A longitudinal transition imputation model for categorical data applied to a large registry dataset.

Imputation of longitudinal categorical covariates with several waves and many predictors is cumbersome in terms of implausible transitions, colinearity, and overfitting. We designed a simulation study with data obtained from a general practitioners' morbidity registry in Belgium for three waves, with smoking as the longitudinal covariate of interest. We set varying proportions of data on smoking to missing completely at random and missing not at random with proportions of missingness equal to 10%, 30%, 50%, and 70%. This study proposed a 3-stage approach that allows flexibility when imputing time-dependent categorical covariates. First, multiple imputation using fully conditional specification or multiple imputation for the predictor variables was deployed using the wide format such that previous and future information of the same patient was utilized. Second, a joint Markov transition model for initial, forward, backward, and intermittent probabilities was developed for each imputed dataset. Finally, this transition model was used for imputation. We compared the performance of this methodology with an analyses of the complete data and with listwise deletion in terms of bias and root mean square error. Next, we applied this methodology in a clinical case for years 2017 to 2021, where we estimated the effect of several covariates on the pneumococcal vaccination. This methodological framework ensures that the plausibility of transitions is preserved, overfitting and colinearity issues are resolved, and confounders can be utilized. Finally, a companion R package was developed to enable the replication and easy application of this methodology.

Pneumococcal vaccination coverage and adherence to recommended dosing schedules in adults: a repeated cross-sectional study of the INTEGO morbidity registry.

BACKGROUND: Since 2014, Belgium's Superior Health Council has recommended pneumococcal vaccination for adults aged 19-85 years at increased risk for pneumococcal diseases with a specific vaccine administration sequence and timing. Currently, Belgium has no publicly funded adult pneumococcal vaccination program. This study investigated the seasonal pneumococcal vaccination trends, evolution of vaccination coverage and adherence to the 2014 recommendations. METHODS: INTEGO is a general practice morbidity registry in Flanders (Belgium) that represents 102 general practice centres and comprised over 300.000 patients in 2021. A repeated cross-sectional study was performed for the period between 2017 and 2021. Using adjusted odds ratios computed via multiple logistic regression, the association between an individual's characteristics (gender, age, comorbidities, influenza vaccination status and socioeconomic status) and schedule-adherent pneumococcal vaccination status was assessed. RESULTS: Pneumococcal vaccination coincided with seasonal flu vaccination. The vaccination coverage in the population at risk decreased from 21% in 2017 to 18.2% in 2018 and then started to increase to 23.6% in 2021. Coverage in 2021 was highest for high-risk adults (33.8%) followed by 50- to 85-year-olds with comorbidities (25.5%) and healthy 65- to 85-year-olds (18.7%). In 2021, 56.3% of the high-risk adults, 74.6% of the 50+ with comorbidities persons, and 74% of the 65+ healthy persons had an adherent vaccination schedule. Persons with a lower socioeconomic status had an adjusted odds ratio of 0.92 (95% Confidence Interval (CI) 0.87-0.97) for primary vaccination, 0.67 (95% CI 0.60-0.75) for adherence to the recommended second vaccination if the 13-valent pneumococcal conjugate vaccine was administered first and 0.86 (95% CI 0.76-0.97) if the 23-valent pneumococcal polysaccharide vaccine was administered first. CONCLUSION: Pneumococcal vaccine coverage is slowly increasing in Flanders, displaying seasonal peaks in sync with influenza vaccination campaigns. However, with less than one-fourth of the target population vaccinated, less than 60% high-risk and approximately 74% of 50 + with comorbidities and 65+ healthy persons with an adherent schedule, there is still much room for improvement. Furthermore, adults with poor socioeconomic status had lower odds of primary vaccination and schedule adherence, demonstrating the need for a publicly funded program in Belgium to ensure equitable access.

Cost of healthcare utilization associated with incident cardiovascular and renal disease in individuals with type 2 diabetes: A multinational, observational study across 12 countries.

