Clinical and Administrative Databases Used in Lower Extremity Arthroplasty Research.

Abstract

The use of national databases in lower extremity arthroplasty research has grown rapidly in recent years. We aimed to better characterize available databases by: (1) quantifying the number of these studies in the highest impact arthroplasty journals; (2) comparing respective sample sizes; and (3) contrasting their measured variables/outcomes. An extensive literature search was conducted to identify all database studies in the top 12 highest impact factor journals that published arthroplasty research between January 1, 2018 and December 31, 2019. A total of 5070 publications were identified. These studies were sorted by both database utilized and journal published. Tables were constructed to compare/contrast databases by metrics and measured outcome parameters including coding, patient sample size, preoperative comorbidities, postoperative complications, and limitations/barriers to their use. Four hundred twenty-six database studies (8.4%, range 0.4%-29.7% per journal) were identified, of which 139 were from non-English-speaking arthroplasty databases. Among English-speaking arthroplasty databases, the 5 most common sources were National Surgical Quality Improvement Project (n = 72), Medicare (n = 62, 39 from Medicare Claims and 23 from PearlDiver), Nationwide Inpatient Sample (n = 35), PearlDiver non-Medicare private insurance (n = 18), and Statewide Planning and Research Cooperative System (n = 18). Metrics, outcome parameters, and features of commonly used registries were reviewed. Database studies constitute an important part of arthroplasty-specific orthopedic research. Their use will continue to grow in the future, and it would be beneficial for clinicians/researchers to be aware of and familiarize themselves with their features to understand which are most appropriate for their work.