Predictors of loss of follow-up in a prospective registry: which patients drop out 12 months after lumbar spine surgery?

BACKGROUND CONTEXT: Long-term patient-reported outcomes (PROMs) are essential in clinical practice and research. Prospective trials and registries often struggle with high rates of loss of follow-up (LOFU), which may bias their findings. Little is known on risk factors for PROM nonresponse, especially for digitally mailed questionnaires. PURPOSE: To elucidate which patients are at high risk for LOFU by identifying associated predictors. STUDY DESIGN: Analysis of a prospective registry. PATIENT SAMPLE: Patients that underwent surgery for degenerative lumbar disease were included. OUTCOME MEASURES: Rate of PROM follow-up response at 12 months postoperatively. METHODS: Preoperatively and at 12 months postoperatively, patients were asked to complete a range of PROM questionnaires using a web-based tool. All patients who successfully completed their baseline questionnaire were included. Patients were not actively reminded upon nonresponse. Univariate and independent predictors of LOFU at 12 months were identified. RESULTS: We included 1,456 patients, of which 861 (59%) were lost to follow-up at 12 months. Univariately, lower age, American Society of Anesthesiologists (ASA) class 1, smoking, lack of prior surgery, higher pain scores and functional disability, and lower quality-of-life were associated with LOFU (all p<.05). Only lower age (OR: 0.98, p=.001), smoking (OR: 1.46, p=.019), lack of prior surgery (OR: 0.59, p=.019), and spondylolisthesis (OR: 0.47, p=.024) independently predicted LOFU. CONCLUSIONS: In a prospective registry of lumbar spine surgery patients based on web-based outcome capturing, younger age, active smoking status, lack of prior surgery, and nonspondylolisthesis surgery were independent predictors of loss of follow-up. In the future, it may become possible to preoperatively identify patients at high-risk for study dropout. As the implementation of prospective registries and the use of automated follow-up methods are on the rise, it is crucial to ensure efficiency and reduce bias of the methods on which all clinical research is based on. (C) 2019 Elsevier Inc. All rights reserved.