Alavi et al. (2019) ( 33 ) which reported that longer injection durations increase the risk of HCV infection. This association is likely due to the prolonged exposure and increased opportunities for risky behaviors, such as sharing needles or other injecting equipment, which heighten the risk of HCV transmission over time. Additionally, longer injection durations may reflect a chronic condition of substance use that is harder to manage and treat, further compounding the risk of infection. However, the Pearson correlation analysis for injection duration over the ten-year period revealed a very weak positive correlation in the C.C. method and a moderate negative correlation in the M.I. method. The justification of this M.I. Pearson correlation results is that while longer injection durations increase the overall risk of HCV infection, the proportion of HCV antibody-positive individuals who are HCV RNA positive may decrease over time. This is because individuals who have been injecting for longer are more likely to have been treated and thus cleared the virus, reducing the proportion of those currently infected despite a longer duration of risky behavior. Therefore, this justification is consistent with studies such as Alavi et. al (2019), which highlighted that prolonged injection duration often leads to better engagement with harm reduction services over time ( 33 ), potentially reducing HCV RNA prevalence among HCV antibody-positive patients. 4.2.2 Needle Exchange Programs and HCV RNA Using needle exchange programs (Exch) showed significant associations in both methods, chi-square results and unadjusted analysis. These results indicate that needle exchange programs are associated with higher odds of HCV RNA positivity. However, in the adjusted analysis, which considers multiple variables simultaneously, the association between needle exchange use and HCV RNA status changed as the C.C. method indicated a non-significant association. In contrast, in the M.I. method, the association remained significant. These findings suggest that even after controlling for other variables, the use of needle exchange programs is still associated with a higher likelihood of HCV RNA positivity in the M.I method. However, this does not align with studies like Des Jarlais et al. (2018) and Grebely et al. (2017), which found that effective needle exchange programs significantly reduce HCV incidence among PWID ( 34 , 35 ). While needle exchange programs are designed to reduce the spread of HCV by providing clean injecting equipment and promoting safer injecting practices, our findings indicate a significant positive association between needle exchange use and HCV RNA positivity. This counter-intuitive result can be attributed to the fact that in our study and for our study population, individuals at higher risk of HCV infection, likely due to factors such as more frequent injecting and sharing of equipment, are more likely to use needle exchange programs, rather than the programs themselves increasing the risk of HCV infection. This suggests the possibility that higher-risk individuals are more likely to use needle exchange programs, as the use of needle exchange itself is unlikely to be a causal factor for infection. 4.2.3 Needle Sharing and HCV RNA The needle-sharing variable had a missing data percentage of 27% in this study. In the unadjusted results, sharing needles had non-significant associations, and the odds ratio dropped between the Complete Case (C.C) and Multiple Imputation (M.I) methods. However, in the adjusted results, sharing needles was found to have a significant association with HCV RNA status in both the C.C and M.I methods, with the change in OR from 24%(C.C) to 13% (M.I) increased likelihood, indicating that the behavior of sharing needles significantly increased the possibility of HCV RNA positivity in this study population in both methods with the M.I results showing a lower odds ratio, possibly the accurate picture after accounting for missing data. This finding aligns with the results of Bruneau et al. (2020), who noted that among injection drug users, sharing syringes is the most significant factor leading to HCV seroconversion ( 36 )​. Similarly, Mateu-Gelabert et al. (2010) emphasized that Injection drug users are more prone to engaging in risky injection behaviors during withdrawal periods ( 37 )​. Moreover, Platt et al. (2017) highlighted that research indicates that using needles and syringes previously used by others is the primary risk factor for contracting HIV and HCV among people who inject drugs ( 38 )​. These studies collectively emphasize needle sharing as a critical risk factor for HCV infection, supporting the findings of this analysis that demonstrate improved accuracy and reliability through multiple imputations. Homelessness and HCV RNA Homeless status (Homeless1) showed a strong significant association with HCV RNA in both methods and in univariate and multivariate analyses, indicating that homeless individuals are significantly more likely to be HCV RNA positive even after adjusting for confounding variables. Homelessness is associated with numerous risk factors for HCV, including unstable living conditions, lack of access to healthcare, and a higher likelihood of engaging in high-risk behaviors such as sharing needles or having unprotected sex ( 39 ). These factors contribute to the higher prevalence of HCV among homeless individuals, highlighting the need for targeted interventions to address the unique challenges faced by this population. These findings collectively highlight the strong association between homelessness and HCV RNA positivity. The significant associations found in both analyses and the positive correlations observed in the Pearson analysis suggest a consistent relationship between homelessness and higher HCV prevalence. Comparing these findings with the literature, Nyamathi et al. (2019) and Beijer et al. (2018) also reported higher HCV prevalence among homeless individuals ( 40 , 41 ), which aligns with our results. This consistency across studies and methods reinforces the need for targeted interventions and comprehensive support services to address the unique challenges faced by homeless individuals to curb the prevalence of hepatitis C. Prison before injection commencement and HCV RNA Prison before injection commencement association with HCV RNA was significant in Complete Case Analysis but not in Multiple Imputation in Chi-square results. The Pearson correlation for prison before injection commencement showed very weak positive correlations in both methods, suggests that other factors, along with imprisonment, might contribute to the risk of HCV infection. In the unadjusted analysis, previous imprisonment before injection commencement was found to have a significant positive association with HCV RNA status in both methods. After adjusting for confounding factors in the multivariate regression analysis, the association between previous imprisonment before injection commencement and HCV RNA status was significant in the C.C. method but not in the M.I. method. This suggests that, after accounting for missing data, the significance of the association diminishes, indicating that other factors may be at play. These results collectively indicate that while previous imprisonment before injection commencement is associated with HCV RNA positivity, this association’s strength and significance can vary depending on the method used to handle missing data. The non-significant association in the M.I. method differs from results by Stone et al. (2021) and Degenhardt et al. (2017), which found higher HCV prevalence among those who have started using drug injections after going to prison ( 42 , 43 ). Rivera Saldana. Et al. study findings indicate a significant relationship between previous imprisonment before injection commencement and various risk behaviors including drug injection, among participants. The increased risk can be attributed to several factors, including the social and economic instability often following release from prison. The study also revealed that the criminalization of drug use and the punitive measures associated with it exacerbate these risks by perpetuating cycles of incarceration and risky behaviors ( 44 ). The differences in significance between the methods in my study might be attributed to the nature of the data handling. The C.C method also may overrepresent specific subgroups due to excluding cases with missing data (10% of data in this variable is missing), while M.I attempt to account for missing data, potentially providing a more balanced view. The positive correlations observed in the Pearson analysis further suggest that while injection commencement after prison is an essential factor, other variables also contribute to the risk of HCV infection, necessitating comprehensive prevention and treatment strategies that address multiple risk factors. 4.2.6 Gender and HCV RNA Gender differences were significant in the chi-square of both methods, with males showing higher HCV prevalence. The Pearson correlation analysis showed a very weak positive correlation in the C.C. method and a weak positive correlation in the M.I. method, suggesting that males are slightly more likely to be HCV positive. In the univariate regression analysis, gender was found to have a significant association with HCV RNA status in both methods. These findings suggest a significant gender difference, and the interpretation of the OR less than 1 indicates that females, in this context, might have lower odds of being HCV RNA positive compared to males, which is consistent with the initial chi-square results and literature findings. In the multivariate regression analysis, gender continued to show significant associations with HCV RNA status in only one of the methods. In the C.C. method, a non-significant p-value of 0.1472 suggests no significant difference in HCV RNA positivity between genders when adjusting for other variables. However, in the M.I. method, a highly significant p-value of > 0.001 indicates that females are significantly less likely to be HCV RNA positive compared to males after adjusting for other variables and with 23.45% lower odds compared to males. This result, again, appears consistent with the initial chi-square findings and unadjusted findings and aligns with findings by Zibbell et al. (2018) and Uusküla et al. (2020), which have shown that males are more likely to be hepatitis C positive ( 45 , 46 ). This trend may reflect higher engagement in risky injecting behaviors among males compared to females, as well as differential access to healthcare and harm reduction services. Additionally, social, and cultural factors might influence the likelihood of males seeking help or accessing services, further contributing to the observed gender disparity in HCV prevalence. 4.2.7 Condom Use and HCV RNA The use of condoms was significant in the Complete Case (C.C) method but not in the Multiple Imputation (M.I) method in the chi-square test. This difference may be due to the higher variability and missing data in the Condom1 variable (20%), which the M.I. method attempted to address by imputation. The Pearson correlation results showed weak positive correlations in both methods, indicating that condom use is generally associated with a lower risk of HCV. However, its impact may be less pronounced compared to other variables, such as needle sharing. In the univariate regression analysis, condom use was found to have significant associations with HCV RNA status in the C.C method. Conversely, in the M.I method, the non-significant p-value indicated no significant association between condom use and HCV RNA positivity after accounting for missing data. After adjusting for potential confounders for individuals who sometimes use condoms, both the C.C. and M.I analyses showed non-significant p-values, with ORs around 1.10, suggesting a 10% higher likelihood but statistically insignificant association with HCV prevalence. For individuals who always use condoms, the C.C analysis indicated a significant association with an OR of 1.31, meaning a 31% higher odds of HCV prevalence. In contrast, the M.I analysis showed a non-significant p-value with an OR of 1.14, indicating 14% higher odds of HCV. These findings suggest that after controlling for other factors, always using condoms is not significantly associated with being HCV positive. Studies by Leyna GH et al. have shown a correlation between condom use and HCV infection risk in People Who Inject Drugs ( 47 ), which doesn’t corroborate the multivariate Multiple Imputation findings. The significant association found in the C.C. method might be due to unaccounted confounding factors and the effect of missing data. Given that most HCV infections occur through injection rather than sexual activity, the relevance of condom use may be limited in this context. Therefore, the association between sexual behavior and injecting behavior may be more critical, and after controlling for variables like gender and age, condom use appears less significant. 4.2.8 Ever Been in Prison/Young Offenders Institute and HCV RNA Ever Been In Prison (Epris1) was significantly associated with HCV RNA in both methods in the Chi-square test, and the Pearson correlation for HCV positivity in those that have been to prison showed very weak positive correlations in both methods, indicating higher HCV prevalence among incarcerated individuals. Prisons are high-risk environments for HCV transmission due to overcrowding, limited access to clean injecting equipment, and inadequate healthcare services. The findings of this study are in line with the research conducted by Stone et al. (2021) and Dolan et al. (2016), which also reported higher HCV prevalence among incarcerated populations ( 42 , 48 ). These studies highlighted similar contributing factors, such as overcrowding and inadequate healthcare. Degenhardt et al. (2017) further corroborated these findings by emphasizing the role of injecting equipment sharing and the lack of harm reduction services in prison ( 43 ). However, it is essential to consider the bidirectional nature of this correlation, as individuals with high-risk injecting behaviors are more likely to be imprisoned. Those with severe addiction may engage in more risky behaviors, such as sharing needles, which increases their likelihood of contracting HCV and subsequently being incarcerated for related crimes. This hypothesis suggests that high-risk individuals are both more susceptible to HCV and more likely to be incarcerated, thus reinforcing the observed correlation between imprisonment and HCV prevalence. STRENGTHS AND LIMITATIONS STRENGTHS The study demonstrates several strengths that significantly enhance its contributions to epidemiology and public health. Firstly, the use of both complete case analysis (CCA) and multiple imputation (MI) techniques to handle missing data provides a comprehensive evaluation of the impact of different data handling methods on estimating Hepatitis C prevalence among people who inject drugs (PWID). This dual-method approach allows for an understanding of how missing data can affect epidemiological findings, thus highlighting potential biases inherent in C.C. analysis and the benefits of MI in producing more reliable results. The study offers valuable insights into effective strategies for managing missing data in public health research by demonstrating how different data handling methods influence prevalence trends. Additionally, the study's large sample size, encompassing approximately 17,000 participants, significantly enhances the statistical power and recruitment from drug services from multiple sites across England, Wales and Northern Ireland, alongside the anonymity of the data collection process, which serves to make it more representative. One of the key strengths of this research is the use of the Predictive Mean Matching (PMM) type of multiple imputation technique, which effectively accounts for non-normal and categorical variables, enhancing the robustness of the imputation process. Additionally, through model-based diagnostics, efforts were made to assess the plausibility of the Missing at Random (MAR) assumption. Specifically, a logistic regression was fitted to examine the relationship between all variables used in the imputation model and the outcome variable. This method helps to identify patterns and relationships in the data, supporting the MAR assumption for the outcome variable. However, assessing the MAR assumption fully requires examining the missingness patterns in all variables, not just the outcome. Little’s test was also conducted to rule out the possibility of the missingness being Missing Completely At Random (MCAR) and tilt the missingness more towards MAR which is the assumption I worked with to implement the multiple imputation. Furthermore, including all variables in the dataset—those that could predict missingness, those influencing the process of missing data, and even those not directly relevant to the analysis—further strengthened the study. This comprehensive approach helps to ensure that the imputation process is as accurate and unbiased as possible, providing more reliable results. Another notable strength of the study is its robust design. It utilises an annually collected, cross-sectional observational approach with primary data from the Unlinked Anonymous Monitoring (UAM) Survey, which effectively captures real-world conditions and trends over time. Using logistic regression analysis to evaluate the relationships between multiple independent variables and Hepatitis C prevalence offers robust insights into the factors influencing HCV infection. The study includes forest plots to visually represent the logistic regression analysis results, providing a clear and concise depiction of the odds ratios and confidence intervals, which facilitates a better understanding of the significant predictors of HCV prevalence. The study also carefully controls for confounding variables, ensuring that other factors do not spuriously influence the associations observed. LIMITATIONS Limitations of the Imputation Process There are notable limitations associated with using the multiple Imputation method. The assumption that data are missing at random (MAR) may not always hold, potentially leading to biased estimates if the missingness is related to unobserved factors. Multiple imputation (MI) could introduce more bias than complete case (CC) analysis if the data are not MAR, as MI might inaccurately impute values based on the observed data patterns. When the missing data is unrelated to the outcome given the covariates, complete case analysis has minimal bias, whereas multiple imputation tends to introduce bias away from the null hypothesis ( 49 ). Due to time constraints, sensitivity analysis, which involves testing the robustness of key inferences by assuming a range of missing not-at-random (MNAR) mechanisms and re-imputing the data under these different scenarios, could not be conducted. This process would have helped ensure that findings are not overly dependent on the assumption that data is MAR. It would have provided a more comprehensive understanding of how results might vary under different missing data assumptions, such as Missing Not At Random (MNAR), as self-reporting bias on risky behaviors is possible, thereby enhancing the validity of the conclusions ( 49 , 50 ). Limitations of the Data One primary limitation is the reliance on retrospective observational data from the Unlinked Anonymous Monitoring (UAM) Survey, which may introduce recall bias and inaccuracies in self-reported behaviors and health status. Additionally, there is potential for residual confounding despite efforts to control for known confounders. The study addresses various socioeconomic and behavioral factors, but there may be other unmeasured