Exploratory Study of Fall Event Recording in Patients with Knee Osteoarthritis using Clinical Practice Research Datalink and Hospital Episode Statistics Linked Data

Exploratory Study of Fall Event Recording in Patients with Knee Osteoarthritis using Clinical Practice Research Datalink and Hospital Episode Statistics Linked Data | 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 Exploratory Study of Fall Event Recording in Patients with Knee Osteoarthritis using Clinical Practice Research Datalink and Hospital Episode Statistics Linked Data Aqila Taqi, Sonia Gran, Roger David Knaggs This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5385374/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 4 You are reading this latest preprint version Abstract Objective The present study explored fall recording in primary care data and in primary- secondary care linked data, to inform the data source adequacy for examining the association between analgesic use and the risk of falls in patients with knee osteoarthritis (KOA). Method Data were obtained from the clinical practice research datalink (CPRD) and Hospital Episode Statistics (HES). The study population included adults with an incident diagnosis of KOA recorded in CPRD between 2000 and 2014. The study captured the proportion of fallers in CPRD, HES or both and measured the gap between fall recording in both databases. Descriptive statistics were used report study measures Results There were 104,082 patients diagnosed with KOA in CPRD standalone data; 3,275 (3.1%) of them had a fall while HES-linked data included 57,383 and 2,384 ( 4.1%) of them had a fall recorded within one year of KOA diagnosis. Among the linked population, 1,852 (3.2%) had a fall record only in CPRD ; however, there were an additional 365 (0.6%) patients who had a fall recorded only within HES data , while167 (0.3%) patients had it recorded in both databases. There was a median of 19.5 days’ difference (IQR 4.5, 91) in fall recording between primary and secondary care data. Conclusion CPRD records included 84.7% of the overall falls experienced by patients within one year of their KOA diagnosis. However, 15.3% of the falls were only recorded in hospital data; although not a substantial proportion, this represents a group that probably suffered a serious event. Figures Figure 1 Figure 2 Figure 3 Introduction Osteoarthritis (OA) is a chronic degenerative joint disorder that affects around 500 million people worldwide and results in pain and disability that compromise quality of life. The condition can affect any joint, however, knees are most commonly affected and the prevalence increases with age[ 1 , 2 ]. Treatment guidelines recommend analgesics including paracetamol, NSAIDs and opioids or antidepressants prescribed in an incremental approach according to pain intensity and individual patient’s risk factors [ 3 ]. Although not included in many OA management guidelines, antiepileptic drugs (AEDs) are commonly prescribed for patients with OA, as an off-label use [ 4 ]. The use of analgesics is common among patients with OA (ranges between 50%-68%) in community dwelling individuals [ 5 , 6 ], which may subject them to adverse drug events such as falls. Falls are a major public health concern in the UK and many countries around the world particularly those with aging populations[ 7 ]. In fact, patients with KOA might be at an increased risk of falls due to joint pathology and chronic pain [ 8 ], nevertheless, the association between analgesic use and risk of fall in patients with KOA is understudied. Therefore, we aimed at examining the association between analgesic use and the risk of falls in patients with KOA, and to obtain accurate estimates, it was important to ensure the selection of an adequate data source to optimize the capture of fall events. Consequently, this study explored and compared fall recording within primary care clinical practice research datalink (CPRD) and secondary care hospital episode statistics (HES) datasets, to inform optimal data source to maximise fall events capture. Falls are mostly managed in primary care settings, where patients usually present following a fall. However, up to 5.3% of the UK population aged 60 years or over attend accident and emergency (A&E) departments with fall-related injuries, and up to 31% of them are hospitalised, resulting in the high cost implication of £1 billion [ 9 ].Patients with KOA who experience fall events resulting in injuries such as hip fractures may be hospitalised without necessarily presenting in primary care. Secondary care encounters are recorded in primary care records; however, a delay is expected, due to the need to feed this information into primary health records. Consequently, a number of studies have reported suboptimal recording of secondary care events in primary care records; for example, 21% of myocardial infarctions (MI) and 80% of clinically relevant bleeding events occurring in the hospital and recorded in the linked datasets did not appear in primary care records [ 10 , 11 ]. The added value of primary-secondary care linked data over single source data in estimating accurate incidence and prevalence of diseases was proved for several conditions, including infectious conditions such as community acquired pneumonia (CAP) [ 12 ] and cardiovascular events [ 10 ]. For exemple, Millet et al. (2016) reported that the identification of population-averaged incidence estimates of CAP over the period from 1997 to 2010 were 39–83% higher using primary–secondary care linked data compared to standalone primary care data. In another study, Harrett et al. (2013) found that there was a significant improvement in the crude incidence of acute MI using linked national health care sources – CPRD, HES and MINAP – with single sources underestimating the crude incidence of acute MI by 25–50% compared to the use of the entire linked sample [ 10 ]. Hence, analysis of such events and their possible management in either primary or secondary care, or in both, will benefit from using primary care data linked to secondary care data [ 13 ]. Findings from these studies suggested the use of linked data (CPRD linked with external datasets) to combine information from different sources, and provide more complete and comprehensive information on all health events and encounters. Nonetheless, the potential benefits of record linkage in maximising the capture of fall events in patients with KOA are unclear. . In summary, the importance of record linkage, particularly between primary care (e.g. CPRD) and hospitalisation records (e.g. HES) has been acknowledged by researchers, as well as policy makers, with regard to deriving accurate estimates of outcomes. Given the importance of record linkage demonstrated in several clinical conditions, the present study was devised to determine the proportion of patients with a fall recorded within one year of KOA diagnosis within CPRD or HES, or both databases, using CPRD–HES-linked patient data. Methods To determine the proportion of patients with a recorded fall within the hospitalisation records, the study used data from linkage-eligible patients. Data Sources: Data were obtained from the clinical practice research datalink (CPRD) and hospital Episode Statistics (HES). CPRD is a large primary care database that contains records of patients from general practices across the UK since 1987 [ 14 ]. Data are collected as part of the normal clinical care when patients consult their general practitioners (GPs), in participating practices. Around the study time, there were 647 practices contributing data to CPRD and records of more than 4.4 million patients were included [ 15 ]. The HES database contains information on all admissions to NHS trusts in England and data are collected during patients stay at hospital. Study Population and Follow-up Time The overall study population constituted patients with an incident diagnosis of KOA recorded in CPRD between 1st January 2000 and 31st December 2014 who were aged 18 years or older. Of the patients with an incident diagnosis of KOA, this study used the data of the CPRD patients who were registered with practices which had consented for linkage. The follow-up period started from the date of incident diagnosis of KOA (index diagnosis). The end of the follow-up period was the earliest date of the following: date of death, transfer out date, the practice’s last collection date, or the end of study follow-up period on 31st December 2015. The follow-up end date (31st December 2015) was selected to ensure at least a one-year follow-up period. Patients who had their incident diagnosis after 31st December 2014 were not included in this study. Definition of Fall in CPRD and HES In CPRD a patient had a fall if a Read code for fall was recorded Read codes are a hierarchical clinical coding system of over 80,000 terms used to record clinical data in UK primary care. Patients who had a hospital episode related to falling in HES admitted patient care (APC) records were identified using corresponding ICD-10 codes. Within HES data, the date of the first episode of care overseen by a healthcare professional was selected to represent the date of fall. The period of one year after KOA was selected to examine the association; this was based on evidence which showed that patients may discontinue, augment or switch treatments within the first year of treatment [ 16 ]. Previous research also showed that the risk of fall is greater within one month of antidepressant initiation [ 17 ]. Outcome Measures Outcome Measures Number and proportion of fallers in CPRD, HES or both Within the period of one year of index KOA diagnosis, the study identified the number and proportion of patients with a record of falling in CPRD, HES or both, and patients were grouped accordingly. The operational definition and the corresponding practical definition along with the resulting patient group are presented in Table 1 . The stepwise process of deriving these patient groups is outlined in Fig. 1 . Table 1 Definitions of Patient Groups According to Fall Records in CPRD, HES or Both Databases Operational Definition Practical Definition of Patient Group Patient Patients who had a fall recorded only in CPRD (no fall recorded in HES data) Those who had no hospitalisation record for fall, i.e. were presented and managed only in primary care Group 1 Patients who had a fall recorded only in HES data (no fall recorded in CPRD) Those who only have hospitalisation records, i.e. were hospitalised straight without presentation in primary care Group 2 Patients who had a fall recorded in both CPRD and HES data Those who had both a primary care and hospital record of fall, i.e. initially presented in primary care and were then hospitalised, or hospitalised, then records fed into GP system Group 3 Patients who had no fall record in CPRD or HES data Non-fallers in both databases: no GP record or hospitalisation record for fall Group 4 Time Gap between Fall Recordings For patients in Group 3, i.e. those who had a record of falling during the first year after KOA diagnosis within both databases, the date of first fall recording was identified, and the gap in days between the two dates was determined. Patients were grouped and categorised according to the gap (number of days) between the two dates, as presented in Table 2 . These definitions were adopted from previous work on the incidence of falling in primary care [ 18 ]. Table 2 Time Gap between Fall Recording Dates in CPRD and HES Datasets: Category and Definition Time Gap Category Definition No gap fall events in CPRD and HES were recorded on the exact same date Very short gap fall dates in CPRD and HES were within 2 days of one another Short gap fall dates in CPRD and HES were within 7 days of one another Intermediate gap fall dates in CPRD and HES were within 14 days of one another Long gap fall dates in CPRD and HES were within 30 days of one another Prolonged gap1 fall dates in CPRD and HES were within 60 days of one another Prolonged gap2 fall dates in CPRD and HES were within 90 days of one another Prolonged gap3 fall dates in CPRD and HES were over 90 days apart Data Management The HES files were provided in .txt format and were imported into a statistical software package (STATA 15.2) for further analysis. The records were checked for inconsistencies in dates; for example, discharge date before admission date or episode start date after discharge date were identified and dropped. Statistical Analysis Descriptive analysis of demographic and clinical characteristics of the study population was performed. The study granted approval by the CPRD Independent Scientific Advisory Committee (ISAC) protocol number (protocol number 18_170R). Results The whole Study Population The overall study population constituted of a total of 108,221 patients with an incident diagnosis of KOA in CPRD, however, 4,139 of them were excluded as their earliest KOA diagnosis date within CPRD records was after 31th December 2014. The remaining 104,082 patients from 660 practices across the UK constituted the whole study population which was selected from CPRD standalone data. A total of 45,678 patients were aged 40–64 years at the start of follow-up (43.9%), while 40,305 (38.7%) of them were aged 65–80 years. Females constituted 58.7% of the study population (n = 61,057) (Table 3 ). HES-Linked Population A total of 59,737 patients diagnosed with KOA were eligible for linkage from the 390 consenting English practices; 2,354 (3.9%) of them were excluded due to missing discharge dates or having inconsistent dates (episode start date later than discharge date, or episode start date later than episode end date). This yielded 57,383 HES-linked patients for further analyses, representing 55.1% of the total patients with a diagnosis of KOA (N = 104,082) as illustrated in Fig. 2 . The demographic and socioeconomic characteristics of the HES-linked patients were largely similar to those of the whole study population (Table 3 ). Table 3 Characteristics of the Study Population Whole Study Population (CPRD standalone data) HES-Linked Patients (CPRD-HES linked data) Number of patients 104,082 57,383 (55.1%) Gender Males 43,025 (41.3%) 23,352 (40.7%) Females 61,057 (58.7%) 34,031 (59.3%) Age in years* Mean (± SD) 66.30 (± 12.76) 67.04 (± 12.82) Range (18.03-106.37) (18.03-104.72) Age ranks, years (% from total) 80 15,956 (15.3%) 9,670 (16.9%) IMD score (% from total) 1 (least deprived) 14,176 (21.6%) 12,100 (21.9%) 2 15,456 (23.5%) 13,425 (23.4%) 3 13,842 (21.2%) 12,153 (21.2%) 4 12,610 (19.2%) 11,062 (19.2%) 5 (most deprived) 9,696 (14.8%) 8,620 (15.0%) * Calculated at KOA diagnosis; SD – standard deviation; IMD – index of multiple deprivation Proportion of Patients with a Recorded Fall in CPRD, HES or Both Within CPRD the whole study population consisted of 104,082 patients, and 3,275 (3.1%) of them had a fall recorded within one year of KOA diagnosis. However, using the HES-linked population (n = 57,383), the number of patients with a fall recorded within one year of KOA diagnosis was 2,384 , representing 4.1% of the total linked patients. Subsequently, the process of identifying each of the four patient groups was followed and the results are summarised in Fig. 2 . Within the linked population (n = 57,383), the number of patients who had a fall record only in CPRD (group 1) was 1,852 (3.2%); however, there were an additional 365 (0.6%) patients who had a fall recorded only within HES data (group 2). There were 167 (0.3%) patients who had a record in both databases (group 3) (Fig. 3 ). Characteristics of Patients with a Fall Record Identified in CPRD, HES or Both The demographic, socioeconomic and clinical characteristics of patients who had a fall record identified in CPRD, HES or both databases were analysed and presented in Table 4 . Table 4 Demographic and Clinical Characteristics of Patients with Fall Records among the HES-linked Population Patient Characteristics Patients with a Fall Record Identified in Respective Database(s), n = 2,384 Only CPRD Only HES CPRD & HES Number of patients (%) from total 1,852 (77.7) 365 (15.3) 167 (7.0) Females (%) 1,391 (75.1) 229 (62.7) 135 (80.8) Age Mean (± SD years) 76.2 (± 11.87) 76.1 (± 12.26) 81.6 (± 10.94) Age range 33.1- 102.7 30.2–103.4 36.9–104.7 Age group, years 80 821 (44.3) 163(44.7) 111(66.5) IMD Score 1 361 (19.5) 75 (20.7) 28 (16.8) 2 430 (23.2) 91(25.1) 41 (24.6) 3 406 (21.9) 75 (20.7) 39 (23.4) 4 386 (20.9) 64 (17.6) 32 (19.2) 5 268 (14.5) 58 (16.0) 27 (16.1) Comorbidity 433 (23.4) 58 (23.3) 38 (22.8) FRIDs 1,398 (75.5) 261 (71.5) 132 (79.0) Previous fall 253 (13.7) 79 (19.2) 26 (15.6) SD standard deviation; IMD score index of multiple deprivation; FRID fall risk increasing drugs Comparison of Fall Recording Dates Records of the patients who had a fall recorded in both databases (CPRD and HES) were further analysed to determine the time gap between the two dates of falling. The date of the first fall after the diagnosis of KOA was selected for the comparison, and the results are summarised in Table 5 . Table 5 Gap between Fall Recording Dates in CPRD and HES and Number of Patients in Each Time Gap Category (Days of Gap in Recording between CPRD and HES) Number of Patients with a Record of Fall (n = 167) No gap (same recording date in CPRD and HES) 45 (26.9%) Very short gap (≤ 2days) 19 (11.3%) Short gap (> 2 days and ≤ 7days) 17 (10.2%) Intermediate gap (> 7 days and ≤ 14 days) 9 (5.4%) Long gap (> 14 days and ≤ 30 days) 16 (9.6%) Prolonged gap 1 (> 30 and ≤ 60 days) 15 (9.0%) Prolonged gap 2 (> 60 and ≤ 90 days) 7(4.2%) Prolonged gap 3 (> 90 days) 39 (23.3%) Number of falls first recorded in HES 58 (34.7%) Number of falls first recorded in CPRD 64 (38.4%) For patients who had a record of fall first recorded in primary care (n = 64), there was a median of 19.5 days’ difference (IQR 4.5, 91) in the recording between primary and secondary care data for fall events. Discussion Main findings This study explored fall event recording in patients with KOA within CPRD, HES, and in both databases. A total of 2,384 falls were identified within the CPRD–HES-linked data; the majority of them had a record in CPRD (2,019 falls, representing 84.6% of all falls). However, a small proportion of the total identified fall events were only recorded in HES (365 falls, representing 15% of all falls). Among those with a fall record in both databases within one year of KOA diagnosis, fall dates were within a month for 64% of the patients, while for 23.3% there were more than 90 days between the dates. Proportion of Fallers within the Whole Study Population vs HES-Linked Population Unlike the whole study population, which included 104,082 patients with a diagnosis of KOA, who were registered with GP practices across the UK, HES-linked practices reside only in England. Hence, the number of linked patients was approximately 55% (57,383 patients) of the total diagnosed with KOA. The proportion of patients with a recorded fall within the HES-linked population was 4.1% (2,384 falls in the 57,383 HES-linked population), and was slightly higher than the proportion within the CPRD standalone data (3.1% patients with falls recorded among 104,082 patients with KOA). Nearly 85% (n = 2,019) of all falls were recorded in CPRD, while 22.3% (n = 532) were recorded in HES data. There were 1,852 patients who were identified only within CPRD (77.7% of all falls) compared to 11.1% identified only within HES. This was in line with findings from a previous study on the prevalence of injury in England among children and young people using record linkage. The study reported that 75% of total fractures were recorded only in CPRD data, compared to 8.6% recorded only in HES data (n = 139,662 and 15,972) [ 19 ]. Results showed that linkage to HES data did not substantially increase the number of identified fallers; in fact, a mere 365 fall events were only recorded in HES, and would have been missed if HES records were not explored. These figures suggest a limited role for record linkage in improving the sensitivity of fall identification. Fall events are almost entirely recorded in primary care, and even those which are directly managed in hospitals (group 2) are eventually recorded in primary care. The substantial overlap in recording could be explained by the nature of both the KOA itself and the outcome of the fall. KOA is a chronic condition which is largely managed in primary care. It is, therefore, common that patients with KOA present, are diagnosed, investigated, followed-up and managed in primary care settings. This is reflected in the annual number of GP visits, where it is estimated that 20% of the annual GP visits in the UK are for musculoskeletal conditions, including KOA [ 20 ] Patients with KOA may have frequent visits to their GPs (for follow-up of KOA or care for other comorbidities), who are likely to record encountered health events such as falls during these routine consultations [ 21 , 22 ]. General Practitioners and other primary care professionals’ key roles are in identifying those who have had a fall, and referring those who are at risk of falling to specialised fall prevention services, for multifactorial assessment and intervention. The components of multifactorial risk assessment and multifactorial intervention are summarised in Table 6. Table ‎6 Key Components of the Multifactorial Risk Assessment and Multifactorial Intervention Multifactorial Risk Assessment Examines the Following: Multifactorial Intervention should: Fall history Provide strength and balance training Gait, balance mobility and muscle strength Address home hazards Osteoporosis/fracture risk Correct visual risk factors Functional ability, fear of falling Provide education Vision, cognition and neurological assessment Address specific underlying medical problems Urinary incontinence Home hazards Cardiovascular problems Medications Although record linkage has shown value in driving accurate (unbiased) assessments of incidence and prevalence of CAP in elderly people and MI events [ 10 , 12 ], it should, however, be noted that these findings are likely to depend on the disease area [ 11 ]. For instance, cancer case recording between CPRD and cancer registries was found to have a high level of concordance, with > 80% of cases being present in both databases [ 23 ]. This level of concordance was achieved despite cancer being managed at different care settings, and one would expect a high improvement in sensitivity with record linkage. Similarly, a high concordance between primary and secondary care records was observed in this study for the record of falls. The limited role of record linkage in improving the sensitivity of diagnoses of conditions or outcomes largely managed in primary care settings was also suggested by researchers in primary care. Yu et al. (2018), in their work on estimating the extent of recording of OA in CPRD – using patients who underwent a total hip replacement (THR) or total knee replacement (TKR) as the reference population – found that OA was under-recorded in CPRD, but suggested that it is unlikely that linkage to hospital data will substantially improve the number of identified patients, since it is a condition that is largely managed in general practice [ 24 ]. Record linkage is recommended in studies aiming to estimate the incidence or prevalence of diseases, particularly those which are managed in primary as well as secondary care settings. For conditions that are managed entirely in primary care, record linkage may not be necessary [ 11 ]; however, it must be explored prior to making any final decisions on the use of the HES-linked population for incidence estimation. In this study, the existence of a fall record in CPRD, HES or both databases suggests the occurrence of falls of varying severity; for example, patients in group 3 are probably those who presented at a GP surgery, but required hospitalisation for further management of the fall they experienced. They could also be those who were hospitalised for a fall, and their GP records were updated. On the other hand, patients constituting group 2 may represent those who were hospitalised without presentation at their GP surgery, with their records not fed into the GP system, or those who have died, and hence have not returned to their GP. Such cases may represent a group of fallers who experienced a serious fall. Group 1 represents patients who required medical attention (GP care) but were probably managed at the GP surgery, and did not require escalated care in a hospital. Study Implications Implication for Future Work of this Research fall events identified for HES-linked patients included a subset of patients for whom escalated care/hospital care was required as a result of a probable severe fall-related injury. And therefore, the use of an HES-linked population was considered for further work on the association between analgesic use and the risk of falls among patients with KOA. Implications of Study Findings for Research in General This study showed that record linkage has a limited role in improving the sensitivity of fall diagnosis in patients with KOA. It had demonstrated an example of the events that are likely to be recorded completely in primary care, hence informing researchers on the completeness of primary care records for falls of patients with KOA. Implications for Practice and Policy This study has demonstrated an example of the completeness of primary care records for falls. This knowledge is essential for health services planners, commissioners and other decision makers, as CPRD data can be reliably used for service planning purposes when required. Information provided by this study on the frequency of falling, and characteristics of those who experienced a fall, could be made available to inform fall risk prevention programmes about possible groups of people within those diagnosed with KOA who are most likely to experience a fall. Falls are a growing public health challenge and it is estimated that the prevalence of falls in the older people of the world was 26.5% (95% CI 23.4%-29.8%) [ 25 ].