Child and Youth Chronic Physical Health Conditions: A Comparison of Survey Data and Linked Administrative Health Data in Ontario

Child and Youth Chronic Physical Health Conditions: A Comparison of Survey Data and Linked Administrative Health Data in Ontario | 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 Research Article Child and Youth Chronic Physical Health Conditions: A Comparison of Survey Data and Linked Administrative Health Data in Ontario Grace Golden, Li Wang, Graham J. Reid This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5089891/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 02 Oct, 2025 Read the published version in BMC Pediatrics → Version 1 posted 3 You are reading this latest preprint version Abstract Background : Population-based studies in Canada and the United States estimate chronic physical health conditions affect between 20% to 30% of children aged 0 to 17. One challenge in measuring chronic conditions is that researchers often use inconsistent definitions. The main objective was to develop a chronic health condition (CHC) algorithm. We identified three main elements that must be incorporated from administrative data to determine whether a child has a CHC: (1) the diagnosis recorded for the visit, (2) the number of visits, and (3) within a specific reference period. Methods : Data were from the cross-sectional 2014 Ontario Child Health Study, linked with Ontario Health Insurance Plan (OHIP) data. Unweighted prevalence estimates and agreement analyses (Cohen’s Kappa, sensitivity, specificity) were used to compare the survey parent-reported and algorithm-based presence of a CHC. Results : 31.59% and 26.7% of children and youth had a CHC based on administrative and survey data, respectively. Agreement between administrative and survey data was poor ( k = 0.16). Among a few specific conditions, agreement varied depending on the type of condition (e.g., diabetes k = 0.79 vs health conditions k = 0.21). Conclusion : We found considerable discrepancies between administrative and survey-reported data. The results highlight the importance of using algorithms developed from multiple datasets to examine complex research questions, such as the measurement of chronicity. mental health chronic health condition child and youth survey data administrative data Ontario Health Insurance Plan 2014 Ontario Child Health Study algorithm Figures Figure 1 Figure 2 Figure 3 Introduction Perrin et al. [1] defined a chronic health condition (CHC) as a long-term condition with a biological basis that has lasted, or is expected to last, for 3 months or longer. Population-based studies in Canada and the United States estimate chronic physical health conditions affect between 20% to 30% of children aged 0 to 17 [2-5]. One challenge in measuring chronic conditions is that researchers often use inconsistent definitions. For example, the 3-month timeframe is often not used, with many researchers using 6- or 12-month timeframes [6]. Recent research advises that using two data sources, when possible (e.g., through data linkage studies), may help to answer complex research questions [7]. One source of data is population-based surveys. Survey data offers detailed data on randomly sampled populations, including extensive sociodemographic information [8]. However, survey data may be subject to bias (e.g., recall bias) [7]. Another valuable data source is administrative data. Many studies have used administrative health billing data to answer important questions about children and youth with CHC. For example, administrative data have been used to determine incidence and prevalence, identify risk factors associated with developing certain conditions (e.g., diabetes), and identify healthcare expenditures and use patterns associated with specific chronic conditions [9-11]. Among each of these studies, the first step required identifying children with a specific chronic condition using an algorithm. However, these studies only examined one or a few chronic conditions. Administrative data are often used for disease surveillance or health service research, despite the initial collection not being intended for this purpose [12]. There are several advantages of administrative data. The data are cost-effective and often readily available [13], encompass large populations, and capture many types of health events (e.g., office visits, emergency department use) [12, 14]. There are concerns and cautions with these data. One issue is data quality -- accuracy and completeness [12, 15, 16, 17]. A review examining electronic medical record data quality for seven chronic diseases found that completeness of records was associated with the type of provider, the patient load, and the clinic or organization’s funding model (e.g., fee-for-service, salary-based) [18]. A second issue is that due to restrictions on the amount of information that can be recorded, some aspects of the visit details may be incomplete [19]. For example, a family doctor may only code a single diagnosis for a visit despite addressing multiple problems. Third, there are unique challenges related to identifying children and youth with chronic conditions based on health administrative data. For example, a physician may assess a patient and submit a billing claim with a diagnostic code indicating an acute or chronic condition. Whether or not that child/youth has a chronic condition depends on various factors. Some conditions may resolve on their own. A chronic condition code may be part of an evaluation, which later reveals a negative result. Thus, for some conditions, it is only when the same diagnostic code is assigned again, can we infer that the condition is chronic. Existing Algorithms There are three main elements that must be incorporated from administrative data to determine whether a child has a CHC: (1) the diagnosis recorded for the visit, (2) the number of visits, and (3) within a specific reference period. The reference period refers to the number of years during which the number of visits must be met to identify a chronic condition [20]. One well-established algorithm was developed by the United States Department of Health and Human Services Chronic Conditions Warehouse (CCW) to identify chronic conditions amongst adults with the greatest impact on population health using administrative claims (CCW) [21]. The algorithm uses the International Classification of Diseases (ICD)-9 and ICD-10 codes for common adult conditions (e.g., diabetes, hypertension, breast cancer). To meet the criteria for a chronic condition, a patient must have one to two outpatient billing claims within a specific reference period of 1 to 3 years [22]. The reference period varies depending on the type of condition (e.g., 1 year for breast cancer vs 3 years for Alzheimer’s) [21]. One criticism of this algorithm concerns the reference period. Shorter periods may not identify patients with more irregular access to health care or individuals with well-managed conditions who rarely have follow-up appointments [22]. Nonetheless, the algorithm has been validated and demonstrated more accurate prevalence estimates of chronic conditions over time compared to self-report [22]. Another criticism of the CCW algorithm is that the algorithm only provides reference periods for 27 chronic conditions; further, there are a number of conditions that are rare or non-existent among children or youth [21]. For example, CCW conditions include Alzheimer’s disease, osteoporosis, and rheumatoid arthritis, all of which are rare in children [23, 24, 25]. Other algorithms for chronic conditions have been used, but they also have limitations. Within the Ontario context, algorithms have been examined for a number of specific conditions (e.g., asthma, dementia, diabetes, rheumatoid arthritis) using Institute for Clinical Evaluative Sciences (ICES) datasets, but similarly, these studies have only been with adults [26-30]. The Clinical Risk Groups (CRG) system has been used to assess children, but the algorithm requires a minimum of 2 claims during a 12-month period [31], which, again, has been criticized as too short of a reference period [22]. Nevertheless, the CRG system was designed to assign a severity level rather than solely to determine the presence or absence of a chronic condition. Another widely used algorithm is the pediatric complex Chronic Condition Classification (CCC) system version 2 [32], which provides a comprehensive list of ICD-9 and ICD-10 diagnostic codes to identify a patient with CCC category (e.g., gastrointestinal, respiratory, and cardiovascular). The current study. No study has attempted to capture chronic health conditions across a wide age range during childhood and adolescence using a standardized algorithm that can be applied to the general population. Additionally, this study examined which sociodemographic variables may influence whether a child is identified as having a CHC. Objectives Objective 1) Develop an algorithm to identify children and youth ages 4 to 17 with chronic health conditions. Objective 2) Compare the relationship between parent-reported vs algorithm-based presence of a chronic health condition. Objective 3) Identify sociodemographic variables that relate to the presence or absence of a CHC as identified only by an algorithm applied to administrative data, parent-reported data, or both. Methods The reporting of our study follows the REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) Statement [33]. See Supplemental Material A, Table A.1 for the checklist. Data Data were from the cross-sectional 2014 Ontario Child Health Study (OCHS-2014) [34], linked with Ontario Ministry of Health (MOH; formerly MOHLTC) administrative data. We describe the survey first, as survey participants formed the basis for the study, and then the health administrative data. Finally, the methods used to develop the algorithm are outlined. Survey Data The methodology for the 2014 OCHS has been reported elsewhere [34]. Briefly, Statistics Canada invited over 12,000 eligible households across Ontario of children and youth between the ages of 4 and 17 and their respective families to participate in the 2014 study [34]. Interviews took place between October 2014 to September 2015. The sampling framework was a clustered, stratified, random sample of households from the 2014 Canada Child Tax Benefit [34]. In total, 10,802 children and youth were in the study. Of those, 9,666 parents and/or youth agreed to link their survey data with administrative health data, and 9,301 were able to be linked. Youth and their parents completed at-home interviews. In families with more than one child, a ‘target child’ was randomly selected for detailed assessments, while data from other children (i.e., siblings) in the family were less extensive. The person most knowledgeable (PMK) completed self-report questionnaires and questionnaires about the target child and siblings. Youth aged 12 to 17 provided their own reports using self-report questionnaires on laptops. (N.B. Parent-reported data will henceforth be referred to as survey data). Administrative Data Ontario administrative health data are available for services provided in hospitals (e.g., emergency room visits) or outpatient services provided by physicians (e.g., walk-in clinic visits) and covered by the Ontario Health Insurance Plan (OHIP) [35]. Family physicians in Ontario may be paid via alternative payment plans (e.g. blended capitation payment models); data are still obtained via shadow billing [36]. Specialist physicians bill directly by submitting claims that include a diagnostic and a service fee code [37]. The diagnostic code in OHIP is a truncated version (3 digits) of the International Classification of Diseases (ICD) [38]. All analyses were performed on outpatient billing claims using ICD-9 codes. ICD-10 codes were not implemented until October 2015, after the study window. (N.B. ICD-9 diagnostic codes will henceforth be referred to as diagnostic codes. Diagnoses associated with each code will henceforth be referred to as conditions). Data were obtained for the reference period up to a maximum of 4 years before the date of the 2014 interview (see Figure 1). Reference periods were unique for each participant and based on their interview data. Current Study and Exclusion Criteria . This study had two main exclusion criteria. 1) Participants were excluded from the study if they had i) a parent-reported developmental delay or intellectual disability in 2014 OCHS or ii) an administrative billing claim with the diagnosis code of intellectual disability (ICD-9 codes 317-319). This paper is part of a larger study, where the overarching aim is to examine how children with comorbid chronic health and mental health conditions access mental health services. Children with developmental delays and intellectual disabilities often require cross-specialty services, thereby making it unclear the primary purpose of a healthcare visit [39]. 2) Participants were excluded if they had missing data related to the primary outcome, which was long-term conditions reported by parents (see Figure 2 Sample Selection). Measures Survey Data The PMK reported whether a health professional diagnosed any of 11 long-term conditions. Specifically, they were asked, “Has a health professional diagnosed any of the following long-term conditions for the child?” The conditions included 1) food and digestive allergies, 2) respiratory allergies, 3) other allergies, 4) bronchitis, 5) diabetes, 6) heart condition or disease, 7) epilepsy, 8) cerebral palsy, 9) kidney condition or disease, 10) asthma, and 11) eczema. Importantly, PMK was also provided with an “Other” option to report “any other long-term condition” and asked to specify the condition in text form. Unfortunately, the text data for specific conditions reported by parents are unavailable and not coded by Statistics Canada. Long-term conditions were based on the National Longitudinal Survey of Children and Youth (NLSCY) cycle 8, the Survey of Young Canadians (SYC), and the 1983 OCHS [40, 41, 42]. The validity of the items, as used in the NLSY or the 1983 or 2014 OCHS, has not been examined. Predictor Variables Table 1 presents the 2014 OCHS items, questions asked to the PMK/Partner/Youth, response options, and how variables were derived. Child Characteristics. Age was grouped as child (age 4-11) or youth (age 12-17), and sex was either male or female. Parent and Family Characteristics. PMKs reported their highest level of education and whether the child lives in a two- or single-parent household. Household income strata were based on items from the 2013 Canadian Community Health Survey (CCHS) [43]. Strata were coded as low = <20 th percentile, medium = 20 th - 80 th percentile Child Clinical Characteristics. Health Utility Index (HUI). PMK reported the health problems for the child using the Health Utilities Index (HUI) [44]. The HUI describes an individual’s functional health status based on eight attributes (vision, speech, hearing, ambulation, dexterity, emotion, pain, and cognition). The HUI uses a coding algorithm [44] to generate a single-attribute utility score ranging from -1.00 (severe disability) to 1.00 (full health). Statistics Canada computed scores for each participant. The HUI has a skewed distribution, with most participants scoring 1.00 (full health). Categories were derived from Statistics Canada scores using data from TC and siblings [45]. HUI was coded as 0 = no problems (1.00), 1 = mild problems (0.90-<1.00), 2 = moderate problems (0.68-<0.90) and 3 = severe problems (<0.68; top 5 th percentile), the HUI has been widely used to understand health status [44, 46]. Table 1 Predictors of Chronic Condition Status Variable (2014 OCHS item) Description or Question asked to PMK/Youth Response Options/Code Recode Child Characteristics Age (MEM_AGE) Based on demographics. Continuous 4-17 0 = 4-11 1 = 12-17 Sex (MEM_SEX) Based on demographics. Male Female 0 = Male 1 = Female Parent and Family Characteristics PMK Educational status (EHGP_04) What is the highest certificate, diploma or degree that you have completed? Less than high school or equivalent High school diploma or equivalency certificate Trade certificate or diploma College CEGEP or other non-university certificate or diploma or below bachelor’s level Bachelor’s degree University certificate diploma or degree above bachelor’s level 0 = Trade certificate or high school or less 1 = College/CEGEP or below bachelor’s level 2 = Bachelor’s degree (BA, BSc) 3 = University certificate diploma or degree above bachelor’s level Parent structure (DEMDV04) Based on demographics. Two parents Single parent 0 = Two parent household 1 = Single parent household Household income strata (PSTRATAH) Income strata were created after data collection. Low Medium High 0 = Low 1 = Medium 2 = High Child Clinical Characteristics Health Utility Index (HUIDHSI) HUI uses a coding algorithm to generate a single-attribute utility score. Continuous scores range from -1 to +1. 0 = No problems 1 = Mild problems 2 = Moderate problems 3 = Severe problems Developing the Chronic Health Condition (CHC) Algorithm Below, we review the sequence of developing the CHC algorithm (see Figure 3 for the algorithm development), which involved (1) selecting diagnostic codes and (2) identifying the billing claim requirements and the reference period for each condition. Selecting diagnostic codes Supplemental Material B, Table B.2 presents 219 diagnostic codes used in the final version of the algorithm. Three sources of information were used to select diagnostic codes. First, we selected 198 diagnostic codes based on the pediatric Chronic Condition Classification (CCC) system version 2 (see Table B.1 Supplemental Material B) [32]. Second, we identified and added 12 common chronic conditions with high prevalence among pediatric populations [47-53]. The 12 conditions are commonly asked about in population-based surveys. Third, we examined the total frequency of diagnostic codes billed for each child within a 1-year time frame before the date of the 2014 interview to ensure we did not miss any conditions that may be considered chronic. Chronicity was based on scientific literature or conditions that require ongoing care (e.g., multiple visits), implying that they are chronic. Codes were examined based on the frequency of visits. Nine additional chronic conditions were added. Establishing Reference Period and Minimum Visit Frequency The CCW codebook criteria [21] were used as the starting point. Where applicable, reference period and visit frequency criteria were applied to similar conditions not in the CCW but in the ICD-9. (1) First, a total of 5 conditions had an established reference period with a specified minimum number of visits (i.e., established algorithm) to be considered a CHC based on the CCW codebook. Table B.2 in Supplemental Material B lists all reference periods considered for conditions. (2) Second, the CCW codebook provides established criteria for some, but not all, cancer-related conditions (e.g., breast, colorectal, endometrial, lung). In the CCW, all cancer conditions used the criteria of 2 billing claims in 1 year; we applied this to 92 other cancer-related conditions not listed in the CCW codebook but included in the ICD-9. (3) Third, the CCW used 1 billing claim within 1 year as the criteria for conditions deemed as chronic due to the incurable or irreversible nature of the condition (e.g., anemia, diabetes). We applied these same criteria to 18 conditions (e.g., HIV, infantile cerebral palsy, chromosomal conditions) that, based on the scientific literature, are incurable or irreversible conditions. No existing coding systems for pediatric or adult chronic conditions provide minimum visit criteria for these conditions. Table B.2 provides references used to make this decision for each diagnostic code (4) There were 104 conditions for which the optimal criteria were uncertain. Based on the literature, these conditions were grouped as A) curable conditions and B) diagnostic codes with multiple conditions/subcategories. A) Conditions with a cure may be acute or chronic. For example, conditions that arise due to premature labour or neonatal complications may resolve on their own, require treatment, or persist throughout adulthood [54-57]. Whether a condition is acute or chronic will depend on whether the person receives treatment, and/or whether the condition can resolve on its own after some time or other factors. B) Lastly, several