Comparison of Machine Learning and Conventional Statistical Modeling for Predicting Readmission Following Acute Heart Failure Hospitalization.

INTRODUCTION: Developing accurate models for predicting the risk of 30-day readmission is a major healthcare interest. Evidence suggests that models developed using machine learning (ML) may have better discrimination than conventional statistical models (CSM), but the calibration of such models is unclear. OBJECTIVES: To compare models developed using ML with those developed using CSM to predict 30-day readmission for cardiovascular and non-cardiovascular causes in HF patients. METHODS: We retrospectively enrolled 10,919 patients with HF (> 18 years) discharged alive from a hospital or emergency department (2004-2007) in Ontario, Canada. The study sample was randomly divided into training and validation sets in a 2:1 ratio. CSMs to predict 30-day readmission were developed using Fine-Gray subdistribution hazards regression (treating death as a competing risk), and the ML algorithm employed random survival forests for competing risks (RSF-CR). Models were evaluated in the validation set using both discrimination and calibration metrics. RESULTS: In the validation sample of 3602 patients, RSF-CR (c-statistic=0.620) showed similar discrimination to the Fine-Gray competing risk model (c-statistic=0.621) for 30-day cardiovascular readmission. In contrast, for 30-day non-cardiovascular readmission, the Fine-Gray model (c-statistic=0.641) slightly outperformed the RSF-CR model (c-statistic=0.632). For both outcomes, The Fine-Gray model displayed better calibration than RSF-CR using calibration plots of observed vs. predicted risks across the deciles of predicted risk. CONCLUSIONS: Fine-Gray models had similar discrimination but superior calibration to the RSF-CR model, highlighting the importance of reporting calibration metrics for ML-based prediction models. The discrimination was modest in all readmission prediction models regardless of the methods used.

Apixaban versus Rivaroxaban in Patients with Atrial Fibrillation at High or Low Bleeding Risk: A Population-Based Cohort Study.

BACKGROUND: Despite many atrial fibrillation (AF) patients being at risk of bleeding, very limited data are available on bleeding rates of different direct oral anticoagulants (DOACs) based on the spectrum of bleeding risk. OBJECTIVE: To compare the risk of major bleeding and thromboembolic events with apixaban versus rivaroxaban among AF patients, stratified by bleeding risk. METHODS: We conducted a population-based, retrospective cohort study of all adult patients (66 years or older) with AF in Ontario, Canada, who were treated with apixaban or rivaroxaban between April 1, 2011 and March 31, 2020. Bleeding risk was estimated using the HAS-BLED score with high bleeding risk defined as a score of 3 or greater. The primary safety outcome was major bleeding and the primary efficacy outcome was thromboembolic events. Comparisons were adjusted for baseline comorbidities using inverse probability of treatment weighting (IPTW). RESULTS: This study included 18,156 AF patients with high bleeding risk and 55,186 AF patients with low bleeding risk. Apixaban use was more common in high bleeding risk patients; 63% of high risk patients used apixaban compared to 56% of low risk patients. Apixaban users had lower rates of major bleeding in high risk patients (2.9% vs 4.2% per year; HR 0.69 [95%CI, 0.58-0.81]) and in low risk patients (1.8% vs 2.9% per year; HR 0.63 [95%CI, 0.56-0.70]), compared with rivaroxaban. There were no differences in rates of thromboembolic events 3.1% vs 3.0% per year (HR 1.02 [95%CI, 0.86-1.22]) in high risk patients and 1.9% vs 1.9% per year (HR 1.00 [95%CI, 0.89-1.14]) in low risk patients. CONCLUSIONS: Among older AF patients with high or low bleeding risk, treatment with apixaban was associated with lower rates of major bleeding with no difference in risk for thromboembolic events compared with rivaroxaban.

Comparative Safety Analysis of Nabilone Versus Opioids: A Population-Based Cohort Study.

