Propensity score matching in otolaryngologic literature: A systematic review and critical appraisal

Abstract

Background Propensity score techniques can reduce confounding and bias in observational studies. Such analyses are able to measure and balance pre-determined covariates between treated and untreated groups, leading to results that can approximate those generated by randomized prospective studies when such trials are not feasible. The most commonly used propensity score -based analytic technique is propensity score matching (PSM). Although PSM popularity has continued to increase in medical literature, improper methodology or methodological reporting may lead to biased interpretation of treatment effects or limited scientific reproducibility and generalizability. In this study, we aim to characterize and assess the quality of PSM methodology reporting in high-impact otolaryngologic literature. Methods PubMed and Embase based systematic review of the top 20 journals in otolaryngology, as measured by impact factor from the Journal Citations Reports from 2012 to 2018, for articles using PSM analysis throughout their publication history. Eligible articles were reviewed and assessed for quality and reporting of PSM methodology. Results Our search yielded 101 studies, of which 92 were eligible for final analysis and review. The proportion of studies utilizing PSM increased significantly over time (p < 0.001). Nearly all studies (96.7%, n = 89) specified the covariates used to calculate propensity scores. Covariate balance was illustrated in 67.4% (n = 62) of studies, most frequently through p-values. A minority (17.4%, n = 16) of studies were found to be fully reproducible according to previously established criteria. Conclusions While PSM analysis is becoming increasingly prevalent in otolaryngologic literature, the quality of PSM methodology reporting can be improved. We provide potential recommendations for authors regarding optimal reporting for analyses using PSM.

Microbial characteristics across different tongue coating types in a healthy population

“…The esophageal squamous cell carcinoma-inducing oral bacteria P. melaninogenica and V. dispar [27] were coenriched in the Y-thick greasy group. Multifunctional P. melaninogenica in saliva is also associated with oral squamous cell carcinoma [28] and is causative of periodontitis [29], Alzheimer's disease mortality [30], hyperglycemia [31], [and MHCII, CD80 and IFN-γ upregulation in salivary gland epithelial cells in Sjögren's syndrome [32]. The species A. parvulum may be associated with sepsis [33].…”

A Systematic Review of Propensity Score Matching in the Orthopedic Literature

“…To assess whether the minimum PSM methodological characteristics required for reproducibility were reported, we graded studies on the following: (1) reporting matching algorithm, (2) reporting matching ratio, (3) reporting use of replacement, and (4) reporting whether the statistical tests used to compare PSM cohorts were independent or paired. These criteria, which have been previously established [17,21], were combined to create a total possible score of 4 for each study. Studies were considered to be fully reproducible only if they satisfied all criteria.…”

Propensity score matching after multiple imputation when a confounder has missing data

“…In this application, when (nb imp ; nb rep ) goes from (20,10) to (20,30), the computational time was multiplied by 3; when it goes from (20,10) to (30,10) it was multiplied by 1.4. As we would expect, the Monte Carlo error is reduced by increasing both nb imp and nb rep .…”

“…Despite some criticism in the literature, 6 the propensity score matching method is the most widely used in practice in medical research. 7 , 8 , 9 , 10 Many different implementations of propensity score matching exist depending on: the matching algorithm, greedy or optimal, the metric used, 11 , 12 the presence and size of a caliper that limits the difference between propensity scores of a matched pair, 13 , 14 , 15 the number of non‐treated patients matched to each treated patient 16 and the presence or absence of replacement in the sampling. 12 , 17 Based on the distance between their estimated propensity scores, treated patients are matched to non‐treated patients in order to create a matched sample in which the two treatment groups have similar characteristics.…”