diagnostic codes list multiple conditions or are too broad to understand the prognosis of related conditions (e.g., code 279 = Disorders involving the immune mechanism, code 348 = Other conditions of the brain). Past studies have used longer “look-back” periods to reduce false negatives (i.e., unobserved chronic conditions) and identify conditions that may have more irregular or intermittent access to health services [58, 59]. As such, we examined a minimum of 2 billing claims within 2, 3, and 4 years for A) curable conditions or B) diagnostic codes with multiple conditions/subcategories. The assumption is that at least 2 visits may reflect the ongoing nature of monitoring and/or treatment for the condition, which can be inferred as a chronic condition. Data Analyses Analyses were conducted using Statistical Package for Social Sciences (SPSS) version 29 and STATA 17.0 [60,61]. We applied sampling weights in all analyses to generate estimates that are representative of the population of children and youth in Ontario. To account for the complex survey design, mean bootstrap weights were applied with an adjustment factor to produce accurate standard errors. Sensitivity Analyses Sensitivity analyses were conducted by adjusting the reference period only for conditions where the reference period was uncertain (see Supplemental Material B, section 2.1 for more information). The aim was to assess how the overall prevalence of CHC and patterns of agreement were affected when we adjusted the reference period to 2, 3, and 4 years. First, we examined the overall prevalence of CHC by assessing whether a child met the criteria for any condition using survey or administrative data. Second, we examined patterns of agreement using Cohen’s kappa [62], sensitivity, and specificity estimates between administrative and survey-reported conditions. Sensitivity and specificity usually refer to the accurate identification in comparison to the gold standard: sensitivity examines the proportion of actual positive cases that a test correctly identifies, while specificity examines the proportion of true negative cases. However, there is no gold standard for our analyses [63]. Sensitivity was calculated as the probability of being classified with a CHC in both administrative and survey data. Specificity was calculated as the probability of not being classified with a CHC in either administrative or survey data. Previous research investigating agreement between administrative and survey data has used similar methods [8]. See Table B.2 in Supplemental Material B for a complete list of conditions that required sensitivity testing (i.e., reference period uncertain). Third, given the poor agreement and high prevalence estimates (e.g., 4-years identified >50% children with CHC), we increased the billing claim requirements for each condition where the reference period was uncertain (i.e., conditions that required sensitivity testing) but not the reference period (henceforth referred to as the incremented billing claim requirement) with the goal of optimizing the algorithm. Finally, after determining which of the reference periods was the optimal algorithm, we examined a few specific conditions we would expect to have strong agreement between survey and administrative data. Table C.1 in Supplemental Material C documents which set(s) of diagnostic codes were analyzed with parent-reported data. Multinomial Logistic Regression A multinomial logistic regression examined the association between sociodemographic variables and whether the child was categorized as: 0 = no long-term chronic condition, 1 = administrative CHC only, 2 = survey CHC only, or 3 = both administrative and survey. The independent variables included 1) child characteristics (age, sex), 2) family and household characteristics (PMK education status, parent structure, income strata), and 3) child clinical characteristics (health utility index). All analyses are presented as relative risk ratios (RRR) with corresponding 95% confidence intervals (CI). Missing Data Multiple imputation by chained equation (MICE) was used to address missing data by imputing 10 sets of data using STATA based on best practice recommendations [64]. The specific methodology has been reported elsewhere [65]. Results The total sample included 8,985 [1] children aged 4 to 17. Characteristics of the sample population, weighted to reflect the general Ontario population, are presented in Table 2. Table 2 Sociodemographic and Clinical Characteristics of Sample Selection Sociodemographic Characteristics % Child Characteristics Age ( n = 8,985) 4-11 12-17 56.11 43.89 Sex ( n = 8,985) Female Male 49.19 50.81 Parent and Family Characteristics PMK Education Status ( n = 7,523) Trade certificate; High school or less College/CEGEP; University certificate below BA Bachelor’s degree (BA, BSc) University certificate above Bachelors Parent Structure ( n = 8,925) Two parent household Single parent household 10.77 44.69 30.04 14.49 80.38 19.62 Income strata ( n = 8,985) Low Medium High 20.73 57.11 22.16 Child Clinical Characteristics Health Utility Index ( n = 8,874) No problems Mild problems Moderate problems Severe problems 57.25 21.59 16.81 4.35 Prevalence and Agreement Between Administrative and Survey Data: Any Condition Prevalence: Any Condition Table 3 presents the unweighted prevalence estimates and concordance measures. The initial billing claim requirement demonstrates that 31.60% of children met the criteria for a CHC using the 2-year reference period, 41.69% of children met the criteria for a CHC using the 3-year reference period, and 50.71% of children met the criteria for a CHC using the 4-year reference period. (N.B. 24% of billing claims had the diagnostic code 999 (i.e., without diagnosis). After the initial round of analyses, we increased the billing claim requirements (i.e., from 2 billing claims to 3 billing claims in 2, 3, and 4 years) for each condition where the reference period was uncertain (referred to as the incremented billing claim requirement). This decision was due to the extremely high prevalence estimates (> 50% using the 4-year reference period), which does not align with previous literature examining the prevalence of pediatric chronic conditions (e.g., 20% - 30%) [2-5]. Also, the prevalence of long-term conditions reported using survey data was 26.7%. The incremented billing claim requirement found that 23.27% of children met the criteria for a CHC using the 2-year reference period, 31.59% of children met the criteria for a CHC using the 3-year reference period, and 39.50% of children met the criteria for a CHC using the 4-year reference period. Agreement: Any Condition Both the initial and incremented billing claim requirements have low sensitivity and specificity. According to Cohen’s [62] criteria, the agreement remains slight (2 years: k = 0.15, 3 years: k = 0.13, 4 years: k = 0.11) using the initial billing claim requirement. The agreement improves but remains slight using the incremented billing claim requirement (2 years: k = 0.16, 3 years: k = 0.16; 4 years: k = 0.15). Table 3 Unweighted Prevalence and Agreement of Chronic Conditions Using Administrative Data and Survey Data Prevalence (%) Cohen's k Sensitivity (%) Specificity (%) 2 years 3 years 4 years 2 years 3 years 4 years 2 years 3 years 4 years 2 years 2 years 4 years 1. 31.60 41.69 51.71 0.15 0.13 0.11 11.55 14.07 16.29 53.23 45.65 38.86 2. 23.27 31.59 39.50 0.16 0.16 0.15 9.28 11.74 13.94 59.28 53.43 47.72 1. Initial billing claim requirement 2. Incremented billing claim requirement. Note . Sensitivity and specificity are measured by comparing Ontario health administrative billing claims to survey data (i.e., the parent report). Based on the findings in Table 3, we determined that the incremented billing claim criteria were optimal, given that kappa values increased across all reference periods. Among the incremented billing claim criteria, the 3-year reference period was optimal. 1) The kappa value was consistent between 2- and 3-years. As previously discussed, it is advised to use a longer look-back or reference period to avoid missing cases [22]. 2) However, in comparison, we selected 3-years over 4-years given that the estimated prevalence of the 3-year reference period fell within the prevalence range for pediatric chronic conditions. Agreement Between Administrative and Survey Data: Specific Conditions The agreement between administrative and survey data for identifying a child as having a CHC was poor. Thus, we examined four conditions we expected would have strong agreement to help clarify the discrepancy. The selected diagnoses were conditions that we expected to be readily apparent to a parent as a “chronic condition:” diabetes, heart condition or disease (henceforth referred to as heart conditions), epilepsy, and cerebral palsy. Sample sizes were adequate only for diabetes and heart conditions. Table 4 shows that diabetes and heart conditions demonstrate substantial (i.e., 0.61-0.80) and fair agreement (i.e., 0.21-0.40), respectively. For details on the diagnostic codes matched with survey codes refer to Table C.1 in Supplemental Material C. Table 4 Unweighted Agreement on Survey Data and Administrative Data using Specific Diagnostic Codes, 3 years Prevalence Estimates Agreement Estimates Administrative (%) Survey (%) Cohen's k Sensitivity (%) Specificity (%) Diabetes 0.5% 0.3% 0.79 0.32% 99.51% Heart Conditions 1.1% 1.1% 0.21 0.23% 98.06% Multinomial Logistic Regression Model Indices The average relative variance increase (RVI) was 0.05. The largest fraction of missing information (FMI) was 0.20. The model was statistically significant ( F (33, 104517.9) = 5.16, p < .001). The relative risk ratios ( RRR ) and confidence intervals (CI) are presented in Table 5. Administrative CHC Younger age was significantly associated with having an administrative CHC ( RRR = 0.41, 95% CI [0.325, 0.517]) compared to no CHC. Adolescents (ages 12-17) showed a 59% relative risk reduction compared to children (ages 4-11). Parent structure was also significant ( RRR = 0.78, 95% CI [0.608, 0.999]). Children of single parents showed a 22% relative risk reduction of having an administrative CHC compared to two-parent households. (N.B. Parent structure results should be interpreted cautiously due to the 95% CI upper bound approaching 1.00). Child sex, PMK education status, income strata and HUI were not statistically significant. Survey CHC Older age was significantly associated with having a survey-reported CHC ( RRR = 1.44, 95% CI [1.148, 1.805]) compared to no CHC. Adolescents (ages 12-17) were 44% more likely to have a survey-reported CHC compared to children (ages 4-11). Health problems, measured by HUI, were significantly associated with whether a child had a survey-reported CHC compared to no CHC (see Table 5). Severe health problems were found to be the strongest predictor of survey-reported CHC ( RRR = 2.99, 95% CI [1.849, 4.827]). Child sex, PMK education status, parent structure, and income strata were not statistically significant. Both Administrative and Survey CHC Younger age was significantly associated with having an administrative and survey-reported CHC ( RRR = 0.65, 95% CI [0.503, 0.847]) compared to no CHC. Adolescents (ages 12-17) showed a 35% relative risk reduction of having an administrative and survey-reported CHC compared to children (ages 4-11). Severe health problems were found to be the strongest predictor of survey-reported CHC ( RRR = 1.97, 95% CI [1.171, 3.305]). Children with severe health problems were 97% more likely to have an administrative and survey-reported CHC compared to children with no health problems. Child sex, PMK education status, parent structure, and income strata were not statistically significant. Table 5 Multinomial Logistic Regression Examining Predictors of Chronic Condition Status using the Incremented Billing Claim Requirements, 3-year reference period Administrative vs. No CHC Survey-Reported vs. No CHC Administrative & Survey-Reported vs. No CHC RRR [ 95% CI] RRR [ 95% CI] RRR [ 95% CI] Child Characteristics Age a : 12-17 0.41 [0.325, 0.517] 1.44 [1.148, 1.805] 0.65 [0.503, 0.847] Sex b : Male 0.99 [0.814, 1.214] 1.02 [0.817, 1.274] 1.28 [0.988, 1.647] Parent and Family Characteristics PMK Education Status c : College/CEGEP; University certificate below BA 0.79 [0.542, 1.164] 0.99 [0.663, 1.474] 1.46 [0.873, 2.443] Bachelor’s degree (BA, BSc) 0.95 [0.644, 1.389] 0.96 [0.621, 1.487] 1.28 [0.763, 2.160] University certificate degree above Bachelors 0.76 [0.492, 1.161] 0.92 [0.560, 1.504] 1.30 [0.736, 2.284] Parent Structure d : Single Parent 0.78 [0.608, 0.999] 0.94 [0.704, 1.258] 0.92 [0.657, 1.276] Income strata e : Medium (20 th -80 th percentile) 0.87 [0.695, 1.082] 1.11 [0.861, 1.439] 0.86 [0.632, 1.170] High (High: >80 th percentile) 0.90 [0.705, 1.143] 1.26 [0.955, 1.659] 1.09 [0.784, 1.507] Child Clinical Characteristics Health Utility Index f : Mild problems 0.95 [0.730, 1.242] 1.75 [1.320, 2.318] 1.53 [1.097, 2.128] Moderate problems 0.97 [0.742, 1.280] 1.65 [1.252, 2.174] 1.86 [1.330, 2.592] Severe problems 0.80 [0.442, 1.449] 2.99 [1.849, 4.827] 1.97 [1.171, 3.305] Significant RRRs ( p < .05) are bolded. RRR = relative risk ratio, CI = confidence interval a. Age – “4-11” is the reference group. b. Sex – “Female” is the reference group. c. PMK Education Status – “Trade certificate or High school or less” is the reference group. d. Parent Structure – “Two parents” is the reference group. e. Income strata – “Low (≤ 20 th percentile)” is the reference group. f. Health Utility Index – “No problems” is the reference group. [1] The dependent variable was not imputed, resulting in a total sample of 8,985 for the regression. Discussion We developed an algorithm to categorize children and youth aged 4 to 17 as having a CHC using health administrative data consisting of ICD-9 diagnosis codes and the number of visits within a specified reference period. First, we found differences in chronic condition prevalence and agreement between survey and administrative data. Prevalence estimates were consistently higher in administrative data. Other studies on physical and mental health have also found that prevalence estimates are higher in administrative data than in survey data [66, 67]. It may be that parents underestimate what constitutes a long-term condition. Importantly, the survey asked parents whether a health professional diagnosed a long-term condition but did not provide an operational definition, such as a condition expected to last 3 months or more. There are different definitions of chronic conditions, ranging from ≥ 3 months to ≥ 1 year [6]. Future surveys could include a reference period – for example, “Within the past year, has your child suffered from…” – when asking about long-term conditions. We examined diabetes and heart conditions as specific conditions we expected parents could readily and easily identify as chronic conditions. Diabetes had a substantial agreement between administrative and survey data. However, it is worth noting the small sample size, with only 31 and 42 participants having diabetes using survey and administrative data, respectively (N.B. The minimum unweighted criteria to disclose is 30). Heart conditions had fair agreement between administrative data and survey data relative to diabetes. Interestingly, both administrative and survey data identified the same prevalence estimate for heart conditions despite having fair agreement. Studies examining agreement between multiple conditions have also found considerable variability depending on the type of condition [66, 68]. Correlates of CHC identification. Age significantly influenced whether administrative, survey, or both capture a CHC compared to no CHC. Administrative data was less likely to capture CHC in older children compared to younger children. The algorithm may not be appropriate for use with older children. However, it may also be that younger children are more likely to have frequent medical visits (e.g., well-baby or well-child care) due to heightened attention to developmental milestones and early detection of potential health issues [69]. Alternatively, older children were more likely to have a survey-reported CHC. The algorithm covered up to three years for conditions, while parents reported any lifetime chronic condition. It may be that the algorithm missed conditions diagnosed in early childhood that no longer require frequent follow-up visits as the child ages. Except for parent structure, all parent and family characteristics did not relate to whether a child had an administrative CHC or administrative/survey CHC compared to no CHC. This suggests that the healthcare billing system effectively captures CHC uniformly across different demographics, such as income, child sex, and PMK education. Finally, health problems were associated with survey CHC and administrative/survey CHC. Previous research indicates that self-reported chronic conditions are associated with higher morbidity levels, also measured by HUI [66]. While the 2014 OCHS is parent-reported, these findings align with the notion that survey data may capture more severe health conditions. Strengths This study had several strengths. First, we approached algorithm development as a multistep process by determining a list of conditions based on previous literature and applying a unique algorithm to each condition. Second, our algorithm includes more possible diagnostic codes than previous studies [21, 22, 66]. Thus, it was more likely to capture less prevalent conditions that may have been overlooked in other studies. This may have implications for studies comparing cases with a CHC vs healthy controls. Unless health controls are categorized based on the absence of all possible CHCs, some children with a CHC may have been included in healthy control groups. Third, we examined a few specific conditions to understand the poor agreement. We found that even among conditions that we may expect to be readily apparent as chronic, there is still considerable disagreement depending on the type of condition. The study provides a basis for improving data collection methods by identifying specific conditions with reporting discrepancies, which can help inform the development of better tools and protocols for reporting and recording health conditions. Limitations The current study has limitations. First, we did not have access to the response options to the “Other” item from parents that specified the specific conditions. If text data had been coded, researchers might have excluded conditions parents consider “chronic conditions.” The “Other” category includes 3.5% of all children and youth with a CHC, which may be a reason for poor agreement. Second, physicians’ ratings of whether or not a child had a chronic condition were not available; such ratings could be useful in comparing both the algorithm coding and parent reports. Nevertheless, achieving consensus in terms of what is considered a “chronic” condition across physicians may be difficult [70], especially across specialties. A review examining the miscoding of diagnostic codes in medical records found that 16% of clinical cases required a diagnosis-related group change (i.e., the grouping of medical conditions into a certain category) [70]. Furthermore, a review examining definitions of chronic conditions used in research found that several characteristics influenced the classification of chronicity (e.g., duration or latency, need for medical attention, departure from well-being, noncontagious nature, multiple risk factors, pathology, and nonamenability to cure) [6]. Third, diagnostic codes were inputted by a coding specialist. Coders may vary in how they code the same condition. For example, while one coder may bill a visit using “Bronchitis NOS” (code = 490), another may bill for “Chronic Bronchitis” (code = 491). It was beyond the scope of this study to examine the ambiguity in coding between specific conditions, but this is a potential area for future research. Fourth, the cell count was very small for many conditions, limiting our analyses and raising the possibility of spurious findings. Finally, this study excluded 316 children with developmental delays and intellectual disabilities, which may result in an underestimation of the prevalence of CHCs. A systematic review found that the prevalence of CHC in children with intellectual disability is much higher than in the general population [71]. Future research could consider