AIM: To examine how the development of cardiovascular and renal disease (CVRD) translates to hospital healthcare costs in individuals with type 2 diabetes (T2D) initially free from CVRD. METHODS: Data were obtained from the digital healthcare systems of 12 nations using a prespecified protocol. A fixed country-specific index date of 1 January was chosen to secure sufficient cohort disease history and maximal follow-up, varying between each nation from 2006 to 2017. At index, all individuals were free from any diagnoses of CVRD (including heart failure [HF], chronic kidney disease [CKD], coronary ischaemic disease, stroke, myocardial infarction [MI], or peripheral artery disease [PAD]). Outcomes during follow-up were hospital visits for CKD, HF, MI, stroke, and PAD. Hospital healthcare costs obtained from six countries, representing 68% of the total study population, were cumulatively summarized for CVRD events occurring during follow-up. RESULTS: In total, 1.2 million CVRD-free individuals with T2D were identified and followed for 4.5 years (mean), that is, 4.9 million patient-years. The proportion of individuals indexed before 2010 was 18% (n = 207 137); 2010-2015, 31% (361 175); and after 2015, 52% (609 095). Overall, 184 420 (15.7%) developed CVRD, of which cardiorenal disease was most frequently the first disease to develop (59.7%), consisting of 23.0% HF and 36.7% CKD, and more common than stroke (16.9%), MI (13.7%), and PAD (9.7%). The total cumulative cost for CVRD was US$1 billion, of which 59.0% was attributed to cardiorenal disease, 3-, 5-, and 6-fold times greater than the costs for stroke, MI, and PAD, respectively. CONCLUSION: Across all nations, HF or CKD was the most frequent CVRD manifestation to develop in a low-risk population with T2D, accounting for the highest proportion of hospital healthcare costs. These novel findings highlight the importance of cardiorenal awareness when planning healthcare.

Fast and optimal algorithm for case-control matching using registry data: application on the antibiotics use of colorectal cancer patients.

BACKGROUND: In case-control studies most algorithms allow the controls to be sampled several times, which is not always optimal. If many controls are available and adjustment for several covariates is necessary, matching without replacement might increase statistical efficiency. Comparing similar units when having observational data is of utter importance, since confounding and selection bias is present. The aim was twofold, firstly to create a method that accommodates the option that a control is not resampled, and second, to display several scenarios that identify changes of Odds Ratios (ORs) while increasing the balance of the matched sample. METHODS: The algorithm was derived in an iterative way starting from the pre-processing steps to derive the data until its application in a study to investigate the risk of antibiotics on colorectal cancer in the INTEGO registry (Flanders, Belgium). Different scenarios were developed to investigate the fluctuation of ORs using the combination of exact and varying variables with or without replacement of controls. To achieve balance in the population, we introduced the Comorbidity Index (CI) variable, which is the sum of chronic diseases as a means to have comparable units for drawing valid associations. RESULTS: This algorithm is fast and optimal. We simulated data and demonstrated that the run-time of matching even with millions of patients is minimal. Optimal, since the closest controls is always captured (using the appropriate ordering and by creating some auxiliary variables), and in the scenario that a case has only one control, we assure that this control will be matched to this case, thus maximizing the cases to be used in the analysis. In total, 72 different scenarios were displayed indicating the fluctuation of ORs, and revealing patterns, especially a drop when balancing the population. CONCLUSIONS: We created an optimal and computationally efficient algorithm to derive a matched case-control sample with and without replacement of controls. The code and the functions are publicly available as an open source in an R package. Finally, we emphasize the importance of displaying several scenarios and assess the difference of ORs while using an index to balance population in observational data.

An automated data cleaning method for Electronic Health Records by incorporating clinical knowledge.

BACKGROUND: The use of Electronic Health Records (EHR) data in clinical research is incredibly increasing, but the abundancy of data resources raises the challenge of data cleaning. It can save time if the data cleaning can be done automatically. In addition, the automated data cleaning tools for data in other domains often process all variables uniformly, meaning that they cannot serve well for clinical data, as there is variable-specific information that needs to be considered. This paper proposes an automated data cleaning method for EHR data with clinical knowledge taken into consideration. METHODS: We used EHR data collected from primary care in Flanders, Belgium during 1994-2015. We constructed a Clinical Knowledge Database to store all the variable-specific information that is necessary for data cleaning. We applied Fuzzy search to automatically detect and replace the wrongly spelled units, and performed the unit conversion following the variable-specific conversion formula. Then the numeric values were corrected and outliers were detected considering the clinical knowledge. In total, 52 clinical variables were cleaned, and the percentage of missing values (completeness) and percentage of values within the normal range (correctness) before and after the cleaning process were compared. RESULTS: All variables were 100% complete before data cleaning. 42 variables had a drop of less than 1% in the percentage of missing values and 9 variables declined by 1-10%. Only 1 variable experienced large decline in completeness (13.36%). All variables had more than 50% values within the normal range after cleaning, of which 43 variables had a percentage higher than 70%. CONCLUSIONS: We propose a general method for clinical variables, which achieves high automation and is capable to deal with large-scale data. This method largely improved the efficiency to clean the data and removed the technical barriers for non-technical people.

