Use and Misuse of Multivariable Approaches in Interventional Cardiology Studies on Drug‐Eluting Stents: A Systematic Review

Abstract

Better exploitation and methodological appraisal of multivariable analysis is needed to improve the clinical and research impact and reliability of nonrandomized studies.

Reporting of covariate selection and balance assessment in propensity score analysis is suboptimal: a systematic review

“…Our study is consistent with previous systematic reviews with respect to low quality of reporting and/or conduct of PS analysis [8,9,11,12]. However, our review included a large number of recently published studies to evaluate the current status in conducting and reporting of PS analysis.…”

“…First, those studies that reported aspects of the PS analysis used inappropriate statistical methods such as significance testing or C-statistic. Second, inconsistency of reporting was observed irrespective of the IF of the journal in which the article was published and even with in a specific journal, which is also in line with a previous systematic review [12].…”

“…Although the use of PS methods has shown a dramatic increase in the medical literature [8], previous literature reviews indicated that most authors do not adequately report information on the PS model development [9,10], the balance of covariates between the treatment groups in PS analysis [8,9,11,12], and those who report, often use inappropriate diagnostics [8,9,11]. In addition, researchers often ignore explicit discussion of the PS estimate (estimand) and its relationship with their research question [5].…”

Transcatheter aortic valve implantation in low ejection fraction/low transvalvular gradient patients

“…Cox proportional hazards regression was performed for the multivariable analyses. 16 The multivariable models included all the covariates of Tables 1 and 2 with a P value of less than 0.10 in univariate analyses, and the model was performed separately for LVEF 40% or less and MTG less than 40 mmHg and for the association of LVEF 40% or less and MTG less than 40 mmHg. Results are presented as hazard ratios for Cox proportional hazards regression analysis with 95% confidence intervals (CIs).…”

High sensitive TROponin levels In Patients with Chest pain and kidney disease: A multicenter registry — The TROPIC study

“…In primary prevention, a recent study involving more than one million patients demonstrated an incidence of myocardial infarction similar in CKD patients compared to diabetic patients [24], introducing the concept of renal disease as another coronary heart disease equivalent. Actually CKD at its different stages is characterized by oxidative stress, inflammation, and dyslipidemia, a combination which promotes accelerated atherosclerosis [25,26].…”