Assessing The Performance of Physician’s Prescribing Preference As An Instrumental Variable in Comparative Effectiveness Research With Moderate and Small Sample Sizes: A Simulation Study

BackgroundInstrumental variable (IV) analyses are used to account for unmeasured confounding in Comparative Effectiveness Research (CER) in pharmacoepidemiology. To date, simulation studies assessing the performance of IV analyses have been based on large samples. However, in many settings, sample sizes are not large.Objective In this simulation study, we assess the utility of Physician’s Prescribing Preference (PPP) as an IV for moderate and smaller sample sizes.MethodsWe designed a simulation study in a CER setting with moderate (around 2500) and small (around 600) sample sizes. The outcome and treatment variables were binary and three variables were used to represent confounding (a binary and a continuous variable representing measured confounding, and a further continuous variable representing unmeasured confounding). We compare the performance of IV and non-IV approaches using two-stage least squares (2SLS) and ordinary least squares (OLS) methods, respectively. Further, we test the performance of different forms of proxies for PPP as an IV.ResultsThe PPP IV approach results in a percent bias of approximately 20%, while the percent bias of OLS is close to 60%. The sample size is not associated with the level of bias for the PPP IV approach. However, smaller sample sizes led to lower statistical power for the PPP IV. Using proxies for PPP based on longer prescription histories result in stronger IVs, partly offsetting the effect on power of smaller sample sizes.Conclusion Irrespective of sample size, the PPP IV approach leads to less biased estimates of treatment effectiveness than conventional multivariable regression adjusting for known confounding only. Particularly for smaller sample sizes, we recommend constructing PPP from long prescribing histories to improve statistical power.

Understanding Source-Related Influences on Physicians’ Translating Comparative Effectiveness Research into Patient Care: Results from a Study of Cardiologists

Comparative effectiveness research (CER), which refers to an evaluation of the clinical effectiveness of two or more medical interventions that are used to treat the same condition, has the potential to inform decision-making in both policy circles and physicians’ exam rooms. The ability of stakeholders to translate that research into practice has important implications for health outcomes, but the impact of information sources on physicians in translating CER remains understudied. This project examines the source-related influences on and motivations of cardiologists with respect to willingness to make changes in their practice based on emerging CER results. The results from this survey of cardiologists (N = 42) indicate that the source of information (including perceived credibility of those sources) matters greatly to cardiologists when deciding whether to make a change in practice. These findings suggest data-based implications for researchers and practitioners that are engaged in closing the CER translation gap.

Analysing and reporting of observational data: a systematic review informing the EULAR points to consider when analysing and reporting comparative effectiveness research with observational data in rheumatology

ObjectivesTo evaluate the analysis and reporting of comparative effectiveness research with observational data in rheumatology, informing European Alliance of Associations for Rheumatology points to consider.MethodsWe performed a systematic literature review searching Ovid MEDLINE for original articles comparing drug effectiveness in longitudinal observational studies, published in key rheumatology journals between 2008 and 2019. The extracted information focused on reporting and types of analyses. We evaluated if year of publication impacted results.ResultsFrom 9969 abstracts reviewed, 211 articles fulfilled the inclusion criteria. Ten per cent of studies did not adjust for confounding factors. Some studies did not explain how they chose covariates for adjustment (9%), used bivariate screening (21%) and/or stepwise selection procedures (18%). Only 33% studies reported the number of patients lost to follow-up and 25% acknowledged attrition (drop-out or treatment cessation). To account for attrition, studies used non-responder imputation, followed by last observation carried forward (LOCF) and complete case (CC) analyses. Most studies did not report the number of missing data on covariates (83%), and when addressed, 49% used CC and 11% LOCF. Date of publication did not influence the results.ConclusionMost studies did not acknowledge missing data and attrition, and a tenth did not adjust for any confounding factors. When attempting to account for them, several studies used methods which potentially increase bias (LOCF, CC analysis, bivariate screening…). This study shows that there is no improvement over the last decade, highlighting the need for recommendations for the assessment and reporting of comparative drug effectiveness in observational data in rheumatology.