Research inefficiencies in cardiovascular clinical prediction rule development: Exploration of causes and implications.
Clinical prediction rules (CPRs) help address uncertainties in decision making using information from patients. Once a CPR is constructed in a derivation study, it should be applied to different populations to confirm its performance in external validation studies. Before it is implemented in practice, whether it improves outcomes should be evaluated in impact studies. Many cardiovascular CPRs have been developed, but most are not externally validated, and few has an impact study.
The aim of my thesis is to describe inefficiencies in cardiovascular CPR research, explore potential causes, and evaluate how they influence the uptake in primary care. Firstly, I reviewed the introduction sections of derivation studies and showed authors often do not justify new cardiovascular CPRs by citing existing CPRs although most authors indicated it was important to do so in the survey.
Secondly, I conducted a time-to-event analysis to show that most cardiovascular CPRs are not independently and externally validated even many years after their derivations. I also explored whether characteristics of derivation, reporting, and publication are associated with time to the first independent external validation.
I am currently assessing whether studies evaluating the impact of cardiovascular CPRs use proper design and methods, and provide clear report. I am also exploring how research inefficiencies influence the uptake of cardiovascular CPRs in practice.