Propensity Scoring in Plastic Surgery Research: An Analysis and Best Practice Guide

Abstract

Summary: Randomized controlled trials, though considered the gold standard in clinical research, are often not feasible in plastic surgery research. Instead, researchers rely heavily on observational studies, leading to potential issues with confounding and selection bias. Propensity scoring—a statistical technique that estimates a patient’s likelihood of having received the exposure of interest—can improve the comparability of study groups by either guiding the selection of study participants or generating a covariate that can be adjusted for in multivariate analyses. In this study, we conducted a comprehensive review of research articles published in three major plastic surgery journals (Plastic and Reconstructive Surgery, Journal of Plastic, Reconstructive, & Aesthetic Surgery, and Annals of Plastic Surgery) to determine the utilization of propensity scoring methods in plastic surgery research from August 2018 to August 2020. We found that propensity scoring was used in only eight (0.8%) of 971 research articles, none of which fully reported all components of their propensity scoring methodology. We provide a brief overview of propensity score techniques and recommend guidelines for accurate reporting of propensity scoring methods for plastic surgery research. Improved understanding of propensity scoring may encourage plastic surgery researchers to incorporate the method in their own work and improve plastic surgeons’ ability to understand and analyze future research studies that utilize propensity score methods.

Quasi-Experimental Design for Health Policy Research: A Methodology Overview

“…Although matching techniques, such as synthetic controls and inverse probability weighting have been applied, PSM is the most widely used in our field. 5 PSM considers that patients are likely to have factors that influence their treatment, termed “confounding by indication.” 50 Propensity scores are weighted measures derived from the degree to which known confounders affect the likelihood of treatment. They are calculated for patients based on comorbidities or other relevant factors.…”

“…They are calculated for patients based on comorbidities or other relevant factors. 5,51 PSM then assigns similarly scored patients to different intervention groups, effectively isolating the intervention outcome.…”

“…Investigators seldom have applied these methods to plastic surgery research. 5 This article aims to build the foundation for health policy research in plastic surgery by presenting quasi-experimental research techniques in an informative style.…”

Eflornithine as Postimmunotherapy Maintenance in High-Risk Neuroblastoma: Externally Controlled, Propensity Score–Matched Survival Outcome Comparisons

“…Thus, we have used patient-level data from ANBL0032 as an external control database to identify a representative population that did not receive DFMO for comparison with DFMO-treated patients with more rigorous statistical approaches, including propensity score-matched analysis. Although randomized control trials (RCTs) remain the gold standard for treatment effect, propensity scorematched analysis is a widely published statistical method for comparing two cohorts [25][26][27][28] by effectively balancing the distribution of patient baseline characteristics and risk categories between treatment groups to reduce potential sources of confounding bias. On the basis of propensity scores (PSs), patients are placed into matched sets, and an estimation of treatment effect is obtained by comparing their outcomes.…”

Discussion: Abdominal Wall Reconstruction with Retrorectus Self-Adhering Mesh: A Single-Center Long-Term Follow-up

“…This could have introduced confounders and selection bias related to mesh selection. 8 The patient demographics in Table 1 demonstrate the limitations of their "manual" matched analysis. Although the differences in patient demographics and characteristics between the two groups were not significant (almost certainly because of the small sample), they did have clinically significant differences in comorbidities, surgical characteristics, and Kanters grade.…”