variables influencing the prevalence and risk factors of Hepatitis C among people who inject drugs (PWID). Furthermore, the study's findings, while comprehensive, are specific to the UK context and may not be fully generalizable to other regions with different healthcare systems, social structures, and HCV prevalence rates. Finally, the cross-sectional nature of the data limits the ability to infer causality between the observed associations, necessitating cautious interpretation of the results. CONCLUSION AND FUTURE DIRECTION This study utilized a more specific and sophisticated imputation type, the Predictive Mean Matching (PMM), to handle the missing values in the dataset, alongside comparing the results to that of the Complete-Case analysis. Multiple Imputation provided a more logically accurate trend of HCV prevalence than the Complete-Case Analysis. In the adjusted regression model, several significant findings were observed related to the influence of various factors on Hepatitis C Virus (HCV) prevalence among people who inject drugs (PWID). After multiple imputations, Homelessness was only significant for participants who weren’t homeless in the last year, suggesting that while it is a critical factor, its influence varies over time and conditions. Interestingly, prior imprisonment before injection commencement was not significant for M.I., highlighting that other factors may play a more substantial role in the spread of HCV among PWID. The use of condoms, whether sometimes or always, did not show a significant impact on HCV infection despite having higher odds ratios. This indicates that while sexual transmission is relevant, it might not be the primary mode of HCV spread in this population, as the higher likelihood observed could be due to chance. Gender emerged as a significant factor in M.I., emphasizing the different risk profiles for men and women. Both methods confirmed that the years 2018 to 2021 were significant, with the OR dropping across each year, marking a period of decreased HCV prevalence. This trend coincides with the widespread use of direct-acting antivirals (DAAs) in the UK, which have significantly improved HCV treatment outcomes, reducing the overall prevalence of the virus during these years ( 51 ). These findings from multiple imputation analyses further raises questions about the authenticity of the results of the complete-case analysis, as well as those already existing studies in the literature. Future research should delve deeper into understanding the specific mechanisms through which homelessness and gender influence HCV prevalence. Longitudinal studies could provide more insight into how these factors interact over time. Additionally, exploring the role of mental health and access to medical care about HCV among PWID could uncover more comprehensive intervention points. Investigating the effectiveness of various harm reduction programs in different socio-economic contexts will also be essential to tailor culturally and regionally appropriate strategies. By focusing on these areas, future research can contribute significantly to reducing HCV prevalence and improving the overall health outcomes of PWID populations. Declarations ETHICS APPROVAL AND CONSENT TO PARTICIPATE. This research did not involve human participants or access to real-world identifiable data. The analysis was conducted using a synthetic dataset, and therefore, ethical approval was not required. ACKNOWLEDGEMENTS. Not Applicable CONSENT FOR PUBLICATION Not Applicable FUNDING STATEMENT. No funding was gotten for this research project. DECLARATION OF INTERESTS All authors declare no competing interest. DATA AVAILABILITY : This study used a synthetic dataset created to simulate the structure and properties of real-world data from the UK Health Security Agency (UKHSA). No individual-level or identifiable data were accessed. As such, the dataset is not publicly available, and no ethical approval was required, as the study did not involve human participants or confidential data. The dataset used and analyzed for this study is available on request from the corresponding author. AUTHORS CONTRIBUTION STATEMENT Adewunmi Akingbola conceptualized, analyzed the data, and drafted the manuscript. Olajumoke Adewole, Abiodun Adegbesan and Joel Chuku edited the manuscript. All authors agreed to the manuscript. References Castaneda D, Gonzalez AJ, Alomari M, Tandon K, Zervos XB. From hepatitis A to E: A critical review of viral hepatitis. World J Gastroenterol. 2021 Apr 28;27(16):1691-1715. doi: 10.3748/wjg.v27.i16.1691. PMID: 33967551; PMCID: PMC8072198. World Health Organization. Hepatitis C; 2020. Available from: https://www.who.int/news-room/fact-sheets/detail/hepatitis-c NHS UK. Hepatitis C nhs.uk. Available from: https://www.nhs.uk/conditions/hepatitis-c/ De Angelis D, Sweeting M, Ades AE, et al. An evidence synthesis approach to estimating hepatitis C prevalence in England and Wales. Stat Methods Med Res. 2008 Nov 26. doi: 10.1177/0962280208094691. Lewer D, Croxford S, Desai M, Emanuel E, Hope VD, McAuley A, et al. The characteristics of people who inject drugs in the United Kingdom: changes in age, duration, and incidence of injecting, 1980–2019, using evidence from repeated cross‐sectional surveys. Addiction. 2022;117(9):2471–80. doi: 10.1111/add.15911. Yuan J-M, Croxford S, Viviani L, Emanuel E, Phipps E, Desai M. Investigating the sociodemographic and behavioural factors associated with hepatitis C virus testing amongst people who inject drugs in England, Wales and Northern Ireland: A quantitative cross-sectional analysis. Int J Drug Policy. 2022;109(103821):103821. doi: 10.1016/j.drugpo.2022.103821. Karahalios A, Baglietto L, Carlin JB, et al. A review of the reporting and handling of missing data in cohort studies with repeated assessment of exposure measures. BMC Med Res Methodol. 2012;12:96. doi: 10.1186/1471-2288-12-96. Kwak S, Kim JH. Statistical data preparation: management of missing values and outliers. Korean J Anesthesiol. 2017 Jul;70(4):407. doi: 10.4097/kjae.2017.70.4.407. Kang H. The prevention and handling of the missing data. Korean J Anesthesiol. 2013 May;64(5):402-6. doi: 10.4097/kjae.2013.64.5.402. PMID: 23741561; PMCID: PMC3668100. Gardner AL, Charlesworth M. How to write a retrospective observational study. Anaesthesia [Internet]. 2023;78(4):521–5. Available from: http://dx.doi.org/10.1111/anae.15831 Cullen KJ, Hope VD, Croxford S, Shute J, Ncube F, Parry JV. Factors associated with recently acquired hepatitis C virus infection in people who inject drugs in England, Wales and Northern Ireland: new findings from an unlinked anonymous monitoring survey. Epidemiology and Infection [Internet]. 2015;143(7):1398–407. Available from: http://dx.doi.org/10.1017/S0950268814002040 Convenience Sampling. Sampling method | descriptive research [Internet]. Ijels.net. [cited 2024 Jul 15]. Available from: https://www.ijels.net/article_162981_06fc448b8914d42316b8d44829c88188.pdf Unlinked Anonymous Monitoring (UAM) survey of HIV and viral hepatitis among people who inject drugs (PWID): 2023 report [Internet]. Gov.uk. [cited 2024 Jul 15]. Available from: https://www.gov.uk/government/publications/people-who-inject-drugs-hiv-and-viral-hepatitis-monitoring/unlinked-anonymous-monitoring-uam-survey-of-hiv-and-viral-hepatitis-among-people-who-inject-drugs-pwid-2023-report Tully DC, Power KA, Sarette J, Stopka TJ, Friedmann PD, Korthuis PT, et al. Validation of dried blood spots for capturing hepatitis C virus diversity for genomic surveillance [Internet]. bioRxiv. 2023. p. 2023.07.06.23292160. Available from: http://medrxiv.org/content/early/2023/07/07/2023.07.06.23292160.abstract Bhandari P. Independent vs dependent variables [Internet]. Scribbr. 2022 [cited 2024 Jul 15]. Available from: https://www.scribbr.co.uk/research-methods/independent-vs-dependent-variables/ Andrade C. A student’s guide to the classification and operationalization of variables in the conceptualization and design of a clinical study: Part 1. Indian J Psychol Med [Internet]. 2021;43(2):177–9. Available from: http://dx.doi.org/10.1177/0253717621994334 Little, R.j.a. (1988) A Test of Missing Completely at Random for Multivariate Data with Missing Values. Journal of the American Statistical Association, 83, 1198-1202. - references - scientific research publishing [Internet]. Scirp.org. [cited 2024 Jul 20]. Available from: https://www.scirp.org/reference/referencespapers?referenceid=1374672 van Buuren S, Groothuis-Oudshoorn K. mice: Multivariate Imputation by Chained Equations in R . J Stat Softw [Internet]. 2011;45(3):1–67. Available from: https://www.jstatsoft.org/index.php/jss/article/view/v045i03 Morris TP, White IR, Royston P. Tuning multiple imputation by predictive mean matching and local residual draws. BMC Med Res Methodol [Internet]. 2014;14(1). Available from: http://dx.doi.org/10.1186/1471-2288-14-75 Sp PS 2013-P. Combining analysis results from multiply imputed categorical data [Internet]. Lexjansen.com. [cited 2024 Jul 15]. Available from: https://www.lexjansen.com/pharmasug/2013/SP/PharmaSUG-2013-SP03.pdf Brand J, van Buuren S, le Cessie S, van den Hout W. Combining multiple imputation and bootstrap in the analysis of cost‐effectiveness trial data. Stat Med [Internet]. 2019;38(2):210–20. Available from: http://dx.doi.org/10.1002/sim.7956 Alam S, Ayub MS, Arora S, Khan MA. An investigation of the imputation techniques for missing values in ordinal data enhancing clustering and classification analysis validity. Decision Analytics Journal [Internet]. 2023;9(100341):100341. Available from: http://dx.doi.org/10.1016/j.dajour.2023.100341 McHugh ML. The Chi-square test of independence. Biochem Med (Zagreb) [Internet]. 2013;143–9. Available from: http://dx.doi.org/10.11613/bm.2013.018 Andrade C. The P value and statistical significance: Misunderstandings, explanations, challenges, and alternatives. Indian J Psychol Med [Internet]. 2019;41(3):210–5. Available from: http://dx.doi.org/10.4103/ijpsym.ijpsym_193_19 McKnight PE, Najab J. Mann‐Whitney U test [Internet]. The Corsini Encyclopedia of Psychology. Wiley; 2010. p. 1–1. Available from: http://dx.doi.org/10.1002/9780470479216.corpsy0524 Mishra P, Singh U, Pandey C, Mishra P, Pandey G. Application of student’s t-test, analysis of variance, and covariance. Ann Card Anaesth [Internet]. 2019 [cited 2024 Jul 15];22(4):407. Available from: https://journals.lww.com/aoca/fulltext/2019/22040/application_of_student_s_t_test,_analysis_of.11.aspx Sperandei S. Understanding logistic regression analysis. Biochem Med (Zagreb) [Internet]. 2014;12–8. Available from: http://dx.doi.org/10.11613/bm.2014.003 Szumilas M. Explaining odds ratios. Journal of the Canadian Academy of Child and Adolescent Psychiatry. 2010;19(3):227. Adult substance misuse treatment statistics 2021 to 2022: report [Internet]. Gov.uk. [cited 2024 Jul 16]. Available from: https://www.gov.uk/government/statistics/substance-misuse-treatment-for-adults-statistics-2021-to-2022/adult-substance-misuse-treatment-statistics-2021-to-2022-report Pradat P, Voirin N, Tillmann HL, Chevallier M, Trépo C. Progression to cirrhosis in hepatitis C patients: an age‐dependent process. Liver Int [Internet]. 2007;27(3):335–9. Available from: http://dx.doi.org/10.1111/j.1478-3231.2006.01430.x Public Health Scotland. Needle Exchange Surveillance Initiative (NESI) - Needle Exchange Surveillance Initiative (NESI) - publications - public health Scotland [Internet]. Publichealthscotland.scot. [cited 2024 Jul 16]. Available from: https://publichealthscotland.scot/publications/needle-exchange-surveillance-initiative-nesi/needle-exchange-surveillance-initiative-nesi/ Hope VD, Eramova I, Capurro D, Donoghoe MC. Prevalence and estimation of hepatitis B and C infections in the WHO European Region: a review of data focusing on the countries outside the European Union and the European Free Trade Association. Epidemiol Infect [Internet]. 2014 [cited 2024 Jul 16];142(2):270–86. Available from: https://www.cambridge.org/core/journals/epidemiology-and-infection/article/prevalence-and-estimation-of-hepatitis-b-and-c-infections-in-the-who-european-region-a-review-of-data-focusing-on-the-countries-outside-the-european-union-and-the-european-free-trade-association/9661546AEC010338F9857A7839D9E6C6 Alavi M, Grebely J, Micallef M, Dunlop AJ, Balcomb AC, Day CA, et al. Assessment and treatment of hepatitis C virus infection among people who inject drugs in the opioid substitution setting: ETHOS study. Clin Infect Dis [Internet]. 2013 [cited 2024 Jul 16];57(suppl_2):S62–9. Available from: https://academic.oup.com/cid/article/57/suppl_2/S62/398725?login=true Des Jarlais DC, Arasteh K, McKnight C, Hagan H, Perlman D, Friedman SR. Using hepatitis C virus and herpes simplex virus-2 to track HIV among injecting drug users in New York City. Drug Alcohol Depend [Internet]. 2009;101(1–2):88–91. Available from: http://dx.doi.org/10.1016/j.drugalcdep.2008.11.007 Grebely J, Dore GJ. Prevention of hepatitis C virus in injecting drug users: A narrow window of opportunity. J Infect Dis [Internet]. 2011 [cited 2024 Jul 16];203(5):571–4. Available from: https://academic.oup.com/jid/article/203/5/571/896768 Bruneau J, Roy É, Arruda N, Zang G, Jutras-Aswad D. The rising prevalence of prescription opioid injection and its association with hepatitis C incidence among street‐drug users. Addiction [Internet]. 2012;107(7):1318–27. Available from: http://dx.doi.org/10.1111/j.1360-0443.2012.03803.x Mateu-Gelabert P, Sandoval M, Meylakhs P, Wendel T, Friedman SR. Strategies to avoid opiate withdrawal: Implications for HCV and HIV risks. Int J Drug Policy [Internet]. 2010;21(3):179–85. Available from: http://dx.doi.org/10.1016/j.drugpo.2009.08.007 Platt L, Minozzi S, Reed J, Vickerman P, Hagan H, French C, et al. Needle and syringe programmes and opioid substitution therapy for preventing HCV transmission among people who inject drugs: findings from a Cochrane Review and meta‐analysis. Addiction [Internet]. 2018;113(3):545–63. Available from: http://dx.doi.org/10.1111/add.14012 Arum C, Fraser H, Artenie AA, Bivegete S, Trickey A, Alary M, et al. Homelessness, unstable housing, and risk of HIV and hepatitis C virus acquisition among people who inject drugs: a systematic review and meta-analysis. Lancet Public Health [Internet]. 2021;6(5):e309–23. Available from: http://dx.doi.org/10.1016/s2468-2667(21)00013-x Nyamathi A, Salem BE, Zhang S, Farabee D, Hall B, Khalilifard F, et al. Nursing case management, peer coaching, and hepatitis A and B vaccine completion among homeless men recently released on parole: Randomized clinical trial. Nurs Res [Internet]. 2015 [cited 2024 Jul 16];64(3):177–89. Available from: https://journals.lww.com/nursingresearchonline/fulltext/2015/05000/nursing_case_management,_peer_coaching,_and.6.aspx Beijer U, Wolf A, Fazel S. Prevalence of tuberculosis, hepatitis C virus, and HIV in homeless people: a systematic review and meta-analysis. Lancet Infect Dis [Internet]. 2012;12(11):859–70. Available from: http://dx.doi.org/10.1016/s1473-3099(12)70177-9 Stone J, Fraser H, Lim AG, Walker JG, Ward Z, MacGregor L, et al. Incarceration history and risk of HIV and hepatitis C virus acquisition among people who inject drugs: a systematic review and meta-analysis. Lancet Infect Dis [Internet]. 2018;18(12):1397–409. Available from: http://dx.doi.org/10.1016/s1473-3099(18)30469-9 Degenhardt L, Peacock A, Colledge S, Leung J, Grebely J, Vickerman P, et al. Global prevalence of injecting drug use and sociodemographic characteristics and prevalence of HIV, HBV, and HCV in people who inject drugs: a multistage systematic review. Lancet Glob Health [Internet]. 2017;5(12):e1192–207. Available from: http://dx.doi.org/10.1016/s2214-109x(17)30375-3 Rivera Saldana CD, Beletsky L, Borquez A, Kiene SM, Strathdee SA, Zúñiga ML, et al. Impact of cumulative incarceration and the post‐release period on syringe‐sharing among people who inject drugs in Tijuana, Mexico: a longitudinal analysis. Addiction [Internet]. 2021;116(10):2724–33. Available from: http://dx.doi.org/10.1111/add.15445 Zibbell JE, Asher AK, Patel RC, Kupronis B, Iqbal K, Ward JW, et al. Increases in acute hepatitis C virus infection related to a growing opioid epidemic and associated injection drug use, United States, 2004 to 2014. Am J Public Health [Internet]. 2018;108(2):175–81. Available from: http://dx.doi.org/10.2105/ajph.2017.304132 Uusküla A, Raag M, Marsh K, Talu A, Vorobjov S, Des Jarlais D. HIV prevalence and gender differences among new injection-drug-users in Tallinn, Estonia: A persisting problem in a stable high prevalence epidemic. PLoS One [Internet]. 2017;12(2):e0170956. Available from: http://dx.doi.org/10.1371/journal.pone.0170956 Leyna GH, Makyao N, Mwijage A, Ramadhan A, Likindikoki S, Mizinduko M, et al. HIV/HCV co-infection and associated risk factors among injecting drug users in Dar es Salaam, Tanzania: potential for HCV elimination. Harm Reduct J [Internet]. 2019;16(1). Available from: http://dx.doi.org/10.1186/s12954-019-0346-y Dolan K, Wirtz AL, Moazen B, Ndeffo-mbah M, Galvani A, Kinner SA, et al. Global burden of HIV, viral hepatitis, and tuberculosis in prisoners and detainees. Lancet [Internet]. 2016;388(10049):1089–102. Available from: http://dx.doi.org/10.1016/s0140-6736(16)30466-4 White IR, Carlin JB. Bias and efficiency of multiple imputation compared with complete‐case analysis for missing covariate values. Stat Med [Internet]. 2010;29(28):2920–31. Available from: http://dx.doi.org/10.1002/sim.3944 Sterne JAC, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ [Internet]. 2009 [cited 2024 Jul 16];338(jun29 1):b2393–b2393. Available from: https://www.bmj.com/content/338/bmj.b2393 Valerio H, Alavi M, Conway A, Silk D, Treloar C, Martinello M, et al. Declining prevalence of current HCV infection and increased treatment uptake among people who inject drugs: The ETHOS Engage study. Int J Drug Policy [Internet]. 2022;105(103706):103706. Available from: http://dx.doi.org/10.1016/j.drugpo.2022.103706 Additional Declarations No competing interests reported. Supplementary Files APPENDICES.docx Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 12 Nov, 2025 Reviews received at journal 05 Nov, 2025 Reviewers agreed at journal 21 Oct, 2025 Reviewers agreed at journal 20 Oct, 2025 Reviews received at journal 15 Sep, 2025 Reviewers agreed at journal 14 Sep, 2025 Reviewers invited by journal 31 Jul, 2025 Editor assigned by journal 31 Jul, 2025 Editor invited by journal 18 Jul, 2025 Submission checks completed at journal 15 Jul, 2025 First submitted to journal 15 Jul, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6994675","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":494704484,"identity":"bb5e85b5-a505-42dc-8b8b-3151f607f5eb","order_by":0,"name":"Adewunmi Akingbola","email":"","orcid":"","institution":"University of Cambridge Old Schools Trinity","correspondingAuthor":false,"prefix":"","firstName":"Adewunmi","middleName":"","lastName":"Akingbola","suffix":""},{"id":494704485,"identity":"3ffbcccc-3105-4b99-a137-a5115529ec22","order_by":1,"name":"Olajumoke Adewole","email":"","orcid":"","institution":"Lagos State University","correspondingAuthor":false,"prefix":"","firstName":"Olajumoke","middleName":"","lastName":"Adewole","suffix":""},{"id":494704486,"identity":"3ce1714a-ee16-4aea-be3c-8486cd6f301a","order_by":2,"name":"Abiodun Adegbesan","email":"","orcid":"","institution":"African Cancer Institute, Stellenbosch University Cape Town Tygerberg","correspondingAuthor":false,"prefix":"","firstName":"Abiodun","middleName":"","lastName":"Adegbesan","suffix":""},{"id":494704488,"identity":"023abcd0-fbe3-4f2f-8c44-41cc604988a1","order_by":3,"name":"Joel Chuku","email":"data:image/png;base64,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","orcid":"","institution":"V.N Karazin Kharkiv National University","correspondingAuthor":true,"prefix":"","firstName":"Joel","middleName":"","lastName":"Chuku","suffix":""}],"badges":[],"createdAt":"2025-06-27 23:23:04","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6994675/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6994675/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":88408461,"identity":"ac86c906-b8b6-4f12-9c71-a77c77545619","added_by":"auto","created_at":"2025-08-06 08:12:48","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":297140,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFlowchart of the Statistical Analysis Methods used.