Methodological and reporting quality of pharmacoepidemiology studies using the target trial emulation framework: a systematic review
2Division of Nephrology, Nanfang Hospital, Southern Medical University, China
3Faculty of Social Sciences, University of Warwick, Coventry, UK, UK
Background: Many pharmacoepidemiology studies adopted the target trial emulation approach in recent years, which was considered as a useful framework for avoiding potential sources of bias. However, there is evidence that key methodological details are missing from some report of such studies.
Objectives: Our aim was to systematically evaluate the methodological and reporting quality of pharmacoepidemiology studies using the target trial emulation framework, and to examine opportunities for improvement.
Methods: We searched PubMed, Embase, and Web of Science database to identify all full-text available pharmacoepidemiology studies using the target trial emulation framework. Information regarding methodological elements were extracted and evaluated.
Results: Fifty-five studies were identified. The most common methodological limitations related to: 1) Inappropriate eligibility criteria: 10 (18.2%) studies used post-baseline events to define baseline eligibility (e.g., use of treatment, no follow-up data), which may introduce selection bias; 2) lack of ‘time zero’ synchronization: In 15 (27.3%) studies, follow-up started at eligibility, but treatment strategy is assigned after follow-up, which may introduce immortal time bias and misclassification bias. Among these, only 7(46.7%) studies described a solution to mitigate these biases; 3) Inappropriate assignment procedures: only 7(12.7%) studies reported a critical approach (e.g., use of directed acyclic graphs) for confounder selection, and only 15(27.3%) studies conducted sensitivity analysis (e.g., use of negative control or E-value method) to investigate the potential impact of residual confounder. In addition, only 13 (23.6%) studies clearly mentioned the standard reporting checklist they adhered to (10 used STRBOE checklist, 1 used ISPOR guideline, 1 used RECORD checklist, and only 1 used RECORD-PE checklist), thus there were substantial variations in methodological and reporting details.
Conclusions: Most studies did not explicitly define their target trial and failed to comply with the basic principles of the target trial emulation framework. The methodological and reporting quality of pharmacoepidemiology studies using the target trial emulation framework needs improvement. Our findings emphasize the pressing need to develop evidence-based guidelines for conducting and reporting such studies.
Patient, public and/or healthcare consumer involvement: Patient, public or healthcare consumer were not involved in the design, or conduct, or reporting, or dissemination plans of our research.