limiting the number of conditions examined, comparing algorithms for specific conditions with disease-specific registries, and examining problem lists in electronic medical records or physician-reported diagnoses as methods to validate future algorithms. Conclusion We aimed to create an algorithm to identify all potential CHC in children and youth aged 4 to 17 and determine which sociodemographic factors are linked. We found considerable discrepancies between administrative and survey-reported data. The results highlight the importance of using algorithms developed from multiple datasets to examine complex research questions, such as the measurement of chronicity. Abbreviations CCC: Chronic Condition Classification CCHS: Canadian Community Health Survey CCW: Chronic Condition Warehouse CHC: chronic health condition CRG: Clinical Risk Groups HUI: Health utility index ICD: International Classification of Disease ICES Institute for Clinical Evaluative Sciences MH: mental health MICE: Multiple imputation by chained equation MOH: Ministry of Health NLSCY: National Longitudinal Survey of Children and Youth OCHS: Ontario Child Health Study OHIP: Ontario health insurance plan PMK: Parent most knowledgeable RECORD: REporting of studies Conducted using Observational Routinely-collected health Data Declarations Ethics approval and consent to participate: The 2014 OCHS was a voluntary survey conducted under the Statistics Act, which provides respondents guarantees of their privacy and confidentially. The study was approved by the Hamilton Integrated Research Ethics Board at McMaster University (no. 13-140). Consent for publication: Not applicable. Availability of data and materials : Data access to the 2014 Ontario Child Health Study (OCHS) is available through Statistics Canada. All physician billing data (i.e., OHIP data) and the linked OCHS-OHIP dataset for the current analyses are stored at McMaster University. Access to the linked dataset cannot be shared due to data sharing agreements with Statistics Canada and the Ontario Ministry of Health (MOH). Parts of this material are based on data and information provided by the MOH and the former Ministry of Children and Youth Services (MCYS). The opinions, results and conclusions reported in this paper are those of the authors and do not necessarily reflect those of MOH. Competing Interests: The authors declare that they have no competing interests. Funding: Grace Golden was supported by a Canada Graduate Master’s Scholarship – Social Sciences and Humanities Research Council (SSHRC) grant The principal study was supported by the Canadian Institutes of Health Research (CIHR #125941), the Ontario Ministry of Health and Long-Term Care (MOHLTC) - Health Services Research Grant (#8-42298), and funding from MOHLTC, the Ontario Ministry of Children and Youth Services, and the Ontario Ministry of Education. Graham Reid is supported by the Children’s Health Research Institute, London, ON. Author Contributions: GG: conceptualization, methodology, investigation, analysis, writing – reviewing and editing. LW: analysis – reviewing and editing. GJR: conceptualization, methodology, investigation, analysis, writing – reviewing and editing, supervision. Acknowledgements : We acknowledge the contribution of the following content experts we consulted to better understand long-term conditions for clinical relevance, validity, and robustness. They provided valuable feedback that significantly improved how we analyzed, interpreted, and communicated the data. Mark Ferro, Ph.D. School of Public Health Sciences, University of Waterloo, Waterloo, ON Shannon Reaume, Ph.D. Candidate, School of Public Health Sciences, University of Waterloo, Waterloo, ON Florence Perquier, Ph.D., Cundill Centre for Child and Youth Depression, Centre for Addiction and Mental Health, Toronto, ON; Department of Psychiatry, University of Toronto, Toronto, ON Peter Szatmari, MD, Cundill Centre for Child and Youth Depression, Centre for Addiction and Mental Health, Toronto, ON; The Hospital for Sick Children, Toronto, ON References Perrin, E. C., Newacheck, P., Pless, I. B., Drotar, D., Gortmaker, S. L., Leventhal, J., Perrin, J. M., Stein, R. E., Walker, D. K., & Weitzman, M. (1993). Issues involved in the definition and classification of chronic health conditions. Pediatrics , 91 (4), 787–793. McDougall, J., King, G., De Wit, D. J., Miller, L. T., Hong, S., Offord, D. R., Laporta, J., & Meyer, K. (2004). Chronic physical health conditions and disability among Canadian school-aged children: A national profile. Disability and Rehabilitation , 26 (1), 35–45. https://doi.org/10.1080/09638280410001645076 Newacheck, P. W. (1991). Prevalence and Impact of Chronic Illness Among Adolescents. Archives of Pediatrics & Adolescent Medicine , 145 (12), 1367. https://doi.org/10.1001/archpedi.1991.02160120035015 Newacheck, P. W., & Stoddard, J. J. (1994). Prevalence and impact of multiple childhood chronic illnesses. The Journal of Pediatrics , 124 (1), 40–48. https://doi.org/10.1016/S0022-3476(94)70252-7 Newacheck, P. W., & Taylor, W. R. (1992). Childhood chronic illness: Prevalence, severity, and impact. American Journal of Public Health , 82 (3), 364–371. https://doi.org/10.2105/ajph.82.3.364 Goodman, R. A., Posner, S. F., Huang, E. S., Parekh, A. K., & Koh, H. K. (2013). Defining and measuring chronic conditions: Imperatives for research, policy, program, and practice. Preventing Chronic Disease , 10 , E66. https://doi.org/10.5888/pcd10.120239 Berete, F., Demarest, S., Charafeddine, R., De Ridder, K., Van Oyen, H., Van Hoof, W., Bruyère, O., & Van Der Heyden, J. (2023). Linking health survey data with health insurance data: Methodology, challenges, opportunities and recommendations for public health research. An experience from the HISlink project in Belgium. Archives of Public Health , 81 (1), 198. https://doi.org/10.1186/s13690-023-01213-0 Duncan, L., Georgiades, K., Wang, L., Edwards, J., & Comeau, J. (2022). Estimating prevalence of child and youth mental disorder and mental health-related service contacts: A comparison of survey data and linked administrative health data. Epidemiology and Psychiatric Sciences , 31 , e35. https://doi.org/10.1017/S204579602200018X Guttmann, A., Nakhla, M., Henderson, M., To, T., Daneman, D., Cauch-Dudek, K., Wang, X., Lam, K., & Hux, J. (2010). Validation of a health administrative data algorithm for assessing the epidemiology of diabetes in Canadian children. Pediatric Diabetes , 11 (2), 122–128. https://doi.org/10.1111/j.1399-5448.2009.00539.x Hoefgen, E. R., Andrews, A. L., Richardson, T., Hall, M., Neff, J. M., Macy, M. L., Bettenhausen, J. L., Shah, S. S., & Auger, K. A. (2017). Health Care Expenditures and Utilization for Children With Noncomplex Chronic Disease. Pediatrics , 140 (3), e20170492. https://doi.org/10.1542/peds.2017-0492 To, T., Dell, S., Dick, P. T., Cicutto, L., Harris, J. K., MacLusky, I. B., & Tassoudji, M. (2006). Case verification of children with asthma in Ontario. Pediatric Allergy and Immunology , 17 (1), 69–76. https://doi.org/10.1111/j.1399-3038.2005.00346.x Peng, M., Lee, S., D’Souza, A. G., Doktorchik, C. T. A., & Quan, H. (2020). Development and validation of data quality rules in administrative health data using association rule mining. BMC Medical Informatics and Decision Making , 20 (1), 75. https://doi.org/10.1186/s12911-020-1089-0 Bhandari, A., & Wagner, T. (2006). Self-Reported Utilization of Health Care Services: Improving Measurement and Accuracy. Medical Care Research and Review , 63 (2), 217–235. https://doi.org/10.1177/1077558705285298 Smith, B., Chu, L. K., Smith, T. C., Amoroso, P. J., Boyko, E. J., Hooper, T. I., Gackstetter, G. D., & Ryan, M. A. (2008). Challenges of self-reported medical conditions and electronic medical records among members of a large military cohort. BMC Medical Research Methodology , 8 (1), 37. https://doi.org/10.1186/1471-2288-8-37 Clark, R. E., Ricketts, S. K., & McHugo, G. J. (1996). Measuring hospital use without claims: A comparison of patient and provider reports. Health Services Research , 31 (2), 153–169. Horwitz, R. I., & Yu, E. C. (1984). Assessing the reliability of epidemiologic data obtained from medical records. Journal of Chronic Diseases , 37 (11), 825–831. https://doi.org/10.1016/0021-9681(84)90015-8 Kopec, J. A. (2022). Estimating Disease Prevalence in Administrative Data. Clinical and Investigative Medicine , 45 (2), E21-27. https://doi.org/10.25011/cim.v45i2.38100 Singer, A., Yakubovich, S., Kroeker, A. L., Dufault, B., Duarte, R., & Katz, A. (2016). Data quality of electronic medical records in Manitoba: Do problem lists accurately reflect chronic disease billing diagnoses? Journal of the American Medical Informatics Association , 23 (6), 1107–1112. https://doi.org/10.1093/jamia/ocw013 Terry, A. L., Stewart, M., Cejic, S., Marshall, J. N., De Lusignan, S., Chesworth, B. M., Chevendra, V., Maddocks, H., Shadd, J., Burge, F., & Thind, A. (2019). A basic model for assessing primary health care electronic medical record data quality. BMC Medical Informatics and Decision Making , 19 (1), 30. https://doi.org/10.1186/s12911-019-0740-0 Gorina, Y., & Kramarow, E. A. (2011). Identifying chronic conditions in Medicare claims data: Evaluating the Chronic Condition Data Warehouse algorithm. Health Services Research , 46 (5), 1610–1627. https://doi.org/10.1111/j.1475-6773.2011.01277.x Chronic Conditions Warehouse. (2022). CODEBOOK: Master Beneficiary Summary File (MBSF) 27 CCW Chronic Conditions (CC) File . https://resdac.org/cms-data/files/mbsf-27-cc/data-documentation Voss, R. W., Schmidt, T. D., Weiskopf, N., Marino, M., Dorr, D. A., Huguet, N., Warren, N., Valenzuela, S., O’Malley, J., & Quiñones, A. R. (2022). Comparing ascertainment of chronic condition status with problem lists versus encounter diagnoses from electronic health records. Journal of the American Medical Informatics Association , 29 (5), 770–778. https://doi.org/10.1093/jamia/ocac016 Masters, C. L., Bateman, R., Blennow, K., Rowe, C. C., Sperling, R. A., & Cummings, J. L. (2015). Alzheimer’s disease. Nature Reviews Disease Primers , 1 (1), 15056. https://doi.org/10.1038/nrdp.2015.56 Salari, N., Ghasemi, H., Mohammadi, L., Behzadi, M. H., Rabieenia, E., Shohaimi, S., & Mohammadi, M. (2021). The global prevalence of osteoporosis in the world: A comprehensive systematic review and meta-analysis. Journal of Orthopaedic Surgery and Research , 16 (1), 609. https://doi.org/10.1186/s13018-021-02772-0 Symmons, D. P. M. (2002). Epidemiology of rheumatoid arthritis: Determinants of onset, persistence and outcome. Best Practice & Research Clinical Rheumatology , 16 (5), 707–722. https://doi.org/10.1053/berh.2002.0257 Gershon, A. S., Wang, C., Guan, J., Vasilevska-Ristovska, J., Cicutto, L., & To, T. (2009). Identifying patients with physician-diagnosed asthma in health administrative databases. Canadian Respiratory Journal , 16 (6), 183–188. https://doi.org/10.1155/2009/963098 Jaakkimainen, L., Duchen, R., Lix, L., Al-Azazi, S., Yu, B., Butt, D., Park, S.-B., & Widdifield,J. (2022). Identification of Early Onset Dementia in Population-Based HealthAdministrative Data: A Validation Study Using Primary Care Electronic Medical Records. Journal of Alzheimer’s Disease , 89 (4), 1463–1472. https://doi.org/10.3233/JAD-220384 Muggah, E., Graves, E., Bennett, C., & Manuel, D. G. (2012). The impact of multiple chronic diseases on ambulatory care use; a population based study in Ontario, Canada. BMC Health Services Research , 12 (1), 452. https://doi.org/10.1186/1472-6963-12-452 Shah, B. R., Booth, G. L., Feig, D. S., & Lipscombe, L. L. (2023). Validation of Algorithms to Identify Gestational Diabetes From Population-level Health-care Administrative Data. Canadian Journal of Diabetes , 47 (1), 25–30. https://doi.org/10.1016/j.jcjd.2022.06.010 Widdifield, J., Bombardier, C., Bernatsky, S., Paterson, J. M., Green, D., Young, J., Ivers, N., Butt, D. A., Jaakkimainen, R. L., Thorne, J. C., & Tu, K. (2014). An administrative data validation study of the accuracy of algorithms for identifying rheumatoid arthritis: The influence of the reference standard on algorithm performance. BMC Musculoskeletal Disorders , 15 (1), 216. https://doi.org/10.1186/1471-2474-15-216 Neff, J. M., Sharp, V. L., Muldoon, J., Graham, J., Popalisky, J., & Gay, J. C. (2002). Identifying and classifying children with chronic conditions using administrative data with the clinical risk group classification system. Ambulatory Pediatrics: The Official Journal of the Ambulatory Pediatric Association , 2 (1), 71–79. https://doi.org/10.1367/1539-4409(2002)0022.0.co;2 Feudtner, C., Feinstein, J. A., Zhong, W., Hall, M., & Dai, D. (2014). Pediatric complex chronic conditions classification system version 2: Updated for ICD-10 and complex medical technology dependence and transplantation. BMC Pediatrics , 14 , 199. https://doi.org/10.1186/1471-2431-14-199 Benchimol, E. I., Smeeth, L., Guttmann, A., Harron, K., Moher, D., Petersen, I., Sørensen, H. T., von Elm, E., Langan, S. M., & RECORD Working Committee. (2015). The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement. PLoS Medicine , 12 (10), e1001885. https://doi.org/10.1371/journal.pmed.1001885 Boyle, M. H., Georgiades, K., Duncan, L., Comeau, J., Wang, L., & 2014 Ontario Child Health Study Team. (2019). The 2014 Ontario Child Health Study—Methodology. The Canadian Journal of Psychiatry , 64 (4), 237–245. https://doi.org/10.1177/0706743719833675 Cadarette, S. M., & Wong, L. (2015). An Introduction to Health Care Administrative Data. The Canadian Journal of Hospital Pharmacy , 68 (3), 232–237. https://doi.org/10.4212/cjhp.v68i3.1457 Canadian Institute for Health Information. (2022). An overview of physicians payments and cost per service . https://www.cihi.ca/en/health-workforce-in-canada-in-focus-including-nurses-and-physicians/an-overview-of-physician Steele, L. S., Glazier, R. H., Lin, E., & Evans, M. (2004). Using Administrative Data to Measure Ambulatory Mental Health Service Provision in Primary Care: Medical Care , 42 (10), 960–965. https://doi.org/10.1097/00005650-200410000-00004 Hwee, J., Sung, L., Kwong, J. C., Sutradhar, R., Tu, K., & Pole, J. D. (2018). Use of physician billing claims to identify infections in children. PLoS ONE , 13 (11), e0207468. https://doi.org/10.1371/journal.pone.0207468 King-Dowling, S., Proudfoot, N. A., & Obeid, J. (2019). Comorbidity Among Chronic Physical Health Conditions and Neurodevelopmental Disorders in Childhood. Current Developmental Disorders Reports , 6 (4), 248–258. https://doi.org/10.1007/s40474-019-00173-x Offord, D. R., Boyle, M. H., Fleming, J. E., Blum, H. M., & Rae Grant, N. I. (1989). Summary of Selected Results. The Canadian Journal of Psychiatry , 34 (6), 483–491. https://doi.org/10.1177/070674378903400602 Statistics Canada NLSCY. (2006, 2007). National Longitudinal Survey of Children and Youth (NLSCY), Cycle 8. Statistics Canada . https://www23.statcan.gc.ca/imdb/p2SV.pl?Function=getSurvey&SDDS=4450 Statistics Canada SYC. (2012). Survey of Young Canadians (SYC) . https://www23.statcan.gc.ca/imdb/p2SV.pl?Function=getSurvey&SDDS=5179#a1 Statistics Canada CCHS. (2015). Canadian Community Health Survey (CCHS)—2013. Canadian Community Health Survey (CCHS) - 2013 . Canadian Community Health Survey (CCHS) – 2013 Horsman, J., Furlong, W., Feeny, D., & Torrance, G. (2003). The Health Utilities Index (HUI®): Concepts, measurement properties and applications. Health and Quality of Life Outcomes , 1 (1), 54. https://doi.org/10.1186/1477-7525-1-54 Feng, Y., Bernier, J., McIntosh, C., & Orpana, H. (2009). Validation of disability categories derived from Health Utilities Index Mark 3 scores. Health Reports , 20 (2), 43–50. Thorrington, D., & Eames, K. (2015). Measuring Health Utilities in Children and Adolescents: A Systematic Review of the Literature. PLOS ONE , 10 (8), e0135672. https://doi.org/10.1371/journal.pone.0135672 Cochrane, S., Beyer, K., Clausen, M., Wjst, M., Hiller, R., Nicoletti, C., Szepfalusi, Z., Savelkoul, H., Breiteneder, H., Manios, Y., Crittenden, R., & Burney, P. (2009). Factors influencing the incidence and prevalence of food allergy. Allergy , 64 (9), 1246–1255. https://doi.org/10.1111/j.1398-9995.2009.02128.x Garner, R., & Kohen, D. (2008). Changes in the prevalence of asthma among Canadian children. Health Reports , 19 (2), 45–50. Karpa, K. D., Paul, I. M., Leckie, J. A., Shung, S., Carkaci-Salli, N., Vrana, K. E., Mauger, D., Fausnight, T., & Poger, J. (2012). A retrospective chart review to identify perinatal factors associated with food allergies. Nutrition Journal , 11 (1), 87. https://doi.org/10.1186/1475-2891-11-87 Lawrence, J. M., Divers, J., Isom, S., Saydah, S., Imperatore, G., Pihoker, C., Marcovina, S. M., Mayer-Davis, E. J., Hamman, R. F., Dolan, L., Dabelea, D., Pettitt, D. J., Liese, A. D., & SEARCH for Diabetes in Youth Study Group. (2021). Trends in Prevalence of Type 1 and Type 2 Diabetes in Children and Adolescents in the US, 2001-2017. JAMA , 326 (8), 717. https://doi.org/10.1001/jama.2021.11165 Perret, J. L., Wurzel, D., Walters, E. H., Lowe, A. J., Lodge, C. J., Bui, D. S., Erbas, B., Bowatte, G., Russell, M. A., Thompson, B. R., Gurrin, L., Thomas, P. S., Hamilton, G., Hopper, J. L., Abramson, M. J., Chang, A. B., & Dharmage, S. C. (2022). Childhood “bronchitis” and respiratory outcomes in middle-age: A prospective cohort study from age 7 to 53 years. BMJ Open Respiratory Research , 9 (1), e001212. https://doi.org/10.1136/bmjresp-2022-001212 Shaw, T. E., Currie, G. P., Koudelka, C. W., & Simpson, E. L. (2011). Eczema Prevalence in the United States: Data from the 2003 National Survey of Children’s Health. Journal of Investigative Dermatology , 131 (1), 67–73. https://doi.org/10.1038/jid.2010.251 Silverberg, J. I., Braunstein, M., & Lee-Wong, M. (2015). Association between climate factors, pollen counts, and childhood hay fever prevalence in the United States. Journal of Allergy and Clinical Immunology , 135 (2), 463-469.e5. https://doi.org/10.1016/j.jaci.2014.08.003 Ahearne, C. E., Boylan, G. B., & Murray, D. M. (2016). Short and long term prognosis in perinatal asphyxia: An update. World Journal of Clinical Pediatrics , 5 (1), 67–74. https://doi.org/10.5409/wjcp.v5.i1.67 Bertrand, G., & Kaplan, C. (2014). How do we treat fetal and neonatal alloimmune thrombocytopenia? Transfusion , 54 (7), 1698–1703. https://doi.org/10.1111/trf.12671 Gupta, R., & Cabacungan, E. T. (2021). Neonatal Birth Trauma: Analysis of Yearly Trends, Risk Factors, and Outcomes. The Journal of Pediatrics , 238 , 174-180.e3. https://doi.org/10.1016/j.jpeds.2021.06.080 Stewart, A. L., & Reynolds, E. O. R. (1974). Improved Prognosis for Infants of Very Low Birthweight. Pediatrics , 54 (6), 724–735. https://doi.org/10.1542/peds.54.6.724 Rector, T. S., Wickstrom, S. L., Shah, M., Thomas Greeenlee, N., Rheault, P., Rogowski, J., Freedman, V., Adams, J., & Escarce, J. J. (2004). Specificity and sensitivity of claims-based algorithms for identifying members of Medicare+Choice health plans that have chronic medical conditions. Health Services Research , 39 (6 Pt 1), 1839–1857. https://doi.org/10.1111/j.1475-6773.2004.00321.x Steffler, M., Li, Y., Weir, S., Shaikh, S., Murtada, F., Wright, J. G., & Kantarevic, J. (2021). Trends in prevalence of chronic disease and multimorbidity in Ontario, Canada. Canadian Medical Association Journal , 193 (8), E270–E277. https://doi.org/10.1503/cmaj.201473 IBM Corp. (2023). IBM SPSS Statistics for Windows, Version 29.0.2.0 [Computer software]. StataCorp. (2023). Stata Statistical Software: Release 18 [Computer software]. StataCorp LLC. Cohen. (1960). A Coefficient of Agreement for Nominal Scales . 