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6994675/v1/946ab63991ed7df4c40cab40.jpeg"},{"id":88408466,"identity":"fd790e54-fabd-4fd0-ae87-b4827c04f9df","added_by":"auto","created_at":"2025-08-06 08:12:48","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":165008,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAge Distribution Comparison: Complete-Case Analysis (left) vs Multiple Imputation (right).\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6994675/v1/c4b7f02bdbaa10f4890d6e2c.jpeg"},{"id":88408467,"identity":"0ad06065-424e-4a86-bf65-5b9c123da945","added_by":"auto","created_at":"2025-08-06 08:12:48","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":181341,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eInjection Duration. Comparison of injection duration density plots using Complete-Case Analysis (left) and Multiple Imputation methods (right).\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6994675/v1/bda0a9342c865738c5805c4a.jpeg"},{"id":88408481,"identity":"33995f52-c6f2-435d-a158-a55096191c84","added_by":"auto","created_at":"2025-08-06 08:12:49","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":264898,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe trend in HCV RNA Prevalence Over 10 Years, Complete-case analysis (left) and Multiple imputation (right).\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6994675/v1/78132532127898c0544d8fe8.jpeg"},{"id":88409843,"identity":"052b8865-1750-46cd-a5fb-d5a52602a40f","added_by":"auto","created_at":"2025-08-06 08:20:49","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":374412,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTrends in HCV, Homelessness, Imprisonment before injection commencement, and Ever Imprisoned Over 10 Years\u003c/strong\u003e. \u003cstrong\u003eThe graphs depict the trends of various factors related to HCV prevalence, comparing complete-case analysis (left) with multiple imputations (right). The variables include homelessness prevalence, imprisonment prevalence, and Ever Imprisoned.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6994675/v1/332812008f855895976c45fa.jpeg"},{"id":88408472,"identity":"9c15bd09-08dc-4bae-b2f8-bbf1a5ef9d08","added_by":"auto","created_at":"2025-08-06 08:12:49","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":309926,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTrends in Injection Practices and Hepatitis C Prevalence Over 10 Years - Complete-Case Analysis vs. Multiple Imputation. The graphs compare trends in HCV prevalence, injection duration, needle exchange program, and needle sharing over ten years, highlighting differences between complete-case analysis (left) and multiple imputations (right). The trends illustrate the impact of different data handling methods on the analysis of these variables among people who inject drugs.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6994675/v1/9c0e0dfefbf85dfeed2339d1.jpeg"},{"id":88408476,"identity":"9bfd4a02-8e1b-4e1f-b20d-ddfe2cfd6c7b","added_by":"auto","created_at":"2025-08-06 08:12:49","extension":"jpeg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":267117,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTrends in HCV RNA and Condom Use Over the Years - Complete-Case Analysis vs. Multiple Imputation: The graphs illustrate the trends in HCV RNA prevalence and condom use over the years, highlighting differences between complete-case analysis (left) and multiple imputation (right).\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6994675/v1/22d4da55aaf22a07f09b67a5.jpeg"},{"id":88408468,"identity":"6dc2eeec-c59b-41fe-bed4-600a2b04445e","added_by":"auto","created_at":"2025-08-06 08:12:48","extension":"jpeg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":313048,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eForest Plot of Logistic Regression Results: Complete-Case Analysis vs. Multiple Imputation\u003c/strong\u003e. \u003cstrong\u003eThe forest plots compare the odds ratios of various variables associated with HCV RNA, using complete-case analysis (left) and multiple imputation (right), showing each variable’s confidence intervals and significance levels.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6994675/v1/68904e2e03ea0e9889a8a6ed.jpeg"},{"id":88411885,"identity":"acbf86a0-7e22-4497-a133-1a40b74ff4cd","added_by":"auto","created_at":"2025-08-06 08:28:50","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4578943,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6994675/v1/eadb1c48-fed5-4ace-9996-1952c2cdd519.pdf"},{"id":88408459,"identity":"24a706c2-d0b5-4c03-8173-1016e70367cc","added_by":"auto","created_at":"2025-08-06 08:12:48","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":20366,"visible":true,"origin":"","legend":"","description":"","filename":"APPENDICES.docx","url":"https://assets-eu.researchsquare.com/files/rs-6994675/v1/32a76941ddcf1d00df75d4fb.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Improving Prevalence Estimates of Hepatitis C in Key Populations: A Simulated Data-Based Comparison of Missing Data Techniques","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eHepatitis C, an infectious disease and a type of viral Hepatitis caused by the Hepatitis C virus accounts for an estimated 15\u0026ndash;20% of acute hepatitis cases worldwide. About 16% of patients develop cirrhosis within 20 years and about 25% of Hepatocellular carcinoma cases have been attributed to this disease (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). It is a major public health concern with an estimated 58\u0026nbsp;million people currently infected and commonly spreads through sharing needles or syringes, amongst other routes, thus people who inject drugs (PWID) account for an estimated 11\u0026nbsp;million cases globally (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). In the United Kingdom, an estimated 118,000 people lived with chronic hepatitis C infection in 2019(\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e), with an estimated 80% of them contracting the disease due to drug use through injections (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThrough an Unlinked Anonymous Monitoring Survey conducted in the United Kingdom between 1990 and 2019, the median age of people who inject psychoactive substances which was previously at 27 years, was shown to have increased to 33 years in 2008 and 40 years in 2019. This indicates that before the 1990s, drug injection was prevalent amongst young people and a shift to middle-aged years has been identified, due to fewer new and a lot of existing old drug users. 74.9% of the 66,545 respondents of the survey were Males, with the North-west region and London having the highest percentage of older respondents in England, Wales, and Northern Ireland, while Glasgow and Clyde had the highest percentage of older people in Scotland. This means that the Male Gender has a higher inclination to use drugs, thus a higher potential of contracting the Hepatitis C virus, and the age-by-location distribution offers insights into the availability of drugs and the effectiveness of public health policies in curbing this trend in those locations (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e). Further extensive analysis of the 2019 Unlinked Anonymous Monitoring Survey for England, Wales, and Northern Ireland, 51.9% of PWIDs respondents alluded to having engaged in sexualized drug use in the last year, 70% of them have been in prison and 50.4% have been homeless. These provide valuable insights into some social and behavioral events associated with PWIDs and can as well, predispose them to contracting the Hepatitis C disease. (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e)\u003c/p\u003e\u003cp\u003eHowever, routinely collected epidemiological data like the Unlinked Anonymous Monitoring Survey has a high propensity for missing values which can occur because of loss to follow-ups of patients, non-response and/or errors during data entry. This results in a biased result with questionable validity and reliability, alongside reducing the statistical power of the data such that the probability of rejecting the null hypothesis when it is false is reduced, if not appropriately handled. (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). Traditionally, missing values in a dataset are handled using the list-wise deletion, also known as complete-case analysis which can cause bias in the estimates and thus, also impact the results and inferences from it. Several alternative statistical methods exist to handle missing data including Multiple Imputation. This method predicts likely values with similar variability and uncertainty with the right values, from other values in the dataset, thus safeguarding the validity of the results obtained from the imputed data (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e). In this research project, complete-case analysis and multiple imputation techniques will be employed in the analysis of the data and a comparison of both methods will be made, insights and differences will be identified, and their implications assessed.\u003c/p\u003e"},{"header":"METHODOLOGY","content":"\u003cp\u003eThis study employed a cross-sectional, observational design using a simulated version of the Unlinked Anonymous Monitoring (UAM) Survey dataset, which is routinely used to monitor the prevalence of Hepatitis C (HCV) among people who inject drugs (PWID) in the UK (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e). The simulated dataset replicates the structure and content of the original UAM survey, allowing for methodological exploration without involving real participants. The original UAM survey gathers anonymous data from PWID recruited through convenience sampling at drug treatment centers, needle exchange programs, and related services, enabling access to a typically hard-to-reach population (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). Data collection is conducted anonymously to encourage truthful responses and minimize stigma, thereby improving data reliability (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e). While the original design offers advantages in studying epidemiological trends and associations efficiently and cost-effectively, it carries inherent limitations such as selection bias, recall bias, and issues related to data completeness (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). These challenges persist in the simulated dataset, making it an appropriate tool for evaluating missing data strategies and analytical approaches in public health research. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e below summarizes the steps in analysis.\u003c/p\u003e\u003cp\u003e\u003cb\u003eVariables of Interest\u003c/b\u003e:\u003c/p\u003e\u003cp\u003eIn this study, the variables of interest were classified into dependent and independent variables (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e) to analyze the prevalence and risk factors associated with Hepatitis C (HCV) among people who inject drugs (PWID).\u003c/p\u003e\u003cp\u003e\u003cb\u003eAnalysis of Missing Data Mechanisms\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThis study conducted a comprehensive analysis to understand the mechanisms underlying the missing data in the dataset. Using Little's MCAR test, which evaluates whether missing data is Missing Completely at Random (MCAR) by comparing the observed data patterns to what would be expected if the data were truly random, using a chi-square statistic to determine significance (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e) with results in Appendix A below, I assessed whether the missingness in each variable was completely at random. To further investigate, we performed a logistic regression analysis on the missingness indicators for the 'HCV RNA' outcome variable (see Appendix B) but did not assess the plausibility of an MNAR scenario.\u003c/p\u003e\u003cp\u003e\u003cb\u003eComplete Case Analysis: -\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo handle missing data in the key variables, I performed a complete case analysis function in R, which involves retaining only those observations with no missing values in the specified variables. The process was as follows: First, we identified the key variables of interest, and I generated a logical vector indicating which rows in the dataset contained complete data across the selected variables in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e above. We then subset the dataset to include only these complete cases.\u003c/p\u003e\u003cp\u003e\u003cb\u003eMultiple Imputation Process\u003c/b\u003e:\u003c/p\u003e\u003cp\u003eMultiple imputation was performed using the Multiple Imputation by Chained Equations (MICE) package in R (version 4.4.0). This method provides a robust way to impute missing values, ensuring that the relationships between variables are preserved and the imputed data accurately reflects the underlying patterns (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe multiple imputation procedure involved the following steps:\u003c/p\u003e\u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eImputation Method Selection: The imputation was conducted using the MICE function, which employs a chained equations approach. The method chosen for imputation was predictive mean matching (PMM), suitable for handling both continuous and categorical data types. PMM matches missing values to observed values that are similar in the predictive model, thus preserving the original distribution of the data (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e).\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eExecution of Multiple Imputation: The MICE function generated multiple imputed datasets. The imputation was configured as follows:\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003edata_for_imputation\u003c/b\u003e represented the dataset with missing values.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003em\u0026thinsp;=\u0026thinsp;5 specified the creation of five imputed datasets, balancing computational efficiency with the reliability of the imputation results.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003emaxit\u0026thinsp;=\u0026thinsp;50 indicated the maximum number of iterations for the imputation algorithm to converge.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eSeed\u0026thinsp;=\u0026thinsp;500 ensured reproducibility by setting a seed for the random number generator.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eMethod\u0026thinsp;=\u0026thinsp;\u003cb\u003eimputation_methods\u003c/b\u003e applied specific imputation methods tailored to each variable, particularly Predictive Mean Matching (PMM).\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003e3. Analysis and Pooling of Imputed Data: Each of the five imputed datasets was analyzed separately to obtain estimates for the statistical analyses. The results from these analyses were then combined using Rubin\u0026rsquo;s rules. This pooling process accounted for the variability between the imputations and provided final estimates, standard errors, and confidence intervals (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e). This comprehensive approach helped integrate the imputed datasets, thus reflecting the uncertainty due to missing data and yielding more robust and reliable statistical inferences (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e).\u003c/p\u003e\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eStatistical Analysis\u003c/h2\u003e\u003cp\u003eThe statistical analysis was employed to summarize and infer key socio-demographic variables and factors related to Hepatitis C Virus (HCV) prevalence within the study population. Descriptive statistics provided a comprehensive overview of the critical variables, while inferential statistics facilitated the identification of significant associations and trends. The analysis included data processing, summary measures, and advanced statistical modelling to draw meaningful conclusions. Descriptive and Inferential analysis included Chi-square tests (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e), Wilcoxon Rank Sum tests (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e), T-test (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e), Logistic regression models (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e). All analyses were conducted using R software version 4.4.0.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"RESULTS","content":"\u003cp\u003e\u003cb\u003eDESCRIPTIVE ANALYSIS\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe complete-case analysis and multiple imputation age distribution in Fig.\u0026nbsp;2 below are bell-shaped curves indicating that they are typically distributed, and their highest frequencies are observed between ages 35 and 45 years with their peak ages approximately 40 years old. However, the frequency of the complete-case analysis is very close to 1200, and that of multiple imputation is close to 4000, which is the number of individuals within the most populous age group. The density plots for injection duration in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e3\u003c/span\u003e illustrate the distribution among participants, comparing Complete-Case Analysis and Multiple Imputation methods. Both plots exhibit a similar shape, peaking around 10 to 15 years of injection duration. The Multiple Imputation plot shows a more continuous distribution and fluctuation, while the Complete-Case Analysis plot has slightly less pronounced peaks and troughs, reflecting differences in data handling between the two methods.\u003c/p\u003e\u003cp\u003eThe analysis included several variables, each with varying percentages of missing data, as listed in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e below. The age variable had a missing percentage of 1.35%, while gender had 0.40%. Condom use (Condom1) was the variable with the highest missing percentage at 45.05%, followed by sharing needles(share) at 27.05%. Previous imprisonment (epris1) had a missing percentage of 2.14%, homeless status (homeless1) was missing for 1.94% of cases, and the exchange of needle (exch) had a 0.92% missing rate. Prison before injection (prisbeforeinj) had 10.00% missing data, and the HCV RNA status (hcvrna) variable had 21.70% missing data. These percentages highlight the need for robust statistical methods to handle incomplete datasets due to the large proportion of data that would be omitted in Complete Case (C.C), which may be used in Multiple Imputation (M.I) methods.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eVariables and Missing Percentage.