BACKGROUND: Some have advocated that nabilone be used rather than opioids to manage chronic, noncancer pain, since the former drug may have a better safety profile. OBJECTIVE: We compared the safety of incident nabilone use relative to incident opioid use with respect to multiple clinically important outcomes. DESIGN: A population-based, retrospective cohort study. SETTING: Province of Ontario, Canada. PARTICIPANTS: Persons aged 12 years and older, diagnosed with a musculoskeletal condition within the past 3 years prior to the index date. EXPOSURES: Incident nabilone use, with incident opioid use serving as the reference group. MEASUREMENTS: Within 3 months following the index date, we separately evaluated for pneumonia, motor vehicle accidents, falls or fractures, mental and behavioral disorder due to psychoactive substance use, and all-cause mortality. RESULTS: A total of 18,863 incident nabilone users were propensity score matched to an equal number of opioid users. In the overall matched analysis, incident nabilone users vs. incident opioid users had significantly lower rates of pneumonia (hazard ratio [HR] 0.78, 95% CI 0.63-0.96), falls or fractures (HR 0.56, 95% CI 0.50-0.64), and all-cause mortality (HR 0.79, 95% CI 0.65-0.95), but significantly higher rate of mental or behavioral disorder (HR 2.23, 95% CI 1.45-3.43). There was no significant difference between groups with respect to rate of motor vehicle accidents. LIMITATIONS: Unmeasured confounding may have influenced results. CONCLUSIONS: While usage of nabilone relative to opioids was associated with reduced rates of pneumonia, falls or fractures, and all-cause mortality, it was simultaneously associated with an increased rate of adverse mental health outcomes. This picture of mixed safety results raises concerns with the policy approach of broadly substituting use of opioids with nabilone. FUNDING SOURCE: Ontario Ministry of Health.

Minimizing confounding in comparative observational studies with time-to-event outcomes: An extensive comparison of covariate balancing methods using Monte Carlo simulation.

Observational studies are frequently used in clinical research to estimate the effects of treatments or exposures on outcomes. To reduce the effects of confounding when estimating treatment effects, covariate balancing methods are frequently implemented. This study evaluated, using extensive Monte Carlo simulation, several methods of covariate balancing, and two methods for propensity score estimation, for estimating the average treatment effect on the treated using a hazard ratio from a Cox proportional hazards model. With respect to minimizing bias and maximizing accuracy (as measured by the mean square error) of the treatment effect, the average treatment effect on the treated weighting, fine stratification, and optimal full matching with a conventional logistic regression model for the propensity score performed best across all simulated conditions. Other methods performed well in specific circumstances, such as pair matching when sample sizes were large (n = 5000) and the proportion treated was < 0.25. Statistical power was generally higher for weighting methods than matching methods, and Type I error rates were at or below the nominal level for balancing methods with unbiased treatment effect estimates. There was also a decreasing effective sample size with an increasing number of strata, therefore for stratification-based weighting methods, it may be important to consider fewer strata. Generally, we recommend methods that performed well in our simulations, although the identification of methods that performed well is necessarily limited by the specific features of our simulation. The methods are illustrated using a real-world example comparing beta blockers and angiotensin-converting enzyme inhibitors among hypertensive patients at risk for incident stroke.

Persistent disparities in insulin pump uptake despite a universal pump programme for type 1 diabetes in Ontario, Canada.

AIM: To evaluate associations between social disadvantage and insulin pump use among adults with type 1 diabetes (T1D) in the context of a universal publicly funded insulin pump programme in Ontario, Canada, and to ascertain whether social disparities in insulin pump programme enrolment have decreased over time. METHODS: Population-based cross-sectional studies were conducted using administrative healthcare data in Ontario, Canada. First, among adults aged older than 18 years diagnosed with T1D before 31 March 2021, logistic regression was used to assess the association between neighbourhood social disadvantage (Ontario marginalization index quintiles) and insulin pump use. Second, among all paediatric and adult applicants to the insulin pump programme from 1 September 2006 to 31 March 2022, ordinal logistic regression was used to evaluate associations between year of insulin pump initiation and social disadvantage. RESULTS: Among 27 453 adults with T1D, 60% used insulin pumps. Greater social disadvantage was associated with lower odds of insulin pump use (adjusted odds ratio [OR] 0.44 [95% confidence interval {CI} 0.39-0.48] for greatest vs. lowest social disadvantage quintile). Among 21 002 paediatric and adult applicants to the insulin pump programme, social disparities in pump use decreased in the first 3 years of the programme, plateaued until 2020, then increased from 2020 to 2022, with no change in the odds of being in a higher social deprivation quintile for 2022 relative to 2007 (OR 1.09 [95% CI 0.83-1.44]). CONCLUSIONS: Despite a universal pump programme for individuals with T1D, disparities by social disadvantage persist. Residual financial and non-financial barriers must be addressed to promote equitable insulin pump uptake.