The health and economic impact of acute gastroenteritis in Belgium, 2010-2014.

Acute gastroenteritis (AGE) remains a common condition in both low- and high-income countries. In Belgium, however, there is currently a lack of information on the societal health and economic impact of AGE. We conducted a retrospective study using mortality and cause-of-death data, hospital data, primary care data, health interview survey data and other published data. We estimated the burden of illness during a 5-year period (2010-2014) in Belgium in terms of deaths, patients admitted to hospitals, patients consulting their general practitioner (GP) and cases occurring in the community. We further quantified the health impact in terms of disability-adjusted life years (DALYs) and the economic impact in terms of cost-of-illness estimates. We estimated 343 deaths, 27 707 hospitalised patients, 464 222 GP consultations and 10 058 741 episodes occurring in the community (0.91 cases/person) on average per year. AGE was associated with 11 855 DALYs per year (107 DALY per 100 000 persons). The economic burden was estimated to represent direct costs of €112 million, indirect costs of €927 million (90% of the total costs) and an average total cost of €103 per case and €94 per person. AGE results in a substantial health and economic impact in Belgium, justifying continued mitigation efforts.

Knowledge and prescribing behaviour of Flemish general practitioners regarding novel glucose-lowering medications: Online cross-sectional survey.

AIMS: To determine the knowledge and prescribing behaviour regarding new type 2 diabetes medication in general practice. Physicians in Belgium are bound by the prescription criteria which do not always correspond to the international guidelines. DESIGN & METHOD: A mixed methods study with an online questionnaire was conducted in Flanders to collect data on demographic characteristics, theoretical knowledge, and prescribing behaviour, using ten theoretical questions and six clinical cases, based on the American Diabetes Association/European Association for the Study of Diabetes (ADA/EASD) guidelines and the Belgian reimbursement criteria. RESULTS: 201 GPs and GPs in training were included in this study with a median age of 30 years and 68 % female participants. On the knowledge questionnaire, the mean test result was 7.15/15 (= 48 %) with a median of 8. Further analysis showed that 90 % of the respondents correctly recommended a sodium-glucose cotransporter 2 (sglt2) inhibitor when the clinical case showed a comorbidity of heart failure, whereas only 42 % suggested correctly a glucagon-like peptide 1 (GLP-1) agonist if presence of cardiovascular disease. Subgroup analysis showed no statistically significant demographic differences in obtained test results. Regarding prescription behaviour, 23 % of the respondents would prescribe medication that did not match the reimbursement criteria in at least one of the 6 proposed clinical cases. CONCLUSION: This study highlights the need for enhanced knowledge and updated prescribing practices among Flemish GPs and Trainee GPs to effectively manage patients with T2DM.

Association between non-registration of chronic kidney disease and mortality and cardiovascular outcome: a time-to-event analysis of retrospective primary care data.

OBJECTIVE: Patients with impaired kidney function and increased albuminuria are at risk of developing cardiovascular disease (CVD). Previous research has revealed that a substantial proportion of patients with chronic kidney disease (CKD) do not get a registered diagnosis in the electronic health record of the general practitioner. The aim of this study was to investigate the association between non-registration of CKD and all-cause mortality and cardiovascular outcome. DESIGN AND SETTING: A retrospective study in primary care. METHODS: The analyses were carried out in the INTEGO database, a general practice-based morbidity registration network in Flanders, Belgium. The study used INTEGO data from the year 2018 for all patients ≥18 years old, including 10 551 patients. To assess the risk of mortality and CVD, a time-to-event analysis was performed. Cox proportional hazard model was used to evaluate the association between non-registration and incidence of all-cause mortality and cardiovascular events with mortality as a competing risk. Subgroup analyses were performed for estimated glomerular filtration rate stages (3A, 3B, 4 and 5). Multiple imputation was done following the methodology of Mamouris et al. RESULTS: Mortality was higher in patients with non-registered CKD compared with patients with registered CKD (HR 1.29, 95% CI 1.19 to 1.41). Non-registration of CKD was not associated with an increased risk for the development of CVD (HR 0.92, 95% CI 0.77 to 1.11). CONCLUSION: An association between non-registration and all-cause mortality was identified, although no such association was apparent for CVD.