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"2\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariable\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMissing Percentage\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.35%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGender(gen)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.40%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUse of Condom (condom1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e45.05%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEver been in Prison (epris1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2.14%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEver Homeless (homeless1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.94%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNeedle Exchange (exch)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.92%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePrison before Injection (prisbeforeinj)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e10.00%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSharing of Needle (Share)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e27.05%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHCV Prevalence (HCV RNA)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e21.70%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInjection Duration (Injdur)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e4.54%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe prevalence of HCV RNA over the years was tracked to understand trends and patterns. Figure\u0026nbsp;4 and Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e below shows the Complete Case (C.C) image, a line graph marked by notable volatility. From 2011 to 2013, the graph shows relatively stable rates, indicating consistent monitoring and recording of very similar prevalence during these years. However, between 2013 and 2014, there was a dramatic spike, suggesting data incongruencies as all negative results were observed to be missing between those years as RNA Testing was done retrospectively. This sharp rise is visually prominent and indicates a heightened concern that continued into 2015 when it peaked. Following this peak, the graph depicts a steep decline starting in 2015 and continuing into the following years, suggesting effective intervention measures were implemented. By the end of the period, from 2018 to 2021, the trend line stabilizes, showing a gradual decrease.\u003c/p\u003e\u003cp\u003eIn contrast, the Multiple Imputation (M.I.) image portrays a smoother and more consistent trend over the same period. The line graph starts with a similar stability from 2011 to 2013. However, unlike the C.C. method, there was no dramatic spike in 2013 or 2015. Instead, the M.I. method shows a gradual increase during these years, reflecting a more moderated rise in prevalence. After 2015, the M.I. graph shows a steady decline starting from 2018, like the C.C method, but with a more gradual slope. The trend continues downward to 2021, presenting a stable decrease and indicating sustained effectiveness of intervention measures.\u003c/p\u003e\u003cp\u003eIn the unadjusted results, as seen in Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e below, the year variables show variations between complete case analysis (C.C) and multiple imputation (M.I). For instance, C.C. indicates higher ORs for 2014 (2.92, 95% CI: 2.19\u0026ndash;3.88) and 2015 (4.02, 95% CI: 2.95\u0026ndash;5.49), which are statistically significant, while the M.I results for these years are closer to 1 and non-significant. Both methods show that years like 2018 (C.C OR: 1.39, 95% CI: 1.06\u0026ndash;1.82, p-value: 0.0162; M.I OR: 0.84, 95% CI: 0.72\u0026ndash;0.98, p-value: 0.0296) and 2021 (C.C OR: 0.47, 95% CI: 0.31\u0026ndash;0.73, p-value: 0.000824; M.I OR: 1.90, 95% CI: 1.46\u0026ndash;2.47, p-value: \u0026lt;0.0001) are statistically significant, with M.I showing consistently lower p-values for these years. In the adjusted results from the multivariate analysis in Table\u0026nbsp;9 below, C.C shows significant ORs for 2014 (2.49, 95% CI: 1.85\u0026ndash;3.34) and 2015 (3.37, 95% CI: 2.45\u0026ndash;4.63), while these years are non-significant in M.I (close to 1). The years 2019, 2020, and 2021 are significant in both methods, with C.C showing ORs of 0.71 (95% CI: 0.54\u0026ndash;0.94), 0.51 (95% CI: 0.31\u0026ndash;0.84), and 0.37 (95% CI: 0.24\u0026ndash;0.58), and M.I showing ORs of 0.77 (95% CI: 0.65\u0026ndash;0.91), 0.47 (95% CI: 0.34\u0026ndash;0.64), and 0.40 (95% CI: 0.31\u0026ndash;0.53), respectively, all with strong statistical significance.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003ePrevalence per Year Table: The prevalence of HCV RNA, as calculated by Complete Case Analysis and Multiple Imputation, shows trends in prevalence rates from 2011 to 2021.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eComplete-Case Analysis Prevalence per Year (%)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eMultiple Imputation Prevalence per Year (%)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2011\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e29.7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e30.7\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2012\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e29.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e30.5\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2013\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e29.9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e30.8\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2014\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e55.2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e30.8\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2015\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e63.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e32.5\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2016\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e30.4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e33.8\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2017\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e28.8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e30.5\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2018\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e37.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e34.4\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2019\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e27.0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e29.2\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2020\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e23.1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e21.1\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2021\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e16.7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e18.9\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eVariables and HCV RNA\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe association between various variables and HCV RNA status was analyzed using different statistical tests as shown in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e below. The Injection Duration and HCV RNA variables showed significant associations in both C.C (Wilcoxon, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001) and M.I (Wilcoxon, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001) analyses. This suggests that the duration of injection use is related to the presence or level of HCV RNA, and the consistency of this result across both the C.C. and M.I. analyses further reinforce the robustness of this finding. The Ever-Used Needle Exchange Program (exch) variable also exhibited significant associations (C.C: Chi-square, p\u0026thinsp;=\u0026thinsp;0.0122; M.I: Chi-square, p\u0026thinsp;=\u0026thinsp;0.0012). The Share variable for Sharing of Needle did not show significant associations in either method (C.C: Chi-square, p\u0026thinsp;=\u0026thinsp;0.3207; M.I: Chi-square, p\u0026thinsp;=\u0026thinsp;1.0000). Ever Homeless status (Homeless1) had significant associations in both methods (C.C: Chi-square, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001; M.I: Chi-square, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001), meaning that current or previous homelessness is related to the presence of HCV RNA/Hepatitis C infection. Prison before Injection Commencement variable was significant in the C.C method (Chi-square, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001), meaning that having been in prison before starting injections is strongly associated with having Hepatitis C, but not in the M.I method (Chi-square, p\u0026thinsp;=\u0026thinsp;0.2434). Gender showed significant associations in both methods (C.C: Chi-square, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001; M.I: Chi-square, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001). Condom Use was significant in the C.C method (Chi-square, p\u0026thinsp;=\u0026thinsp;0.0170) but not in the M.I. method (Chi-square, p\u0026thinsp;=\u0026thinsp;0.3547). Finally, Ever in Prison or Young Offenders Institute was significant in both methods (Chi-square, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001). The age variable also showed significant associations (C.C: t-test, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001; M.I: t-test, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003e\u003cb\u003eAssociation Between Variables and HCV RNA: Table summarizing the association between various variables and HCV RNA status, using different statistical tests, and comparing Complete Case (C.C) and Multiple Imputation (M.I) methods.\u003c/b\u003e\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"6\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariable\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTest\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eComplete Case Analysis (Statistic, p-value)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eInterpretation (C.C)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eMultiple Imputation (Statistic, p-value)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eInterpretation (M.I)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInjection Duration (injdur)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWilcoxon\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eW\u0026thinsp;=\u0026thinsp;2057993, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSignificant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eW\u0026thinsp;=\u0026thinsp;14000, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eSignificant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNeedle Exchange (exch)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eChi-square\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eX\u0026sup2; = 6.28, p\u0026thinsp;=\u0026thinsp;0.0122\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSignificant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eX\u0026sup2; = 22.87, p\u0026thinsp;=\u0026thinsp;0.0012\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eSignificant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSharing of Needle (Share)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eChi-square\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eX\u0026sup2; = 0.99, p\u0026thinsp;=\u0026thinsp;0.3207\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNot Significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eX\u0026sup2; = 0.60, p\u0026thinsp;=\u0026thinsp;1.0000\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eNot Significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEver Homeless (homeless1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eChi-square\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eX\u0026sup2; = 35.73, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSignificant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eX\u0026sup2; = 104.49, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eSignificant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePrison before Injection commencement (prisbeforeinj)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eChi-square\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eX\u0026sup2; = 16.80, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSignificant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eX\u0026sup2; = 5.78, p\u0026thinsp;=\u0026thinsp;0.2434\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eNot Significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGender (gen)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eChi-square\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eX\u0026sup2; = 16.956, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSignificant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eX\u0026sup2; = 83.71, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eSignificant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCondom Use (condom1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eChi-square\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eX\u0026sup2; = 8.16, p\u0026thinsp;=\u0026thinsp;0.0170\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSignificant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eX\u0026sup2; = 9.61, p\u0026thinsp;=\u0026thinsp;0.3547\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eNot Significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEver been in Prison (epris1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eChi-square\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eX\u0026sup2; = 101.68, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSignificant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eX\u0026sup2; = 101.68, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eSignificant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eT-test\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003et = -7.33, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSignificant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003et\u0026thinsp;=\u0026thinsp;13.86, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eSignificant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eAs shown in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, the Pearson correlation results provide insights into the strength and direction of relationships between variables. For injection duration, the C.C. method showed a very weak positive correlation (0.0805), while the M.I. method indicated a moderate negative correlation (-0.5253). The exchange variable had a moderate positive correlation in the C.C. method (0.5284) and a strong positive correlation in the M.I. method (0.8045). The sharing of injection showed a moderate negative correlation in the C.C. method (-0.5896) and a strong negative correlation in the M.I. method (-0.7759). Condom use had a very weak positive correlation in the C.C. method (0.1484) and a weak positive correlation in the M.I. method (0.2768). Homeless Status showed weak positive correlations in both methods (C.C: 0.3287; M.I: 0.3681). Previous imprisonment before Injection had very weak positive correlations in both methods (C.C: 0.1544; M.I: 0.1728). The Ever in Prison or Young Offenders Institute had a strong positive correlation in the C.C. method (0.6776) and a moderate positive correlation in the M.I. method (0.5117).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003e\u003cb\u003ePearson Correlation Table: Pearson correlation coefficients and interpretations for Complete Case (C.C) and Multiple Imputation (M.I) methods, illustrating the strength and direction of relationships between variables and HCV RNA status.\u003c/b\u003e\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariable\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003ePearson Corr (C.C)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eInterpretation (C.C)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003ePearson Corr (M.I)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eInterpretation (M.I)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInjection Duration (injdur)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.0805\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eVery weak positive\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-0.5253\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eModerate negative\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNeedle Exchange Programs (exch)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.5284\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eModerate positive\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.8045\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eStrong positive\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSharing of Needle (Share)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e-0.5896\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eModerate negative\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e-0.7759\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eStrong negative\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUse of Condom (condom1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.1484\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eVery weak positive\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.2768\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eWeak positive\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEver Homeless (homeless1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.3287\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eWeak positive\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.3681\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eWeak positive\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePrison before Injection Commencement (prisbeforeinj)(prisbeforeinj)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.1544\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eVery weak positive\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.1728\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eVery weak positive\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEver in Prison (epris1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.6776\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eStrong positive\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.5117\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eModerate positive\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eEpris1 (\u0026ldquo;Ever in Prison or Young Offenders Institute\u0026rdquo;)\u003c/b\u003e\u003c/p\u003e\u003cp\u003eIn the Complete Case (C.C) part of Fig.\u0026nbsp;5, Ever in Prison/Young Offenders Institute, represented by the blue line, shows significant variability. From 2011 to 2021, the trend starts high and remains relatively high throughout the period, with sharp fluctuation from 2014 and 2016 through 2021. A notable peak occurred around 2013, indicating increased reported cases of previous imprisonment. Following this peak, there is a gradual decline, although the values remain elevated compared to other variables. The sharp rise and fall suggest data inconsistencies or heightened reporting during these years. The Multiple Imputation (M.I) image shows a more stable trend for epris1, represented by the blue line. It started at a high level in 2011 and peaked in 2013, but the trend remains relatively constant with fewer fluctuations. There has been a slight decline over the years, but the overall trend is smoother and more stable.\u003c/p\u003e\u003cp\u003e\u003cb\u003eHomeless Status\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe homeless Status variable in the C.C. image of Fig.\u0026nbsp;5, depicted by the orange line, shows moderate fluctuations. From 2011, the trend starts at a lower level compared to epris1for \u0026ldquo;Ever in Prison or Young Offenders Institute\u0026rdquo; but shows an increase around 2014\u0026ndash;2015, drops around 2016\u0026ndash;2017, and followed by stabilization in subsequent years. In the M.I. image of Fig.\u0026nbsp;5, the homeless status variable, shown by the orange line, presents a similar trend. The prevalence starts higher than in the C.C. method and remains relatively constant. It also records the decline around 2016\u0026ndash;2017, similarly seen in the Complete Case Analysis method.