[ 26 ]. [ 27 ]. In the UK, falls are estimated to cost the NHS more than £2.3 billion per year. Falls are the leading cause of injury-related hospitalisation, and exert a burden on hospital services, in terms of both A&E visits and inpatient admissions [ 28 ]. Consequently, fall prevention is one of the three priorities for optimisation highlighted in the NHS RightCare Falls and Fragility Fractures pathway, in addition to detecting and managing osteoporosis and fracture risk, and optimal support following a fragility fracture [ 29 ]. Although CPRD can be used as a standalone data source for information on the absolute number of patients with falls requiring medical care from GPs, there were some patients (n = 365) with more serious falls which required hospitalisation for more intense medical care. Observing changes in the proportion of those hospitalised following falls may inform future resource planning and hospital bed demand/allocation planning for potential fall injury-related hospitalisation. Any reduction in the proportion of those hospitalised may serve as a quality improvement indicator for fall prevention or fracture liaison services. Strengths and Limitations The study used data from both primary care and secondary care, and maximised the number of identified fall cases using code lists which were applied by other researchers in published work [ 8 , 30 ]. However, the study has some limitations that should be mentioned. Falls identified through electronic health records (EHR) represent just those for which some form of medical attention was required, It was revealed that up to 30% of participants in a population-based survey reported at least one fall in the prior 12 months [ 31 ]. Hence, compared to surveys, the number of falls within the EHR database is potentially underestimated. This study did not include codes for fall-related injuries, resulting in possible underestimation of the true number of falls in HES data (fall events may have been coded with the resulting injury codes, rather than fall codes). Additionally, there is the potential for under- or over-estimation of falls in HES due to inaccurate coding. Conclusion This study has explored the potential benefits of record linkage in improving the sensitivity of fall identification. Using the CPRD–HES-linked patient data, this study found that CPRD records included approximately 84.7% of the overall falls experienced by patients within one year of their KOA diagnosis. However, 15.3% of the falls were only recorded in hospital data; although not a substantial proportion, this represents a group of patients who probably suffered a serious event. The study addressed the concern that EHR from one part of the health system, such as primary care, may not capture health events occurring in other parts of the system, such as hospital care Declarations Funding The authors declare that no funds, grants, or other support were received during the preparation of this manuscript. Competing Interests The authors have no relevant financial or non-financial interests to disclose Disclosures and declarations The authors declare no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work; no other relationships or activities that could appear to have influenced the submitted work. Data Availability Statement The data that support the findings of this study are owned by the CPRD and may be available subject to data sharing regulations of the CPRD Human Ethics and Consent to Participate declarations Human Ethics and Consent to Participate declarations: Not applicable Ethics Approval declaration The study was approved by the CPRD Independent Scientific Advisory Committee (ISAC) (protocol number 18_170R). Consent to Participate declarations: Not applicable. Clinical trial number : Not applicable. References J. G. Quicke, P. G. Conaghan, N. Corp, and G. Peat, "Osteoarthritis year in review 2021: epidemiology & therapy," Osteoarthritis and Cartilage, vol. 30, no. 2, pp. 196-206, 2022, doi: 10.1016/j.joca.2021.10.003. Q. Yao et al. , "Osteoarthritis: pathogenic signaling pathways and therapeutic targets," (in eng), Signal Transduct Target Ther, vol. 8, no. 1, p. 56, Feb 3 2023, doi: 10.1038/s41392-023-01330-w. NICE. "Osteoarthritis: Care and Management." National Institute for Clinical and Care Excellence. https://www.nice.org.uk/guidance/CG177 (accessed 15 July, 2020). A. Taqi, S. Gran, and R. D. Knaggs, "Analgesic utilization in people with knee osteoarthritis: A population-based study using primary care data," (in eng), Pain Pract, vol. 23, no. 5, pp. 523-534, Jun 2023, doi: 10.1111/papr.13212. N. Wilson et al. , "Drug utilization in patients with OA: a population-based study," Rheumatology, vol. 54, no. 5, pp. 860-867, 2015, doi: 10.1093/rheumatology/keu403. S. R. Kingsbury, E. M. A. Hensor, C. A. E. Walsh, M. C. Hochberg, and P. G. Conaghan, "How do people with knee osteoarthritis use osteoarthritis pain medications and does this change over time? Data from the Osteoarthritis Initiative," (in eng), Arthritis Res Ther, vol. 15, no. 5, pp. R106-R106, 2013, doi: 10.1186/ar4286. S. L. James et al. , "The global burden of falls: global, regional and national estimates of morbidity and mortality from the Global Burden of Disease Study 2017," (in eng), Inj Prev, vol. 26, no. Supp 1, pp. i3-i11, Oct 2020, doi: 10.1136/injuryprev-2019-043286. N. K. Arden et al. , "Knee pain, knee osteoarthritis, and the risk of fracture," Arthritis Care & Research, vol. 55, no. 4, pp. 610-615, 2006/08/15 2006, doi: 10.1002/art.22088. P. Scuffham, S. Chaplin, and R. Legood, "Incidence and costs of unintentional falls in older people in the United Kingdom," Journal of Epidemiology and Community Health, vol. 57, no. 9, p. 740, 2003, doi: 10.1136/jech.57.9.740. E. Herrett et al. , "Completeness and diagnostic validity of recording acute myocardial infarction events in primary care, hospital care, disease registry, and national mortality records: cohort study," (in eng), BMJ (Clinical research ed.), vol. 346, pp. f2350-f2350, 2013, doi: 10.1136/bmj.f2350. L. McDonald, C. J. Sammon, M. Samnaliev, and S. Ramagopalan, "Under-recording of hospital bleeding events in UK primary care: a linked Clinical Practice Research Datalink and Hospital Episode Statistics study," (in eng), Clinical epidemiology, vol. 10, pp. 1155-1168, 2018, doi: 10.2147/CLEP.S170304. E. R. C. Millett, J. K. Quint, B. L. De Stavola, L. Smeeth, and S. L. Thomas, "Improved incidence estimates from linked vs. stand-alone electronic health records," (in eng), Journal of clinical epidemiology, vol. 75, pp. 66-69, 2016, doi: 10.1016/j.jclinepi.2016.01.005. K. J. Rothnie et al. , "Validation of the Recording of Acute Exacerbations of COPD in UK Primary Care Electronic Healthcare Records," PLOS ONE, vol. 11, no. 3, p. e0151357, 2016, doi: 10.1371/journal.pone.0151357. E. Herrett et al. , "Data Resource Profile: Clinical Practice Research Datalink (CPRD)," (in eng), Int J Epidemiol, vol. 44, no. 3, pp. 827-836, 2015, doi: 10.1093/ije/dyv098. E. Herrett, S. L. Thomas, W. M. Schoonen, L. Smeeth, and A. J. Hall, "Validation and validity of diagnoses in the General Practice Research Database: a systematic review," British Journal of Clinical Pharmacology, vol. 69, no. 1, pp. 4-14, 04/09/received 08/06/accepted 2010, doi: 10.1111/j.1365-2125.2009.03537.x. M. Gore, A. B. Sadosky, D. L. Leslie, K.-S. Tai, and P. Emery, "Therapy Switching, Augmentation, and Discontinuation in Patients with Osteoarthritis and Chronic Low Back Pain," Pain Practice, vol. 12, no. 6, pp. 457-468, 2012/07/01 2012, doi: 10.1111/j.1533-2500.2011.00524.x. C. Coupland, P. Dhiman, R. Morriss, A. Arthur, G. Barton, and J. Hippisley-Cox, "Antidepressant use and risk of adverse outcomes in older people: population based cohort study," The BMJ, vol. 343, p. d4551, 08/02 06/13/accepted 2011, doi: 10.1136/bmj.d4551. J. GRIBBIN, "FALLS IN OLDER PEOPLE AND THE ROLE OF COMMONLY PRESCRIBED ANTIDEPRESSANT AND ANTIHYPERTENSIVE MEDICATIONS " PhD School of Medicine University of Nottingham UK 2013. [Online]. Available: http://eprints.nottingham.ac.uk/28451/1/594820%20minus%20articles.pdf R. Baker, L. J. Tata, D. Kendrick, and E. Orton, "Identification of incident poisoning, fracture and burn events using linked primary care, secondary care and mortality data from England: implications for research and surveillance," Injury Prevention, vol. 22, no. 1, p. 59, 2016, doi: 10.1136/injuryprev-2015-041561. VersusArthritis. "The State of Musculoskeletal Health 2019." Versus Arthritis. https://www.versusarthritis.org/about-arthritis/data-and-statistics/state-of-musculoskeletal-health-2019/ (accessed 30 November 2019). NICE. "Falls in older people: assessing risk and prevention." https://www.nice.org.uk/guidance/cg161 (accessed 15th November 2019). NICE. "Falls in older people." https://www.nice.org.uk/guidance/qs86 (accessed 15 November 2019). R. Boggon, T. P. van Staa, M. Chapman, A. M. Gallagher, T. A. Hammad, and M. A. Richards, "Cancer recording and mortality in the General Practice Research Database and linked cancer registries," Pharmacoepidemiology and Drug Safety, vol. 22, no. 2, pp. 168-175, 2013/02/01 2013, doi: 10.1002/pds.3374. D. Yu, K. P. Jordan, and G. Peat, "Underrecording of osteoarthritis in United Kingdom primary care electronic health record data," (in eng), Clinical epidemiology, vol. 10, pp. 1195-1201, 2018, doi: 10.2147/CLEP.S160059. N. Salari, N. Darvishi, M. Ahmadipanah, S. Shohaimi, and M. Mohammadi, "Global prevalence of falls in the older adults: a comprehensive systematic review and meta-analysis," (in eng), J Orthop Surg Res, vol. 17, no. 1, p. 334, Jun 28 2022, doi: 10.1186/s13018-022-03222-1. ONS. "Overview of the UK population." https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/articles/overvie woftheukpopulation/august2019 (accessed 30 November 2019). J. T. Yang Tian, David Buck, Lara Sonola, "Exploring the system-wide costs of falls in older people in Torbay," The King's Fund London 2013. Accessed: 15th November 2019 [Online]. Available: https://www.kingsfund.org.uk/sites/default/files/field/field_publication_file/exploring-system-wide-costs-of-falls-in-torbay-kingsfund-aug13.pdf BGS, "NICE impact report on falls and fragility fractures," National Institute fo Health and Care Excellence British Geriatrics Society, 2018. Accessed: 15th November 2019. [Online]. Available: https://www.bgs.org.uk/resources/2018-nice-impact-report-on-falls-and-fragility-fractures NHS. "RightCare Pathway: Falls and Fragility Fractures." https://www.england.nhs.uk/rightcare/wp-content/uploads/sites/40/2017/12/falls-fragility-fractures-pathway-v18.pdf (accessed 15th November 2019). A. Taqi, S. Gran, and R. D. Knaggs, ""Application of five different strategies to define a cohort of patients with knee osteoarthritis in a large primary care database"," (in eng), J Eval Clin Pract, Jun 25 2024, doi: 10.1111/jep.14045. T. Gill, A. W. Taylor, and A. Pengelly, "A Population-Based Survey of Factors Relating to the Prevalence of Falls in Older People," Gerontology, vol. 51, no. 5, pp. 340-345, 2005, doi: 10.1159/000086372. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 14 Nov, 2024 Editor assigned by journal 11 Nov, 2024 Submission checks completed at journal 11 Nov, 2024 First submitted to journal 04 Nov, 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-5385374","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":378210933,"identity":"7e9c04a6-64fd-4e0f-8d67-c14286602ae9","order_by":0,"name":"Aqila Taqi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAx0lEQVRIiWNgGAWjYLCCBww2EAYP0VoSGNJI13KYBC387ckHHyTUnLfnu5HA+OBtG4OcbgMBLRJnniUbJBy7nTjzRgKz4dw2BmOzA4SsuZFjJpHAdjvB4EYCmzRvG0PiNkJa5G/kf/+R8O+cPVAL+2+itBjcyGFjSGw7wLgBaAszUVoMzzwzlkjsS06ceeZhs+SccxKE/SJ3PPnhhw/f7Oz5jicf/PCmzEaOsPeBkQIBBxgbgKQEQfXIWohRPApGwSgYBSMSAADQlUW4Ph/XHAAAAABJRU5ErkJggg==","orcid":"","institution":"University of Nottingham","correspondingAuthor":true,"prefix":"","firstName":"Aqila","middleName":"","lastName":"Taqi","suffix":""},{"id":378210936,"identity":"772a3432-d3fe-4af9-b30a-f9f602f07d37","order_by":1,"name":"Sonia Gran","email":"","orcid":"","institution":"University of Nottingham","correspondingAuthor":false,"prefix":"","firstName":"Sonia","middleName":"","lastName":"Gran","suffix":""},{"id":378210940,"identity":"3d4e8c61-bfcd-4856-9a93-462a96d2eff7","order_by":2,"name":"Roger David Knaggs","email":"","orcid":"","institution":"University of Nottingham","correspondingAuthor":false,"prefix":"","firstName":"Roger","middleName":"David","lastName":"Knaggs","suffix":""}],"badges":[],"createdAt":"2024-11-04 06:38:29","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5385374/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5385374/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":71560948,"identity":"18e56e30-8361-4b84-aa5c-f16c51e85e09","added_by":"auto","created_at":"2024-12-16 16:55:55","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":542997,"visible":true,"origin":"","legend":"\u003cp\u003eFlow Diagram Outlining the Process of Patient Group Identification According to the Existence of Fall\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-5385374/v1/1bd99e89699177f8f554e93e.png"},{"id":71560945,"identity":"d38cb1c6-2149-4fd7-a4dd-e83074009240","added_by":"auto","created_at":"2024-12-16 16:55:55","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":58738,"visible":true,"origin":"","legend":"\u003cp\u003eFlow Diagram Outlining the Process of Patient Group Identification According to the Existence of a Fall Record in CPRD, HES or Both.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-5385374/v1/a57b706fe5aabf0ebdf8f3a7.png"},{"id":71560946,"identity":"49739e88-6df6-45bb-8ac3-1f2d96b276e6","added_by":"auto","created_at":"2024-12-16 16:55:55","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":640092,"visible":true,"origin":"","legend":"\u003cp\u003eNumber of Patients with a Fall Record Identified within Primary Care (CPRD Standalone Data) and Secondary Care (HES).\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-5385374/v1/e0e4c8aa9543c5da18b8b47c.png"},{"id":71561511,"identity":"7746d0cb-7c80-49b4-847e-1534ab71ec18","added_by":"auto","created_at":"2024-12-16 17:03:55","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2038780,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5385374/v1/5ecf01d9-4635-40bb-993d-340a14ab9ffb.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Exploratory Study of Fall Event Recording in Patients with Knee Osteoarthritis using Clinical Practice Research Datalink and Hospital Episode Statistics Linked Data","fulltext":[{"header":"Introduction","content":"\u003cp\u003eOsteoarthritis (OA) is a chronic degenerative joint disorder that affects around 500\u0026nbsp;million people worldwide and results in pain and disability that compromise quality of life. The condition can affect any joint, however, knees are most commonly affected and the prevalence increases with age[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eTreatment guidelines recommend analgesics including paracetamol, NSAIDs and opioids or antidepressants prescribed in an incremental approach according to pain intensity and individual patient\u0026rsquo;s risk factors [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Although not included in many OA management guidelines, antiepileptic drugs (AEDs) are commonly prescribed for patients with OA, as an off-label use [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe use of analgesics is common among patients with OA (ranges between 50%-68%) in community dwelling individuals [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e], which may subject them to adverse drug events such as falls. Falls are a major public health concern in the UK and many countries around the world particularly those with aging populations[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. In fact, patients with KOA might be at an increased risk of falls due to joint pathology and chronic pain [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], nevertheless, the association between analgesic use and risk of fall in patients with KOA is understudied.\u003c/p\u003e \u003cp\u003eTherefore, we aimed at examining the association between analgesic use and the risk of falls in patients with KOA, and to obtain accurate estimates, it was important to ensure the selection of an adequate data source to optimize the capture of fall events. Consequently, this study explored and compared fall recording within primary care clinical practice research datalink (CPRD) and secondary care hospital episode statistics (HES) datasets, to inform optimal data source to maximise fall events capture.\u003c/p\u003e \u003cp\u003eFalls are mostly managed in primary care settings, where patients usually present following a fall. However, up to 5.3% of the UK population aged 60 years or over attend accident and emergency (A\u0026amp;E) departments with fall-related injuries, and up to 31% of them are hospitalised, resulting in the high cost implication of \u0026pound;1\u0026nbsp;billion [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e].Patients with KOA who experience fall events resulting in injuries such as hip fractures may be hospitalised without necessarily presenting in primary care. Secondary care encounters are recorded in primary care records; however, a delay is expected, due to the need to feed this information into primary health records. Consequently, a number of studies have reported suboptimal recording of secondary care events in primary care records; for example, 21% of myocardial infarctions (MI) and 80% of clinically relevant bleeding events occurring in the hospital and recorded in the linked datasets did not appear in primary care records [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe added value of primary-secondary care linked data over single source data in estimating accurate incidence and prevalence of diseases was proved for several conditions, including infectious conditions such as community acquired pneumonia (CAP) [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] and cardiovascular events [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eFor exemple, Millet et al. (2016) reported that the identification of population-averaged incidence estimates of CAP over the period from 1997 to 2010 were 39\u0026ndash;83% higher using primary\u0026ndash;secondary care linked data compared to standalone primary care data. In another study, Harrett et al. (2013) found that there was a significant improvement in the crude incidence of acute MI using linked national health care sources \u0026ndash; CPRD, HES and MINAP \u0026ndash; with single sources underestimating the crude incidence of acute MI by 25\u0026ndash;50% compared to the use of the entire linked sample [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Hence, analysis of such events and their possible management in either primary or secondary care, or in both, will benefit from using primary care data linked to secondary care data [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Findings from these studies suggested the use of linked data (CPRD linked with external datasets) to combine information from different sources, and provide more complete and comprehensive information on all health events and encounters. Nonetheless, the potential benefits of record linkage in maximising the capture of fall events in patients with KOA are unclear.\u003c/p\u003e \u003cp\u003e.\u003c/p\u003e \u003cp\u003eIn summary, the importance of record linkage, particularly between primary care (e.g. CPRD) and hospitalisation records (e.g. HES) has been acknowledged by researchers, as well as policy makers, with regard to deriving accurate estimates of outcomes. Given the importance of record linkage demonstrated in several clinical conditions, the present study was devised to determine the proportion of patients with a fall recorded within one year of KOA diagnosis within CPRD or HES, or both databases, using CPRD\u0026ndash;HES-linked patient data.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eTo determine the proportion of patients with a recorded fall within the hospitalisation records, the study used data from linkage-eligible patients.\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eData Sources:\u003c/h2\u003e \u003cp\u003eData were obtained from the clinical practice research datalink (CPRD) and hospital Episode Statistics (HES). CPRD is a large primary care database that contains records of patients from general practices across the UK since 1987 [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Data are collected as part of the normal clinical care when patients consult their general practitioners (GPs), in participating practices. Around the study time, there were 647 practices contributing data to CPRD and records of more than 4.4\u0026nbsp;million patients were included [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. The HES database contains information on all admissions to NHS trusts in England and data are collected during patients stay at hospital.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eStudy Population and Follow-up Time\u003c/h3\u003e\n\u003cp\u003eThe overall study population constituted patients with an incident diagnosis of KOA recorded in CPRD between 1st January 2000 and 31st December 2014 who were aged 18 years or older. Of the patients with an incident diagnosis of KOA, this study used the data of the CPRD patients who were registered with practices which had consented for linkage.\u003c/p\u003e \u003cp\u003eThe follow-up period started from the date of incident diagnosis of KOA (index diagnosis). The end of the follow-up period was the earliest date of the following: date of death, transfer out date, the practice\u0026rsquo;s last collection date, or the end of study follow-up period on 31st December 2015. The follow-up end date (31st December 2015) was selected to ensure at least a one-year follow-up period. Patients who had their incident diagnosis after 31st December 2014 were not included in this study.\u003c/p\u003e\n\u003ch3\u003eDefinition of Fall in CPRD and HES\u003c/h3\u003e\n\u003cp\u003eIn CPRD a patient had a fall if a Read code for fall was recorded Read codes are a hierarchical clinical coding system of over 80,000 terms used to record clinical data in UK primary care. Patients who had a hospital episode related to falling in HES admitted patient care (APC) records were identified using corresponding ICD-10 codes. Within HES data, the date of the first episode of care overseen by a healthcare professional was selected to represent the date of fall.\u003c/p\u003e \u003cp\u003eThe period of one year after KOA was selected to examine the association; this was based on evidence which showed that patients may discontinue, augment or switch treatments within the first year of treatment [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Previous research also showed that the risk of fall is greater within one month of antidepressant initiation [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e].