Trends in oral anticoagulant prescription in patients with nonvalvular atrial fibrillation in Flanders and the impact of switching patients from vitamin K antagonists to DOACS in terms of the burden caused by complications of the disease: a registry-based study.

CONTEXT: Since the introduction of direct oral anticoagulants (DOACs) for the treatment of nonvalvular atrial fibrillation (Afib), oral anticoagulants (OACs) prescription has evolved. AIM: We aim first to explore the OACs prescription behaviour in Flanders from 2002 to 2019 before exploring the impact of switching patients from vitamin K antagonists (VKAs) to DOACs in terms of the burden caused by stroke as a complication of non-valvular Afib. METHODS: Data were obtained from INTEGO, a Flemish, general practice-based morbidity registration network. Comprised patients had at least one visit to their GP per year between 2002 and 2019 and a follow-up of at least 1 year after the diagnosis of Afib. Public prices were retrieved from the Belgian Centre for Pharmacotherapeutic Information (BCFI) and the National Institute for Health and Disability Insurance (RIZIV/ INAMI) sites. The number of Disability-Adjusted Life Years (DALYs) was based on the Global Burden of Disease (GBD) literature. The calculation of the Number Needed to Switch (NNSw) was the basis for conducting cost-utility analyses accounting for the global benefit in terms of the cost of prevented stroke/DALY and the cost of switching Flemish ≥ 65 years patients from VKAs to DOACs in two scenarios. RESULTS: Increased DOAC use has been observed since 2012. The incremental cost effectiveness ratios (ICERs) yielded 553 to 824 €/DALY of prevented stroke. CONCLUSION: In this registry-based study, we found a significant positive trend in OAC use in Flanders between 2002 and 2019. Switching to DOACs seems cost-effective compared to a threshold of 20000€/DALY.

Trends in the registration of anxiety in Belgian primary care from 2000 to 2021: a registry-based study.

BACKGROUND: Anxiety is frequently encountered in general practice, but figures regarding prevalence and incidence in this healthcare setting remain scarce. AIM: To provide insight about the trends in prevalence and incidence of anxiety in Belgian general practice, as well as the comorbidities and treatment of anxiety in this context. DESIGN AND SETTING: Retrospective cohort study using the INTEGO morbidity registration network, with clinical data from over 600 000 patients in Flanders, Belgium. METHOD: Trends in age-standardised prevalence and incidence of anxiety from 2000 to 2021 as well as prescriptions in patients with prevalent anxiety were analysed with joinpoint regression. Comorbidity profiles were analysed using the Cochran-Armitage test and the Jonckheere-Terpstra test. RESULTS: During the 22-year study period, 8451 unique patients with anxiety were identified. The prevalence of anxiety diagnoses rose significantly during this period, from 1.1% in 2000 to 4.8% in 2021. The overall incidence rate rose from 1.1/1000 patient-years (PY) in 2000 to 9.9/1000 PY in 2021. The average chronic disease count per patient increased significantly during the study period, from 1.5 to 2.3 chronic conditions. The most frequent comorbidities in patients with anxiety in 2017-2021 were malignancy (20.1%), hypertension (18.2%), and irritable bowel syndrome (13.5%). The proportion of patients treated with psychoactive medication rose from 25.7% to almost 40% over the study period. CONCLUSION: A significantly increasing prevalence and incidence of physician-registered anxiety was found in the study. Patients with anxiety tend to become more complex, with more comorbidities. Treatment for anxiety in Belgian primary care is very dependent on medication.

Prevalence of Atherosclerotic Cardiovascular Disease, Heart Failure, and Chronic Kidney Disease in Patients with Type 2 Diabetes Mellitus: A Primary Care Research Network-based Study.