\u003c/p\u003e\u003cp\u003e\u003cb\u003ePrison Before Injection Commencement\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe prison before injection variable in the C.C. image of Fig.\u0026nbsp;5, represented by the green line, exhibits a relatively stable trend with minor fluctuations. From 2011, there was a sharp increase around 2013 and a corresponding decrease in 2014, but overall, the trend does not show significant changes. This suggests that this variable's data might be more consistently reported or less affected by missing data than the other variables. In the M.I. image of Fig.\u0026nbsp;5, the prison before injection variable, depicted by the brown line, also shows a stable trend but a generally slight decline over the years, with its nadir in 2014. The trend is smoother than the C.C method, indicating that the imputation method effectively fills in missing data, leading to a more reliable representation of the variable over time. The smoother trend in the M.I. image suggests that imputation helps create a more accurate and less volatile depiction of the data.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eNeedle Exchange Programs\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\u003cp\u003eIn the Complete Case (C.C) image of Fig.\u0026nbsp;6, the needle exchange prevalence trend (purple line) is consistently high from 2011 to 2021. The values remain close to the maximum (100%), indicating a high prevalence of needle exchange programs use amongst participants or consistent data collection practices. The trend line shows minor fluctuations but maintains a steady trajectory, suggesting reliable reporting and a low impact of missing data on this variable. The multiple imputation (M.I.) image in Fig.\u0026nbsp;6 shows the needle exchange prevalence trend (purple line) as equally high. The trend line in the M.I. image is almost like that of the Complete Case Analysis method. The prevalence remains high across all years, identical to the C.C. method, which implies that the data for this variable is consistently robust in both methods.\u003c/p\u003e\u003cp\u003e\u003cb\u003eNeedle Sharing Prevalence\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe needle-sharing prevalence trend (Blue line) in the C.C. image of Fig.\u0026nbsp;6 starts at a moderate level in 2011 and shows noticeable variability over the years. Initially, the trend is relatively low, but it increases gradually but irregularly, peaking significantly around 2019\u0026ndash;2021. This increase suggests variability in data collection or an actual rise in needle-sharing practices during these years. In the M.I. image, the needle-sharing prevalence trend (Blue line) is more stable and smoother than the C.C method. The trend started at a similar moderate level in 2011 but exhibited a more gradual increase over the years.\u003c/p\u003e\u003cp\u003e\u003cb\u003eInjection Duration\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe injection duration trend (red line) in the C.C. image of Fig.\u0026nbsp;6 shows moderate stability with some fluctuations. The trend starts at a low level in 2011 and gradually increases, peaking around 2014\u0026ndash;2015. After this peak, the trend stabilized somewhat but still showed minor fluctuations in the later years, especially a decline between 2017 and 2019. This pattern suggests variability in data reporting or an actual change in injection practices over time, with the peak around 2014\u0026ndash;2015 potentially reflecting changes in behavior or reporting accuracy. In the M.I. image of Fig.\u0026nbsp;6, the injection duration trend (red line) shows a more consistent and smoother pattern. Starting at a low level in 2011, the trend gradually increased with less pronounced fluctuations than the C.C method. The trend line remains relatively stable, with a slight upward trajectory, indicating that the imputation method provides a more precise and reliable depiction of injection duration trends over time, addressing the missing data.\u003c/p\u003e\u003cdiv id=\"Sec5\" class=\"Section3\"\u003e\u003cdiv class=\"Heading\"\u003e3.2.7 Condom Use Prevalence\u003c/div\u003e\u003cp\u003eIn the Complete Case (C.C) image of Fig.\u0026nbsp;7 below, the condom use prevalence trend (blue line) started at a moderate level around 2011 and maintained a relatively stable pattern with slight fluctuations throughout the period. The multiple Imputation (M.I) method in Fig.\u0026nbsp;7 shows the condom use prevalence trend (blue line) to have a similar trend and pattern as the Complete case method image. The trend started at a similar moderate level in 2011 and continued with fluctuations throughout the period and a few sharp peaks and significant increases, like the C.C. method.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e3.3 Comparative Analysis of Univariate Regression: Complete Case (C.C) vs. Multiple Imputation (M.I)\u003c/h2\u003e\u003cp\u003eThe univariate regression analysis, as shown in Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, examines the relationships between the independent and dependent variables. This section compares the results obtained from the Complete Case (C.C) and the Multiple Imputation (M.I) methods, highlighting the differences in odds ratios, confidence intervals, and p-values.\u003c/p\u003e\u003cp\u003eFor injection duration (injdur), the odds ratio (OR) in the C.C method is 1.03 (95% CI: 1.02\u0026ndash;1.04) with a highly significant p-value of \u0026lt;\u0026thinsp;0.0001. In the M.I. method, the OR is slightly higher at 1.03 (95% CI: 1.03\u0026ndash;1.04) with an equally significant p-value of \u0026lt;\u0026thinsp;0.0001. Both methods indicate a significant positive association between injury duration and the dependent variable, HCVRNA. The C.C. method\u0026rsquo;s injection exchange programs (exch) variable shows an OR of 1.86 (95% CI: 1.18\u0026ndash;3.07), with a significant p-value of 0.0101. The M.I. method provides an OR of 1.55 (95% CI: 1.29\u0026ndash;1.87) with a p-value of \u0026lt;\u0026thinsp;0.0001. Both methods demonstrate a significant positive association, with similar ORs and overlapping confidence intervals, indicating consistent findings and less impact of missing values.\u003c/p\u003e\u003cp\u003eFor sharing of injection (share), the C.C. method shows an OR of 1.08 (95% CI: 0.93\u0026ndash;1.26) with a non-significant p-value of 0.3019. The M.I. method reports an OR of 1.02 (95% CI: 0.91\u0026ndash;1.14) with a non-significant p-value of 0.738. Both methods indicate a non-significant association, with close ORs and overlapping confidence intervals, confirming the lack of a significant effect. The homeless status variable in the C.C. method has an OR of 1.38 (95% CI: 1.15\u0026ndash;1.65) with a highly significant p-value of 0.0005. In the M.I. method, the OR is higher at 1.47 (95% CI: 1.31\u0026ndash;1.65) with a p-value of \u0026lt;\u0026thinsp;0.0001. Both methods show a significant positive association, with the M.I. method showing a marginally higher Odds Ratio. Ever in Prison, or Young Offenders Institute variable (prisbeforeinj) reveals an OR of 1.31 (95% CI: 1.15\u0026ndash;1.49) in the C.C method, with a significant p-value of \u0026lt;\u0026thinsp;0.0001. The M.I. method shows a significantly higher OR of 2.03 (95% CI: 1.86\u0026ndash;2.22) with a highly significant p-value of \u0026lt;\u0026thinsp;0.0001. This suggests that imputing missing data uncovers a stronger relationship between previous imprisonment and the dependent variable.\u003c/p\u003e\u003cp\u003eFor age, the C.C. method reports an OR of 1.03 (95% CI: 1.02\u0026ndash;1.04) with a highly significant p-value of \u0026lt;\u0026thinsp;0.0001. The M.I. method provides an OR of 1.03 (95% CI: 1.02\u0026ndash;1.03) with a p-value of \u0026lt;\u0026thinsp;0.0001. Both methods indicate a significant positive association, with very close ORs and confidence intervals, confirming the robustness of the association. The gender variable in the C.C method shows an OR of 0.75 (95% CI: 0.65\u0026ndash;0.86) with a significant p-value of \u0026lt;\u0026thinsp;0.0001. The M.I method has an OR of 0.69 (95% CI: 0.64\u0026ndash;0.75) with a p-value of \u0026lt;\u0026thinsp;0.0001. Both methods demonstrate a significant negative association, with the M.I. method indicating a slightly weaker effect but consistent findings due to overlapping confidence intervals.\u003c/p\u003e\u003cp\u003eFor condom use, the C.C. method shows an OR of 1.05 (95% CI: 0.92\u0026ndash;1.21) with a non-significant p-value of 0.4519. The M.I. method reports an OR of 1.08 (95% CI: 0.97\u0026ndash;1.20) with a non-significant p-value of 0.1721. This discrepancy highlights the impact of handling missing data, where both methods suggest the relationship is insignificant.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eUnivariate Table: Univariate analysis showing odds ratios, confidence intervals, and p-values for Complete Case (C.C) and Multiple Imputation (M.I) methods, identifying key variables associated with HCV RNA status.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariable\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eC.C OR (CI)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eC.C P-Value\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eInterpretation (C.C)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eM.I OR (CI)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eM.I P-Value\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003eInterpretation (M.I)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInjection Duration (injdur)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.03 (1.02\u0026ndash;1.04)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.03 (1.03\u0026ndash;1.04)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNeedle Exchange Programs (exch)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.86 (1.16-3.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.0101\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.55 (1.29\u0026ndash;1.87)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSharing of Needle (Share)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.08 (0.93\u0026ndash;1.26)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.302\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.02 (0.91\u0026ndash;1.14)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.738\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePrison before injection Commencement (prisbeforeinj1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.31 (1.15\u0026ndash;1.49)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e2.03 (1.86\u0026ndash;2.22)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.03 (1.02\u0026ndash;1.04)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.03 (1.02\u0026ndash;1.03)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGender (gen1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.75 (0.65\u0026ndash;0.86)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.69 (0.64\u0026ndash;0.76)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2012\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.00 (0.79\u0026ndash;1.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.974\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.01 (0.87\u0026ndash;1.18)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.9015\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2013\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.01 (0.79\u0026ndash;1.29)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.934\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.99 (0.85\u0026ndash;1.16)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.9343\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear - 2014\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2.92 (2.19\u0026ndash;3.88)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.00 (0.85\u0026ndash;1.17)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.9503\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2015\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e4.02 (2.95\u0026ndash;5.49)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.92 (0.78\u0026ndash;1.08)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.2993\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2016\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.03 (0.79\u0026ndash;1.35)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.814\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.87 (0.74\u0026ndash;1.02)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.0777\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2017\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.96 (0.73\u0026ndash;1.26)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.754\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.01 (0.85\u0026ndash;1.19)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.9519\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2018\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.39 (1.06\u0026ndash;1.82)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.0162\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.84 (0.72\u0026ndash;0.98)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.0296\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2019\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.88 (0.67\u0026ndash;1.15)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.339\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.07 (0.91\u0026ndash;1.26)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.3918\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2020\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.71 (0.44\u0026ndash;1.16)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.169\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.66 (1.23\u0026ndash;2.23)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.00281\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2021\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.47 (0.31\u0026ndash;0.73)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.000824\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.90 (1.46\u0026ndash;2.47)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEver Homeless (homeless1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2.30 (1.62\u0026ndash;3.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.000469\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.47 (1.31\u0026ndash;1.65)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCondom Use (condom1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.36 (0.99\u0026ndash;1.87)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.452\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.08 (0.97\u0026ndash;1.20)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.1721\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNon-statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eComparative Analysis of Multivariate Logistic Regression: Complete Case (C.C) vs. Multiple Imputation (M.I)\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe multivariate logistic regression analysis results, as shown in Table\u0026nbsp;\u003cspan refid=\"Tab6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, provide insights into the relationships between multiple independent variables and the dependent variable simultaneously. This analysis compares the results obtained from the Complete Case (C.C) method and the Multiple Imputation (M.I) method, using forest plots to visually represent the odds ratios and confidence intervals.\u003c/p\u003e\u003cp\u003eIn the Complete Case (C.C) method, the intercept shows an odds ratio (OR) of 0.10 (95% CI: 0.05\u0026ndash;0.19), with a highly significant p-value of \u0026lt;\u0026thinsp;0.0001. This suggests a strong baseline effect when all other variables are at zero. In contrast, the Multiple Imputation (M.I) method provides an OR of 0.12 (95% CI: 0.09\u0026ndash;0.16), with a significant p-value of \u0026lt;\u0026thinsp;0.0001, indicating a robust baseline effect with imputed data. For injection duration (injdur), the C.C. method reports an OR of 1.02 (95% CI: 1.01\u0026ndash;1.04) with a significant p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.0001, indicating a positive association. The M.I. method shows a slightly lower OR of 1.03 (95% CI: 1.02\u0026ndash;1.04) with a p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.0001, suggesting a consistently significant association but slightly reduced effect size. The Injection Exchange Programs (exch1) variable in the C.C method has an OR of 1.51 (95% CI: 0.92\u0026ndash;2.46) and a non-significant p-value of 0.102. The M.I. method reports a lower OR of 1.30 (95% CI: 1.07\u0026ndash;1.56) with a significant p-value of 0.0069, indicating a positive association with imputed data.\u003c/p\u003e\u003cp\u003eSharing of injection (share1) in the C.C method has an OR of 1.24 (95% CI: 1.06\u0026ndash;1.46) with a p-value of 0.0081, suggesting a significant association. The M.I. method provides an OR of 1.13 (95% CI: 1.01\u0026ndash;1.27) with a significant p-value of 0.035, indicating that the association is also significant when missing data is accounted for. For people experiencing homelessness but not last year (homeless11), the C.C. method shows an OR of 1.18 (95% CI: 0.98\u0026ndash;1.43) with a p-value of 0.0811, indicating a non-significant association. The M.I. method reports a higher OR of 1.35 (95% CI: 1.20\u0026ndash;1.52) with a highly significant p-value of \u0026lt;\u0026thinsp;0.0001, suggesting a significant association with imputed data. The second category for Homeless last year (homeless12) in the C.C method shows an OR of 1.59 (95% CI: 1.30\u0026ndash;1.86) with a highly significant p-value of \u0026lt;\u0026thinsp;0.0001. The M.I. method has a similar OR of 1.61 (95% CI: 1.44\u0026ndash;1.81) with a significant p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.0001. Both methods indicate a strong positive association, with the M.I. method providing a slightly narrower confidence interval.\u003c/p\u003e\u003cp\u003eFor condom use (condom11), the C.C. method shows an OR of 1.10 (95% CI: 0.95\u0026ndash;1.28) with a non-significant p-value of 0.1880. The M.I. method has an OR of 1.11 (95% CI: 0.99\u0026ndash;1.23) with a p-value of 0.0929, which is also non-significant. Both methods indicate that the association between condom use and HCVRNA is not significant. In the second category of condom use (condom12), the C.C. method shows an OR of 1.26 (95% CI: 1.04\u0026ndash;1.51) with a significant p-value of 0.0155. The M.I. method reports an OR of 1.14 (95% CI: 0.97\u0026ndash;1.34) with a non-significant p-value of 0.1522. This indicates that the initial significant association observed in the C.C. method is not supported when missing data is imputed. For previous imprisonment before injection commencement (prisbeforeinj1), the C.C method shows an OR of 1.30 (95% CI: 1.13\u0026ndash;1.51) with a significant p-value of 0.0004. The M.I. method reports an OR of 1.01 (95% CI: 0.91\u0026ndash;1.13) with a non-significant p-value of 0.8133, suggesting that the significant association in the C.C. method is not present when missing data is accounted for.\u003c/p\u003e\u003cp\u003eAge in the C.C. method has an OR of 1.01 (95% CI: 1.00\u0026ndash;1.02) with a non-significant p-value of 0.0596. The M.I. method shows an OR of 1.01 (95% CI: 1.01\u0026ndash;1.02) with a significant p-value of 0.0001. This indicates that age becomes a significant factor after imputation. Finally, for gender (gen1), the C.C. method shows an OR of 0.89 (95% CI: 0.77\u0026ndash;1.04) with a non-significant p-value of 0.147. The M.I. method reports a lower OR of 0.77 (95% CI: 0.71\u0026ndash;0.84) with a highly significant p-value of \u0026lt;\u0026thinsp;0.0001, suggesting a stronger and more significant association when missing data is accounted for.