How Often Are Study Design and Level of Evidence Misreported in the Pediatric Orthopaedic Literature?

Abstract

Observational studies are the most commonly used study designs in the pediatric orthopaedic literature. The differences between observational study designs are important but not widely understood, leading to potential discrepancies between the reported and actual study design. Study design misclassification is associated with a potential for misreporting level of evidence (LOE). The purpose of this study was to determine the degree of study design and LOE misclassification in the pediatric orthopaedic literature. The Institute for Scientific Information (ISI) Web of Science was queried to identify all pediatric orthopaedic observational studies published from 2014 to 2017. Reported study design and LOE were recorded for each study. The actual study design and LOE were determined on the basis of established clinical epidemiological criteria by reviewers with advanced epidemiological training. Studies with a discrepancy between reported versus actual study design and LOE were identified. The following covariates were recorded for each study: subspecialty, inclusion of a statistician coauthor, sample size, journal, and journal impact factor. χ test was used to identify factors associated with study design and LOE misreporting. In total, 1000 articles were screened, yielding 647 observational studies. A total of 335 publications (52%) did not clearly report a study design in the abstract or manuscript text. Of those that did, 59/312 (19%) reported the incorrect study design. The largest discrepancy was in the 109 studies that were reported to be case series, among which 30 (27.5%) were actually retrospective cohort studies. In total, 313 publications (48%) did not report a LOE. Of those that did, 95/334 (28%) reported the incorrect LOE. In total, 33 studies (19%) reported a LOE that was higher than the actual LOE and 62 (35%) under-reported the LOE. The majority of observational pediatric orthopaedic studies did not report a study design or reported the wrong study design. Similarly, the majority of studies did not report or misreported their LOE. Greater epidemiological rigor in evaluating observational studies is required on the part of investigators, reviewers, and editors. Level II.