\u003c/p\u003e\n\u003ch3\u003eOutcome Measures\u003c/h3\u003e\n\u003cdiv class=\"Heading\"\u003eOutcome Measures\u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eNumber and proportion of fallers in CPRD, HES or both\u003c/h2\u003e \u003cp\u003eWithin the period of one year of index KOA diagnosis, the study identified the number and proportion of patients with a record of falling in CPRD, HES or both, and patients were grouped accordingly. The operational definition and the corresponding practical definition along with the resulting patient group are presented in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. The stepwise process of deriving these patient groups is outlined in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDefinitions of Patient Groups According to Fall Records in CPRD, HES or Both Databases\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOperational Definition\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePractical Definition of Patient Group\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePatient\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePatients who had a fall recorded only in CPRD\u003c/p\u003e \u003cp\u003e(no fall recorded in HES data)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eThose who had \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eno hospitalisation\u003c/span\u003e record for fall,\u003c/p\u003e \u003cp\u003ei.e. were presented and managed only in primary care\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGroup 1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePatients who had a fall recorded only in HES data\u003c/p\u003e \u003cp\u003e(no fall recorded in CPRD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eThose who only have hospitalisation records, i.e. were hospitalised straight without presentation in primary care\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGroup 2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePatients who had a fall recorded in both CPRD and HES data\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eThose who had both a primary care and hospital record of fall,\u003c/p\u003e \u003cp\u003ei.e. initially presented in primary care and were then hospitalised, or\u003c/p\u003e \u003cp\u003ehospitalised, then records fed into GP system\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGroup 3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePatients who had no fall record in CPRD or HES data\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNon-fallers in both databases: no GP record or hospitalisation record for fall\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGroup 4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eTime Gap between Fall Recordings\u003c/h2\u003e \u003cp\u003eFor patients in Group 3, i.e. those who had a record of falling during the first year after KOA diagnosis within both databases, the date of first fall recording was identified, and the gap in days between the two dates was determined. Patients were grouped and categorised according to the gap (number of days) between the two dates, as presented in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. These definitions were adopted from previous work on the incidence of falling in primary care [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e].\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eTime Gap between Fall Recording Dates in CPRD and HES Datasets: Category and Definition\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTime Gap Category\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDefinition\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo gap\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003efall events in CPRD and HES were recorded on the exact same date\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVery short gap\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003efall dates in CPRD and HES were within 2 days of one another\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShort gap\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003efall dates in CPRD and HES were within 7 days of one another\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIntermediate gap\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003efall dates in CPRD and HES were within 14 days of one another\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLong gap\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003efall dates in CPRD and HES were within 30 days of one another\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProlonged gap1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003efall dates in CPRD and HES were within 60 days of one another\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProlonged gap2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003efall dates in CPRD and HES were within 90 days of one another\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProlonged gap3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003efall dates in CPRD and HES were over 90 days apart\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eData Management\u003c/h3\u003e\n\u003cp\u003eThe HES files were provided in .txt format and were imported into a statistical software package (STATA 15.2) for further analysis. The records were checked for inconsistencies in dates; for example, discharge date before admission date or episode start date after discharge date were identified and dropped.\u003c/p\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eDescriptive analysis of demographic and clinical characteristics of the study population was performed. The study granted approval by the CPRD Independent Scientific Advisory Committee (ISAC) protocol number (protocol number 18_170R).\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eThe whole Study Population\u003c/h2\u003e \u003cp\u003eThe overall study population constituted of a total of 108,221 patients with an incident diagnosis of KOA in CPRD, however, 4,139 of them were excluded as their earliest KOA diagnosis date within CPRD records was after 31th December 2014. The remaining 104,082 patients from 660 practices across the UK constituted the whole study population which was selected from CPRD standalone data.\u003c/p\u003e \u003cp\u003eA total of 45,678 patients were aged 40\u0026ndash;64 years at the start of follow-up (43.9%), while 40,305 (38.7%) of them were aged 65\u0026ndash;80 years. Females constituted 58.7% of the study population (n\u0026thinsp;=\u0026thinsp;61,057) (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eHES-Linked Population\u003c/h2\u003e \u003cp\u003eA total of 59,737 patients diagnosed with KOA were eligible for linkage from the 390 consenting English practices; 2,354 (3.9%) of them were excluded due to missing discharge dates or having inconsistent dates (episode start date later than discharge date, or episode start date later than episode end date).\u003c/p\u003e \u003cp\u003eThis yielded 57,383 HES-linked patients for further analyses, representing 55.1% of the total patients with a diagnosis of KOA (N\u0026thinsp;=\u0026thinsp;104,082) as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. The demographic and socioeconomic characteristics of the HES-linked patients were largely similar to those of the whole study population (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCharacteristics of the Study Population\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWhole Study Population (CPRD standalone data)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHES-Linked Patients\u003c/p\u003e \u003cp\u003e(CPRD-HES linked data)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNumber of patients\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e104,082\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e57,383 (55.1%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eGender\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMales\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e43,025 (41.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e23,352 (40.7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemales\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e61,057 (58.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e34,031 (59.3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003eAge in years*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean (\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e66.30 (\u0026plusmn;\u0026thinsp;12.76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e67.04 (\u0026plusmn;\u0026thinsp;12.82)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRange\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(18.03-106.37)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(18.03-104.72)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAge ranks, years (% from total)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2,143 (2.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1,157 (2.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e40\u0026ndash;64\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e45,678 (43.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e23,766 (41.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e65\u0026ndash;80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e40,305 (38.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e22,790 (39.7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e15,956 (15.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9,670 (16.9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eIMD score (% from total)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1 (least deprived)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14,176 (21.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e12,100 (21.9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e15,456 (23.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13,425 (23.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e13,842 (21.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e12,153 (21.2%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12,610 (19.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11,062 (19.2%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5 (most deprived)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9,696 (14.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8,620 (15.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"3\"\u003e* Calculated at KOA diagnosis; SD \u0026ndash; standard deviation; IMD \u0026ndash; index of multiple deprivation\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eProportion of Patients with a Recorded Fall in CPRD, HES or Both\u003c/h2\u003e \u003cp\u003eWithin CPRD the whole study population consisted of 104,082 patients, and \u003cb\u003e3,275\u003c/b\u003e (3.1%) of them had a fall recorded within one year of KOA diagnosis. However, using the HES-linked population (n\u0026thinsp;=\u0026thinsp;57,383), the number of patients with a fall recorded within one year of KOA diagnosis was \u003cb\u003e2,384\u003c/b\u003e, representing 4.1% of the total linked patients. Subsequently, the process of identifying each of the four patient groups was followed and the results are summarised in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eWithin the linked population (n\u0026thinsp;=\u0026thinsp;57,383), the number of patients who had a fall record \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eonly in CPRD (group 1)\u003c/span\u003e was 1,852 (3.2%); however, there were an additional 365 (0.6%) patients who had a fall recorded \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eonly within HES data (group 2).