20 (1). https://doi.org/10.1177/001316446002000104 Trevethan, R. (2017). Sensitivity, Specificity, and Predictive Values: Foundations, Pliabilities, and Pitfalls in Research and Practice. Frontiers in Public Health , 5 , 307. https://doi.org/10.3389/fpubh.2017.00307 Johnson, D. R., & Young, R. (2011). Toward Best Practices in Analyzing Datasets with Missing Data: Comparisons and Recommendations. Journal of Marriage and Family , 73 (5), 926–945. https://doi.org/10.1111/j.1741-3737.2011.00861.x Golden, G. (2024). Mental Health Service Use among Children and Youth with Co-Occurring Health and Mental Health Disorders in Ontario. Electronic Thesis and Dissertation Repository . https://ir.lib.uwo.ca/etd/10280 Muggah, E., Graves, E., Bennett, C., & Manuel, D. G. (2013). Ascertainment of chronic diseases using population health data: A comparison of health administrative data and patient self-report. BMC Public Health , 13 (1), 16. https://doi.org/10.1186/1471-2458-13-16 O’Donnell, S., Vanderloo, S., McRae, L., Onysko, J., Patten, S. B., & Pelletier, L. (2016). Comparison of the estimated prevalence of mood and/or anxiety disorders in Canada between self-report and administrative data. Epidemiology and Psychiatric Sciences , 25 (4), 360–369. https://doi.org/10.1017/S2045796015000463 Jiang, L., Zhang, B., Smith, M. L., Lorden, A. L., Radcliff, T. A., Lorig, K., Howell, B. L.,Whitelaw, N., & Ory, M. G. (2015). Concordance between Self-Reports and Medicare Claims among Participants in a National Study of Chronic Disease Self-Management Program. Frontiers in Public Health , 3 , 222. https://doi.org/10.3389/fpubh.2015.00222 Canadian Pediatric Society. (2021, July). Schedule of well-child visits . https://caringforkids.cps.ca/handouts/pregnancy-and-babies/schedule_of_well_child_visits Cheng, P., Gilchrist, A., Robinson, K. M., & Paul, L. (2009). The Risk and Consequences of Clinical Miscoding Due to Inadequate Medical Documentation: A Case Study of the Impact on Health Services Funding. Health Information Management Journal , 38 (1), 35–46. https://doi.org/10.1177/183335830903800105 Oeseburg, B., Dijkstra, G. J., Groothoff, J. W., Reijneveld, S. A., & Jansen, D. E. M. C. (2011). Prevalence of Chronic Health Conditions in Children With Intellectual Disability: A Systematic Literature Review. Intellectual and Developmental Disabilities , 49 (2), 59–85. https://doi.org/10.1352/1934-9556-49.2.59 Additional Declarations No competing interests reported. Supplementary Files SupplementalMaterial2024.docx Cite Share Download PDF Status: Published Journal Publication published 02 Oct, 2025 Read the published version in BMC Pediatrics → Version 1 posted Editorial decision: Revision requested 30 Sep, 2024 Submission checks completed at journal 27 Sep, 2024 First submitted to journal 27 Sep, 2024 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-5089891","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":360435795,"identity":"28e7eaa5-1d5f-4230-8bb9-9153ecd6b011","order_by":0,"name":"Grace Golden","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAqklEQVRIiWNgGAWjYHACxgM8DAxypOkBaTEmXUtiA9HK5f0XHzjwpuJO+oYbyQcYftQQocXwxrOEg3POPMvdcCMtgbHnGDFaZpwxOMzbdjh3w+0cA2YGNqK1/DucbgDW8o8ILfL8PUAtDYcTwFoY24jQYiDBBvTLscOGM+8DPdXbR4wt/YcPPnhTc1ie7wyQ8eMbMbbcSEBwDhChAWQLcepGwSgYBaNgJAMA5DRARnN9tlAAAAAASUVORK5CYII=","orcid":"","institution":"The University of Western Ontario","correspondingAuthor":true,"prefix":"","firstName":"Grace","middleName":"","lastName":"Golden","suffix":""},{"id":360435796,"identity":"5be83c13-840f-4ec4-99bc-6bc4cfc44cec","order_by":1,"name":"Li Wang","email":"","orcid":"","institution":"McMaster University","correspondingAuthor":false,"prefix":"","firstName":"Li","middleName":"","lastName":"Wang","suffix":""},{"id":360435797,"identity":"8846c7af-5d37-4f2f-b578-06f1ef2b6ede","order_by":2,"name":"Graham J. Reid","email":"","orcid":"","institution":"The University of Western Ontario","correspondingAuthor":false,"prefix":"","firstName":"Graham","middleName":"J.","lastName":"Reid","suffix":""}],"badges":[],"createdAt":"2024-09-14 15:51:31","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5089891/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5089891/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12887-025-05747-w","type":"published","date":"2025-10-02T15:58:09+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":80201190,"identity":"c4ad5db5-978e-4c5c-bd1f-9bf6382eb581","added_by":"auto","created_at":"2025-04-09 06:49:49","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":107582,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eAdministrative Data Reference Period\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eNote. \u003c/em\u003eFigure 1 demonstrates an example reference period for a participant with interview data in May 2015. Thus, this parent reported whether the child had any long-term conditions up until May 2015. Administrative data was examined for a reference period up to a maximum of 4 years before the date of the May 2015 interview.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-5089891/v1/365a3ff66e146d3d6886cb9c.png"},{"id":80202313,"identity":"b40a3d12-315e-417f-b38d-bc62959b18a6","added_by":"auto","created_at":"2025-04-09 06:57:49","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":86287,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eFlow Chart Showing Children’s Inclusion in Data Analysis\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e1. Criteria applied sequentially in the order shown. 2. Missing indicates 1) no response, 2) not stated, and 3) don’t know.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-5089891/v1/40ecf2dce834a51e120cb62f.png"},{"id":80201188,"identity":"32d240c9-8967-4b59-a21f-63a53f35e99d","added_by":"auto","created_at":"2025-04-09 06:49:49","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":71692,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eAlgorithm Development: Selection of Diagnostic Codes, Billing Claim Requirements, and Reference Periods\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eCCC System: pediatric complex Chronic Condition Classification (CCC) system version 2 [32]. \u003cem\u003eNote\u003c/em\u003e. Figure 3 displays the process for selecting diagnostic codes. After determining the list of codes, each condition was specified with a billing claim requirement and reference period.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-5089891/v1/85bd410aa1df86a99a6d0a46.png"},{"id":92884470,"identity":"1cb077b9-70de-4f24-823b-5160f76e1f63","added_by":"auto","created_at":"2025-10-06 16:13:12","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1654381,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5089891/v1/740f016e-b311-41b1-8939-000fcc9feaba.pdf"},{"id":80201191,"identity":"fcd9df74-a7a6-4041-ae29-1fa3b76cb227","added_by":"auto","created_at":"2025-04-09 06:49:50","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":110909,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementalMaterial2024.docx","url":"https://assets-eu.researchsquare.com/files/rs-5089891/v1/07afe3e641522482c604af4b.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Child and Youth Chronic Physical Health Conditions: A Comparison of Survey Data and Linked Administrative Health Data in Ontario","fulltext":[{"header":"Introduction","content":"\u003cp\u003ePerrin et al. [1] defined a chronic health condition (CHC) as a long-term condition with a biological basis that has lasted, or is expected to last, for 3 months or longer. Population-based studies in Canada and the United States estimate chronic physical health conditions affect between 20% to 30% of children aged 0 to 17 [2-5]. One challenge in measuring chronic conditions is that researchers often use inconsistent definitions. For example, the 3-month timeframe is often not used, with many researchers using 6- or 12-month timeframes [6].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRecent research advises that using two data sources, when possible (e.g., through data linkage studies), may help to answer complex research questions [7]. One source of data is population-based surveys. Survey data offers detailed data on randomly sampled populations, including extensive sociodemographic information [8]. However, survey data may be subject to bias (e.g., recall bias) [7].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAnother valuable data source is administrative data. Many studies have used administrative health billing data to answer important questions about children and youth with CHC. For example, administrative data have been used to determine incidence and prevalence, identify risk factors associated with developing certain conditions (e.g., diabetes), and identify healthcare expenditures and use patterns associated with specific chronic conditions [9-11]. Among each of these studies, the first step required identifying children with a specific chronic condition using an algorithm. However, these studies only examined one or a few chronic conditions. \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAdministrative data are often used for disease surveillance or health service research, despite the initial collection not being intended for this purpose [12]. There are several advantages of administrative data. The data are cost-effective and often readily available [13], \u0026nbsp;encompass large populations, and capture many types of health events (e.g., office visits, emergency department use) [12, 14]. There are concerns and cautions with these data. One issue is data quality -- \u0026nbsp; accuracy and completeness [12, 15, 16, 17]. A review examining electronic medical record data quality for seven chronic diseases found that completeness of records was associated with the type of provider, the patient load, and the clinic or organization’s funding model (e.g., fee-for-service, salary-based) [18]. A second issue is that due to restrictions on the amount of information that can be recorded, some aspects of the visit details may be incomplete [19]. For example, a family doctor may only code a single diagnosis for a visit despite addressing multiple problems. Third, there are unique challenges related to identifying children and youth with chronic conditions based on health administrative data. For example, a physician may assess a patient and submit a billing claim with a diagnostic code indicating an acute or chronic condition. Whether or not that child/youth has a chronic condition depends on various factors. Some conditions may resolve on their own. A chronic condition code may be part of an evaluation, which later reveals a negative result. Thus, for some conditions, it is only when the same diagnostic code is assigned again, can we infer that the condition is chronic.\u003c/p\u003e\n\u003cp id=\"_Toc174959013\"\u003e\u003cstrong\u003eExisting Algorithms\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThere are three main elements that must be incorporated from administrative data to determine whether a child has a CHC: (1) the diagnosis recorded for the visit, (2) the number of visits, and (3) within a specific reference period. The reference period refers to the number of years during which the number of visits must be met to identify a chronic condition [20]. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOne well-established algorithm was developed by the United States Department of Health and Human Services Chronic Conditions Warehouse (CCW) to identify chronic conditions amongst adults with the greatest impact on population health using administrative claims (CCW) [21]. The algorithm uses the International Classification of Diseases (ICD)-9 and ICD-10 codes for common adult conditions (e.g., diabetes, hypertension, breast cancer). To meet the criteria for a chronic condition, a patient must have one to two outpatient billing claims within a specific reference period of 1 to 3 years [22]. The reference period varies depending on the type of condition (e.g., 1 year for breast cancer vs 3 years for Alzheimer’s) [21]. One criticism of this algorithm concerns the reference period. Shorter periods may not identify patients with more irregular access to health care or individuals with well-managed conditions who rarely have follow-up appointments [22]. Nonetheless, the algorithm has been validated and demonstrated more accurate prevalence estimates of chronic conditions over time compared to self-report [22].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAnother criticism of the CCW algorithm is that the algorithm only provides reference periods for 27 chronic conditions; further, there are a number of conditions that are rare or non-existent among children or youth [21]. For example, CCW conditions include Alzheimer’s disease, osteoporosis, and rheumatoid arthritis, all of which are rare in children [23, 24, 25]. Other algorithms for chronic conditions have been used, but they also have limitations. Within the Ontario context, algorithms have been examined for a number of specific conditions (e.g., asthma, dementia, diabetes, rheumatoid arthritis) using Institute for Clinical Evaluative Sciences (ICES) datasets, but similarly, these studies have only been with adults [26-30]. The Clinical Risk Groups (CRG) system has been used to assess children, but the algorithm requires a minimum of 2 claims during a 12-month period [31], which, again, has been criticized as too short of a reference period [22]. Nevertheless, the CRG system was designed to assign a severity level rather than solely to determine the presence or absence of a chronic condition. \u0026nbsp; Another widely used algorithm is the pediatric complex Chronic Condition Classification (CCC) system version 2 [32], which provides a comprehensive list of ICD-9 and ICD-10 diagnostic codes to identify a patient with CCC category (e.g., gastrointestinal, respiratory, and cardiovascular). \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe current study.\u003c/strong\u003e No study has attempted to capture chronic health conditions across a wide age range during childhood and adolescence using a standardized algorithm that can be applied to the general population. Additionally, this study examined which sociodemographic variables may influence whether a child is identified as having a CHC.\u0026nbsp;\u003c/p\u003e\n\u003cp id=\"_Toc174959014\"\u003e\u003cstrong\u003eObjectives\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eObjective 1) Develop an algorithm to identify children and youth ages 4 to 17 with chronic health conditions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eObjective 2) Compare the relationship between parent-reported vs algorithm-based presence of a chronic health condition.\u003c/p\u003e\n\u003cp\u003eObjective 3) Identify sociodemographic variables that relate to the presence or absence of a CHC as identified only by an algorithm applied to administrative data, parent-reported data, or both. \u0026nbsp; \u0026nbsp;\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eThe reporting of our study follows the REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) Statement [33]. See Supplemental Material A, Table A.1 for the checklist.\u0026nbsp;\u003c/p\u003e\n\u003cp id=\"_Toc174959016\"\u003e\u003cstrong\u003eData\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData were from the cross-sectional 2014 Ontario Child Health Study (OCHS-2014) [34], linked with Ontario Ministry of Health (MOH; formerly MOHLTC) administrative data. We describe the survey first, as survey participants formed the basis for the study, and then the health administrative data. Finally, the methods used to develop the algorithm are outlined.\u003c/p\u003e\n\u003cp id=\"_Toc174959017\"\u003e\u003cstrong\u003eSurvey Data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe methodology for the 2014 OCHS has been reported elsewhere [34]. Briefly, Statistics Canada invited over 12,000 eligible households across Ontario of children and youth between the ages of 4 and 17 and their respective families to participate in the 2014 study [34]. Interviews took place between October 2014 to September 2015. The sampling framework was a clustered, stratified, random sample of households from the 2014 Canada Child Tax Benefit [34]. In total, 10,802 children and youth were in the study. Of those, 9,666 parents and/or youth agreed to link their survey data with administrative health data, and 9,301 were able to be linked. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eYouth and their parents completed at-home interviews. In families with more than one child, a \u0026lsquo;target child\u0026rsquo; was randomly selected for detailed assessments, while data from other children (i.e., siblings) in the family were less extensive. The person most knowledgeable (PMK) completed self-report questionnaires and questionnaires about the target child and siblings. Youth aged 12 to 17 provided their own reports using self-report questionnaires on laptops. (N.B. Parent-reported data will henceforth be referred to as survey data).\u003c/p\u003e\n\u003cp id=\"_Toc174959018\"\u003e\u003cstrong\u003eAdministrative Data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOntario administrative health data are available for services provided in hospitals (e.g., emergency room visits) or outpatient services provided by physicians (e.g., walk-in clinic visits) and covered by the Ontario Health Insurance Plan (OHIP) [35]. Family physicians in Ontario may be paid via alternative payment plans (e.g. blended capitation payment models); data are still obtained via shadow billing [36]. Specialist physicians bill directly by submitting claims that include a diagnostic and a service fee code [37]. The diagnostic code in OHIP is a truncated version (3 digits) of the International Classification of Diseases (ICD) [38]. All analyses were performed on outpatient billing claims using ICD-9 codes. ICD-10 codes were not implemented until October 2015, after the study window. (N.B. ICD-9 diagnostic codes will henceforth be referred to as diagnostic codes. Diagnoses associated with each code will henceforth be referred to as conditions).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eData were obtained for the reference period up to a maximum of 4 years before the date of the 2014 interview (see Figure 1). Reference periods were unique for each participant and based on their interview data.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCurrent Study and Exclusion Criteria\u003c/strong\u003e\u003cstrong\u003e.\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study had two main exclusion criteria. 1) Participants were excluded from the study if they had i) a parent-reported developmental delay or intellectual disability in 2014 OCHS or ii) an administrative billing claim with the diagnosis code of intellectual disability (ICD-9 codes 317-319). This paper is part of a larger study, where the overarching aim is to examine how children with comorbid chronic health and mental health conditions access mental health services. Children with developmental delays and intellectual disabilities often require cross-specialty services, thereby making it unclear the primary purpose of a healthcare visit [39]. 2) Participants were excluded if they had missing data related to the primary outcome, which was long-term conditions reported by parents (see Figure 2 Sample Selection).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMeasures\u003c/strong\u003e\u003c/p\u003e\n\u003cp id=\"_Toc174959021\"\u003e\u003cstrong\u003eSurvey Data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe PMK reported whether a health professional diagnosed any of 11 long-term conditions. Specifically, they were asked, \u0026ldquo;Has a health professional diagnosed any of the following long-term conditions for the child?\u0026rdquo; The conditions included 1) food and digestive allergies, 2) respiratory allergies, 3) other allergies, 4) bronchitis, 5) diabetes, 6) heart condition or disease, 7) epilepsy, 8) cerebral palsy, 9) kidney condition or disease, 10) asthma, and 11) eczema. Importantly, PMK was also provided with an \u0026ldquo;Other\u0026rdquo; option to report \u0026ldquo;any other long-term condition\u0026rdquo; and asked to specify the condition in text form. Unfortunately, the text data for specific conditions reported by parents are unavailable and not coded by Statistics Canada. Long-term conditions were based on the National Longitudinal Survey of Children and Youth (NLSCY) cycle 8, the Survey of Young Canadians (SYC), and the 1983 OCHS [40, 41, 42]. The validity of the items, as used in the NLSY or the 1983 or 2014 OCHS, has not been examined.\u0026nbsp;\u003c/p\u003e\n\u003cp id=\"_Toc174959022\"\u003e\u003cstrong\u003ePredictor Variables\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTable 1 presents the 2014 OCHS items, questions asked to the PMK/Partner/Youth, response options, and how variables were derived.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eChild Characteristics.\u003c/strong\u003e Age was grouped as child (age 4-11) or youth (age 12-17), and sex was either male or female.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eParent and Family Characteristics.