Collection of Patient-Reported Outcome Measures: Comparing Paper to Online Data Collection Among Knee Arthroplasty Patients.

Abstract

At our institution, patient-reported outcome measures were completed on paper forms until 2021, when we began sending emails to knee arthroplasty patients so they could complete surveys electronically. This study evaluated our transition from paper-based to electronic collection of the Knee injury and Osteoarthritis Outcome Score for Joint Replacement (KOOS JR). We compared 276 knee arthroplasty procedures performed from March 2020 through June 2020 that were eligible to complete paper KOOS JR surveys with 490 knee arthroplasty procedures performed from March 2022 through June 2022 that were eligible to complete electronic surveys. Survey completion rates at preoperative and 1-year follow-up were evaluated as well as the relative frequency of patients achieving a substantial clinical benefit (SCB) KOOS JR score increase of 20 points or more. Multivariate regressions were used to assess the potential influence of covariates, including age at surgery, sex, body mass index, type of insurance, surgery site, and the distance patients traveled to our institution. Response rates for preoperative surveys completed within 90 days of surgery increased from 53% (146 of 276) with paper to 83% (406 of 490) with electronic surveys, while 1-year follow-up response rates improved from 38% (105 of 276) to 65% (320 of 490). Multivariate analyses indicated that only the survey type (paper or electronic) was associated with response rates. Electronic data collection also reduced incomplete (13 to 0.4%) and unnecessary (38 to 0.4%) surveys. The annual cost of data collection decreased from $140,696 with paper-based forms to $105,742 with electronic surveys. However, patients achieving a SCB declined from 81% (42 of 52) with paper to 64% (176 of 276) with electronic surveys (P = 0.02). Compared to paper forms, electronic data collection at our institution increased follow-up rates and improved data quality at lower costs, but the relative frequency of patients reporting a SCB decreased.