\u003c/span\u003e There were 167 (0.3%) patients who had a record in \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eboth databases (group 3)\u003c/span\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eCharacteristics of Patients with a Fall Record Identified in CPRD, HES or Both\u003c/h2\u003e \u003cp\u003eThe demographic, socioeconomic and clinical characteristics of patients who had a fall record identified in CPRD, HES or both databases were analysed and presented in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDemographic and Clinical Characteristics of Patients with Fall Records among the HES-linked Population\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePatient Characteristics\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003ePatients with a Fall Record Identified in Respective Database(s), n\u0026thinsp;=\u0026thinsp;2,384\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOnly CPRD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOnly HES\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCPRD \u0026amp; HES\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNumber of patients\u003c/p\u003e \u003cp\u003e(%) from total\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1,852\u003c/p\u003e \u003cp\u003e(77.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e365\u003c/p\u003e \u003cp\u003e(15.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e167\u003c/p\u003e \u003cp\u003e(7.0)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemales (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1,391 (75.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e229 (62.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e135 (80.8)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003eAge\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean (\u0026plusmn;\u0026thinsp;SD years)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e76.2 (\u0026plusmn;\u0026thinsp;11.87)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e76.1 (\u0026plusmn;\u0026thinsp;12.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e81.6 (\u0026plusmn;\u0026thinsp;10.94)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge range\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e33.1- 102.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e30.2\u0026ndash;103.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e36.9\u0026ndash;104.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003eAge group, years\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7 (0.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2 (0.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1 (0.6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e40\u0026ndash;64\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e331 (17.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e66 (18.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e14 (8.4)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e65\u0026ndash;80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e693 (37.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e134 (36.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e41(24.5)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e821 (44.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e163(44.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e111(66.5)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003eIMD Score\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e361 (19.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e75 (20.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e28 (16.8)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e430 (23.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e91(25.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e41 (24.6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e406 (21.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e75 (20.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e39 (23.4)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e386 (20.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e64 (17.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e32 (19.2)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e268 (14.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e58 (16.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e27 (16.1)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eComorbidity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e433 (23.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e58 (23.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e38 (22.8)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFRIDs\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1,398 (75.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e261 (71.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e132 (79.0)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePrevious fall\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e253 (13.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e79 (19.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e26 (15.6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eSD standard deviation; IMD score index of multiple deprivation; FRID fall risk increasing drugs\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eComparison of Fall Recording Dates\u003c/h2\u003e \u003cp\u003eRecords of the patients who had a fall recorded in both databases (CPRD and HES) were further analysed to determine the time gap between the two dates of falling. The date of the first fall after the diagnosis of KOA was selected for the comparison, and the results are summarised in Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eGap between Fall Recording Dates in CPRD and HES and Number of Patients in Each\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTime Gap Category\u003c/p\u003e \u003cp\u003e(Days of Gap in Recording between CPRD and HES)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumber of Patients with a Record of Fall (n\u0026thinsp;=\u0026thinsp;167)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo gap (same recording date in CPRD and HES)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e45 (26.9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVery short gap (\u0026le;\u0026thinsp;2days)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e19 (11.3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShort gap (\u0026gt;\u0026thinsp;2 days and \u0026le;\u0026thinsp;7days)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e17 (10.2%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIntermediate gap (\u0026gt;\u0026thinsp;7 days and \u0026le;\u0026thinsp;14 days)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e9 (5.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLong gap (\u0026gt;\u0026thinsp;14 days and \u0026le;\u0026thinsp;30 days)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e16 (9.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProlonged gap 1 (\u0026gt;\u0026thinsp;30 and \u0026le;\u0026thinsp;60 days)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e15 (9.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProlonged gap 2 (\u0026gt;\u0026thinsp;60 and \u0026le;\u0026thinsp;90 days)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e7(4.2%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProlonged gap 3 (\u0026gt;\u0026thinsp;90 days)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e39 (23.3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNumber of falls first recorded in HES\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e58 (34.7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNumber of falls first recorded in CPRD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e64 (38.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eFor patients who had a record of fall first recorded in primary care (n\u0026thinsp;=\u0026thinsp;64), there was a median of 19.5 days\u0026rsquo; difference (IQR 4.5, 91) in the recording between primary and secondary care data for fall events.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eMain findings\u003c/h2\u003e \u003cp\u003eThis study explored fall event recording in patients with KOA within CPRD, HES, and in both databases. A total of 2,384 falls were identified within the CPRD\u0026ndash;HES-linked data; the majority of them had a record in CPRD (2,019 falls, representing 84.6% of all falls). However, a small proportion of the total identified fall events were only recorded in HES (365 falls, representing 15% of all falls).\u003c/p\u003e \u003cp\u003eAmong those with a fall record in both databases within one year of KOA diagnosis, fall dates were within a month for 64% of the patients, while for 23.3% there were more than 90 days between the dates.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eProportion of Fallers within the Whole Study Population vs HES-Linked Population\u003c/h2\u003e \u003cp\u003eUnlike the whole study population, which included 104,082 patients with a diagnosis of KOA, who were registered with GP practices across the UK, HES-linked practices reside only in England. Hence, the number of linked patients was approximately 55% (57,383 patients) of the total diagnosed with KOA.\u003c/p\u003e \u003cp\u003eThe proportion of patients with a recorded fall within the HES-linked population was 4.1% (2,384 falls in the 57,383 HES-linked population), and was slightly higher than the proportion within the CPRD standalone data (3.1% patients with falls recorded among 104,082 patients with KOA). Nearly 85% (n\u0026thinsp;=\u0026thinsp;2,019) of all falls were recorded in CPRD, while 22.3% (n\u0026thinsp;=\u0026thinsp;532) were recorded in HES data.\u003c/p\u003e \u003cp\u003eThere were 1,852 patients who were identified only within CPRD (77.7% of all falls) compared to 11.1% identified only within HES. This was in line with findings from a previous study on the prevalence of injury in England among children and young people using record linkage. The study reported that 75% of total fractures were recorded only in CPRD data, compared to 8.6% recorded only in HES data (n\u0026thinsp;=\u0026thinsp;139,662 and 15,972) [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eResults showed that linkage to HES data did not substantially increase the number of identified fallers; in fact, a mere 365 fall events were only recorded in HES, and would have been missed if HES records were not explored. These figures suggest a limited role for record linkage in improving the sensitivity of fall identification.\u003c/p\u003e \u003cp\u003eFall events are almost entirely recorded in primary care, and even those which are directly managed in hospitals (group 2) are eventually recorded in primary care.\u003c/p\u003e \u003cp\u003eThe substantial overlap in recording could be explained by the nature of both the KOA itself and the outcome of the fall. KOA is a chronic condition which is largely managed in primary care. It is, therefore, common that patients with KOA present, are diagnosed, investigated, followed-up and managed in primary care settings. This is reflected in the annual number of GP visits, where it is estimated that 20% of the annual GP visits in the UK are for musculoskeletal conditions, including KOA [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]\u003c/p\u003e \u003cp\u003ePatients with KOA may have frequent visits to their GPs (for follow-up of KOA or care for other comorbidities), who are likely to record encountered health events such as falls during these routine consultations [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. General Practitioners and other primary care professionals\u0026rsquo; key roles are in identifying those who have had a fall, and referring those who are at risk of falling to specialised fall prevention services, for multifactorial assessment and intervention. The components of multifactorial risk assessment and multifactorial intervention are summarised in Table\u0026nbsp;6.