Timing of follow-up visits after hospital discharge for COPD: Application of a new method.

RATIONALE: A common strategy to reduce COPD readmissions is to encourage patient follow-up with a physician within 1 to 2 weeks of discharge, yet evidence confirming its benefit is lacking. We used a new study design called target randomized trial emulation to determine the impact of follow-up visit timing on patient outcomes. METHODS: All Ontario residents aged 35 or older discharged from a COPD hospitalization were identified using health administrative data and randomly assigned to those who received and did not receive physician visit follow-up by within seven days. They were followed to all-cause emergency department visits, readmissions or death. Targeted randomized trial emulation was used to adjust for differences between the groups. COPD emergency department visits, readmissions or death was also considered. RESULTS: There were 94,034 patients hospitalized with COPD, of whom 73.5% had a physician visit within 30 days of discharge. Adjusted hazard ratio for all-cause readmission, emergency department visits or death for people with a visit within seven days post discharge was 1.03 (95% Confidence Interval [CI]: 1.01-1.05) and remained around 1 for subsequent days; adjusted hazard ratio for the composite COPD events was 0.97 (95% CI 0.95-1.00) and remained significantly lower than 1 for subsequent days. CONCLUSION: While a physician visit after discharge was found to reduce COPD events, a specific time period when a physician visit was most beneficial was not found. This suggests that follow-up visits should not occur at a predetermined time but be based on factors such as anticipated medical need.

Wait-times Benchmarks for risk-based prioritization in Transcatheter Aortic Valve Implantation: a simulation study.

BACKGROUND: Demand for transcatheter aortic valve implantation (TAVI) has increased in the last decade, resulting in prolonged wait-times and undesirable health outcomes in many health systems. Risk-based prioritization and wait-times benchmarks can improve equitable access to patients. METHODS: We used simulation models to follow-up a synthetic population of 50,000 individuals from referral to completion of TAVI. Based on their risk of adverse events, patients could be classified as "low-", "medium-" and "high-risk", and shorter wait-times were assigned for the higher risk groups. We assessed the impacts of the size and wait-times for each risk group on waitlist mortality, hospitalization and urgent TAVIs. All scenarios had the same resource constraints, allowing us to explore the trade-offs between faster access for prioritized patients and deferred access for non-prioritized groups. RESULTS: Increasing the proportion of patients categorized as high-risk, and providing more rapid access to the higher-risk groups achieved the greatest reductions in mortality, hospitalizations and urgent TAVIs (relative reductions of up to 29%, 23% and 38%, respectively). However, this occurs at the expense of excessive wait-times in the non-prioritized low-risk group (up to 25 weeks). We propose wait-times of up to 3 weeks for high-risk patients and 7 weeks for medium-risk patients. CONCLUSIONS: Prioritizing higher-risk patients with faster access leads to better health outcomes, however this also results in unacceptably long wait-times for the non-prioritized groups in settings with limited capacity. Decision-makers must be aware of these implications when developing and implementing waitlist prioritization strategies.

Disparities in diabetes processes of care among people experiencing homelessness: An opportunity for intervention.