\u003c/strong\u003e PMKs reported their highest level of education and whether the child lives in a two- or single-parent household. Household income strata were based on items from the 2013 Canadian Community Health Survey (CCHS) [43]. Strata were coded as low = \u0026lt;20\u003csup\u003eth\u003c/sup\u003e percentile, medium = 20\u003csup\u003eth\u003c/sup\u003e - \u0026lt;80\u003csup\u003eth\u003c/sup\u003e percentile, high = \u0026gt; 80\u003csup\u003eth\u003c/sup\u003e percentile\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eChild Clinical Characteristics.\u003c/strong\u003e \u003cem\u003eHealth Utility Index (HUI).\u003c/em\u003e PMK reported the health problems for the child using the Health Utilities Index (HUI) [44]. The HUI describes an individual\u0026rsquo;s functional health status based on eight attributes (vision, speech, hearing, ambulation, dexterity, emotion, pain, and cognition). The HUI uses a coding algorithm [44] to generate a single-attribute utility score ranging from -1.00 (severe disability) to 1.00 (full health). Statistics Canada computed scores for each participant. The HUI has a skewed distribution, with most participants scoring 1.00 (full health). Categories were derived from Statistics Canada scores using data from TC and siblings [45]. HUI was coded as 0 = no problems (1.00), 1 = mild problems (0.90-\u0026lt;1.00), 2 = moderate problems (0.68-\u0026lt;0.90) and 3\u003cem\u003e=\u003c/em\u003e severe problems (\u0026lt;0.68; top 5\u003csup\u003eth\u003c/sup\u003e percentile), the HUI has been widely used to understand health status [44, 46].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1\u0026nbsp;\u003c/strong\u003e\u003cem\u003ePredictors of Chronic Condition Status\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"852\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eVariable (2014 OCHS item)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDescription or Question asked to PMK/Youth\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\" style=\"width: 288px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eResponse Options/Code\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRecode\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"4\" valign=\"top\" style=\"width: 366px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eChild Characteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 131px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 204px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003eAge (MEM_AGE)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003eBased on demographics.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\" style=\"width: 288px;\"\u003e\n \u003cp\u003eContinuous 4-17 \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003e0 = 4-11\u003c/p\u003e\n \u003cp\u003e1 = 12-17\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003eSex (MEM_SEX)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003eBased on demographics.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\" style=\"width: 288px;\"\u003e\n \u003cp\u003eMale\u003c/p\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003e0 = Male\u003c/p\u003e\n \u003cp\u003e1 = Female\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"3\" valign=\"top\" style=\"width: 360px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eParent and Family Characteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\" style=\"width: 288px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003ePMK Educational status (EHGP_04)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003eWhat is the highest certificate, diploma or degree that you have completed?\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\" style=\"width: 288px;\"\u003e\n \u003cp\u003eLess than high school or equivalent\u003c/p\u003e\n \u003cp\u003eHigh school diploma or equivalency certificate\u003c/p\u003e\n \u003cp\u003eTrade certificate or diploma\u003c/p\u003e\n \u003cp\u003eCollege CEGEP or other non-university certificate or diploma or below bachelor\u0026rsquo;s level\u003c/p\u003e\n \u003cp\u003eBachelor\u0026rsquo;s degree\u003c/p\u003e\n \u003cp\u003eUniversity certificate diploma or degree above bachelor\u0026rsquo;s level\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003e0 = Trade certificate or high school or less\u003c/p\u003e\n \u003cp\u003e1 = College/CEGEP or below bachelor\u0026rsquo;s level\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e2 = Bachelor\u0026rsquo;s degree (BA, BSc)\u003c/p\u003e\n \u003cp\u003e3 = University certificate diploma or degree above bachelor\u0026rsquo;s level\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003eParent structure (DEMDV04)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003eBased on demographics.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\" style=\"width: 288px;\"\u003e\n \u003cp\u003eTwo parents\u003c/p\u003e\n \u003cp\u003eSingle parent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003e0 = Two parent household\u003c/p\u003e\n \u003cp\u003e1 = Single parent household\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003eHousehold income strata (PSTRATAH)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003eIncome strata were created after data collection.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\" style=\"width: 288px;\"\u003e\n \u003cp\u003eLow\u0026nbsp;\u003cbr\u003e\u0026nbsp;Medium\u003cbr\u003e\u0026nbsp;High\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003e0 = Low\u003cbr\u003e\u0026nbsp;1 = Medium\u003cbr\u003e\u0026nbsp;2 = High\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 224px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003cstrong\u003eChild Clinical Characteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"5\" style=\"width: 628px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 156px;\"\u003e\n \u003cp\u003eHealth Utility Index (HUIDHSI)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003eHUI uses a coding algorithm to generate a single-attribute utility score.\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\" style=\"width: 288px;\"\u003e\n \u003cp\u003eContinuous scores range from -1 to +1.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 204px;\"\u003e\n \u003cp\u003e0 = No problems\u003c/p\u003e\n \u003cp\u003e1 = Mild problems\u003c/p\u003e\n \u003cp\u003e2 = Moderate problems\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e3 = Severe problems\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp id=\"_Toc174959023\"\u003e\u003cstrong\u003eDeveloping the Chronic Health Condition (CHC) Algorithm\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBelow, we review the sequence of developing the CHC algorithm (see Figure 3 for the algorithm development), which involved (1) selecting diagnostic codes and (2) identifying the billing claim requirements and the reference period for each condition. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eSelecting diagnostic codes\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSupplemental Material B, Table B.2 presents 219 diagnostic codes used in the final version of the algorithm. Three sources of information were used to select diagnostic codes. First, we selected 198 diagnostic codes based on the pediatric Chronic Condition Classification (CCC) system version 2 (see Table B.1 Supplemental Material B) [32]. Second, we identified and added 12 common chronic conditions with high prevalence among pediatric populations [47-53]. The 12 conditions are commonly asked about in population-based surveys. Third, we examined the total frequency of diagnostic codes billed for each child within a 1-year time frame before the date of the 2014 interview to ensure we did not miss any conditions that may be considered chronic. Chronicity was based on scientific literature or conditions that require ongoing care (e.g., multiple visits), implying that they are chronic. Codes were examined based on the frequency of visits. Nine additional chronic conditions were added.\u003c/p\u003e\n\u003cp id=\"_Toc174959024\"\u003e\u003cstrong\u003e\u003cem\u003eEstablishing Reference Period and Minimum Visit Frequency\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe CCW codebook criteria [21] were used as the starting point. Where applicable, reference period and visit frequency criteria were applied to similar conditions not in the CCW but in the ICD-9.\u003c/p\u003e\n\u003cp\u003e(1) First, a total of 5 conditions had an established reference period with a specified minimum number of visits (i.e., established algorithm) to be considered a CHC based on the CCW codebook. Table B.2 in Supplemental Material B lists all reference periods considered for conditions. (2) Second, the CCW codebook provides established criteria for some, but not all, cancer-related conditions (e.g., breast, colorectal, endometrial, lung). In the CCW, all cancer conditions used the criteria of 2 billing claims in 1 year; we applied this to 92 other cancer-related conditions not listed in the CCW codebook but included in the ICD-9. (3) Third, the CCW used 1 billing claim within 1 year as the criteria for conditions deemed as chronic due to the incurable or irreversible nature of the condition (e.g., anemia, diabetes). We applied these same criteria to 18 conditions (e.g., HIV, infantile cerebral palsy, chromosomal conditions) that, based on the scientific literature, are incurable or irreversible conditions. No existing coding systems for pediatric or adult chronic conditions provide minimum visit criteria for these conditions. Table B.2 provides references used to make this decision for each diagnostic code (4) There were 104 conditions for which the optimal criteria were uncertain. Based on the literature, these conditions were grouped as A) curable conditions and B) diagnostic codes with multiple conditions/subcategories.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eA) Conditions with a cure may be acute or chronic. For example, conditions that arise due to premature labour or neonatal complications may resolve on their own, require treatment, or persist throughout adulthood [54-57]. Whether a condition is acute or chronic will depend on whether the person receives treatment, and/or whether the condition can resolve on its own after some time or other factors.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eB) Lastly, several diagnostic codes list multiple conditions or are too broad to understand the prognosis of related conditions (e.g., code 279 = Disorders involving the immune mechanism, code 348 = Other conditions of the brain). Past studies have used longer \u0026ldquo;look-back\u0026rdquo; periods to reduce false negatives (i.e., unobserved chronic conditions) and identify conditions that may have more irregular or intermittent access to health services [58, 59]. As such, we examined a minimum of 2 billing claims within 2, 3, and 4 years for A) curable conditions or B) diagnostic codes with multiple conditions/subcategories. The assumption is that at least 2 visits may reflect the ongoing nature of monitoring and/or treatment for the condition, which can be inferred as a chronic condition.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Analyses\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAnalyses were conducted using Statistical Package for Social Sciences (SPSS) version 29 and STATA 17.0 [60,61]. We applied sampling weights in all analyses to generate estimates that are representative of the population of children and youth in Ontario. To account for the complex survey design, mean bootstrap weights were applied with an adjustment factor to produce accurate standard errors. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSensitivity Analyses\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSensitivity analyses were conducted by adjusting the reference period only for conditions where the reference period was uncertain (see Supplemental Material B, section 2.1 for more information). The aim was to assess how the overall prevalence of CHC and patterns of agreement were affected when we adjusted the reference period to 2, 3, and 4 years. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFirst, we examined the overall prevalence of CHC by assessing whether a child met the criteria for any condition using survey or administrative data.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSecond, we examined patterns of agreement using Cohen\u0026rsquo;s kappa [62], sensitivity, and specificity estimates between administrative and survey-reported conditions. Sensitivity and specificity usually refer to the accurate identification in comparison to the gold standard: sensitivity examines the proportion of actual positive cases that a test correctly identifies, while specificity examines the proportion of true negative cases. However, there is no gold standard for our analyses [63]. Sensitivity was calculated as the probability of being classified with a CHC in both administrative and survey data. Specificity was calculated as the probability of not being classified with a CHC in either administrative or survey data. Previous research investigating agreement between administrative and survey data has used similar methods [8]. See Table B.2 in Supplemental Material B for a complete list of conditions that required sensitivity testing (i.e., reference period uncertain).\u0026nbsp;\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThird, given the poor agreement and high prevalence estimates (e.g., 4-years identified \u0026gt;50% children with CHC), we increased the billing claim requirements for each condition where the reference period was uncertain (i.e., conditions that required sensitivity testing) but not the reference period (henceforth referred to as the incremented billing claim requirement) with the goal of optimizing the algorithm.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFinally, after determining which of the reference periods was the optimal algorithm, we examined a few specific conditions we would expect to have strong agreement between survey and administrative data. Table C.1 in Supplemental Material C documents which set(s) of diagnostic codes were analyzed with parent-reported data.\u0026nbsp;\u003c/p\u003e\n\u003cp id=\"_Toc174959027\"\u003e\u003cstrong\u003eMultinomial Logistic Regression\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA multinomial logistic regression examined the association between sociodemographic variables and whether the child was categorized as: 0 = no long-term chronic condition, 1 = administrative CHC only, 2 = survey CHC only, or 3 = both administrative and survey. The independent variables included 1) child characteristics (age, sex), 2) family and household characteristics (PMK education status, parent structure, income strata), and 3) child clinical characteristics (health utility index). All analyses are presented as relative risk ratios (RRR) with corresponding 95% confidence intervals (CI).\u0026nbsp;\u003c/p\u003e\n\u003cp id=\"_Toc174959028\"\u003e\u003cstrong\u003eMissing Data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMultiple imputation by chained equation (MICE) was used to address missing data by imputing 10 sets of data using STATA based on best practice recommendations [64]. The specific methodology has been reported elsewhere [65].\u0026nbsp;\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eThe total sample included 8,985\u003csup\u003e[1] \u003c/sup\u003echildren aged 4 to 17. Characteristics of the sample population, weighted to reflect the general Ontario population, are presented in Table 2.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2\u0026nbsp;\u003c/strong\u003e\u003cem\u003eSociodemographic and Clinical Characteristics of Sample Selection\u003c/em\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"624\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 510px;\"\u003e\n \u003cp\u003eSociodemographic Characteristics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 114px;\"\u003e\n \u003cp\u003e% \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 624px;\"\u003e\n \u003cp\u003eChild Characteristics\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 510px;\"\u003e\n \u003cp\u003eAge (\u003cem\u003en\u003c/em\u003e = 8,985)\u003c/p\u003e\n \u003cp\u003e4-11\u003c/p\u003e\n \u003cp\u003e12-17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e56.11\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e43.89\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 510px;\"\u003e\n \u003cp\u003eSex (\u003cem\u003en\u003c/em\u003e = 8,985)\u003c/p\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003cp\u003eMale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 114px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e49.19\u003c/p\u003e\n \u003cp\u003e50.81\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 624px;\"\u003e\n \u003cp\u003eParent and Family Characteristics\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 510px;\"\u003e\n \u003cp\u003ePMK Education Status (\u003cem\u003en\u003c/em\u003e = 7,523)\u003c/p\u003e\n \u003cp\u003eTrade certificate; High school or less\u003c/p\u003e\n \u003cp\u003eCollege/CEGEP; University certificate below BA\u003c/p\u003e\n \u003cp\u003eBachelor\u0026rsquo;s degree (BA, BSc)\u003c/p\u003e\n \u003cp\u003eUniversity certificate above Bachelors\u003c/p\u003e\n \u003cp\u003eParent Structure (\u003cem\u003en\u003c/em\u003e = 8,925)\u003c/p\u003e\n \u003cp\u003eTwo parent household\u003c/p\u003e\n \u003cp\u003eSingle parent household\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 114px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e10.77\u003c/p\u003e\n \u003cp\u003e44.69\u003c/p\u003e\n \u003cp\u003e30.04 \u0026nbsp;\u003c/p\u003e\n \u003cp\u003e14.49\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e80.38 \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e19.62\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 510px;\"\u003e\n \u003cp\u003eIncome strata (\u003cem\u003en\u003c/em\u003e = 8,985)\u003c/p\u003e\n \u003cp\u003eLow\u003c/p\u003e\n \u003cp\u003eMedium\u003c/p\u003e\n \u003cp\u003eHigh\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 114px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e20.73\u003c/p\u003e\n \u003cp\u003e57.11\u003c/p\u003e\n \u003cp\u003e22.16\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 510px;\"\u003e\n \u003cp\u003eChild Clinical Characteristics\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 114px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 510px;\"\u003e\n \u003cp\u003eHealth Utility Index (\u003cem\u003en\u003c/em\u003e = 8,874)\u003c/p\u003e\n \u003cp\u003eNo problems\u003c/p\u003e\n \u003cp\u003eMild problems\u003c/p\u003e\n \u003cp\u003eModerate problems\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eSevere problems\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 114px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e57.25\u003c/p\u003e\n \u003cp\u003e21.59\u003c/p\u003e\n \u003cp\u003e16.81\u003c/p\u003e\n \u003cp\u003e4.35\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003ePrevalence and Agreement Between Administrative and Survey Data: Any Condition\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003ePrevalence: Any Condition\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTable 3 presents the unweighted prevalence estimates and concordance measures. The initial billing claim requirement demonstrates that 31.60% of children met the criteria for a CHC using the 2-year reference period, 41.69% of children met the criteria for a CHC using the 3-year reference period, and 50.71% of children met the criteria for a CHC using the 4-year reference period. (N.B. 24% of billing claims had the diagnostic code 999 (i.e., without diagnosis). \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAfter the initial round of analyses, we increased the billing claim requirements (i.e., from 2 billing claims to 3 billing claims in 2, 3, and 4 years) for each condition where the reference period was uncertain (referred to as the incremented billing claim requirement). This decision was due to the extremely high prevalence estimates (\u0026gt; 50% using the 4-year reference period), which does not align with previous literature examining the prevalence of pediatric chronic conditions (e.g., 20% - 30%) [2-5]. Also, the prevalence of long-term conditions reported using survey data was 26.7%.