AIMS: This study aims to assess the prevalence of atherosclerotic cardiovascular disease (ASCVD), heart failure (HF), chronic kidney disease (CKD), and their combined presence in type 2 diabetes (T2D) patients in primary care for whom the 2019 ADA/EASD consensus update "Management of Hyperglycemia in Type 2 Diabetes" recommends GLP-1 receptor agonists (GLP-1RA) or sodium-glucose cotransporter-2 inhibitors (SGLT-I) as first-line medications after metformin. METHODS: Data were obtained in 2015 from Intego, a morbidity registration network of 111 general practitioners (GPs) working in 48 practices and including 123 261 registered patients. RESULTS: Of 123 261 patients, 9616 had T2D. Of these patients, 4200 (43.7%) presented with ASCVD and/or CKD and/or HF. Specifically, 3348 (34.8%) patients had ASCVD, 388 (4.0%) had heart failure, and 1402 (14.6%) had CKD. Compared to patients without any of these comorbidities, patients with at least 1 of these conditions were older (69.7 ±12.6 vs. 63.1±12.5 years), had higher LDL-C values (104.2±35.8 mg/dl vs. 97.2±37.7) and less frequently achieved the systolic blood pressure target of 140 mm Hg (53 vs. 61%) (all p<0.001). Comorbid patients also had significantly more other comorbidities, such as dementia or cancer; received more recommended medications, such as statins; and received less metformin. Most patients with HF (325; 3.4%) had ASCVD (114; 1.2%), CKD (76; 0.8%), or both (135; 1.4%). In total, 478 patients with CKD (5.0%) also had ASCVD. CONCLUSIONS: At the primary care level, 44% of T2D patients suffer from ASCVD, CKD, and/or HF, and thus qualify for GLP-1RA or SGLT2-I therapy.

An exploration of under-registration of chronic kidney disease stages 3-5 in Belgian general practices using logistic regression.

BACKGROUND: Early detection and treatment of chronic kidney disease (CKD) can prevent further deterioration and complications. Previous studies suggested that the diagnosis is often made when advanced renal failure occurs. The aims of this study were to describe the prevalence of unregistered CKD stages 3-5 in a Belgian General Practitioner population, to determine risk factors for under-registration and to investigate the diagnostic delay. METHODS: The analyses were carried out in the INTEGO database, a Flanders general practice-based morbidity registration network. The study used INTEGO data from the year 2018 for all patients ≥18 years old. CKD was defined as two consecutive eGFR laboratory measurements (eGFR <60 mL/min/1.73m2) at least three months apart during the baseline period. Registered CKD was characterised by a documented diagnosis of CKD (ICPC2 U99) during the ≥12-month lookback period before the first eGFR measurement and up to six months after the second eGFR in the EHR. The prevalence of unregistered CKD and the median time of diagnostic delay were estimated. Baseline characteristics were described. A multivariate cross-sectional logistic regression analysis was conducted to identify determinants of unregistered CKD. We estimated the odds ratios and their 95% confidence interval. RESULTS: Among included patients, there were 10 551 patients (5.5%) meeting the criteria of CKD. The prevalence of unregistered CKD was 68%. The mean diagnostic delay was 1.94 years (Standard deviation 0.93). Being a male, a concurrent diagnosis of diabetes, stroke, heart failure and hypertension, and more severe CKD (stages 3b, 4 and 5) independently increased the chance on registered CKD. CONCLUSION: The proportion of patients who had no registered CKD code in the EHR was substantial. The differences between registered and unregistered patients make thinking about solutions to facilitate registration in the EHR imperative.

Evolution of benzodiazepine receptor agonist prescriptions in general practice: A registry-based study.