\u003c/p\u003e \u003cp\u003eTable \u0026lrm;6 Key Components of the Multifactorial Risk Assessment and Multifactorial Intervention\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Taba\" border=\"1\"\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMultifactorial Risk Assessment Examines the Following:\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMultifactorial Intervention should:\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFall history\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eProvide strength and balance training\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGait, balance mobility and muscle strength\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAddress home hazards\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOsteoporosis/fracture risk\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCorrect visual risk factors\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFunctional ability, fear of falling\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eProvide education\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVision, cognition and neurological assessment\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAddress specific underlying medical problems\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUrinary incontinence\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHome hazards\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCardiovascular problems\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMedications\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eAlthough record linkage has shown value in driving accurate (unbiased) assessments of incidence and prevalence of CAP in elderly people and MI events [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], it should, however, be noted that these findings are likely to depend on the disease area [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. For instance, cancer case recording between CPRD and cancer registries was found to have a high level of concordance, with \u0026gt;\u0026thinsp;80% of cases being present in both databases [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. This level of concordance was achieved despite cancer being managed at different care settings, and one would expect a high improvement in sensitivity with record linkage. Similarly, a high concordance between primary and secondary care records was observed in this study for the record of falls.\u003c/p\u003e \u003cp\u003eThe limited role of record linkage in improving the sensitivity of diagnoses of conditions or outcomes largely managed in primary care settings was also suggested by researchers in primary care. Yu et al. (2018), in their work on estimating the extent of recording of OA in CPRD \u0026ndash; using patients who underwent a total hip replacement (THR) or total knee replacement (TKR) as the reference population \u0026ndash; found that OA was under-recorded in CPRD, but suggested that it is unlikely that linkage to hospital data will substantially improve the number of identified patients, since it is a condition that is largely managed in general practice [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eRecord linkage is recommended in studies aiming to estimate the incidence or prevalence of diseases, particularly those which are managed in primary as well as secondary care settings. For conditions that are managed entirely in primary care, record linkage may not be necessary [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]; however, it must be explored prior to making any final decisions on the use of the HES-linked population for incidence estimation.\u003c/p\u003e \u003cp\u003eIn this study, the existence of a fall record in CPRD, HES or both databases suggests the occurrence of falls of varying severity; for example, patients in group 3 are probably those who presented at a GP surgery, but required hospitalisation for further management of the fall they experienced. They could also be those who were hospitalised for a fall, and their GP records were updated. On the other hand, patients constituting group 2 may represent those who were hospitalised without presentation at their GP surgery, with their records not fed into the GP system, or those who have died, and hence have not returned to their GP. Such cases may represent a group of fallers who experienced a serious fall.\u003c/p\u003e \u003cp\u003eGroup 1 represents patients who required medical attention (GP care) but were probably managed at the GP surgery, and did not require escalated care in a hospital.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eStudy Implications\u003c/h2\u003e \u003cdiv id=\"Sec21\" class=\"Section3\"\u003e \u003ch2\u003eImplication for Future Work of this Research\u003c/h2\u003e \u003cp\u003efall events identified for HES-linked patients included a subset of patients for whom escalated care/hospital care was required as a result of a probable severe fall-related injury. And therefore, the use of an HES-linked population was considered for further work on the association between analgesic use and the risk of falls among patients with KOA.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003eImplications of Study Findings for Research in General\u003c/h2\u003e \u003cp\u003eThis study showed that record linkage has a limited role in improving the sensitivity of fall diagnosis in patients with KOA. It had demonstrated an example of the events that are likely to be recorded completely in primary care, hence informing researchers on the completeness of primary care records for falls of patients with KOA.\u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003ch2\u003eImplications for Practice and Policy\u003c/h2\u003e \u003cp\u003eThis study has demonstrated an example of the completeness of primary care records for falls. This knowledge is essential for health services planners, commissioners and other decision makers, as CPRD data can be reliably used for service planning purposes when required.\u003c/p\u003e \u003cp\u003eInformation provided by this study on the frequency of falling, and characteristics of those who experienced a fall, could be made available to inform fall risk prevention programmes about possible groups of people within those diagnosed with KOA who are most likely to experience a fall.\u003c/p\u003e \u003cp\u003eFalls are a growing public health challenge and it is estimated that the prevalence of falls in the older people of the world was 26.5% (95% CI 23.4%-29.8%) [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e].[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. In the UK, falls are estimated to cost the NHS more than \u0026pound;2.3\u0026nbsp;billion per year. Falls are the leading cause of injury-related hospitalisation, and exert a burden on hospital services, in terms of both A\u0026amp;E visits and inpatient admissions [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Consequently, fall prevention is one of the three priorities for optimisation highlighted in the NHS RightCare Falls and Fragility Fractures pathway, in addition to detecting and managing osteoporosis and fracture risk, and optimal support following a fragility fracture [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eAlthough CPRD can be used as a standalone data source for information on the absolute number of patients with falls requiring medical care from GPs, there were some patients (n\u0026thinsp;=\u0026thinsp;365) with more serious falls which required hospitalisation for more intense medical care. Observing changes in the proportion of those hospitalised following falls may inform future resource planning and hospital bed demand/allocation planning for potential fall injury-related hospitalisation. Any reduction in the proportion of those hospitalised may serve as a quality improvement indicator for fall prevention or fracture liaison services.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003eStrengths and Limitations\u003c/h2\u003e \u003cp\u003eThe study used data from both primary care and secondary care, and maximised the number of identified fall cases using code lists which were applied by other researchers in published work [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eHowever, the study has some limitations that should be mentioned. Falls identified through electronic health records (EHR) represent just those for which some form of medical attention was required, It was revealed that up to 30% of participants in a population-based survey reported at least one fall in the prior 12 months [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Hence, compared to surveys, the number of falls within the EHR database is potentially underestimated.\u003c/p\u003e \u003cp\u003eThis study did not include codes for fall-related injuries, resulting in possible underestimation of the true number of falls in HES data (fall events may have been coded with the resulting injury codes, rather than fall codes). Additionally, there is the potential for under- or over-estimation of falls in HES due to inaccurate coding.\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study has explored the potential benefits of record linkage in improving the sensitivity of fall identification. Using the CPRD\u0026ndash;HES-linked patient data, this study found that CPRD records included approximately 84.7% of the overall falls experienced by patients within one year of their KOA diagnosis. However, 15.3% of the falls were only recorded in hospital data; although not a substantial proportion, this represents a group of patients who probably suffered a serious event.\u003c/p\u003e \u003cp\u003eThe study addressed the concern that EHR from one part of the health system, such as primary care, may not capture health events occurring in other parts of the system, such as hospital care\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eFunding\u003c/h2\u003e\n\u003cp\u003e\u003cem\u003eThe authors declare that no funds, grants, or other support were received during the preparation of this manuscript.\u003c/em\u003e\u003c/p\u003e\n\u003ch2\u003e\u003cem\u003eCompeting Interests\u003c/em\u003e\u003c/h2\u003e\n\u003cp\u003e\u003cem\u003eThe authors have no relevant financial or non-financial interests to disclose\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003eDisclosures and declarations\u003c/p\u003e\n\u003cp\u003eThe authors declare no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work; no other relationships or activities that could appear to have influenced the submitted work.\u003c/p\u003e\n\u003ch2\u003eData Availability Statement\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eThe data that support the findings of this study are owned by the CPRD and may be available subject to data sharing regulations of the CPRD\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHuman Ethics and Consent to Participate declarations\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHuman Ethics and Consent to Participate declarations: Not applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics Approval declaration\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eThe study was approved by the CPRD Independent\u0026nbsp;\u003c/em\u003eScientific Advisory Committee (ISAC) (protocol number 18_170R).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to Participate declarations:\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical trial number\u003c/strong\u003e: Not applicable.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eJ. G. Quicke, P. G. Conaghan, N. Corp, and G. Peat, \u0026quot;Osteoarthritis year in review 2021: epidemiology \u0026amp;amp; therapy,\u0026quot; \u003cem\u003eOsteoarthritis and Cartilage, \u003c/em\u003evol. 30, no. 2, pp. 196-206, 2022, doi: 10.1016/j.joca.2021.10.003.\u003c/li\u003e\n\u003cli\u003eQ. Yao\u003cem\u003e et al.\u003c/em\u003e, \u0026quot;Osteoarthritis: pathogenic signaling pathways and therapeutic targets,\u0026quot; (in eng), \u003cem\u003eSignal Transduct Target Ther, \u003c/em\u003evol. 8, no. 1, p. 56, Feb 3 2023, doi: 10.1038/s41392-023-01330-w.\u003c/li\u003e\n\u003cli\u003eNICE. \u0026quot;Osteoarthritis: Care and Management.\u0026quot; National Institute for Clinical and Care Excellence. https://www.nice.org.uk/guidance/CG177 (accessed 15 July, 2020).\u003c/li\u003e\n\u003cli\u003eA. Taqi, S. Gran, and R. D. Knaggs, \u0026quot;Analgesic utilization in people with knee osteoarthritis: A population-based study using primary care data,\u0026quot; (in eng), \u003cem\u003ePain Pract, \u003c/em\u003evol. 23, no. 5, pp. 523-534, Jun 2023, doi: 10.1111/papr.13212.\u003c/li\u003e\n\u003cli\u003eN. Wilson\u003cem\u003e et al.\u003c/em\u003e, \u0026quot;Drug utilization in patients with OA: a population-based study,\u0026quot; \u003cem\u003eRheumatology, \u003c/em\u003evol. 54, no. 5, pp. 860-867, 2015, doi: 10.1093/rheumatology/keu403.