\u003c/p\u003e\n\u003cp\u003eThe incremented billing claim requirement found that 23.27% of children met the criteria for a CHC using the 2-year reference period, 31.59% of children met the criteria for a CHC using the 3-year reference period, and 39.50% of children met the criteria for a CHC using the 4-year reference period.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAgreement: Any Condition\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBoth the initial and incremented billing claim requirements have low sensitivity and specificity. According to Cohen\u0026rsquo;s [62] criteria, the agreement remains slight (2 years: \u003cem\u003ek\u003c/em\u003e = 0.15, 3 years: \u003cem\u003ek\u003c/em\u003e = 0.13, 4 years: \u003cem\u003ek\u003c/em\u003e = 0.11) using the initial billing claim requirement. The agreement improves but remains slight using the incremented billing claim requirement (2 years: \u003cem\u003ek\u003c/em\u003e = 0.16, 3 years: \u003cem\u003ek\u003c/em\u003e = 0.16; 4 years: \u003cem\u003ek\u003c/em\u003e = 0.15).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 3\u0026nbsp;\u003c/strong\u003e\u003cem\u003eUnweighted Prevalence and Agreement of Chronic Conditions Using Administrative Data and Survey Data\u003c/em\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"810\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 40px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"top\" style=\"width: 193px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePrevalence (%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"bottom\" style=\"width: 193px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCohen\u0026apos;s \u003cem\u003ek\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"bottom\" style=\"width: 192px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSensitivity (%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"bottom\" style=\"width: 192px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSpecificity (%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 40px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2 years\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e3 years\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e4 years\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2 years\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e3 years\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e4 years\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2 years\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e3 years\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e4 years\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2 years\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2 years\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e4 years\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 40px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e31.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e41.69\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e51.71\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e11.55\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e14.07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e16.29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e53.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e45.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e38.86\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 40px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2.\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e23.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e31.59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e39.50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e9.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e11.74\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e13.94\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e59.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e53.43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e47.72\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e1. Initial billing claim requirement\u003c/p\u003e\n\u003cp\u003e2. Incremented billing claim requirement.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eNote\u003c/em\u003e. Sensitivity and specificity are measured by comparing Ontario health administrative billing claims to survey data (i.e., the parent report).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBased on the findings in Table 3, we determined that the incremented billing claim criteria were optimal, given that kappa values increased across all reference periods. Among the incremented billing claim criteria, the 3-year reference period was optimal. 1) The kappa value was consistent between 2- and 3-years. As previously discussed, it is advised to use a longer look-back or reference period to avoid missing cases [22]. 2) However, in comparison, we selected 3-years over 4-years given that the estimated prevalence of the 3-year reference period fell within the prevalence range for pediatric chronic conditions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAgreement Between Administrative and Survey Data: Specific Conditions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe agreement between administrative and survey data for identifying a child as having a CHC was poor. Thus, we examined four conditions we expected would have strong agreement to help clarify the discrepancy. The selected diagnoses were conditions that we expected to be readily apparent to a parent as a \u0026ldquo;chronic condition:\u0026rdquo; diabetes, heart condition or disease (henceforth referred to as heart conditions), epilepsy, and cerebral palsy. Sample sizes were adequate only for diabetes and heart conditions. Table 4 shows that diabetes and heart conditions demonstrate substantial (i.e., 0.61-0.80) and fair agreement (i.e., 0.21-0.40), respectively. For details on the diagnostic codes matched with survey codes refer to Table C.1 in Supplemental Material C.\u0026nbsp;\u003c/p\u003e\n\u003cp id=\"_Toc172553270\"\u003e\u003cstrong\u003eTable 4\u0026nbsp;\u003c/strong\u003e\u003cem\u003eUnweighted Agreement on Survey Data and Administrative Data using Specific Diagnostic Codes, 3 years\u003c/em\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"861\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 302px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePrevalence Estimates\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" valign=\"bottom\" style=\"width: 417px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAgreement Estimates\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 161px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAdministrative (%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSurvey (%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 105px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCohen\u0026apos;s \u003cem\u003ek\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 178px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSensitivity (%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSpecificity (%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDiabetes\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 161px;\"\u003e\n \u003cp\u003e0.5%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e0.3%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 105px;\"\u003e\n \u003cp\u003e0.79\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 178px;\"\u003e\n \u003cp\u003e0.32%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e99.51%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eHeart Conditions\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 161px;\"\u003e\n \u003cp\u003e1.1%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 142px;\"\u003e\n \u003cp\u003e1.1%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 105px;\"\u003e\n \u003cp\u003e0.21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 178px;\"\u003e\n \u003cp\u003e0.23%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e98.06%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eMultinomial Logistic Regression\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel Indices\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe average relative variance increase (RVI) was 0.05. The largest fraction of missing information (FMI) was 0.20. The model was statistically significant (\u003cem\u003eF\u0026nbsp;\u003c/em\u003e(33, 104517.9) = 5.16, \u003cem\u003ep\u003c/em\u003e \u0026lt; .001). The relative risk ratios (\u003cem\u003eRRR\u003c/em\u003e) and confidence intervals (CI) are presented in Table 5.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAdministrative CHC\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYounger age was significantly associated with having an administrative CHC (\u003cem\u003eRRR\u003c/em\u003e = 0.41, 95% CI [0.325, 0.517]) compared to no CHC. Adolescents (ages 12-17) showed a 59% relative risk reduction compared to children (ages 4-11). Parent structure was also significant (\u003cem\u003eRRR\u003c/em\u003e = 0.78, 95% CI [0.608, 0.999]). Children of single parents showed a 22% relative risk reduction of having an administrative CHC compared to two-parent households. (N.B. Parent structure results should be interpreted cautiously due to the 95% CI upper bound approaching 1.00). Child sex, PMK education status, income strata and HUI were not statistically significant.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSurvey CHC\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOlder age was significantly associated with having a survey-reported CHC (\u003cem\u003eRRR\u003c/em\u003e = 1.44, 95% CI [1.148, 1.805]) compared to no CHC. Adolescents (ages 12-17) were 44% more likely to have a survey-reported CHC compared to children (ages 4-11). Health problems, measured by HUI, were significantly associated with whether a child had a survey-reported CHC compared to no CHC (see Table 5). Severe health problems were found to be the strongest predictor of survey-reported CHC (\u003cem\u003eRRR\u003c/em\u003e = 2.99, 95% CI [1.849, 4.827]). Child sex, PMK education status, parent structure, and income strata were not statistically significant.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBoth Administrative and Survey CHC\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYounger age was significantly associated with having an administrative and survey-reported CHC (\u003cem\u003eRRR\u003c/em\u003e = 0.65, 95% CI [0.503, 0.847]) compared to no CHC. Adolescents (ages 12-17) showed a 35% relative risk reduction of having an administrative and survey-reported CHC compared to children (ages 4-11). Severe health problems were found to be the strongest predictor of survey-reported CHC (\u003cem\u003eRRR\u003c/em\u003e = 1.97, 95% CI [1.171, 3.305]). Children with severe health problems were 97% more likely to have an administrative and survey-reported CHC compared to children with no health problems. Child sex, PMK education status, parent structure, and income strata were not statistically significant. \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 5\u0026nbsp;\u003c/strong\u003e\u003cem\u003eMultinomial Logistic Regression Examining Predictors of Chronic Condition Status using the Incremented Billing Claim Requirements, 3-year reference period\u003c/em\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"864\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.268%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 26.6618%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAdministrative vs.\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eNo CHC\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSurvey-Reported vs.\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eNo CHC\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.34%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAdministrative \u0026amp; Survey-Reported vs. No CHC\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.268%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 26.6618%;\"\u003e\n \u003cp\u003e\u003cem\u003eRRR [\u003c/em\u003e 95% CI]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e\u003cem\u003eRRR [\u003c/em\u003e 95% CI]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.34%;\"\u003e\n \u003cp\u003e\u003cem\u003eRRR [\u003c/em\u003e 95% CI]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.268%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eChild Characteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 26.6618%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 20.34%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.268%;\"\u003e\n \u003cp\u003eAge \u003csup\u003ea\u003c/sup\u003e: 12-17\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.41 [0.325, 0.517]\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.44 [1.148, 1.805]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.34%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.65 [0.503, 0.847]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.268%;\"\u003e\n \u003cp\u003eSex \u003csup\u003eb\u003c/sup\u003e: Male\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\n \u003cp\u003e0.99 [0.814, 1.214]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e1.02 [0.817, 1.274]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.34%;\"\u003e\n \u003cp\u003e1.28 [0.988, 1.647]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.268%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eParent and Family Characteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.34%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.268%;\"\u003e\n \u003cp\u003ePMK Education Status \u003csup\u003ec\u003c/sup\u003e:\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.34%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 24.268%;\"\u003e\n \u003cp\u003eCollege/CEGEP; University certificate below BA\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\n \u003cp\u003e0.79 [0.542, 1.164]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e0.99 [0.663, 1.474]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.34%;\"\u003e\n \u003cp\u003e1.46 [0.873, 2.443]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 24.268%;\"\u003e\n \u003cp\u003eBachelor\u0026rsquo;s degree (BA, BSc)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\n \u003cp\u003e0.95 [0.644, 1.389]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e0.96 [0.621, 1.487]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.34%;\"\u003e\n \u003cp\u003e1.28 [0.763, 2.160]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 24.268%;\"\u003e\n \u003cp\u003eUniversity certificate degree above Bachelors\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\n \u003cp\u003e0.76 [0.492, 1.161]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e0.92 [0.560, 1.504]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.34%;\"\u003e\n \u003cp\u003e1.30 [0.736, 2.284]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 24.268%;\"\u003e\n \u003cp\u003eParent Structure \u003csup\u003ed\u003c/sup\u003e: Single Parent\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.78 [0.608, 0.999]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e0.94 [0.704, 1.258]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.34%;\"\u003e\n \u003cp\u003e0.92 [0.657, 1.276]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 24.268%;\"\u003e\n \u003cp\u003eIncome strata \u003csup\u003ee\u003c/sup\u003e:\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.34%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 24.268%;\"\u003e\n \u003cp\u003eMedium (20\u003csup\u003eth\u003c/sup\u003e-80\u003csup\u003eth\u003c/sup\u003e percentile)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\n \u003cp\u003e0.87 [0.695, 1.082]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e1.11 [0.861, 1.439]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.34%;\"\u003e\n \u003cp\u003e0.86 [0.632, 1.170]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 24.268%;\"\u003e\n \u003cp\u003eHigh (High: \u0026gt;80\u003csup\u003eth\u003c/sup\u003e percentile)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\n \u003cp\u003e0.90 [0.705, 1.143]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e1.26 [0.955, 1.659]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 20.34%;\"\u003e\n \u003cp\u003e1.09 [0.784, 1.507]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.4394%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eChild Clinical Characteristics\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.34%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.4394%;\"\u003e\n \u003cp\u003eHealth Utility Index \u003csup\u003ef\u003c/sup\u003e:\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.34%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.4394%;\"\u003e\n \u003cp\u003eMild problems\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\u003cbr\u003e0.95 [0.730, 1.242]\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.75 [1.320, 2.318]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.34%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003cstrong\u003e1.53 [1.097, 2.128]\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.4394%;\"\u003e\n \u003cp\u003eModerate problems\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\u003cbr\u003e0.97 [0.742, 1.280]\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1.65 [1.252, 2.174]\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.34%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003cstrong\u003e1.86 [1.330, 2.592]\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 21.4394%;\"\u003e\n \u003cp\u003eSevere problems\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 26.6618%;\"\u003e\u003cbr\u003e0.80 [0.442, 1.449]\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 24.8752%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2.99 [1.849, 4.827]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.34%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003cstrong\u003e1.97 [1.171, 3.305]\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eSignificant RRRs (\u003cem\u003ep\u003c/em\u003e \u0026lt; .05) are bolded. \u003cem\u003eRRR\u003c/em\u003e = relative risk ratio, CI = confidence interval\u003c/p\u003e\n\u003cp\u003ea. Age \u0026ndash; \u0026ldquo;4-11\u0026rdquo; is the reference group.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eb. Sex \u0026ndash; \u0026ldquo;Female\u0026rdquo; is the reference group.\u003c/p\u003e\n\u003cp\u003ec. PMK Education Status \u0026ndash; \u0026ldquo;Trade certificate or High school or less\u0026rdquo; is the reference group.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003ed. Parent Structure \u0026ndash; \u0026ldquo;Two parents\u0026rdquo; is the reference group.\u003c/p\u003e\n\u003cp\u003ee. Income strata \u0026ndash; \u0026ldquo;Low (\u0026le; 20\u003csup\u003eth\u003c/sup\u003e percentile)\u0026rdquo; is the reference group.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003ef. Health Utility Index \u0026ndash; \u0026ldquo;No problems\u0026rdquo; is the reference group.\u003c/p\u003e\n\u003cp\u003e[1] The dependent variable was not imputed, resulting in a total sample of 8,985 for the regression.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eWe developed an algorithm to categorize children and youth aged 4 to 17 as having a CHC using health administrative data consisting of ICD-9 diagnosis codes and the number of visits within a specified reference period.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFirst, we found differences in chronic condition prevalence and agreement between survey and administrative data. Prevalence estimates were consistently higher in administrative data. Other studies on physical and mental health have also found that prevalence estimates are higher in administrative data than in survey data [66, 67]. It may be that parents underestimate what constitutes a long-term condition. Importantly, the survey asked parents whether a health professional diagnosed a long-term condition but did not provide an operational definition, such as a condition expected to last 3 months or more. There are different definitions of chronic conditions, ranging from \u0026nbsp;\u0026ge; 3 months to \u0026ge; 1 year [6]. Future surveys could include a reference period \u0026ndash; for example, \u0026ldquo;Within the past year, has your child suffered from\u0026hellip;\u0026rdquo; \u0026ndash; when asking about long-term conditions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe examined diabetes and heart conditions as specific conditions we expected parents could readily and easily identify as chronic conditions. Diabetes had a substantial agreement between administrative and survey data. However, it is worth noting the small sample size, with only 31 and 42 participants having diabetes using survey and administrative data, respectively (N.B. The minimum unweighted criteria to disclose is 30). Heart conditions had fair agreement between administrative data and survey data relative to diabetes. Interestingly, both administrative and survey data identified the same prevalence estimate for heart conditions despite having fair agreement. Studies examining agreement between multiple conditions have also found considerable variability depending on the type of condition [66, 68].