AIMS: To compare processes of diabetes care by homeless status. METHODS: A population-based propensity matched cohort study was conducted in Ontario, Canada. People with diabetes were identified in administrative healthcare data between April 2006 and March 2019. Those with a documented history of homelessness were matched to non-homeless controls. Data on processes of care measures included glucose monitoring tests, screening for microvascular complications, and physician follow-up. Differences in processes of care were compared by homeless status using proportions, risk ratios, and rate ratios. RESULTS: Of the 1,076,437 people with diabetes, 5219 matched pairs were identified. Homelessness was associated with fewer tests for glycated hemoglobin (RR = 0.63; 95 %CI: 0.60-0.67), LDL cholesterol (RR = 0.80; 95 %CI: 0.78-0.82), serum creatinine (RR = 0.94; 95 %CI: 0.92-0.97), urine protein quantification (RR = 0.62; 95 %CI: 0.59-0.66), and eye examinations (RR = 0.74; 95 %CI: 0.71-0.77). People with a history of homelessness were less likely to use primary care for diabetes management (RR = 0.62; 95 %CI: 0.59-0.66) or specialist care (RR = 0.87; 95 %CI: 0.83-0.91) compared to non-homeless controls. CONCLUSIONS: Disparities in diabetes care are evident for people with a history of homelessness and contribute to excess morbidity in this population. These data provide an impetus for investment in tailored interventions to improve healthcare equity and prevent long-term complications.

Antipsychotic Discontinuation and New Trazodone Use in Ontario Nursing Homes: Evidence of Medication Substitution.

OBJECTIVES: An unintended consequence of efforts to reduce antipsychotic medications in nursing homes is the increase in use of other psychotropic medications; however, evidence of substitution remains limited. Our objective was to measure individual-level prescribing patterns consistent with substitution of trazodone for antipsychotics. DESIGN: Retrospective cohort study. SETTING AND PARTICIPANTS: Residents of Ontario nursing homes aged 66-105 years with an admission assessment between April 1, 2010, and March 31, 2019, who were receiving an antipsychotic and had no antidepressant medication use at admission to the nursing home. METHODS: We used linked health administrative data to examine changes in medication use over three quarterly assessments following admission. Antipsychotic and trazodone use were measured at each assessment. The rate of trazodone initiation was compared between residents no longer dispensed an antipsychotic (discontinued) and those with an ongoing antipsychotic (continued) using discrete time survival analysis, controlling for baseline resident characteristics. RESULTS: We identified 13,306 residents dispensed an antipsychotic with no antidepressant use at admission (mean age 84 years, 61.5% women, 82.8% with dementia). As of the first quarterly assessment, nearly 20% of residents no longer received an antipsychotic and 9% received a new trazodone medication. Over time, residents who discontinued antipsychotics had a rate of trazodone initiation that was 82% higher compared to residents who continued (adjusted hazard ratio 1.82, 95% CI 1.66-2.00). CONCLUSIONS AND IMPLICATIONS: Residents admitted to a nursing home with antipsychotic use had a higher rate of trazodone initiation if they discontinued (vs continued) an antipsychotic. These findings suggest antipsychotic substitution with trazodone after entering a nursing home.

Comparative effectiveness and safety of apixaban and rivaroxaban in older patients with atrial fibrillation: A population-based cohort study.

BACKGROUND: There are no clinical trials with a head-to-head comparison between the 2 most commonly used oral anticoagulants (apixaban and rivaroxaban) in patients with atrial fibrillation (AF). The comparative efficacy and safety between these drugs remain unclear, especially in older patients who are at the highest risk for stroke and bleeding. OBJECTIVE: The purpose of this study was to compare the risk of major bleeding and thromboembolic events between apixaban and rivaroxaban in older patients with AF. METHODS: We conducted a population-based retrospective cohort study of all adult patients (66 years or older) with AF in Ontario, Canada, who were treated with apixaban or rivaroxaban between April 1, 2011, and March 31, 2020. The primary safety outcome was major bleeding, and the primary efficacy outcome was thromboembolic events. Secondary outcomes included any bleeding. Rates and hazard ratios (HRs) were adjusted for baseline comorbidities with inverse probability of treatment weighting. RESULTS: This study included 42,617 patients with AF treated with apixaban and 30,725 patients treated with rivaroxaban. After inverse probability of treatment weighting using the propensity score, patients in the apixaban and rivaroxaban groups were well balanced for baseline values of demographic characteristics, comorbidities, and medications; both groups had a similar mean age of 77.4 years, and 49.9% were female. At 1 year, the apixaban group had a lower risk for both major bleeding with an absolute risk reduction at 1 year of 1.1% (2.1% vs 3.2%; HR 0.65; 95% confidence interval [CI] 0.59-0.71]) and any bleeding (8.1% vs 10.9%; HR 0.73; 95% CI 0.69-0.77), with no difference in the risk for thromboembolic events (2.2% vs 2.2%; HR 1.02; 95% CI 0.92-1.13). CONCLUSION: In patients with AF, 66 years or older, treatment with apixaban was associated with lower risk for major bleeding, with no difference in the risk for thromboembolic events compared with rivaroxaban.