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab6\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 6\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003e\u003cb\u003eMultivariate Regression Table: This multivariate regression analysis compares odds ratios, confidence intervals, and p-values for Complete Case (C.C) and Multiple Imputation (M.I) methods, highlighting significant associations with HCV RNA status.\u003c/b\u003e\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariable\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eOddsRatio (95% CI) C.C\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003ep_value C.C\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eInterpretation C.C\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eOddsRatio (95% CI) M.M\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003ep_value M.M\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003eInterpretation M.M\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e(Intercept)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.10 (0.05\u0026ndash;0.19)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.12 (0.09\u0026ndash;0.16)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eInjection Duration (injdur)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.02 (1.01\u0026ndash;1.04)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.03 (1.02\u0026ndash;1.04)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNeedle exchange programs (exch1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.51 (0.92\u0026ndash;2.46)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.1020\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.30 (1.07\u0026ndash;1.56)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.0069\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSharing of Needle (Share)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.24 (1.06\u0026ndash;1.46)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.0081\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.13 (1.01\u0026ndash;1.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.0359\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHomeless, but not in the last year (homeless11)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.18 (0.98\u0026ndash;1.43)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.0811\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.35 (1.20\u0026ndash;1.52)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHomeless in the last year (homeless12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.56 (1.30\u0026ndash;1.86)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.61 (1.44\u0026ndash;1.81)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePrison before injection commencement (prisbeforeinj1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.30 (1.13\u0026ndash;1.51)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.0004\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.01 (0.91\u0026ndash;1.13)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.8133\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSometimes Used Condoms (condom11)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.10 (0.95\u0026ndash;1.28)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.1880\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.11 (0.99\u0026ndash;1.23)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.0929\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAlways Used Condoms (condom12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.26 (1.04\u0026ndash;1.51)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.0155\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.14 (0.97\u0026ndash;1.34)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.1522\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.01 (1.00\u0026ndash;1.02)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.0596\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eBorderline Non-statistical significance\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.01 (1.01\u0026ndash;1.02)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGender (gen1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.89 (0.77\u0026ndash;1.04)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.1420\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.77 (0.71\u0026ndash;0.84)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2012\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.00 (0.78\u0026ndash;1.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.9830\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.00 (0.86\u0026ndash;1.17)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.9992\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2013\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.95 (0.74\u0026ndash;1.22)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.6730\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.95 (0.81\u0026ndash;1.10)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.4812\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2014\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2.49 (1.85\u0026ndash;3.34)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.92 (0.78\u0026ndash;1.09)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.3233\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2015\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e3.37 (2.45\u0026ndash;4.63)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.98 (0.83\u0026ndash;1.16)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.8209\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2016\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.88 (0.67\u0026ndash;1.16)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.3560\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.01 (0.86\u0026ndash;1.19)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.8739\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2017\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.81 (0.61\u0026ndash;1.07)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.1360\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.85 (0.71\u0026ndash;1.01)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.0717\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2018\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.20 (0.91\u0026ndash;1.58)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.1980\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.00 (0.85\u0026ndash;1.17)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.9862\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNot statistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2019\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.71 (0.54\u0026ndash;0.94)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.0172\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.77 (0.65\u0026ndash;0.91)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.0023\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2020\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.51 (0.31\u0026ndash;0.84)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.0081\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.47 (0.34\u0026ndash;0.64)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYear \u0026minus;\u0026thinsp;2021\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.37 (0.24\u0026ndash;0.58)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.40 (0.31\u0026ndash;0.53)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eStatistically significant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eForest Plots\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe forest plots in Fig.\u0026nbsp;8 below provide a visual representation of the odds ratios and confidence intervals for the variables in both the C.C. and M.I. methods. The plots clearly illustrate the differences in the magnitude and significance of associations between the two methods. In general, the M.I. method shows tighter confidence intervals and more significant associations for several variables.\u003c/p\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eThe discussion section interprets the results of our study, comparing the implications of using complete case analysis versus multiple imputation methods for estimating HCV prevalence and identifying key risk factors. We also explore the broader impact of our findings on public health strategies and policies aimed at controlling and preventing HCV among high-risk populations.\u003c/p\u003e\u003cp\u003eThe Complete Case analysis results of this study revealed a bell-shaped age distribution curve with the highest frequencies between ages 35 and 45, a median age of 35.5 years, and a mean age of 35.86 years. In comparison, the multiple imputation analysis results also had the highest frequencies between ages 35 and 45, with a slightly higher median age of 37 and a mean age of 37.68 years, reflecting the inclusion of missing data. These findings align with trends observed in the literature and data from the Office for Health Improvements and Disparities, where the age of people in treatment for substance use are the older age groups. More than half of the individuals in treatment were over 40 years old, with significant proportions in the 40 to 44, 45 to 49, and 50 to 54 age brackets. This trend is attributed to many opiate users having initiated heroin use during the epidemics of the 1980s and 1990s, resulting in a current treatment population that is predominantly older. The median age for those in treatment for opiates was 43, reinforcing the pattern of an ageing cohort within this group. This comparison highlights the importance of considering age distribution trends when analyzing and interpreting data related to substance use treatment populations (\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e) and brings to the fore their susceptibility to developing liver cirrhosis (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e)\u003c/p\u003e\u003cp\u003eOur study's findings on the mean duration of injection practices using the C.C. and M.I. methods align with observations from the Needle Exchange Surveillance Initiative (NESI) report. In our study, the mean injection duration was 13.78 years for the C.C. method and 14.83 years for the M.I. method. These figures reflect long-term injection behaviors, which is a common trend among people who inject drugs (PWID). The NESI report does not explicitly provide an average duration of drug injection in years. However, it does highlight the ageing cohort of PWID, indicating long-term injection practices. The report mentions an increasing average age of participants, rising from 33 years old in 2008–09 to 41 years old in 2019–20, suggesting prolonged engagement in injection drug use. The NESI report also discusses the static nature of the age at first injection, reinforcing the idea of long-term injection behaviors (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e). This consistency aligns with our findings, as it implies sustained injection practices among older PWID rather than new injectors starting at an older age and brings to the fore the need for ongoing support services to mitigate health risks, such as blood-borne viruses, that increase with prolonged injection use.\u003c/p\u003e\u003cp\u003e\u003cb\u003eHCV Prevalence\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe C.C method showed a volatile trend with a spike between 2013 and 2014 and a peak in 2015 which may have occurred because HCV RNA data was missing for all antibody-negative results in 2013/2014, while the M.I method indicated a smoother trend with a gradual increase from 2011 to 2016 (30.7% -33.8%), a sharp drop in 2017 (30.5%) then another increase in 2018 (34.4%) which was followed subsequently by a steady decline from 2017 to 2021. For the years 2017–2021, the unadjusted results, C.C. indicates higher ORs for 2014 (2.92) and 2015 (4.02), which are statistically significant, while M.I results for these years are closer to 1 and non-significant. Both methods show significance for 2018 and 2021, but M.I. shows consistently lower p-values, indicating higher reliability and higher statistical power. In the adjusted results, C.C. shows significant ORs for 2014 (2.49) and 2015 (3.37), which are non-significant in M.I. Both methods indicate significance for 2019, 2020, and 2021, but the ORs are closer in M.I (e.g., 2021: C.C OR 0.37, M.I OR 0.40), suggesting more consistent and reliable estimates.\u003c/p\u003e\u003cp\u003e\u003cb\u003eAssociation of variables with HCV RNA\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eInjection Duration and HCV RNA\u003c/b\u003e\u003c/p\u003e\u003cp\u003eInjection duration (Injdur) was significantly associated with HCV RNA status in both methods in the chi-square analysis, adjusted and unadjusted regression analysis, which is consistent with findings from Hope et al. (2020) (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e), Alavi et al. (2019) (\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e) which reported that longer injection durations increase the risk of HCV infection. This association is likely due to the prolonged exposure and increased opportunities for risky behaviors, such as sharing needles or other injecting equipment, which heighten the risk of HCV transmission over time. Additionally, longer injection durations may reflect a chronic condition of substance use that is harder to manage and treat, further compounding the risk of infection.\u003c/p\u003e\u003cp\u003eHowever, the Pearson correlation analysis for injection duration over the ten-year period revealed a very weak positive correlation in the C.C. method and a moderate negative correlation in the M.I. method. The justification of this M.I. Pearson correlation results is that while longer injection durations increase the overall risk of HCV infection, the proportion of HCV antibody-positive individuals who are HCV RNA positive may decrease over time. This is because individuals who have been injecting for longer are more likely to have been treated and thus cleared the virus, reducing the proportion of those currently infected despite a longer duration of risky behavior. Therefore, this justification is consistent with studies such as Alavi et. al (2019), which highlighted that prolonged injection duration often leads to better engagement with harm reduction services over time (\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e), potentially reducing HCV RNA prevalence among HCV antibody-positive patients.\u003c/p\u003e\u003cdiv id=\"Sec8\" class=\"Section3\"\u003e\u003ch2\u003e4.2.2 Needle Exchange Programs and HCV RNA\u003c/h2\u003e\u003cp\u003eUsing needle exchange programs (Exch) showed significant associations in both methods, chi-square results and unadjusted analysis. These results indicate that needle exchange programs are associated with higher odds of HCV RNA positivity. However, in the adjusted analysis, which considers multiple variables simultaneously, the association between needle exchange use and HCV RNA status changed as the C.C. method indicated a non-significant association. In contrast, in the M.I. method, the association remained significant. These findings suggest that even after controlling for other variables, the use of needle exchange programs is still associated with a higher likelihood of HCV RNA positivity in the M.I method. However, this does not align with studies like Des Jarlais et al. (2018) and Grebely et al. (2017), which found that effective needle exchange programs significantly reduce HCV incidence among PWID (\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e). While needle exchange programs are designed to reduce the spread of HCV by providing clean injecting equipment and promoting safer injecting practices, our findings indicate a significant positive association between needle exchange use and HCV RNA positivity. This counter-intuitive result can be attributed to the fact that in our study and for our study population, individuals at higher risk of HCV infection, likely due to factors such as more frequent injecting and sharing of equipment, are more likely to use needle exchange programs, rather than the programs themselves increasing the risk of HCV infection. This suggests the possibility that higher-risk individuals are more likely to use needle exchange programs, as the use of needle exchange itself is unlikely to be a causal factor for infection.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section3\"\u003e\u003ch2\u003e4.2.3 Needle Sharing and HCV RNA\u003c/h2\u003e\u003cp\u003eThe needle-sharing variable had a missing data percentage of 27% in this study. In the unadjusted results, sharing needles had non-significant associations, and the odds ratio dropped between the Complete Case (C.C) and Multiple Imputation (M.I) methods. However, in the adjusted results, sharing needles was found to have a significant association with HCV RNA status in both the C.C and M.I methods, with the change in OR from 24%(C.C) to 13% (M.I) increased likelihood, indicating that the behavior of sharing needles significantly increased the possibility of HCV RNA positivity in this study population in both methods with the M.I results showing a lower odds ratio, possibly the accurate picture after accounting for missing data. This finding aligns with the results of Bruneau et al. (2020), who noted that among injection drug users, sharing syringes is the most significant factor leading to HCV seroconversion (\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e)​. Similarly, Mateu-Gelabert et al. (2010) emphasized that Injection drug users are more prone to engaging in risky injection behaviors during withdrawal periods (\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e)​. Moreover, Platt et al. (2017) highlighted that research indicates that using needles and syringes previously used by others is the primary risk factor for contracting HIV and HCV among people who inject drugs (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e)​. These studies collectively emphasize needle sharing as a critical risk factor for HCV infection, supporting the findings of this analysis that demonstrate improved accuracy and reliability through multiple imputations.\u003c/p\u003e\u003cp\u003e\u003cb\u003eHomelessness and HCV RNA\u003c/b\u003e\u003c/p\u003e\u003cp\u003eHomeless status (Homeless1) showed a strong significant association with HCV RNA in both methods and in univariate and multivariate analyses, indicating that homeless individuals are significantly more likely to be HCV RNA positive even after adjusting for confounding variables. Homelessness is associated with numerous risk factors for HCV, including unstable living conditions, lack of access to healthcare, and a higher likelihood of engaging in high-risk behaviors such as sharing needles or having unprotected sex (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e). These factors contribute to the higher prevalence of HCV among homeless individuals, highlighting the need for targeted interventions to address the unique challenges faced by this population.\u003c/p\u003e\u003cp\u003eThese findings collectively highlight the strong association between homelessness and HCV RNA positivity. The significant associations found in both analyses and the positive correlations observed in the Pearson analysis suggest a consistent relationship between homelessness and higher HCV prevalence. Comparing these findings with the literature, Nyamathi et al. (2019) and Beijer et al. (2018) also reported higher HCV prevalence among homeless individuals (\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e), which aligns with our results. This consistency across studies and methods reinforces the need for targeted interventions and comprehensive support services to address the unique challenges faced by homeless individuals to curb the prevalence of hepatitis C.