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorrelates of CHC identification.\u0026nbsp;\u003c/strong\u003eAge significantly influenced whether administrative, survey, or both capture a CHC compared to no CHC. Administrative data was less likely to capture CHC in older children compared to younger children. The algorithm may not be appropriate for use with older children. However, it may also be that younger children are more likely to have frequent medical visits (e.g., well-baby or well-child care) due to heightened attention to developmental milestones and early detection of potential health issues [69]. Alternatively, older children were more likely to have a survey-reported CHC. The algorithm covered up to three years for conditions, while parents reported any lifetime chronic condition. It may be that the algorithm missed conditions diagnosed in early childhood that no longer require frequent follow-up visits as the child ages.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eExcept for parent structure, all parent and family characteristics did not relate to whether a child had an administrative CHC or administrative/survey CHC compared to no CHC. This suggests that the healthcare billing system effectively captures CHC uniformly across different demographics, such as income, child sex, and PMK education. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFinally, health problems were associated with survey CHC and administrative/survey CHC. Previous research indicates that self-reported chronic conditions are associated with higher morbidity levels, also measured by HUI [66]. While the 2014 OCHS is parent-reported, these findings align with the notion that survey data may capture more severe health conditions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStrengths\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study had several strengths. First, we approached algorithm development as a multistep process by determining a list of conditions based on previous literature and applying a unique algorithm to each condition. Second, our algorithm includes more possible diagnostic codes than previous studies [21, 22, 66]. Thus, it was more likely to capture less prevalent conditions that may have been overlooked in other studies. This may have implications for studies comparing cases with a CHC vs healthy controls. Unless health controls are categorized based on the absence of all possible CHCs, some children with a CHC may have been included in healthy control groups. Third, we examined a few specific conditions to understand the poor agreement. We found that even among conditions that we may expect to be readily apparent as chronic, there is still considerable disagreement depending on the type of condition. The study provides a basis for improving data collection methods by identifying specific conditions with reporting discrepancies, which can help inform the development of better tools and protocols for reporting and recording health conditions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLimitations\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe current study has limitations. First, we did not have access to the response options to the \u0026ldquo;Other\u0026rdquo; item from parents that specified the specific conditions. If text data \u003cem\u003ehad\u003c/em\u003e been coded, researchers might have excluded conditions parents consider \u0026ldquo;chronic conditions.\u0026rdquo; The \u0026ldquo;Other\u0026rdquo; category includes 3.5% of all children and youth with a CHC, which may be a reason for poor agreement.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSecond, physicians\u0026rsquo; ratings of whether or not a child had a chronic condition were not available; such ratings could be useful in comparing both the algorithm coding and parent reports. Nevertheless, achieving consensus in terms of what is considered a \u0026ldquo;chronic\u0026rdquo; condition across physicians may be difficult [70], especially across specialties. A review examining the miscoding of diagnostic codes in medical records found that 16% of clinical cases required a diagnosis-related group change (i.e., the grouping of medical conditions into a certain category) [70]. Furthermore, a review examining definitions of chronic conditions used in research found that several characteristics influenced the classification of chronicity (e.g., duration or latency, need for medical attention, departure from well-being, noncontagious nature, multiple risk factors, pathology, and nonamenability to cure) [6].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThird, diagnostic codes were inputted by a coding specialist. Coders may vary in how they code the same condition. For example, while one coder may bill a visit using \u0026ldquo;Bronchitis NOS\u0026rdquo; (code = 490), another may bill for \u0026ldquo;Chronic Bronchitis\u0026rdquo; (code = 491). It was beyond the scope of this study to examine the ambiguity in coding between specific conditions, but this is a potential area for future research. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFourth, the cell count was very small for many conditions, limiting our analyses and raising the possibility of spurious findings.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFinally, this study excluded 316 children with developmental delays and intellectual disabilities, which may result in an underestimation of the prevalence of CHCs. A systematic review found that the prevalence of CHC in children with intellectual disability is much higher than in the general population [71].\u003c/p\u003e\n\u003cp\u003eFuture research could consider limiting the number of conditions examined, comparing algorithms for specific conditions with disease-specific registries, and examining problem lists in electronic medical records or physician-reported diagnoses as methods to validate future algorithms. \u0026nbsp;\u0026nbsp;\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eWe aimed to create an algorithm to identify all potential CHC in children and youth aged 4 to 17 and determine which sociodemographic factors are linked. We found considerable discrepancies between administrative and survey-reported data. The results highlight the importance of using algorithms developed from multiple datasets to examine complex research questions, such as the measurement of chronicity.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eCCC: Chronic Condition Classification\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCCHS: Canadian Community Health Survey\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCCW: Chronic Condition Warehouse\u003c/p\u003e\n\u003cp\u003eCHC: chronic health condition\u003c/p\u003e\n\u003cp\u003eCRG: Clinical Risk Groups\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eHUI: Health utility index\u003c/p\u003e\n\u003cp\u003eICD: International Classification of Disease\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eICES Institute for Clinical Evaluative Sciences\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMH: mental health\u003c/p\u003e\n\u003cp\u003eMICE: Multiple imputation by chained equation\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMOH: Ministry of Health\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eNLSCY: National Longitudinal Survey of Children and Youth\u003c/p\u003e\n\u003cp\u003eOCHS: Ontario Child Health Study\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOHIP: Ontario health insurance plan\u003c/p\u003e\n\u003cp\u003ePMK: Parent most knowledgeable\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRECORD: REporting of studies Conducted using Observational Routinely-collected health Data\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate:\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe 2014 OCHS was a voluntary survey conducted under the Statistics Act, which provides respondents guarantees of their privacy and confidentially. The study was approved by the Hamilton Integrated Research Ethics Board at McMaster University (no. 13-140).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication:\u003c/strong\u003e Not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e: Data access to the 2014 Ontario Child Health Study (OCHS) is available through Statistics Canada. All physician billing data (i.e., OHIP data) and the linked OCHS-OHIP dataset for the current analyses are stored at McMaster University. Access to the linked dataset cannot be shared due to data sharing agreements with Statistics Canada and the Ontario Ministry of Health (MOH). \u0026nbsp;Parts of this material are based on data and information provided by the MOH and the former Ministry of Children and Youth Services (MCYS). The opinions, results and conclusions reported in this paper are those of the authors and do not necessarily reflect those of MOH.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests:\u003c/strong\u003e The authors declare that they have no competing interests.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding:\u003c/strong\u003e\u0026nbsp; Grace Golden was supported by a Canada Graduate Master’s Scholarship – Social Sciences and Humanities Research Council (SSHRC) grant \u0026nbsp;The principal study was supported by the Canadian Institutes of Health Research (CIHR #125941), the Ontario Ministry of Health and Long-Term Care (MOHLTC) - Health Services Research Grant (#8-42298), and funding from MOHLTC, the Ontario Ministry of Children and Youth Services, and the Ontario Ministry of Education. Graham Reid is supported by the Children’s Health Research Institute, London, ON.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions:\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGG: \u0026nbsp;conceptualization, methodology, investigation, analysis, writing – reviewing and editing.\u003c/p\u003e\n\u003cp\u003eLW: analysis – reviewing and editing.\u003c/p\u003e\n\u003cp\u003eGJR: conceptualization, methodology, investigation, analysis, writing – reviewing and editing, supervision.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003cstrong\u003e:\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe acknowledge the contribution of the following content experts we consulted to better understand long-term conditions for clinical relevance, validity, and robustness. They provided valuable feedback that significantly improved how we analyzed, interpreted, and communicated the data.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMark Ferro, Ph.D. School of Public Health Sciences, University of Waterloo, Waterloo, ON\u003c/p\u003e\n\u003cp\u003eShannon Reaume, Ph.D. Candidate, School of Public Health Sciences, University of Waterloo, Waterloo, ON\u003c/p\u003e\n\u003cp\u003eFlorence Perquier, Ph.D., Cundill Centre for Child and Youth Depression, Centre for Addiction and Mental Health, Toronto, ON; Department of Psychiatry, University of Toronto, Toronto, ON\u003c/p\u003e\n\u003cp\u003ePeter Szatmari, MD, Cundill Centre for Child and Youth Depression, Centre for Addiction and Mental Health, Toronto, ON; The Hospital for Sick Children, Toronto, ON\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003ePerrin, E. C., Newacheck, P., Pless, I. B., Drotar, D., Gortmaker, S. L., Leventhal, J., Perrin, J. M., Stein, R. E., Walker, D. K., \u0026amp; Weitzman, M. (1993). Issues involved in the definition and classification of chronic health conditions. \u003cem\u003ePediatrics\u003c/em\u003e, \u003cem\u003e91\u003c/em\u003e(4), 787\u0026ndash;793.\u003c/li\u003e\n\u003cli\u003eMcDougall, J., King, G., De Wit, D. J., Miller, L. T., Hong, S., Offord, D. R., Laporta, J., \u0026amp; Meyer, K. (2004). Chronic physical health conditions and disability among Canadian school-aged children: A national profile. \u003cem\u003eDisability and Rehabilitation\u003c/em\u003e, \u003cem\u003e26\u003c/em\u003e(1), 35\u0026ndash;45. https://doi.org/10.1080/09638280410001645076 \u003c/li\u003e\n\u003cli\u003eNewacheck, P. W. (1991). Prevalence and Impact of Chronic Illness Among Adolescents. \u003cem\u003eArchives of Pediatrics \u0026amp; Adolescent Medicine\u003c/em\u003e, \u003cem\u003e145\u003c/em\u003e(12), 1367. https://doi.org/10.1001/archpedi.1991.02160120035015 \u003c/li\u003e\n\u003cli\u003eNewacheck, P. W., \u0026amp; Stoddard, J. J. (1994). Prevalence and impact of multiple childhood chronic illnesses. \u003cem\u003eThe Journal of Pediatrics\u003c/em\u003e, \u003cem\u003e124\u003c/em\u003e(1), 40\u0026ndash;48. https://doi.org/10.1016/S0022-3476(94)70252-7 \u003c/li\u003e\n\u003cli\u003eNewacheck, P. W., \u0026amp; Taylor, W. R. (1992). Childhood chronic illness: Prevalence, severity, and impact. \u003cem\u003eAmerican Journal of Public Health\u003c/em\u003e, \u003cem\u003e82\u003c/em\u003e(3), 364\u0026ndash;371. https://doi.org/10.2105/ajph.82.3.364 \u003c/li\u003e\n\u003cli\u003eGoodman, R. A., Posner, S. F., Huang, E. S., Parekh, A. K., \u0026amp; Koh, H. K. (2013). Defining and measuring chronic conditions: Imperatives for research, policy, program, and practice. \u003cem\u003ePreventing Chronic Disease\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e, E66. https://doi.org/10.5888/pcd10.120239 \u003c/li\u003e\n\u003cli\u003eBerete, F., Demarest, S., Charafeddine, R., De Ridder, K., Van Oyen, H., Van Hoof, W., Bruy\u0026egrave;re, O., \u0026amp; Van Der Heyden, J. (2023). Linking health survey data with health insurance data: Methodology, challenges, opportunities and recommendations for public health research. An experience from the HISlink project in Belgium. \u003cem\u003eArchives of Public Health\u003c/em\u003e, \u003cem\u003e81\u003c/em\u003e(1), 198. https://doi.org/10.1186/s13690-023-01213-0 \u003c/li\u003e\n\u003cli\u003eDuncan, L., Georgiades, K., Wang, L., Edwards, J., \u0026amp; Comeau, J. (2022). Estimating prevalence of child and youth mental disorder and mental health-related service contacts: A comparison of survey data and linked administrative health data. \u003cem\u003eEpidemiology and Psychiatric Sciences\u003c/em\u003e, \u003cem\u003e31\u003c/em\u003e, e35. https://doi.org/10.1017/S204579602200018X\u003c/li\u003e\n\u003cli\u003eGuttmann, A., Nakhla, M., Henderson, M., To, T., Daneman, D., Cauch-Dudek, K., Wang, X., Lam, K., \u0026amp; Hux, J. (2010). Validation of a health administrative data algorithm for assessing the epidemiology of diabetes in Canadian children. \u003cem\u003ePediatric Diabetes\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(2), 122\u0026ndash;128. https://doi.org/10.1111/j.1399-5448.2009.00539.x\u003c/li\u003e\n\u003cli\u003eHoefgen, E. R., Andrews, A. L., Richardson, T., Hall, M., Neff, J. M., Macy, M. L., Bettenhausen, J. L., Shah, S. S., \u0026amp; Auger, K. A. (2017). Health Care Expenditures and Utilization for Children With Noncomplex Chronic Disease. \u003cem\u003ePediatrics\u003c/em\u003e, \u003cem\u003e140\u003c/em\u003e(3), e20170492. https://doi.org/10.1542/peds.2017-0492\u003c/li\u003e\n\u003cli\u003eTo, T., Dell, S., Dick, P. T., Cicutto, L., Harris, J. K., MacLusky, I. B., \u0026amp; Tassoudji, M. (2006). Case verification of children with asthma in Ontario. \u003cem\u003ePediatric Allergy and Immunology\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(1), 69\u0026ndash;76. https://doi.org/10.1111/j.1399-3038.2005.00346.x\u003c/li\u003e\n\u003cli\u003ePeng, M., Lee, S., D\u0026rsquo;Souza, A. G., Doktorchik, C. T. A., \u0026amp; Quan, H. (2020). Development and validation of data quality rules in administrative health data using association rule mining. \u003cem\u003eBMC Medical Informatics and Decision Making\u003c/em\u003e, \u003cem\u003e20\u003c/em\u003e(1), 75. https://doi.org/10.1186/s12911-020-1089-0\u003c/li\u003e\n\u003cli\u003eBhandari, A., \u0026amp; Wagner, T. (2006). Self-Reported Utilization of Health Care Services: Improving Measurement and Accuracy. \u003cem\u003eMedical Care Research and Review\u003c/em\u003e, \u003cem\u003e63\u003c/em\u003e(2), 217\u0026ndash;235. https://doi.org/10.1177/1077558705285298\u003c/li\u003e\n\u003cli\u003eSmith, B., Chu, L. K., Smith, T. C., Amoroso, P. J., Boyko, E. J., Hooper, T. I., Gackstetter, G. D., \u0026amp; Ryan, M. A. (2008). Challenges of self-reported medical conditions and electronic medical records among members of a large military cohort. \u003cem\u003eBMC Medical Research Methodology\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e(1), 37. https://doi.org/10.1186/1471-2288-8-37\u003c/li\u003e\n\u003cli\u003eClark, R. E., Ricketts, S. K., \u0026amp; McHugo, G. J. (1996). Measuring hospital use without claims: A comparison of patient and provider reports. \u003cem\u003eHealth Services Research\u003c/em\u003e, \u003cem\u003e31\u003c/em\u003e(2), 153\u0026ndash;169.\u003c/li\u003e\n\u003cli\u003eHorwitz, R. I., \u0026amp; Yu, E. C. (1984). Assessing the reliability of epidemiologic data obtained from medical records. \u003cem\u003eJournal of Chronic Diseases\u003c/em\u003e, \u003cem\u003e37\u003c/em\u003e(11), 825\u0026ndash;831. https://doi.org/10.1016/0021-9681(84)90015-8\u003c/li\u003e\n\u003cli\u003eKopec, J. A. (2022). Estimating Disease Prevalence in Administrative Data. \u003cem\u003eClinical and Investigative Medicine\u003c/em\u003e, \u003cem\u003e45\u003c/em\u003e(2), E21-27. https://doi.org/10.25011/cim.v45i2.38100\u003c/li\u003e\n\u003cli\u003eSinger, A., Yakubovich, S., Kroeker, A. L., Dufault, B., Duarte, R., \u0026amp; Katz, A. (2016). Data quality of electronic medical records in Manitoba: Do problem lists accurately reflect chronic disease billing diagnoses? \u003cem\u003eJournal of the American Medical Informatics Association\u003c/em\u003e, \u003cem\u003e23\u003c/em\u003e(6), 1107\u0026ndash;1112. https://doi.org/10.1093/jamia/ocw013\u003c/li\u003e\n\u003cli\u003eTerry, A. L., Stewart, M., Cejic, S., Marshall, J. N., De Lusignan, S., Chesworth, B. M., Chevendra, V., Maddocks, H., Shadd, J., Burge, F., \u0026amp; Thind, A. (2019). A basic model for assessing primary health care electronic medical record data quality. \u003cem\u003eBMC Medical Informatics and Decision Making\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(1), 30. https://doi.org/10.1186/s12911-019-0740-0\u003c/li\u003e\n\u003cli\u003eGorina, Y., \u0026amp; Kramarow, E. A. (2011). Identifying chronic conditions in Medicare claims data: Evaluating the Chronic Condition Data Warehouse algorithm. \u003cem\u003eHealth Services Research\u003c/em\u003e, \u003cem\u003e46\u003c/em\u003e(5), 1610\u0026ndash;1627. https://doi.org/10.1111/j.1475-6773.2011.01277.x\u003c/li\u003e\n\u003cli\u003eChronic Conditions Warehouse. (2022). \u003cem\u003eCODEBOOK: Master Beneficiary Summary File (MBSF) 27 CCW Chronic Conditions (CC) File\u003c/em\u003e. https://resdac.org/cms-data/files/mbsf-27-cc/data-documentation\u003c/li\u003e\n\u003cli\u003eVoss, R. W., Schmidt, T. D., Weiskopf, N., Marino, M., Dorr, D. A., Huguet, N., Warren, N., Valenzuela, S., O\u0026rsquo;Malley, J., \u0026amp; Qui\u0026ntilde;ones, A. R. (2022). Comparing ascertainment of chronic condition status with problem lists versus encounter diagnoses from electronic health records. \u003cem\u003eJournal of the American Medical Informatics Association\u003c/em\u003e, \u003cem\u003e29\u003c/em\u003e(5), 770\u0026ndash;778. https://doi.org/10.1093/jamia/ocac016\u003c/li\u003e\n\u003cli\u003eMasters, C. L., Bateman, R., Blennow, K., Rowe, C. C., Sperling, R. A., \u0026amp; Cummings, J. L. (2015). Alzheimer\u0026rsquo;s disease. \u003cem\u003eNature Reviews Disease Primers\u003c/em\u003e, \u003cem\u003e1\u003c/em\u003e(1), 15056. https://doi.org/10.1038/nrdp.2015.56\u003c/li\u003e\n\u003cli\u003eSalari, N., Ghasemi, H., Mohammadi, L., Behzadi, M. H., Rabieenia, E., Shohaimi, S., \u0026amp; Mohammadi, M. (2021). The global prevalence of osteoporosis in the world: A comprehensive systematic review and meta-analysis. \u003cem\u003eJournal of Orthopaedic Surgery and Research\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(1), 609. https://doi.org/10.1186/s13018-021-02772-0\u003c/li\u003e\n\u003cli\u003eSymmons, D. P. M. (2002). Epidemiology of rheumatoid arthritis: Determinants of onset, persistence and outcome. \u003cem\u003eBest Practice \u0026amp; Research Clinical Rheumatology\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(5), 707\u0026ndash;722. https://doi.org/10.1053/berh.2002.0257\u003c/li\u003e\n\u003cli\u003eGershon, A. S., Wang, C., Guan, J., Vasilevska-Ristovska, J., Cicutto, L., \u0026amp; To, T. (2009). Identifying patients with physician-diagnosed asthma in health administrative databases. \u003cem\u003eCanadian Respiratory Journal\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(6), 183\u0026ndash;188. https://doi.org/10.1155/2009/963098\u003c/li\u003e\n\u003cli\u003eJaakkimainen, L., Duchen, R., Lix, L., Al-Azazi, S., Yu, B., Butt, D., Park, S.