Association Between Immigration Status and Ambulatory Secondary Stroke Preventive Care in Ontario, Canada.

BACKGROUND AND OBJECTIVES: Secondary stroke preventive care includes evaluation and control of vascular risk factors to prevent stroke recurrence. Our objective was to evaluate the quality of ambulatory stroke preventive care and its variation by immigration status in adult stroke survivors in Ontario, Canada. METHODS: We conducted a population-based administrative database-derived retrospective cohort study in Ontario, Canada. Using immigration records, we defined immigrants as those immigrating after 1985 and long-term residents as those arriving before 1985 or those born in Canada. We included community-dwelling stroke survivors 40 years and older with a first-ever stroke between 2011 and 2017. In the year following their stroke, we evaluated the following metrics of stroke prevention: testing for hyperlipidemia and diabetes; among those with the condition, control of diabetes (hemoglobin A1c ≤7%) and hyperlipidemia (low-density lipoprotein <2 mmol/L); medication use to control hypertension, diabetes, and atrial fibrillation; and visit to a family physician and a specialist (neurologist, cardiologist, or geriatrician). We determined age and sex-adjusted absolute prevalence difference (APD) between immigrants and long-term residents for each metric using generalized linear models with binomial distribution and an identity link function. RESULTS: We included 34,947 stroke survivors (median age 70 years, 46.9% women) of whom 12.4% were immigrants. The receipt of each metric ranged from 68% to 90%. Compared with long-term residents, after adjusting for age and sex, immigrants were slightly more likely to receive screening for hyperlipidemia (APD 5.58%; 95% CI 4.18-6.96) and diabetes (5.49%; 3.76-7.23), have visits to family physicians (1.19%; 0.49-1.90), receive a prescription for antihypertensive (3.12%; 1.76-4.49) and antihyperglycemic medications (9.51%; 6.46-12.57), and achieve control of hyperlipidemia (3.82%; 1.01-6.63). By contrast, they were less likely to achieve diabetes control (-4.79%; -7.86 to -1.72) or have visits to a specialist (-1.68%; -3.12 to -0.24). There was minimal variation by region of origin or time since immigration in immigrants. DISCUSSION: Compared with long-term residents, many metrics of secondary stroke preventive care were better in immigrants, albeit with small absolute differences. However, future work is needed to identify and mitigate the factors associated with the suboptimal quality of stroke preventive care for all stroke survivors.

The effect of number of clusters and magnitude of within-cluster homogeneity in outcomes on the performance of four variance estimators for a marginal multivariable Cox regression model fit to clustered data in the context of observational research.

Researchers often estimate the association between the hazard of a time-to-event outcome and the characteristics of individuals and the clusters in which individuals are nested. Lin and Wei's robust variance estimator is often used with a Cox regression model fit to clustered data. Recently, alternative variance estimators have been proposed: the Fay-Graubard estimator, the Kauermann-Carroll estimator, and the Mancl-DeRouen estimator. Using Monte Carlo simulations, we found that, when fitting a marginal Cox regression model with both individual-level and cluster-level covariates: (i) in the presence of weak to moderate within-cluster homogeneity of outcomes, the Lin-Wei variance estimator can result in estimates of the SE with moderate bias when the number of clusters is fewer than 20-30, while in the presence of strong within-cluster homogeneity, it can result in biased estimation even when the number of clusters is as large as 100; (ii) when the number of clusters was less than approximately 20, the Fay-Graubard variance estimator tended to result in estimates of SE with the lowest bias; (iii) when the number of clusters exceeded approximately 20, the Mancl-DeRouen estimator tended to result in estimated standard errors with the lowest bias; (iv) the Mancl-DeRouen estimator used with a t-distribution tended to result in 95% confidence that had the best performance of the estimators; (v) when the magnitude of within-cluster homogeneity in outcomes was strong or very strong, all methods resulted in confidence intervals with lower than advertised coverage rates even when the number of clusters was very large.