Background: Contrary to most European guidelines, benzodiazepine receptor agonists (BZRA) are often used continuously at a low dosage, being the most common form of long-term use. In Belgium, BZRA use is monitored by analyzing self-report data about medication use in the last 24 h. This method provides insufficient insight into the terms of use of these psychoactive drugs. Aim: To describe trends in BZRA prescribing in Flanders, Belgium, between 2000 and 2019. Design and setting: Population-based trend analysis and a case-control study for the year 2019 were done with data from a morbidity registry in general practice. Methods: Repeated cross-sectional and joinpoint regression analyses revealed trends in sex- and age-standardized prescription rates among adult patients (18+). Results: Overall, BZRA prescriptions increased. The highest overall increase was found among male patients 18-44 years old, with an average annual percentage change of 2.5 (95% CI: 0.9, 4.3). Among 65+ female patients, a decrease was found since 2006, with an annual percentage change of -0.7 (95% CI: -1.3, -0.1). In 2019, 12% of registered patients received minimally one prescription, long-term use was observed in 5%, back pain was the most common morbidity significantly associated with a rise in BZRA prescriptions, and zolpidem was the most prescribed BZRA (22%). Conclusion: Despite some statistically significant decreasing trends, an overall increase in BZRA prescriptions was observed throughout the 19-year study period, especially among long-term users of 18-44 years and 65-plus. Zolpidem became the most prescribed BZRA and warrants more attention.

Pneumococcal vaccination prevented severe LRTIs in adults: a causal inference framework applied in registry data.

OBJECTIVES: We estimated the effect of pneumococcal vaccination (PV) on acute lower respiratory tract infections (LRTIs) in various age and risk groups using different methods within a causal inference methodological framework. STUDY DESIGN AND SETTING: We used data from a general practitioners' morbidity registry for the year 2019. Both traditional statistical methods (regression-based and propensity score methods) and machine learning techniques were deployed. Multiple imputation was used to account for missing data. Relative risks (RRs) with 95% confidence intervals were estimated. Sensitivity analyses were performed to account for the severity of LRTIs and differences in vaccination registration. RESULTS: All methods showed a standardized mean difference below 0.1 for each covariate. No method was found to be superior to another. PV (combination of conjugate and polysaccharide vaccine) had an overall protective effect for severe LRTIs. PV was protective in different age and risk groups, especially in people aged 50-84 years with an intermediate risk group. CONCLUSION: Using several techniques, PV was found to prevent severe LRTIs and confirmed the recommendations of the Belgian Superior Health Council.

Trends in the epidemiology of depression and comorbidities from 2000 to 2019 in Belgium.

BACKGROUND: Depression is a common mental disorder in family practice with an impact on global health. The aim of this study is to provide insight in the trends of epidemiological measures as well as pharmacological treatments and comorbidities of depression. METHODS: A study using data from INTEGO, a family practice registration network in Flanders, Belgium. Trends in age-standardized prevalence and incidence of depression from 2000 to 2019 as well as antidepressant prescriptions in prevalent depression cases were analyzed with join point regression. Comorbidity profiles were explored using the Cochran-Armitage test and the Jonckheere-Terpstra test. RESULTS: We identified 538 299 patients older than 15 years during the study period. We found an increasing trend in the age-standardized prevalence of depression from 6.73 % in 2000 to 9.20 % in 2019. For the incidence of depression, a decreasing trend was observed from 2000 to 2015 with an incidence of 9.42/1000 in 2000 and 6.89/1000 in 2015, followed by an increasing trend from 2015 to 2019 (incidence of 13.64/1000 in 2019). The average number of chronic diseases per patient with depression increased significantly during the study period (from 1.2 to 1.8), and the proportion of patients relative to the whole study population that received at least one antidepressant prescription per year increased between 2000 and 2019 from 26.44% to 40.16%. CONCLUSIONS: The prevalence of depression increases while the incidence sharply rises, but only in recent years. Patients with depression tend to have more comorbidities, making a multi-faceted approach to these patients more important.

Use of antibiotics and colorectal cancer risk: a primary care nested case-control study in Belgium.

OBJECTIVES: To examine the association between the use of oral antibiotics and subsequent colorectal cancer risk. DESIGN: Matched case-control study. SETTING: General practice centres participating in the Integrated Computerised Network database in Flanders, Belgium. PARTICIPANTS: In total, 1705 cases of colorectal cancer diagnosed between 01 January 2010 and 31 December 2015 were matched to 6749 controls by age, sex, comorbidity and general practice centre. PRIMARY OUTCOME MEASURE: The association between the number of prescriptions for oral antibiotics and the incidence of colorectal cancer over a period of 1-10 years, estimated by a conditional logistic regression model. RESULTS: A significantly increased risk of colorectal cancer (OR 1.25, 95% CI 1.10 to 1.44) was found in subjects with one or more prescriptions compared with those with none after correction for diabetes mellitus. No dose-response relationship was found. CONCLUSIONS: This study resulted in a modestly higher risk of having colorectal cancer diagnosed after antibiotic exposure. The main limitation was missing data on known risk factors, in particular smoking behaviour. This study did not allow us to examine the causality of the relationship, indicating the need of further investigation.