-B., \u0026amp; Widdifield,J. (2022). Identification of Early Onset Dementia in Population-Based HealthAdministrative Data: A Validation Study Using Primary Care Electronic Medical Records. \u003cem\u003eJournal of Alzheimer\u0026rsquo;s Disease\u003c/em\u003e, \u003cem\u003e89\u003c/em\u003e(4), 1463\u0026ndash;1472. https://doi.org/10.3233/JAD-220384\u003c/li\u003e\n\u003cli\u003eMuggah, E., Graves, E., Bennett, C., \u0026amp; Manuel, D. G. (2012). The impact of multiple chronic diseases on ambulatory care use; a population based study in Ontario, Canada. \u003cem\u003eBMC Health Services Research\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(1), 452. https://doi.org/10.1186/1472-6963-12-452\u003c/li\u003e\n\u003cli\u003eShah, B. R., Booth, G. L., Feig, D. S., \u0026amp; Lipscombe, L. L. (2023). Validation of Algorithms to Identify Gestational Diabetes From Population-level Health-care Administrative Data. \u003cem\u003eCanadian Journal of Diabetes\u003c/em\u003e, \u003cem\u003e47\u003c/em\u003e(1), 25\u0026ndash;30. https://doi.org/10.1016/j.jcjd.2022.06.010\u003c/li\u003e\n\u003cli\u003eWiddifield, J., Bombardier, C., Bernatsky, S., Paterson, J. M., Green, D., Young, J., Ivers, N., Butt, D. A., Jaakkimainen, R. L., Thorne, J. C., \u0026amp; Tu, K. (2014). An administrative data validation study of the accuracy of algorithms for identifying rheumatoid arthritis: The influence of the reference standard on algorithm performance. \u003cem\u003eBMC Musculoskeletal Disorders\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(1), 216. https://doi.org/10.1186/1471-2474-15-216\u003c/li\u003e\n\u003cli\u003eNeff, J. M., Sharp, V. L., Muldoon, J., Graham, J., Popalisky, J., \u0026amp; Gay, J. C. (2002). Identifying and classifying children with chronic conditions using administrative data with the clinical risk group classification system. \u003cem\u003eAmbulatory Pediatrics: The Official Journal of the Ambulatory Pediatric Association\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e(1), 71\u0026ndash;79. https://doi.org/10.1367/1539-4409(2002)002\u0026lt;0071:iaccwc\u0026gt;2.0.co;2\u003c/li\u003e\n\u003cli\u003eFeudtner, C., Feinstein, J. A., Zhong, W., Hall, M., \u0026amp; Dai, D. (2014). Pediatric complex chronic conditions classification system version 2: Updated for ICD-10 and complex medical technology dependence and transplantation. \u003cem\u003eBMC Pediatrics\u003c/em\u003e, \u003cem\u003e14\u003c/em\u003e, 199. https://doi.org/10.1186/1471-2431-14-199\u003c/li\u003e\n\u003cli\u003eBenchimol, E. I., Smeeth, L., Guttmann, A., Harron, K., Moher, D., Petersen, I., S\u0026oslash;rensen, H. T., von Elm, E., Langan, S. M., \u0026amp; RECORD Working Committee. (2015). The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement. \u003cem\u003ePLoS Medicine\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(10), e1001885. https://doi.org/10.1371/journal.pmed.1001885\u003c/li\u003e\n\u003cli\u003eBoyle, M. H., Georgiades, K., Duncan, L., Comeau, J., Wang, L., \u0026amp; 2014 Ontario Child Health Study Team. (2019). The 2014 Ontario Child Health Study\u0026mdash;Methodology. \u003cem\u003eThe Canadian Journal of Psychiatry\u003c/em\u003e, \u003cem\u003e64\u003c/em\u003e(4), 237\u0026ndash;245. https://doi.org/10.1177/0706743719833675 \u003c/li\u003e\n\u003cli\u003eCadarette, S. M., \u0026amp; Wong, L. (2015). An Introduction to Health Care Administrative Data. \u003cem\u003eThe Canadian Journal of Hospital Pharmacy\u003c/em\u003e, \u003cem\u003e68\u003c/em\u003e(3), 232\u0026ndash;237. https://doi.org/10.4212/cjhp.v68i3.1457\u003c/li\u003e\n\u003cli\u003eCanadian Institute for Health Information. (2022). \u003cem\u003eAn overview of physicians payments and cost per service\u003c/em\u003e. https://www.cihi.ca/en/health-workforce-in-canada-in-focus-including-nurses-and-physicians/an-overview-of-physician\u003c/li\u003e\n\u003cli\u003eSteele, L. S., Glazier, R. H., Lin, E., \u0026amp; Evans, M. (2004). Using Administrative Data to Measure Ambulatory Mental Health Service Provision in Primary Care: \u003cem\u003eMedical Care\u003c/em\u003e, \u003cem\u003e42\u003c/em\u003e(10), 960\u0026ndash;965. https://doi.org/10.1097/00005650-200410000-00004\u003c/li\u003e\n\u003cli\u003eHwee, J., Sung, L., Kwong, J. C., Sutradhar, R., Tu, K., \u0026amp; Pole, J. D. (2018). Use of physician billing claims to identify infections in children. \u003cem\u003ePLoS ONE\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(11), e0207468. https://doi.org/10.1371/journal.pone.0207468\u003c/li\u003e\n\u003cli\u003eKing-Dowling, S., Proudfoot, N. A., \u0026amp; Obeid, J. (2019). Comorbidity Among Chronic Physical Health Conditions and Neurodevelopmental Disorders in Childhood. \u003cem\u003eCurrent Developmental Disorders Reports\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(4), 248\u0026ndash;258. https://doi.org/10.1007/s40474-019-00173-x\u003c/li\u003e\n\u003cli\u003eOfford, D. R., Boyle, M. H., Fleming, J. E., Blum, H. M., \u0026amp; Rae Grant, N. I. (1989). Summary of Selected Results. \u003cem\u003eThe Canadian Journal of Psychiatry\u003c/em\u003e, \u003cem\u003e34\u003c/em\u003e(6), 483\u0026ndash;491. https://doi.org/10.1177/070674378903400602\u003c/li\u003e\n\u003cli\u003eStatistics Canada NLSCY. (2006, 2007). National Longitudinal Survey of Children and Youth (NLSCY), Cycle 8. \u003cem\u003eStatistics Canada\u003c/em\u003e. https://www23.statcan.gc.ca/imdb/p2SV.pl?Function=getSurvey\u0026amp;SDDS=4450\u003c/li\u003e\n\u003cli\u003eStatistics Canada SYC. (2012). \u003cem\u003eSurvey of Young Canadians (SYC)\u003c/em\u003e. https://www23.statcan.gc.ca/imdb/p2SV.pl?Function=getSurvey\u0026amp;SDDS=5179#a1\u003c/li\u003e\n\u003cli\u003eStatistics Canada CCHS. (2015). Canadian Community Health Survey (CCHS)\u0026mdash;2013. \u003cem\u003eCanadian Community Health Survey (CCHS) - 2013\u003c/em\u003e. Canadian Community Health Survey (CCHS) \u0026ndash; 2013\u003c/li\u003e\n\u003cli\u003eHorsman, J., Furlong, W., Feeny, D., \u0026amp; Torrance, G. (2003). The Health Utilities Index (HUI\u0026reg;): Concepts, measurement properties and applications. \u003cem\u003eHealth and Quality of Life Outcomes\u003c/em\u003e, \u003cem\u003e1\u003c/em\u003e(1), 54. https://doi.org/10.1186/1477-7525-1-54\u003c/li\u003e\n\u003cli\u003eFeng, Y., Bernier, J., McIntosh, C., \u0026amp; Orpana, H. (2009). Validation of disability categories derived from Health Utilities Index Mark 3 scores. \u003cem\u003eHealth Reports\u003c/em\u003e, \u003cem\u003e20\u003c/em\u003e(2), 43\u0026ndash;50.\u003c/li\u003e\n\u003cli\u003eThorrington, D., \u0026amp; Eames, K. (2015). Measuring Health Utilities in Children and Adolescents: A Systematic Review of the Literature. \u003cem\u003ePLOS ONE\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(8), e0135672. https://doi.org/10.1371/journal.pone.0135672\u003c/li\u003e\n\u003cli\u003eCochrane, S., Beyer, K., Clausen, M., Wjst, M., Hiller, R., Nicoletti, C., Szepfalusi, Z., Savelkoul, H., Breiteneder, H., Manios, Y., Crittenden, R., \u0026amp; Burney, P. (2009). Factors influencing the incidence and prevalence of food allergy. \u003cem\u003eAllergy\u003c/em\u003e, \u003cem\u003e64\u003c/em\u003e(9), 1246\u0026ndash;1255. https://doi.org/10.1111/j.1398-9995.2009.02128.x\u003c/li\u003e\n\u003cli\u003eGarner, R., \u0026amp; Kohen, D. (2008). Changes in the prevalence of asthma among Canadian children. \u003cem\u003eHealth Reports\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(2), 45\u0026ndash;50.\u003c/li\u003e\n\u003cli\u003eKarpa, K. D., Paul, I. M., Leckie, J. A., Shung, S., Carkaci-Salli, N., Vrana, K. E., Mauger, D., Fausnight, T., \u0026amp; Poger, J. (2012). A retrospective chart review to identify perinatal factors associated with food allergies. \u003cem\u003eNutrition Journal\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(1), 87. https://doi.org/10.1186/1475-2891-11-87\u003c/li\u003e\n\u003cli\u003eLawrence, J. M., Divers, J., Isom, S., Saydah, S., Imperatore, G., Pihoker, C., Marcovina, S. M., Mayer-Davis, E. J., Hamman, R. F., Dolan, L., Dabelea, D., Pettitt, D. J., Liese, A. D., \u0026amp; SEARCH for Diabetes in Youth Study Group. (2021). Trends in Prevalence of Type 1 and Type 2 Diabetes in Children and Adolescents in the US, 2001-2017. \u003cem\u003eJAMA\u003c/em\u003e, \u003cem\u003e326\u003c/em\u003e(8), 717. https://doi.org/10.1001/jama.2021.11165\u003c/li\u003e\n\u003cli\u003ePerret, J. L., Wurzel, D., Walters, E. H., Lowe, A. J., Lodge, C. J., Bui, D. S., Erbas, B., Bowatte, G., Russell, M. A., Thompson, B. R., Gurrin, L., Thomas, P. S., Hamilton, G., Hopper, J. L., Abramson, M. J., Chang, A. B., \u0026amp; Dharmage, S. C. (2022). Childhood \u0026ldquo;bronchitis\u0026rdquo; and respiratory outcomes in middle-age: A prospective cohort study from age 7 to 53 years. \u003cem\u003eBMJ Open Respiratory Research\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(1), e001212. https://doi.org/10.1136/bmjresp-2022-001212\u003c/li\u003e\n\u003cli\u003eShaw, T. E., Currie, G. P., Koudelka, C. W., \u0026amp; Simpson, E. L. (2011). Eczema Prevalence in the United States: Data from the 2003 National Survey of Children\u0026rsquo;s Health. \u003cem\u003eJournal of Investigative Dermatology\u003c/em\u003e, \u003cem\u003e131\u003c/em\u003e(1), 67\u0026ndash;73. https://doi.org/10.1038/jid.2010.251\u003c/li\u003e\n\u003cli\u003eSilverberg, J. I., Braunstein, M., \u0026amp; Lee-Wong, M. (2015). Association between climate factors, pollen counts, and childhood hay fever prevalence in the United States. \u003cem\u003eJournal of Allergy and Clinical Immunology\u003c/em\u003e, \u003cem\u003e135\u003c/em\u003e(2), 463-469.e5. https://doi.org/10.1016/j.jaci.2014.08.003\u003c/li\u003e\n\u003cli\u003eAhearne, C. E., Boylan, G. B., \u0026amp; Murray, D. M. (2016). Short and long term prognosis in perinatal asphyxia: An update. \u003cem\u003eWorld Journal of Clinical Pediatrics\u003c/em\u003e, \u003cem\u003e5\u003c/em\u003e(1), 67\u0026ndash;74. https://doi.org/10.5409/wjcp.v5.i1.67\u003c/li\u003e\n\u003cli\u003eBertrand, G., \u0026amp; Kaplan, C. (2014). How do we treat fetal and neonatal alloimmune thrombocytopenia? \u003cem\u003eTransfusion\u003c/em\u003e, \u003cem\u003e54\u003c/em\u003e(7), 1698\u0026ndash;1703. https://doi.org/10.1111/trf.12671\u003c/li\u003e\n\u003cli\u003eGupta, R., \u0026amp; Cabacungan, E. T. (2021). Neonatal Birth Trauma: Analysis of Yearly Trends, Risk Factors, and Outcomes. \u003cem\u003eThe Journal of Pediatrics\u003c/em\u003e, \u003cem\u003e238\u003c/em\u003e, 174-180.e3. https://doi.org/10.1016/j.jpeds.2021.06.080\u003c/li\u003e\n\u003cli\u003eStewart, A. L., \u0026amp; Reynolds, E. O. R. (1974). Improved Prognosis for Infants of Very Low Birthweight. \u003cem\u003ePediatrics\u003c/em\u003e, \u003cem\u003e54\u003c/em\u003e(6), 724\u0026ndash;735. https://doi.org/10.1542/peds.54.6.724\u003c/li\u003e\n\u003cli\u003eRector, T. S., Wickstrom, S. L., Shah, M., Thomas Greeenlee, N., Rheault, P., Rogowski, J., Freedman, V., Adams, J., \u0026amp; Escarce, J. J. (2004). Specificity and sensitivity of claims-based algorithms for identifying members of Medicare+Choice health plans that have chronic medical conditions. \u003cem\u003eHealth Services Research\u003c/em\u003e, \u003cem\u003e39\u003c/em\u003e(6 Pt 1), 1839\u0026ndash;1857. https://doi.org/10.1111/j.1475-6773.2004.00321.x\u003c/li\u003e\n\u003cli\u003eSteffler, M., Li, Y., Weir, S., Shaikh, S., Murtada, F., Wright, J. G., \u0026amp; Kantarevic, J. (2021). Trends in prevalence of chronic disease and multimorbidity in Ontario, Canada. \u003cem\u003eCanadian Medical Association Journal\u003c/em\u003e, \u003cem\u003e193\u003c/em\u003e(8), E270\u0026ndash;E277. https://doi.org/10.1503/cmaj.201473\u003c/li\u003e\n\u003cli\u003eIBM Corp. (2023). \u003cem\u003eIBM SPSS Statistics for Windows, Version 29.0.2.0\u003c/em\u003e [Computer software].\u003c/li\u003e\n\u003cli\u003eStataCorp. (2023). \u003cem\u003eStata Statistical Software: Release 18\u003c/em\u003e [Computer software]. StataCorp LLC.\u003c/li\u003e\n\u003cli\u003eCohen. (1960). \u003cem\u003eA Coefficient of Agreement for Nominal Scales\u003c/em\u003e. \u003cem\u003e20\u003c/em\u003e(1). https://doi.org/10.1177/001316446002000104\u003c/li\u003e\n\u003cli\u003eTrevethan, R. (2017). Sensitivity, Specificity, and Predictive Values: Foundations, Pliabilities, and Pitfalls in Research and Practice. \u003cem\u003eFrontiers in Public Health\u003c/em\u003e, \u003cem\u003e5\u003c/em\u003e, 307. https://doi.org/10.3389/fpubh.2017.00307\u003c/li\u003e\n\u003cli\u003eJohnson, D. R., \u0026amp; Young, R. (2011). Toward Best Practices in Analyzing Datasets with Missing Data: Comparisons and Recommendations. \u003cem\u003eJournal of Marriage and Family\u003c/em\u003e, \u003cem\u003e73\u003c/em\u003e(5), 926\u0026ndash;945. https://doi.org/10.1111/j.1741-3737.2011.00861.x\u003c/li\u003e\n\u003cli\u003eGolden, G. (2024). Mental Health Service Use among Children and Youth with Co-Occurring Health and Mental Health Disorders in Ontario. \u003cem\u003eElectronic Thesis and Dissertation Repository\u003c/em\u003e. https://ir.lib.uwo.ca/etd/10280\u003c/li\u003e\n\u003cli\u003eMuggah, E., Graves, E., Bennett, C., \u0026amp; Manuel, D. G. (2013). Ascertainment of chronic diseases using population health data: A comparison of health administrative data and patient self-report. \u003cem\u003eBMC Public Health\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(1), 16. https://doi.org/10.1186/1471-2458-13-16\u003c/li\u003e\n\u003cli\u003eO\u0026rsquo;Donnell, S., Vanderloo, S., McRae, L., Onysko, J., Patten, S. B., \u0026amp; Pelletier, L. (2016). Comparison of the estimated prevalence of mood and/or anxiety disorders in Canada between self-report and administrative data. \u003cem\u003eEpidemiology and Psychiatric Sciences\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e(4), 360\u0026ndash;369. https://doi.org/10.1017/S2045796015000463\u003c/li\u003e\n\u003cli\u003eJiang, L., Zhang, B., Smith, M. L., Lorden, A. L., Radcliff, T. A., Lorig, K., Howell, B. L.,Whitelaw, N., \u0026amp; Ory, M. G. (2015). Concordance between Self-Reports and Medicare Claims among Participants in a National Study of Chronic Disease Self-Management Program. \u003cem\u003eFrontiers in Public Health\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e, 222. https://doi.org/10.3389/fpubh.2015.00222\u003c/li\u003e\n\u003cli\u003eCanadian Pediatric Society. (2021, July). \u003cem\u003eSchedule of well-child visits\u003c/em\u003e. https://caringforkids.cps.ca/handouts/pregnancy-and-babies/schedule_of_well_child_visits\u003c/li\u003e\n\u003cli\u003eCheng, P., Gilchrist, A., Robinson, K. M., \u0026amp; Paul, L. (2009). The Risk and Consequences of Clinical Miscoding Due to Inadequate Medical Documentation: A Case Study of the Impact on Health Services Funding. \u003cem\u003eHealth Information Management Journal\u003c/em\u003e, \u003cem\u003e38\u003c/em\u003e(1), 35\u0026ndash;46. https://doi.org/10.1177/183335830903800105\u003c/li\u003e\n\u003cli\u003eOeseburg, B., Dijkstra, G. J., Groothoff, J. W., Reijneveld, S. A., \u0026amp; Jansen, D. E. M. C. (2011). Prevalence of Chronic Health Conditions in Children With Intellectual Disability: A Systematic Literature Review. \u003cem\u003eIntellectual and Developmental Disabilities\u003c/em\u003e, \u003cem\u003e49\u003c/em\u003e(2), 59\u0026ndash;85. https://doi.org/10.1352/1934-9556-49.2.59\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-pediatrics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bped","sideBox":"Learn more about [BMC Pediatrics](http://bmcpediatr.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bped/default.aspx","title":"BMC Pediatrics","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"mental health, chronic health condition, child and youth, survey data, administrative data, Ontario Health Insurance Plan, 2014 Ontario Child Health Study, algorithm","lastPublishedDoi":"10.21203/rs.3.rs-5089891/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5089891/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground\u003c/strong\u003e: Population-based studies in Canada and the United States estimate chronic physical health conditions affect between 20% to 30% of children aged 0 to 17. One challenge in measuring chronic conditions is that researchers often use inconsistent definitions. The main objective was to develop a chronic health condition (CHC) algorithm. We identified three main elements that must be incorporated from administrative data to determine whether a child has a CHC: (1) the diagnosis recorded for the visit, (2) the number of visits, and (3) within a specific reference period.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods\u003c/strong\u003e: Data were from the cross-sectional 2014 Ontario Child Health Study, linked with Ontario Health Insurance Plan (OHIP) data. Unweighted prevalence estimates and agreement analyses (Cohen’s Kappa, sensitivity, specificity) were used to compare the survey parent-reported and algorithm-based presence of a CHC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults\u003c/strong\u003e: 31.59% and 26.7% of children and youth had a CHC based on administrative and survey data, respectively. Agreement between administrative and survey data was poor (\u003cem\u003ek\u003c/em\u003e = 0.16). Among a few specific conditions, agreement varied depending on the type of condition (e.g., diabetes \u003cem\u003ek\u003c/em\u003e = 0.79 vs health conditions \u003cem\u003ek\u003c/em\u003e = 0.21).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion\u003c/strong\u003e: We found considerable discrepancies between administrative and survey-reported data. The results highlight the importance of using algorithms developed from multiple datasets to examine complex research questions, such as the measurement of chronicity.\u003c/p\u003e","manuscriptTitle":"Child and Youth Chronic Physical Health Conditions: A Comparison of Survey Data and Linked Administrative Health Data in Ontario","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-09 06:49:45","doi":"10.21203/rs.3.rs-5089891/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-09-30T06:04:33+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-09-27T15:00:59+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Pediatrics","date":"2024-09-27T14:59:53+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-pediatrics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bped","sideBox":"Learn more about [BMC Pediatrics](http://bmcpediatr.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bped/default.aspx","title":"BMC Pediatrics","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"c9a37fa1-be00-400a-b5d6-d03e5e141814","owner":[],"postedDate":"April 9th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-10-06T16:10:27+00:00","versionOfRecord":{"articleIdentity":"rs-5089891","link":"https://doi.org/10.1186/s12887-025-05747-w","journal":{"identity":"bmc-pediatrics","isVorOnly":false,"title":"BMC Pediatrics"},"publishedOn":"2025-10-02 15:58:09","publishedOnDateReadable":"October 2nd, 2025"},"versionCreatedAt":"2025-04-09 06:49:45","video":"","vorDoi":"10.1186/s12887-025-05747-w","vorDoiUrl":"https://doi.org/10.1186/s12887-025-05747-w","workflowStages":[]},"version":"v1","identity":"rs-5089891","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5089891","identity":"rs-5089891","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}