\u003c/p\u003e\u003cp\u003e\u003cb\u003ePrison before injection commencement and HCV RNA\u003c/b\u003e\u003c/p\u003e\u003cp\u003ePrison before injection commencement association with HCV RNA was significant in Complete Case Analysis but not in Multiple Imputation in Chi-square results. The Pearson correlation for prison before injection commencement showed very weak positive correlations in both methods, suggests that other factors, along with imprisonment, might contribute to the risk of HCV infection. In the unadjusted analysis, previous imprisonment before injection commencement was found to have a significant positive association with HCV RNA status in both methods. After adjusting for confounding factors in the multivariate regression analysis, the association between previous imprisonment before injection commencement and HCV RNA status was significant in the C.C. method but not in the M.I. method. This suggests that, after accounting for missing data, the significance of the association diminishes, indicating that other factors may be at play.\u003c/p\u003e\u003cp\u003eThese results collectively indicate that while previous imprisonment before injection commencement is associated with HCV RNA positivity, this association’s strength and significance can vary depending on the method used to handle missing data. The non-significant association in the M.I. method differs from results by Stone et al. (2021) and Degenhardt et al. (2017), which found higher HCV prevalence among those who have started using drug injections after going to prison (\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e). Rivera Saldana. Et al. study findings indicate a significant relationship between previous imprisonment before injection commencement and various risk behaviors including drug injection, among participants. The increased risk can be attributed to several factors, including the social and economic instability often following release from prison. The study also revealed that the criminalization of drug use and the punitive measures associated with it exacerbate these risks by perpetuating cycles of incarceration and risky behaviors (\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe differences in significance between the methods in my study might be attributed to the nature of the data handling. The C.C method also may overrepresent specific subgroups due to excluding cases with missing data (10% of data in this variable is missing), while M.I attempt to account for missing data, potentially providing a more balanced view. The positive correlations observed in the Pearson analysis further suggest that while injection commencement after prison is an essential factor, other variables also contribute to the risk of HCV infection, necessitating comprehensive prevention and treatment strategies that address multiple risk factors.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section3\"\u003e\u003ch2\u003e4.2.6 Gender and HCV RNA\u003c/h2\u003e\u003cp\u003eGender differences were significant in the chi-square of both methods, with males showing higher HCV prevalence. The Pearson correlation analysis showed a very weak positive correlation in the C.C. method and a weak positive correlation in the M.I. method, suggesting that males are slightly more likely to be HCV positive.\u003c/p\u003e\u003cp\u003eIn the univariate regression analysis, gender was found to have a significant association with HCV RNA status in both methods. These findings suggest a significant gender difference, and the interpretation of the OR less than 1 indicates that females, in this context, might have lower odds of being HCV RNA positive compared to males, which is consistent with the initial chi-square results and literature findings. In the multivariate regression analysis, gender continued to show significant associations with HCV RNA status in only one of the methods. In the C.C. method, a non-significant p-value of 0.1472 suggests no significant difference in HCV RNA positivity between genders when adjusting for other variables. However, in the M.I. method, a highly significant p-value of \u0026gt; 0.001 indicates that females are significantly less likely to be HCV RNA positive compared to males after adjusting for other variables and with 23.45% lower odds compared to males. This result, again, appears consistent with the initial chi-square findings and unadjusted findings and aligns with findings by Zibbell et al. (2018) and Uusküla et al. (2020), which have shown that males are more likely to be hepatitis C positive (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e). This trend may reflect higher engagement in risky injecting behaviors among males compared to females, as well as differential access to healthcare and harm reduction services. Additionally, social, and cultural factors might influence the likelihood of males seeking help or accessing services, further contributing to the observed gender disparity in HCV prevalence.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section3\"\u003e\u003ch2\u003e4.2.7 Condom Use and HCV RNA\u003c/h2\u003e\u003cp\u003eThe use of condoms was significant in the Complete Case (C.C) method but not in the Multiple Imputation (M.I) method in the chi-square test. This difference may be due to the higher variability and missing data in the Condom1 variable (20%), which the M.I. method attempted to address by imputation. The Pearson correlation results showed weak positive correlations in both methods, indicating that condom use is generally associated with a lower risk of HCV. However, its impact may be less pronounced compared to other variables, such as needle sharing. In the univariate regression analysis, condom use was found to have significant associations with HCV RNA status in the C.C method. Conversely, in the M.I method, the non-significant p-value indicated no significant association between condom use and HCV RNA positivity after accounting for missing data. After adjusting for potential confounders for individuals who sometimes use condoms, both the C.C. and M.I analyses showed non-significant p-values, with ORs around 1.10, suggesting a 10% higher likelihood but statistically insignificant association with HCV prevalence. For individuals who always use condoms, the C.C analysis indicated a significant association with an OR of 1.31, meaning a 31% higher odds of HCV prevalence. In contrast, the M.I analysis showed a non-significant p-value with an OR of 1.14, indicating 14% higher odds of HCV. These findings suggest that after controlling for other factors, always using condoms is not significantly associated with being HCV positive. Studies by Leyna GH et al. have shown a correlation between condom use and HCV infection risk in People Who Inject Drugs (\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e), which doesn’t corroborate the multivariate Multiple Imputation findings. The significant association found in the C.C. method might be due to unaccounted confounding factors and the effect of missing data. Given that most HCV infections occur through injection rather than sexual activity, the relevance of condom use may be limited in this context. Therefore, the association between sexual behavior and injecting behavior may be more critical, and after controlling for variables like gender and age, condom use appears less significant.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section3\"\u003e\u003ch2\u003e4.2.8 Ever Been in Prison/Young Offenders Institute and HCV RNA\u003c/h2\u003e\u003cp\u003eEver Been In Prison (Epris1) was significantly associated with HCV RNA in both methods in the Chi-square test, and the Pearson correlation for HCV positivity in those that have been to prison showed very weak positive correlations in both methods, indicating higher HCV prevalence among incarcerated individuals. Prisons are high-risk environments for HCV transmission due to overcrowding, limited access to clean injecting equipment, and inadequate healthcare services. The findings of this study are in line with the research conducted by Stone et al. (2021) and Dolan et al. (2016), which also reported higher HCV prevalence among incarcerated populations (\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e). These studies highlighted similar contributing factors, such as overcrowding and inadequate healthcare. Degenhardt et al. (2017) further corroborated these findings by emphasizing the role of injecting equipment sharing and the lack of harm reduction services in prison (\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e). However, it is essential to consider the bidirectional nature of this correlation, as individuals with high-risk injecting behaviors are more likely to be imprisoned. Those with severe addiction may engage in more risky behaviors, such as sharing needles, which increases their likelihood of contracting HCV and subsequently being incarcerated for related crimes. This hypothesis suggests that high-risk individuals are both more susceptible to HCV and more likely to be incarcerated, thus reinforcing the observed correlation between imprisonment and HCV prevalence.\u003c/p\u003e\u003cp\u003e\u003cb\u003eSTRENGTHS AND LIMITATIONS\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eSTRENGTHS\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe study demonstrates several strengths that significantly enhance its contributions to epidemiology and public health. Firstly, the use of both complete case analysis (CCA) and multiple imputation (MI) techniques to handle missing data provides a comprehensive evaluation of the impact of different data handling methods on estimating Hepatitis C prevalence among people who inject drugs (PWID). This dual-method approach allows for an understanding of how missing data can affect epidemiological findings, thus highlighting potential biases inherent in C.C. analysis and the benefits of MI in producing more reliable results. The study offers valuable insights into effective strategies for managing missing data in public health research by demonstrating how different data handling methods influence prevalence trends. Additionally, the study's large sample size, encompassing approximately 17,000 participants, significantly enhances the statistical power and recruitment from drug services from multiple sites across England, Wales and Northern Ireland, alongside the anonymity of the data collection process, which serves to make it more representative.\u003c/p\u003e\u003cp\u003eOne of the key strengths of this research is the use of the Predictive Mean Matching (PMM) type of multiple imputation technique, which effectively accounts for non-normal and categorical variables, enhancing the robustness of the imputation process. Additionally, through model-based diagnostics, efforts were made to assess the plausibility of the Missing at Random (MAR) assumption. Specifically, a logistic regression was fitted to examine the relationship between all variables used in the imputation model and the outcome variable. This method helps to identify patterns and relationships in the data, supporting the MAR assumption for the outcome variable. However, assessing the MAR assumption fully requires examining the missingness patterns in all variables, not just the outcome. Little’s test was also conducted to rule out the possibility of the missingness being Missing Completely At Random (MCAR) and tilt the missingness more towards MAR which is the assumption I worked with to implement the multiple imputation. Furthermore, including all variables in the dataset—those that could predict missingness, those influencing the process of missing data, and even those not directly relevant to the analysis—further strengthened the study. This comprehensive approach helps to ensure that the imputation process is as accurate and unbiased as possible, providing more reliable results.\u003c/p\u003e\u003cp\u003eAnother notable strength of the study is its robust design. It utilises an annually collected, cross-sectional observational approach with primary data from the Unlinked Anonymous Monitoring (UAM) Survey, which effectively captures real-world conditions and trends over time. Using logistic regression analysis to evaluate the relationships between multiple independent variables and Hepatitis C prevalence offers robust insights into the factors influencing HCV infection. The study includes forest plots to visually represent the logistic regression analysis results, providing a clear and concise depiction of the odds ratios and confidence intervals, which facilitates a better understanding of the significant predictors of HCV prevalence. The study also carefully controls for confounding variables, ensuring that other factors do not spuriously influence the associations observed.\u003c/p\u003e\u003cp\u003e\u003cb\u003eLIMITATIONS\u003c/b\u003e\u003c/p\u003e\u003cp\u003eLimitations of the Imputation Process\u003c/p\u003e\u003cp\u003eThere are notable limitations associated with using the multiple Imputation method. The assumption that data are missing at random (MAR) may not always hold, potentially leading to biased estimates if the missingness is related to unobserved factors. Multiple imputation (MI) could introduce more bias than complete case (CC) analysis if the data are not MAR, as MI might inaccurately impute values based on the observed data patterns. When the missing data is unrelated to the outcome given the covariates, complete case analysis has minimal bias, whereas multiple imputation tends to introduce bias away from the null hypothesis (\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e). Due to time constraints, sensitivity analysis, which involves testing the robustness of key inferences by assuming a range of missing not-at-random (MNAR) mechanisms and re-imputing the data under these different scenarios, could not be conducted. This process would have helped ensure that findings are not overly dependent on the assumption that data is MAR. It would have provided a more comprehensive understanding of how results might vary under different missing data assumptions, such as Missing Not At Random (MNAR), as self-reporting bias on risky behaviors is possible, thereby enhancing the validity of the conclusions (\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eLimitations of the Data\u003c/p\u003e\u003cp\u003eOne primary limitation is the reliance on retrospective observational data from the Unlinked Anonymous Monitoring (UAM) Survey, which may introduce recall bias and inaccuracies in self-reported behaviors and health status. Additionally, there is potential for residual confounding despite efforts to control for known confounders. The study addresses various socioeconomic and behavioral factors, but there may be other unmeasured variables influencing the prevalence and risk factors of Hepatitis C among people who inject drugs (PWID). Furthermore, the study's findings, while comprehensive, are specific to the UK context and may not be fully generalizable to other regions with different healthcare systems, social structures, and HCV prevalence rates. Finally, the cross-sectional nature of the data limits the ability to infer causality between the observed associations, necessitating cautious interpretation of the results.\u003c/p\u003e\u003c/div\u003e"},{"header":"CONCLUSION AND FUTURE DIRECTION","content":"\u003cp\u003eThis study utilized a more specific and sophisticated imputation type, the Predictive Mean Matching (PMM), to handle the missing values in the dataset, alongside comparing the results to that of the Complete-Case analysis. Multiple Imputation provided a more logically accurate trend of HCV prevalence than the Complete-Case Analysis. In the adjusted regression model, several significant findings were observed related to the influence of various factors on Hepatitis C Virus (HCV) prevalence among people who inject drugs (PWID). After multiple imputations, Homelessness was only significant for participants who weren’t homeless in the last year, suggesting that while it is a critical factor, its influence varies over time and conditions. Interestingly, prior imprisonment before injection commencement was not significant for M.I., highlighting that other factors may play a more substantial role in the spread of HCV among PWID. The use of condoms, whether sometimes or always, did not show a significant impact on HCV infection despite having higher odds ratios. This indicates that while sexual transmission is relevant, it might not be the primary mode of HCV spread in this population, as the higher likelihood observed could be due to chance. Gender emerged as a significant factor in M.I., emphasizing the different risk profiles for men and women. Both methods confirmed that the years 2018 to 2021 were significant, with the OR dropping across each year, marking a period of decreased HCV prevalence. This trend coincides with the widespread use of direct-acting antivirals (DAAs) in the UK, which have significantly improved HCV treatment outcomes, reducing the overall prevalence of the virus during these years (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e). These findings from multiple imputation analyses further raises questions about the authenticity of the results of the complete-case analysis, as well as those already existing studies in the literature.\u003c/p\u003e\u003cp\u003eFuture research should delve deeper into understanding the specific mechanisms through which homelessness and gender influence HCV prevalence. Longitudinal studies could provide more insight into how these factors interact over time. Additionally, exploring the role of mental health and access to medical care about HCV among PWID could uncover more comprehensive intervention points. Investigating the effectiveness of various harm reduction programs in different socio-economic contexts will also be essential to tailor culturally and regionally appropriate strategies. By focusing on these areas, future research can contribute significantly to reducing HCV prevalence and improving the overall health outcomes of PWID populations.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eETHICS APPROVAL AND CONSENT TO PARTICIPATE.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research did not involve human participants or access to real-world identifiable data. The analysis was conducted using a synthetic dataset, and therefore, ethical approval was not required.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eACKNOWLEDGEMENTS.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot Applicable\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCONSENT FOR PUBLICATION\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot Applicable\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFUNDING STATEMENT.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo funding was gotten for this research project.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDECLARATION OF INTERESTS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors declare no competing interest.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDATA AVAILABILITY\u003c/strong\u003e:\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis study used a synthetic dataset created to simulate the structure and properties of real-world data from the UK Health Security Agency (UKHSA). No individual-level or identifiable data were accessed. As such, the dataset is not publicly available, and no ethical approval was required, as the study did not involve human participants or confidential data. The dataset used and analyzed for this study is available on request from the corresponding author.