Comorbidity burden in the first three years after diagnosis in patients with rheumatoid arthritis, psoriatic arthritis or spondyloarthritis: a general practice registry-based study.

OBJECTIVES: Rheumatoid arthritis (RA), psoriatic arthritis (PsA) and spondyloarthritis (SpA) are chronic inflammatory rheumatic conditions with high levels of comorbidity requiring additional therapeutic attention. We aimed to compare the 3-year comorbidity incidence and pain medication prescription in patients diagnosed with RA, PsA or SpA versus controls. METHODS: Data between 1999 and 2012 were obtained from Intego, a general practitioner (GP) morbidity registry in Flanders, Belgium. Cases were identified by International Classification of Primary Care (ICPC-2) codes representing 'rheumatoid/seropositive arthritis (L88)' or 'musculoskeletal disease other (L99)'. The registered keywords mapped to these ICPC-2 codes were further verified and mapped to a RA/SpA/PsA diagnosis. Controls were matched on age, gender, GP practice and diagnosis date. We analysed the 3-year comorbidity burden in cases and controls, measured by the Rheumatic Diseases Comorbidity Index (RDCI). All electronically GP-prescribed drugs were registered. RESULTS: In total, 738, 229 and 167 patients were included with a diagnosis of RA, SpA or PsA, respectively. Patients with RA or PsA had comparable median RDCI scores at baseline, but higher scores at year 3 compared with controls (RA: p=0.010; PsA: p=0.008). At baseline, depression was more prevalent in PsA patients vs controls (p<0.003). RA patients had a higher 3-year incidence of cardiovascular disease including myocardial infarction than controls (p<0.035). All disease population were given more prescriptions than controls for any pain medication type, even opioids excluding tramadol. CONCLUSIONS: This study highlights the increasing comorbidity burden of patients with chronic inflammatory rheumatic conditions, especially for individuals with RA or PsA. The high opioid use in all populations was remarkable.

The effect of automated audit and feedback on data completeness in the electronic health record of the general physician: protocol for a cluster randomized controlled trial.

BACKGROUND: The electronic health record (EHR) of the general physician (GP) is an important tool that can be used to assess and improve the quality of healthcare. However, there are some problems when (re) using the data gathered in the EHR for quality assessments. One problem is the lack of data completeness in the EHR. Audit and feedback (A&F) is a well-known quality intervention that can improve the quality of healthcare. We hypothesize that an automated A&F intervention can be adapted to improve the data completeness of the EHR of the GP, more specifically, the number of correctly registered diagnoses of type 2 diabetes and chronic kidney disease. METHODS: This study is a pragmatic cluster randomized controlled trial with an intervention at the level of GP practice. The intervention consists of an audit and extended electronically delivered feedback with multiple components that will be delivered 4 times electronically to general practices over 12 months. The data will be analyzed on an aggregated level (per GP practice). The primary outcome is the percentage of correctly registered diagnoses of type 2 diabetes. The key secondary outcome is the registration of chronic kidney disease. Exploratory secondary outcomes are the registration of heart failure, biometric data and lifestyle habits, and the evolution of 4 different EHR-extractable quality indicators. DISCUSSION: This cluster randomized controlled trial intends to primarily improve the registration of type 2 diabetes in the EHR of the GP and to secondarily improve the registration of chronic kidney disease. In addition, the registration of heart failure, lifestyle parameters, and biometric data in the EHR of the GP are explored together with 4 EHR-extractable quality indicators. By doing so, this study aims to improve the data completeness of the EHR, paving the way for future quality assessments. TRIAL REGISTRATION: ClinicalTrials.gov NCT04388228 . Registered on May 14, 2020.

Epidemiology of knee osteoarthritis in general practice: a registry-based study.