\u003c/li\u003e\n\u003cli\u003eS. R. Kingsbury, E. M. A. Hensor, C. A. E. Walsh, M. C. Hochberg, and P. G. Conaghan, \u0026quot;How do people with knee osteoarthritis use osteoarthritis pain medications and does this change over time? Data from the Osteoarthritis Initiative,\u0026quot; (in eng), \u003cem\u003eArthritis Res Ther, \u003c/em\u003evol. 15, no. 5, pp. R106-R106, 2013, doi: 10.1186/ar4286.\u003c/li\u003e\n\u003cli\u003eS. L. James\u003cem\u003e et al.\u003c/em\u003e, \u0026quot;The global burden of falls: global, regional and national estimates of morbidity and mortality from the Global Burden of Disease Study 2017,\u0026quot; (in eng), \u003cem\u003eInj Prev, \u003c/em\u003evol. 26, no. Supp 1, pp. i3-i11, Oct 2020, doi: 10.1136/injuryprev-2019-043286.\u003c/li\u003e\n\u003cli\u003eN. K. Arden\u003cem\u003e et al.\u003c/em\u003e, \u0026quot;Knee pain, knee osteoarthritis, and the risk of fracture,\u0026quot; \u003cem\u003eArthritis Care \u0026amp; Research, \u003c/em\u003evol. 55, no. 4, pp. 610-615, 2006/08/15 2006, doi: 10.1002/art.22088.\u003c/li\u003e\n\u003cli\u003eP. Scuffham, S. Chaplin, and R. Legood, \u0026quot;Incidence and costs of unintentional falls in older people in the United Kingdom,\u0026quot; \u003cem\u003eJournal of Epidemiology and Community Health, \u003c/em\u003evol. 57, no. 9, p. 740, 2003, doi: 10.1136/jech.57.9.740.\u003c/li\u003e\n\u003cli\u003eE. Herrett\u003cem\u003e et al.\u003c/em\u003e, \u0026quot;Completeness and diagnostic validity of recording acute myocardial infarction events in primary care, hospital care, disease registry, and national mortality records: cohort study,\u0026quot; (in eng), \u003cem\u003eBMJ (Clinical research ed.), \u003c/em\u003evol. 346, pp. f2350-f2350, 2013, doi: 10.1136/bmj.f2350.\u003c/li\u003e\n\u003cli\u003eL. McDonald, C. J. Sammon, M. Samnaliev, and S. Ramagopalan, \u0026quot;Under-recording of hospital bleeding events in UK primary care: a linked Clinical Practice Research Datalink and Hospital Episode Statistics study,\u0026quot; (in eng), \u003cem\u003eClinical epidemiology, \u003c/em\u003evol. 10, pp. 1155-1168, 2018, doi: 10.2147/CLEP.S170304.\u003c/li\u003e\n\u003cli\u003eE. R. C. Millett, J. K. Quint, B. L. De Stavola, L. Smeeth, and S. L. Thomas, \u0026quot;Improved incidence estimates from linked vs. stand-alone electronic health records,\u0026quot; (in eng), \u003cem\u003eJournal of clinical epidemiology, \u003c/em\u003evol. 75, pp. 66-69, 2016, doi: 10.1016/j.jclinepi.2016.01.005.\u003c/li\u003e\n\u003cli\u003eK. J. Rothnie\u003cem\u003e et al.\u003c/em\u003e, \u0026quot;Validation of the Recording of Acute Exacerbations of COPD in UK Primary Care Electronic Healthcare Records,\u0026quot; \u003cem\u003ePLOS ONE, \u003c/em\u003evol. 11, no. 3, p. e0151357, 2016, doi: 10.1371/journal.pone.0151357.\u003c/li\u003e\n\u003cli\u003eE. Herrett\u003cem\u003e et al.\u003c/em\u003e, \u0026quot;Data Resource Profile: Clinical Practice Research Datalink (CPRD),\u0026quot; (in eng), \u003cem\u003eInt J Epidemiol, \u003c/em\u003evol. 44, no. 3, pp. 827-836, 2015, doi: 10.1093/ije/dyv098.\u003c/li\u003e\n\u003cli\u003eE. Herrett, S. L. Thomas, W. M. Schoonen, L. Smeeth, and A. J. Hall, \u0026quot;Validation and validity of diagnoses in the General Practice Research Database: a systematic review,\u0026quot; \u003cem\u003eBritish Journal of Clinical Pharmacology, \u003c/em\u003evol. 69, no. 1, pp. 4-14, 04/09/received 08/06/accepted 2010, doi: 10.1111/j.1365-2125.2009.03537.x.\u003c/li\u003e\n\u003cli\u003eM. Gore, A. B. Sadosky, D. L. Leslie, K.-S. Tai, and P. Emery, \u0026quot;Therapy Switching, Augmentation, and Discontinuation in Patients with Osteoarthritis and Chronic Low Back Pain,\u0026quot; \u003cem\u003ePain Practice, \u003c/em\u003evol. 12, no. 6, pp. 457-468, 2012/07/01 2012, doi: 10.1111/j.1533-2500.2011.00524.x.\u003c/li\u003e\n\u003cli\u003eC. Coupland, P. Dhiman, R. Morriss, A. Arthur, G. Barton, and J. Hippisley-Cox, \u0026quot;Antidepressant use and risk of adverse outcomes in older people: population based cohort study,\u0026quot; \u003cem\u003eThe BMJ, \u003c/em\u003evol. 343, p. d4551, 08/02 06/13/accepted 2011, doi: 10.1136/bmj.d4551.\u003c/li\u003e\n\u003cli\u003eJ. GRIBBIN, \u0026quot;FALLS IN OLDER PEOPLE AND THE ROLE OF COMMONLY PRESCRIBED ANTIDEPRESSANT AND ANTIHYPERTENSIVE MEDICATIONS \u0026quot; PhD School of Medicine University of Nottingham UK 2013. [Online]. Available: http://eprints.nottingham.ac.uk/28451/1/594820%20minus%20articles.pdf\u003c/li\u003e\n\u003cli\u003eR. Baker, L. J. Tata, D. Kendrick, and E. Orton, \u0026quot;Identification of incident poisoning, fracture and burn events using linked primary care, secondary care and mortality data from England: implications for research and surveillance,\u0026quot; \u003cem\u003eInjury Prevention, \u003c/em\u003evol. 22, no. 1, p. 59, 2016, doi: 10.1136/injuryprev-2015-041561.\u003c/li\u003e\n\u003cli\u003eVersusArthritis. \u0026quot;The State of Musculoskeletal Health 2019.\u0026quot; Versus Arthritis. https://www.versusarthritis.org/about-arthritis/data-and-statistics/state-of-musculoskeletal-health-2019/ (accessed 30 November 2019).\u003c/li\u003e\n\u003cli\u003eNICE. \u0026quot;Falls in older people: assessing risk and prevention.\u0026quot; https://www.nice.org.uk/guidance/cg161 (accessed 15th November 2019).\u003c/li\u003e\n\u003cli\u003eNICE. \u0026quot;Falls in older people.\u0026quot; https://www.nice.org.uk/guidance/qs86 (accessed 15 November 2019).\u003c/li\u003e\n\u003cli\u003eR. Boggon, T. P. van Staa, M. Chapman, A. M. Gallagher, T. A. Hammad, and M. A. Richards, \u0026quot;Cancer recording and mortality in the General Practice Research Database and linked cancer registries,\u0026quot; \u003cem\u003ePharmacoepidemiology and Drug Safety, \u003c/em\u003evol. 22, no. 2, pp. 168-175, 2013/02/01 2013, doi: 10.1002/pds.3374.\u003c/li\u003e\n\u003cli\u003eD. Yu, K. P. Jordan, and G. Peat, \u0026quot;Underrecording of osteoarthritis in United Kingdom primary care electronic health record data,\u0026quot; (in eng), \u003cem\u003eClinical epidemiology, \u003c/em\u003evol. 10, pp. 1195-1201, 2018, doi: 10.2147/CLEP.S160059.\u003c/li\u003e\n\u003cli\u003eN. Salari, N. Darvishi, M. Ahmadipanah, S. Shohaimi, and M. Mohammadi, \u0026quot;Global prevalence of falls in the older adults: a comprehensive systematic review and meta-analysis,\u0026quot; (in eng), \u003cem\u003eJ Orthop Surg Res, \u003c/em\u003evol. 17, no. 1, p. 334, Jun 28 2022, doi: 10.1186/s13018-022-03222-1.\u003c/li\u003e\n\u003cli\u003eONS. \u0026quot;Overview of the UK population.\u0026quot; https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/articles/overvie\u003cbr\u003ewoftheukpopulation/august2019 (accessed 30 November 2019).\u003c/li\u003e\n\u003cli\u003eJ. T. Yang Tian, David Buck, Lara Sonola, \u0026quot;Exploring the system-wide costs of falls in older people in Torbay,\u0026quot; The King\u0026apos;s Fund London 2013. Accessed: 15th November 2019 [Online]. Available: https://www.kingsfund.org.uk/sites/default/files/field/field_publication_file/exploring-system-wide-costs-of-falls-in-torbay-kingsfund-aug13.pdf\u003c/li\u003e\n\u003cli\u003eBGS, \u0026quot;NICE impact report on falls and fragility fractures,\u0026quot; National Institute fo Health and Care Excellence British Geriatrics Society, 2018. Accessed: 15th November 2019. [Online]. Available: https://www.bgs.org.uk/resources/2018-nice-impact-report-on-falls-and-fragility-fractures\u003c/li\u003e\n\u003cli\u003eNHS. \u0026quot;RightCare Pathway: Falls and Fragility Fractures.\u0026quot; https://www.england.nhs.uk/rightcare/wp-content/uploads/sites/40/2017/12/falls-fragility-fractures-pathway-v18.pdf (accessed 15th November 2019).\u003c/li\u003e\n\u003cli\u003eA. Taqi, S. Gran, and R. D. Knaggs, \u0026quot;\u0026quot;Application of five different strategies to define a cohort of patients with knee osteoarthritis in a large primary care database\u0026quot;,\u0026quot; (in eng), \u003cem\u003eJ Eval Clin Pract, \u003c/em\u003eJun 25 2024, doi: 10.1111/jep.14045.\u003c/li\u003e\n\u003cli\u003eT. Gill, A. W. Taylor, and A. Pengelly, \u0026quot;A Population-Based Survey of Factors Relating to the Prevalence of Falls in Older People,\u0026quot; \u003cem\u003eGerontology, \u003c/em\u003evol. 51, no. 5, pp. 340-345, 2005, doi: 10.1159/000086372.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-musculoskeletal-disorders","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bmsd","sideBox":"Learn more about [BMC Musculoskeletal Disorders](http://bmcmusculoskeletdisord.biomedcentral.com/)","snPcode":"","submissionUrl":"https://author-welcome.nature.com/12891","title":"BMC Musculoskeletal Disorders","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5385374/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5385374/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eObjective\u003c/h2\u003e \u003cp\u003eThe present study explored fall recording in primary care data and in primary- secondary care linked data, to inform the data source adequacy for examining the association between analgesic use and the risk of falls in patients with knee osteoarthritis (KOA).\u003c/p\u003e\u003ch2\u003eMethod\u003c/h2\u003e \u003cp\u003eData were obtained from the clinical practice research datalink (CPRD) and Hospital Episode Statistics (HES). The study population included adults with an incident diagnosis of KOA recorded in CPRD between 2000 and 2014. The study captured the proportion of fallers in CPRD, HES or both and measured the gap between fall recording in both databases. Descriptive statistics were used report study measures\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eThere were 104,082 patients diagnosed with KOA in CPRD standalone data; \u003cb\u003e3,275\u003c/b\u003e (3.1%) of them had a fall while HES-linked data included 57,383 and \u003cb\u003e2,384 (\u003c/b\u003e4.1%) of them had a fall recorded within one year of KOA diagnosis. Among the linked population, 1,852 (3.2%) had a fall record \u003cspan type=\"ItalicUnderline\" class=\"ItalicUnderline\" name=\"Emphasis\"\u003eonly in CPRD\u003c/span\u003e; however, there were an additional 365 (0.6%) patients who had a fall recorded \u003cspan type=\"ItalicUnderline\" class=\"ItalicUnderline\" name=\"Emphasis\"\u003eonly within HES data\u003c/span\u003e, while167 (0.3%) patients had it recorded in \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eboth databases.\u003c/span\u003e There was a median of 19.5 days\u0026rsquo; difference (IQR 4.5, 91) in fall recording between primary and secondary care data.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eCPRD records included 84.7% of the overall falls experienced by patients within one year of their KOA diagnosis. However, 15.3% of the falls were only recorded in hospital data; although not a substantial proportion, this represents a group that probably suffered a serious event.\u003c/p\u003e","manuscriptTitle":"Exploratory Study of Fall Event Recording in Patients with Knee Osteoarthritis using Clinical Practice Research Datalink and Hospital Episode Statistics Linked Data","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-12-16 16:55:50","doi":"10.21203/rs.3.rs-5385374/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-11-14T15:10:22+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-11-11T07:09:09+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-11-11T07:06:50+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Musculoskeletal Disorders","date":"2024-11-04T06:29:31+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-musculoskeletal-disorders","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bmsd","sideBox":"Learn more about [BMC Musculoskeletal Disorders](http://bmcmusculoskeletdisord.biomedcentral.com/)","snPcode":"","submissionUrl":"https://author-welcome.nature.com/12891","title":"BMC Musculoskeletal Disorders","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"e520d5cb-5082-47ff-91d7-7706c529ae25","owner":[],"postedDate":"December 16th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2024-12-16T16:55:50+00:00","versionOfRecord":[],"versionCreatedAt":"2024-12-16 16:55:50","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5385374","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5385374","identity":"rs-5385374","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}