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAUTHORS CONTRIBUTION STATEMENT\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAdewunmi Akingbola conceptualized, analyzed the data, and drafted the manuscript. Olajumoke Adewole, Abiodun Adegbesan and Joel Chuku edited the manuscript. All authors agreed to the manuscript.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eCastaneda D, Gonzalez AJ, Alomari M, Tandon K, Zervos XB. From hepatitis A to E: A critical review of viral hepatitis. World J Gastroenterol. 2021 Apr 28;27(16):1691-1715. doi: 10.3748/wjg.v27.i16.1691. PMID: 33967551; PMCID: PMC8072198.\u003c/li\u003e\n\u003cli\u003eWorld Health Organization. Hepatitis C; 2020. Available from: https://www.who.int/news-room/fact-sheets/detail/hepatitis-c\u003c/li\u003e\n\u003cli\u003eNHS UK. Hepatitis C nhs.uk. Available from: https://www.nhs.uk/conditions/hepatitis-c/\u003c/li\u003e\n\u003cli\u003eDe Angelis D, Sweeting M, Ades AE, et al. An evidence synthesis approach to estimating hepatitis C prevalence in England and Wales. Stat Methods Med Res. 2008 Nov 26. doi: 10.1177/0962280208094691.\u003c/li\u003e\n\u003cli\u003eLewer D, Croxford S, Desai M, Emanuel E, Hope VD, McAuley A, et al. The characteristics of people who inject drugs in the United Kingdom: changes in age, duration, and incidence of injecting, 1980\u0026ndash;2019, using evidence from repeated cross‐sectional surveys. Addiction. 2022;117(9):2471\u0026ndash;80. doi: 10.1111/add.15911.\u003c/li\u003e\n\u003cli\u003eYuan J-M, Croxford S, Viviani L, Emanuel E, Phipps E, Desai M. Investigating the sociodemographic and behavioural factors associated with hepatitis C virus testing amongst people who inject drugs in England, Wales and Northern Ireland: A quantitative cross-sectional analysis. Int J Drug Policy. 2022;109(103821):103821. doi: 10.1016/j.drugpo.2022.103821.\u003c/li\u003e\n\u003cli\u003eKarahalios A, Baglietto L, Carlin JB, et al. A review of the reporting and handling of missing data in cohort studies with repeated assessment of exposure measures. BMC Med Res Methodol. 2012;12:96. doi: 10.1186/1471-2288-12-96.\u003c/li\u003e\n\u003cli\u003eKwak S, Kim JH. Statistical data preparation: management of missing values and outliers. Korean J Anesthesiol. 2017 Jul;70(4):407. doi: 10.4097/kjae.2017.70.4.407.\u003c/li\u003e\n\u003cli\u003eKang H. The prevention and handling of the missing data. Korean J Anesthesiol. 2013 May;64(5):402-6. doi: 10.4097/kjae.2013.64.5.402. PMID: 23741561; PMCID: PMC3668100.\u003c/li\u003e\n\u003cli\u003eGardner AL, Charlesworth M. How to write a retrospective observational study. Anaesthesia [Internet]. 2023;78(4):521\u0026ndash;5. Available from: http://dx.doi.org/10.1111/anae.15831\u003c/li\u003e\n\u003cli\u003eCullen KJ, Hope VD, Croxford S, Shute J, Ncube F, Parry JV. Factors associated with recently acquired hepatitis C virus infection in people who inject drugs in England, Wales and Northern Ireland: new findings from an unlinked anonymous monitoring survey. Epidemiology and Infection [Internet]. 2015;143(7):1398\u0026ndash;407. Available from: http://dx.doi.org/10.1017/S0950268814002040\u003c/li\u003e\n\u003cli\u003eConvenience Sampling. Sampling method | descriptive research [Internet]. Ijels.net. [cited 2024 Jul 15]. Available from: https://www.ijels.net/article_162981_06fc448b8914d42316b8d44829c88188.pdf\u003c/li\u003e\n\u003cli\u003eUnlinked Anonymous Monitoring (UAM) survey of HIV and viral hepatitis among people who inject drugs (PWID): 2023 report [Internet]. Gov.uk. [cited 2024 Jul 15]. Available from: https://www.gov.uk/government/publications/people-who-inject-drugs-hiv-and-viral-hepatitis-monitoring/unlinked-anonymous-monitoring-uam-survey-of-hiv-and-viral-hepatitis-among-people-who-inject-drugs-pwid-2023-report\u003c/li\u003e\n\u003cli\u003eTully DC, Power KA, Sarette J, Stopka TJ, Friedmann PD, Korthuis PT, et al. Validation of dried blood spots for capturing hepatitis C virus diversity for genomic surveillance [Internet]. bioRxiv. 2023. p. 2023.07.06.23292160. Available from: http://medrxiv.org/content/early/2023/07/07/2023.07.06.23292160.abstract \u003c/li\u003e\n\u003cli\u003eBhandari P. Independent vs dependent variables [Internet]. Scribbr. 2022 [cited 2024 Jul 15]. Available from: https://www.scribbr.co.uk/research-methods/independent-vs-dependent-variables/\u003c/li\u003e\n\u003cli\u003eAndrade C. A student\u0026rsquo;s guide to the classification and operationalization of variables in the conceptualization and design of a clinical study: Part 1. Indian J Psychol Med [Internet]. 2021;43(2):177\u0026ndash;9. Available from: http://dx.doi.org/10.1177/0253717621994334\u003c/li\u003e\n\u003cli\u003eLittle, R.j.a. (1988) A Test of Missing Completely at Random for Multivariate Data with Missing Values. Journal of the American Statistical Association, 83, 1198-1202. - references - scientific research publishing [Internet]. Scirp.org. [cited 2024 Jul 20]. Available from: https://www.scirp.org/reference/referencespapers?referenceid=1374672 \u003c/li\u003e\n\u003cli\u003evan Buuren S, Groothuis-Oudshoorn K. mice: Multivariate Imputation by Chained Equations in\u003cem\u003eR\u003c/em\u003e. J Stat Softw [Internet]. 2011;45(3):1\u0026ndash;67. Available from: https://www.jstatsoft.org/index.php/jss/article/view/v045i03 \u003c/li\u003e\n\u003cli\u003eMorris TP, White IR, Royston P. Tuning multiple imputation by predictive mean matching and local residual draws. BMC Med Res Methodol [Internet]. 2014;14(1). Available from: http://dx.doi.org/10.1186/1471-2288-14-75 \u003c/li\u003e\n\u003cli\u003eSp PS 2013-P. Combining analysis results from multiply imputed categorical data [Internet]. Lexjansen.com. [cited 2024 Jul 15]. Available from: https://www.lexjansen.com/pharmasug/2013/SP/PharmaSUG-2013-SP03.pdf\u003c/li\u003e\n\u003cli\u003eBrand J, van Buuren S, le Cessie S, van den Hout W. Combining multiple imputation and bootstrap in the analysis of cost‐effectiveness trial data. Stat Med [Internet]. 2019;38(2):210\u0026ndash;20. Available from: http://dx.doi.org/10.1002/sim.7956 \u003c/li\u003e\n\u003cli\u003eAlam S, Ayub MS, Arora S, Khan MA. An investigation of the imputation techniques for missing values in ordinal data enhancing clustering and classification analysis validity. Decision Analytics Journal [Internet]. 2023;9(100341):100341. Available from: http://dx.doi.org/10.1016/j.dajour.2023.100341 \u003c/li\u003e\n\u003cli\u003eMcHugh ML. The Chi-square test of independence. Biochem Med (Zagreb) [Internet]. 2013;143\u0026ndash;9. Available from: http://dx.doi.org/10.11613/bm.2013.018 \u003c/li\u003e\n\u003cli\u003eAndrade C. The P value and statistical significance: Misunderstandings, explanations, challenges, and alternatives. Indian J Psychol Med [Internet]. 2019;41(3):210\u0026ndash;5. Available from: http://dx.doi.org/10.4103/ijpsym.ijpsym_193_19 \u003c/li\u003e\n\u003cli\u003eMcKnight PE, Najab J. Mann‐Whitney U test [Internet]. The Corsini Encyclopedia of Psychology. Wiley; 2010. p. 1\u0026ndash;1. Available from: http://dx.doi.org/10.1002/9780470479216.corpsy0524 \u003c/li\u003e\n\u003cli\u003eMishra P, Singh U, Pandey C, Mishra P, Pandey G. Application of student\u0026rsquo;s t-test, analysis of variance, and covariance. Ann Card Anaesth [Internet]. 2019 [cited 2024 Jul 15];22(4):407. Available from: https://journals.lww.com/aoca/fulltext/2019/22040/application_of_student_s_t_test,_analysis_of.11.aspx \u003c/li\u003e\n\u003cli\u003eSperandei S. Understanding logistic regression analysis. Biochem Med (Zagreb) [Internet]. 2014;12\u0026ndash;8. Available from: http://dx.doi.org/10.11613/bm.2014.003 \u003c/li\u003e\n\u003cli\u003eSzumilas M. Explaining odds ratios. Journal of the Canadian Academy of Child and Adolescent Psychiatry. 2010;19(3):227.\u003c/li\u003e\n\u003cli\u003eAdult substance misuse treatment statistics 2021 to 2022: report [Internet]. Gov.uk. [cited 2024 Jul 16]. Available from: https://www.gov.uk/government/statistics/substance-misuse-treatment-for-adults-statistics-2021-to-2022/adult-substance-misuse-treatment-statistics-2021-to-2022-report \u003c/li\u003e\n\u003cli\u003ePradat P, Voirin N, Tillmann HL, Chevallier M, Tr\u0026eacute;po C. Progression to cirrhosis in hepatitis C patients: an age‐dependent process. Liver Int [Internet]. 2007;27(3):335\u0026ndash;9. Available from: http://dx.doi.org/10.1111/j.1478-3231.2006.01430.x \u003c/li\u003e\n\u003cli\u003ePublic Health Scotland. Needle Exchange Surveillance Initiative (NESI) - Needle Exchange Surveillance Initiative (NESI) - publications - public health Scotland [Internet]. Publichealthscotland.scot. [cited 2024 Jul 16]. Available from: https://publichealthscotland.scot/publications/needle-exchange-surveillance-initiative-nesi/needle-exchange-surveillance-initiative-nesi/\u003c/li\u003e\n\u003cli\u003eHope VD, Eramova I, Capurro D, Donoghoe MC. Prevalence and estimation of hepatitis B and C infections in the WHO European Region: a review of data focusing on the countries outside the European Union and the European Free Trade Association. Epidemiol Infect [Internet]. 2014 [cited 2024 Jul 16];142(2):270\u0026ndash;86. Available from: https://www.cambridge.org/core/journals/epidemiology-and-infection/article/prevalence-and-estimation-of-hepatitis-b-and-c-infections-in-the-who-european-region-a-review-of-data-focusing-on-the-countries-outside-the-european-union-and-the-european-free-trade-association/9661546AEC010338F9857A7839D9E6C6\u003c/li\u003e\n\u003cli\u003eAlavi M, Grebely J, Micallef M, Dunlop AJ, Balcomb AC, Day CA, et al. Assessment and treatment of hepatitis C virus infection among people who inject drugs in the opioid substitution setting: ETHOS study. Clin Infect Dis [Internet]. 2013 [cited 2024 Jul 16];57(suppl_2):S62\u0026ndash;9. Available from: https://academic.oup.com/cid/article/57/suppl_2/S62/398725?login=true \u003c/li\u003e\n\u003cli\u003eDes Jarlais DC, Arasteh K, McKnight C, Hagan H, Perlman D, Friedman SR. Using hepatitis C virus and herpes simplex virus-2 to track HIV among injecting drug users in New York City. Drug Alcohol Depend [Internet]. 2009;101(1\u0026ndash;2):88\u0026ndash;91. Available from: http://dx.doi.org/10.1016/j.drugalcdep.2008.11.007 \u003c/li\u003e\n\u003cli\u003eGrebely J, Dore GJ. Prevention of hepatitis C virus in injecting drug users: A narrow window of opportunity. J Infect Dis [Internet]. 2011 [cited 2024 Jul 16];203(5):571\u0026ndash;4. Available from: https://academic.oup.com/jid/article/203/5/571/896768\u003c/li\u003e\n\u003cli\u003eBruneau J, Roy \u0026Eacute;, Arruda N, Zang G, Jutras-Aswad D. The rising prevalence of prescription opioid injection and its association with hepatitis C incidence among street‐drug users. Addiction [Internet]. 2012;107(7):1318\u0026ndash;27. Available from: http://dx.doi.org/10.1111/j.1360-0443.2012.03803.x \u003c/li\u003e\n\u003cli\u003eMateu-Gelabert P, Sandoval M, Meylakhs P, Wendel T, Friedman SR. Strategies to avoid opiate withdrawal: Implications for HCV and HIV risks. Int J Drug Policy [Internet]. 2010;21(3):179\u0026ndash;85. Available from: http://dx.doi.org/10.1016/j.drugpo.2009.08.007 \u003c/li\u003e\n\u003cli\u003ePlatt L, Minozzi S, Reed J, Vickerman P, Hagan H, French C, et al. Needle and syringe programmes and opioid substitution therapy for preventing HCV transmission among people who inject drugs: findings from a Cochrane Review and meta‐analysis. Addiction [Internet]. 2018;113(3):545\u0026ndash;63. Available from: http://dx.doi.org/10.1111/add.14012 \u003c/li\u003e\n\u003cli\u003eArum C, Fraser H, Artenie AA, Bivegete S, Trickey A, Alary M, et al. Homelessness, unstable housing, and risk of HIV and hepatitis C virus acquisition among people who inject drugs: a systematic review and meta-analysis. Lancet Public Health [Internet]. 2021;6(5):e309\u0026ndash;23. Available from: http://dx.doi.org/10.1016/s2468-2667(21)00013-x \u003c/li\u003e\n\u003cli\u003eNyamathi A, Salem BE, Zhang S, Farabee D, Hall B, Khalilifard F, et al. Nursing case management, peer coaching, and hepatitis A and B vaccine completion among homeless men recently released on parole: Randomized clinical trial. Nurs Res [Internet]. 2015 [cited 2024 Jul 16];64(3):177\u0026ndash;89. Available from: https://journals.lww.com/nursingresearchonline/fulltext/2015/05000/nursing_case_management,_peer_coaching,_and.6.aspx \u003c/li\u003e\n\u003cli\u003eBeijer U, Wolf A, Fazel S. Prevalence of tuberculosis, hepatitis C virus, and HIV in homeless people: a systematic review and meta-analysis. Lancet Infect Dis [Internet]. 2012;12(11):859\u0026ndash;70. Available from: http://dx.doi.org/10.1016/s1473-3099(12)70177-9 \u003c/li\u003e\n\u003cli\u003eStone J, Fraser H, Lim AG, Walker JG, Ward Z, MacGregor L, et al. Incarceration history and risk of HIV and hepatitis C virus acquisition among people who inject drugs: a systematic review and meta-analysis. Lancet Infect Dis [Internet]. 2018;18(12):1397\u0026ndash;409. Available from: http://dx.doi.org/10.1016/s1473-3099(18)30469-9 \u003c/li\u003e\n\u003cli\u003eDegenhardt L, Peacock A, Colledge S, Leung J, Grebely J, Vickerman P, et al. Global prevalence of injecting drug use and sociodemographic characteristics and prevalence of HIV, HBV, and HCV in people who inject drugs: a multistage systematic review. Lancet Glob Health [Internet]. 2017;5(12):e1192\u0026ndash;207. Available from: http://dx.doi.org/10.1016/s2214-109x(17)30375-3 \u003c/li\u003e\n\u003cli\u003eRivera Saldana CD, Beletsky L, Borquez A, Kiene SM, Strathdee SA, Z\u0026uacute;\u0026ntilde;iga ML, et al. Impact of cumulative incarceration and the post‐release period on syringe‐sharing among people who inject drugs in Tijuana, Mexico: a longitudinal analysis. Addiction [Internet]. 2021;116(10):2724\u0026ndash;33. Available from: http://dx.doi.org/10.1111/add.15445 \u003c/li\u003e\n\u003cli\u003eZibbell JE, Asher AK, Patel RC, Kupronis B, Iqbal K, Ward JW, et al. Increases in acute hepatitis C virus infection related to a growing opioid epidemic and associated injection drug use, United States, 2004 to 2014. Am J Public Health [Internet]. 2018;108(2):175\u0026ndash;81. Available from: http://dx.doi.org/10.2105/ajph.2017.304132 \u003c/li\u003e\n\u003cli\u003eUusk\u0026uuml;la A, Raag M, Marsh K, Talu A, Vorobjov S, Des Jarlais D. HIV prevalence and gender differences among new injection-drug-users in Tallinn, Estonia: A persisting problem in a stable high prevalence epidemic. PLoS One [Internet]. 2017;12(2):e0170956. Available from: http://dx.doi.org/10.1371/journal.pone.0170956 \u003c/li\u003e\n\u003cli\u003eLeyna GH, Makyao N, Mwijage A, Ramadhan A, Likindikoki S, Mizinduko M, et al. HIV/HCV co-infection and associated risk factors among injecting drug users in Dar es Salaam, Tanzania: potential for HCV elimination. Harm Reduct J [Internet]. 2019;16(1). Available from: http://dx.doi.org/10.1186/s12954-019-0346-y \u003c/li\u003e\n\u003cli\u003eDolan K, Wirtz AL, Moazen B, Ndeffo-mbah M, Galvani A, Kinner SA, et al. Global burden of HIV, viral hepatitis, and tuberculosis in prisoners and detainees. Lancet [Internet]. 2016;388(10049):1089\u0026ndash;102. Available from: http://dx.doi.org/10.1016/s0140-6736(16)30466-4 \u003c/li\u003e\n\u003cli\u003eWhite IR, Carlin JB. Bias and efficiency of multiple imputation compared with complete‐case analysis for missing covariate values. Stat Med [Internet]. 2010;29(28):2920\u0026ndash;31. Available from: http://dx.doi.org/10.1002/sim.3944\u003c/li\u003e\n\u003cli\u003eSterne JAC, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ [Internet]. 2009 [cited 2024 Jul 16];338(jun29 1):b2393\u0026ndash;b2393. Available from: https://www.bmj.com/content/338/bmj.b2393 \u003c/li\u003e\n\u003cli\u003eValerio H, Alavi M, Conway A, Silk D, Treloar C, Martinello M, et al. Declining prevalence of current HCV infection and increased treatment uptake among people who inject drugs: The ETHOS Engage study. Int J Drug Policy [Internet]. 2022;105(103706):103706. Available from: http://dx.doi.org/10.1016/j.drugpo.2022.103706 \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Hepatitis C, People Who Inject Drugs, Missing Data, Multiple Imputation, Complete-Case Analysis, Surveillance Data","lastPublishedDoi":"10.21203/rs.3.rs-6994675/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6994675/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground:\u003c/h2\u003e\u003cp\u003eHepatitis C virus (HCV) remains a major public health challenge, particularly among People Who Inject Drugs (PWIDs). Missing data in surveillance systems can bias prevalence estimates, affecting decision-making. This study compares Complete-Case Analysis (CCA) and Multiple Imputation (MI) for handling missing data in the estimation of HCV prevalence using a simulated dataset derived from the UK\u0026rsquo;s Unlinked Anonymous Monitoring (UAM) Survey.\u003c/p\u003e\u003ch2\u003eMethodology:\u003c/h2\u003e\u003cp\u003eWe conducted a cross-sectional analysis using a simulated version of the UAM dataset, focusing on key demographic and behavioural variables. HCV prevalence was estimated using both CCA and MI approaches. MI was performed using chained equations with five imputations. The effect of missing data handling on prevalence estimates and associated confidence intervals was compared between methods.\u003c/p\u003e\u003ch2\u003eResults:\u003c/h2\u003e\u003cp\u003eHCV prevalence estimates obtained via MI were consistently higher than those from CCA, with narrower confidence intervals. The CCA approach excluded a substantial proportion of cases due to missing data, introducing potential bias. MI preserved sample size and yielded more robust estimates, particularly among subgroups with higher missingness.\u003c/p\u003e\u003ch2\u003eConclusion:\u003c/h2\u003e\u003cp\u003eMultiple Imputation outperformed Complete-Case Analysis in estimating HCV prevalence from the simulated UAM data. These findings highlight the importance of appropriate missing data methods in epidemiological surveillance and public health research.\u003c/p\u003e","manuscriptTitle":"Improving Prevalence Estimates of Hepatitis C in Key Populations: A Simulated Data-Based Comparison of Missing Data Techniques","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-06 08:12:43","doi":"10.21203/rs.3.rs-6994675/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-11-12T07:46:43+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-05T17:39:07+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"127231073628295444503477489913800846581","date":"2025-10-22T02:25:21+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"266009622462055567070001214957337185165","date":"2025-10-20T11:52:31+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-15T16:29:35+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"179827714449335677509015703134407847980","date":"2025-09-14T18:11:07+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-07-31T08:53:57+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-07-31T08:48:33+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-07-18T05:50:58+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-07-15T16:14:45+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-07-15T15:05:12+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"cf62e0cf-6d27-4c0c-b958-592c3f5ea2a7","owner":[],"postedDate":"August 6th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[{"id":52541861,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":52541862,"name":"Health sciences/Diseases"},{"id":52541863,"name":"Health sciences/Gastroenterology"},{"id":52541864,"name":"Health sciences/Health care"},{"id":52541865,"name":"Health sciences/Medical research"},{"id":52541866,"name":"Health sciences/Risk factors"}],"tags":[],"updatedAt":"2025-11-12T07:53:51+00:00","versionOfRecord":[],"versionCreatedAt":"2025-08-06 08:12:43","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6994675","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6994675","identity":"rs-6994675","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}