OBJECTIVES: The present study investigated (1) trends in the prevalence and incidence of knee osteoarthritis over a 20-year period (1996-2015); (2) trends in multimorbidity and (3) trends in drug prescriptions. DESIGN: Registry-based study. SETTING: Primary healthcare, Flanders, Belgium. PARTICIPANTS: Data were collected from Intego, a general practice-based morbidity registration network. In the study period between 1996 and 2015, data from 440 140 unique patients were available. OUTCOME MEASURES: Trends in prevalence and incidence rate of knee osteoarthritis were computed using joinpoint regression analysis. The mean disease count was calculated to assess trends in multimorbidity. In addition, the number of drug prescriptions was identified by the Anatomical Therapeutic Chemical Classification code and trends were equally recorded with joinpoint regression. RESULTS: The total age-standardised prevalence of knee osteoarthritis increased from 2.0% in 1996 to 3.6% in 2015. An upward trend was observed with an average annual percentage change (AAPC) of 2.5 (95% CI 2.2 to 2.9). In 2015, the prevalence rates in the 10 year age groups from the 45-54 years age group onwards were 3.1%, 5.6%, 9.0% and 13.9%, to reach 15.0% in people aged 85 years and older. The incidence remained stable with 3.75‰ in 2015 (AAPC=-0.5, 95% CI -1.4 to 0.5). The mean disease count significantly increased from 1.63 to 2.34 (p<0.001) for incident cases with knee osteoarthritis. Finally, we observed a significantly positive trend in the overall prescription of acetaminophen (AAPC=6.7, 95% CI 5.6 to 7.7), weak opioids (AAPC=4.0, 95% CI 0.9 to 7.3) and glucosamine (AAPC=8.6, 95% CI 2.4 to 15.1). Oral non-steroidal anti-inflammatory drugs were most prescribed, with a prevalence rate of 29.8% in 2015, but remained stable during the study period (AAPC=0.0, 95% CI -1.1 to 1.1). CONCLUSIONS: Increased prevalence, multimorbidity, and number of drug prescriptions confirm an increased burden of knee osteoarthritis. In future, these trends can be used to prioritise initiatives for improvement in care.

Clinical decision support improves the appropriateness of laboratory test ordering in primary care without increasing diagnostic error: the ELMO cluster randomized trial.

BACKGROUND: Inappropriate laboratory test ordering poses an important burden for healthcare. Clinical decision support systems (CDSS) have been cited as promising tools to improve laboratory test ordering behavior. The objectives of this study were to evaluate the effects of an intervention that integrated a clinical decision support service into a computerized physician order entry (CPOE) on the appropriateness and volume of laboratory test ordering, and on diagnostic error in primary care. METHODS: This study was a pragmatic, cluster randomized, open-label, controlled clinical trial. SETTING: Two hundred eighty general practitioners (GPs) from 72 primary care practices in Belgium. PATIENTS: Patients aged ≥ 18 years with a laboratory test order for at least one of 17 indications: cardiovascular disease management, hypertension, check-up, chronic kidney disease (CKD), thyroid disease, type 2 diabetes mellitus, fatigue, anemia, liver disease, gout, suspicion of acute coronary syndrome (ACS), suspicion of lung embolism, rheumatoid arthritis, sexually transmitted infections (STI), acute diarrhea, chronic diarrhea, and follow-up of medication. INTERVENTIONS: The CDSS was integrated into a computerized physician order entry (CPOE) in the form of evidence-based order sets that suggested appropriate tests based on the indication provided by the general physician. MEASUREMENTS: The primary outcome of the ELMO study was the proportion of appropriate tests over the total number of ordered tests and inappropriately not-requested tests. Secondary outcomes of the ELMO study included diagnostic error, test volume, and cascade activities. RESULTS: CDSS increased the proportion of appropriate tests by 0.21 (95% CI 0.16-0.26, p < 0.0001) for all tests included in the study. GPs in the CDSS arm ordered 7 (7.15 (95% CI 3.37-10.93, p = 0.0002)) tests fewer per panel. CDSS did not increase diagnostic error. The absolute difference in proportions was a decrease of 0.66% (95% CI 1.4% decrease-0.05% increase) in possible diagnostic error. CONCLUSIONS: A CDSS in the form of order sets, integrated within the CPOE improved appropriateness and decreased volume of laboratory test ordering without increasing diagnostic error. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT02950142 , registered on October 25, 2016.