Ambulatory Cardiology or General Internal Medicine Assessment Prior to Scheduled Major Vascular Surgery is Associated with Improved Outcomes.

OBJECTIVE: To characterize the association between ambulatory cardiology or general internal medicine (GIM) assessment prior to surgery and outcomes following scheduled major vascular surgery. BACKGROUND: Cardiovascular risk assessment and management prior to high-risk surgery remains an evolving area of care. METHODS: This is population-based retrospective cohort study of all adults who underwent scheduled major vascular surgery in Ontario, Canada, April 1, 2004-March 31, 2019. Patients who had an ambulatory cardiology and/or GIM assessment within 6 months prior to surgery were compared to those who did not. The primary outcome was 30-day mortality. Secondary outcomes included: composite of 30-day mortality, myocardial infarction or stroke; 30-day cardiovascular death; 1-year mortality; composite of 1-year mortality, myocardial infarction or stroke; and 1-year cardiovascular death. Cox proportional hazard regression using inverse probability of treatment weighting (IPTW) was used to mitigate confounding by indication. RESULTS: Among 50,228 patients, 20,484 (40.8%) underwent an ambulatory assessment prior to surgery: 11,074 (54.1%) with cardiology, 8,071 (39.4%) with GIM and 1,339 (6.5%) with both. Compared to patients who did not, those who underwent an assessment had a higher Revised Cardiac Risk Index (N with Index over 2= 4,989[24.4%] vs. 4,587[15.4%], P<0.001) and more frequent pre-operative cardiac testing (N=7,772[37.9%] vs. 6,113[20.6%], P<0.001) but, lower 30-day mortality (N=551[2.7%] vs. 970[3.3%], P<0.001). After application of IPTW, cardiology or GIM assessment prior to surgery remained associated with a lower 30-day mortality (weighted Hazard Ratio [95%CI] = 0.73 [0.65-0.82]) and a lower rate of all secondary outcomes. CONCLUSIONS: Major vascular surgery patients assessed by a cardiology or GIM physician prior to surgery have better outcomes than those who are not. Further research is needed to better understand potential mechanisms of benefit.

Derivation and internal validation of the multivariate toxigenic C. difficile diarrhea model and risk score for emergency room and hospitalized patients with diarrhea.

Background: Many factors have been associated with the risk of toxigenic C. difficile diarrhea (TCdD). This study derived and internally validated a multivariate model for estimating the risk of TCdD in patients with diarrhea using readily available clinical factors. Methods: A random sample of 3,050 symptomatic emergency department or hospitalized patients undergoing testing for toxigenic C. difficile at a single teaching hospital between 2014 and 2018 was created. Unformed stool samples positive for both glutamate dehydrogenase antigen by enzyme immunoassay and tcdB gene by polymerase chain reaction were classified as TCdD positive. The TCdD Model was created using logistic regression and was modified to the TCdD Risk Score to facilitate its use. Results: 8.1% of patients were TCdD positive. TCdD risk increased with abdominal pain (adjusted odds ratio 1.3; 95% CI, 1.0-1.8), previous C. difficile diarrhea (2.5, 1.1-6.1), and prior antibiotic exposure, especially when sampled in the emergency department (4.2, 2.5-7.0) versus the hospital (1.7, 1.3-2.3). TCdD risk also increased when testing occurred earlier during the hospitalization encounter, when age and white cell count increased concurrently, and with decreased eosinophil count. In internal validation, the TCdD Model had moderate discrimination (optimism-corrected C-statistic 0.65, 0.62-0.68) and good calibration (optimism-corrected Integrated Calibration Index [ICI] 0.017, 0.001-0.022). Performance decreased slightly for the TCdD Risk Score (C-statistic 0.63, 0.62-0.63; ICI 0.038, 0.004-0.038). Conclusions: TCdD risk can be predicted using readily available clinical risk factors with modest accuracy.

Implications of Five Different Risk Models In Primary Prevention Guidelines.

BACKGROUND: A lack of consensus exists across guidelines as to which risk model should be used for the primary prevention of cardiovascular disease (CVD). Our objective was to determine potential improvements in the number needed to treat (NNT) and number of events prevented (NEP) using different risk models in patients eligible for risk stratification. METHODS: A retrospective observational cohort was assembled from primary care patients in Ontario, Canada between January 1st, 2010, to December 31st, 2014 and followed for up to 5 years. Risk estimation was undertaken in patients 40-75 years of age, without CVD, diabetes, or chronic kidney disease using the Framingham Risk Score (FRS), Pooled Cohort Equations (PCEs), a recalibrated FRS (R-FRS), Systematic Coronary Risk Evaluation 2 (SCORE2), and the low-risk region recalibrated SCORE2 (LR-SCORE2). RESULTS: The cohort consisted of 47,399 patients (59% women, mean age 54). The NNT with statins was lowest for SCORE2 at 40, followed by LR-SCORE2 at 41, R-FRS at 43, PCEs at 55, and FRS at 65. Models that selected for individuals with a lower NNT recommended statins to fewer, but higher risk patients. For instance, SCORE2 recommended statins to 7.9% of patients (5-year CVD incidence 5.92%). The FRS, however, recommended statins to 34.6% of patients (5-year CVD incidence 4.01%). Accordingly, the NEP was highest for the FRS at 406 and lowest for SCORE2 at 156. CONCLUSIONS: Newer models such as SCORE2 may improve statin allocation to higher risk groups with a lower NNT but prevent fewer events at the population level.

Accounting for Competing Risks in Clinical Research.

This JAMA Guide to Statistics and Methods article discusses accounting for competing risks in clinical research.

Population Trends of New Prescriptions for Antihyperglycemics and Antihypertensives Between 2014 and 2022.

BACKGROUND: In the wake of pandemic-related health decline and health care disruptions, there are concerns that previous gains for cardiovascular risk factors may have stalled or reversed. Population-level excess burden of drug-treated diabetes and hypertension during the pandemic compared with baseline is not well characterized. We evaluated the change in incident prescription claims for antihyperglycemics and antihypertensives before versus during the pandemic. METHODS AND RESULTS: In this retrospective, serial, cross-sectional, population-based study, we used interrupted time series analyses to examine changes in the age- and sex-standardized monthly rate of incident prescriptions for antihyperglycemics and antihypertensives in patients aged ≥66 years in Ontario, Canada, before the pandemic (April 2014 to March 2020) compared with during the pandemic (July 2020 to November 2022). Incident claim was defined as the first prescription filled for any medication in these classes. The characteristics of patients with incident prescriptions of antihyperglycemics (n=151 888) or antihypertensives (n=368 123) before the pandemic were comparable with their pandemic counterparts (antihyperglycemics, n=97 015; antihypertensives, n=146 524). Before the pandemic, monthly rates of incident prescriptions were decreasing (-0.03 per 10 000 individuals [95% CI, -0.04 to -0.01] for antihyperglycemics; -0.14 [95% CI, -0.18 to -0.10] for antihypertensives). After July 2020, monthly rates increased (postinterruption trend 0.31 per 10 000 individuals [95% CI, 0.28-0.34] for antihyperglycemics; 0.19 [95% CI, 0.14-0.23] for antihypertensives). CONCLUSIONS: Population-level increases in new antihyperglycemic and antihypertensive prescriptions during the pandemic reversed prepandemic declines and were sustained for >2 years. Our findings are concerning for current and future cardiovascular health.