Loss to follow-up after lumbar decompressive surgery: effect of collection method and perioperative factors on patient-reported outcome measures.

Abstract

To identify collection methods and perioperative factors linked to 12-month patient-reported outcome measure (PROM) response rate and evaluate the effect of loss to follow-up on PROM results. The study included 2588 elective lumbar decompressive surgery without spinal fusion during 2008-2023. The primary PROM was the Oswestry Disability Index (ODI). The independent effect of PROM collection methods and perioperative factors on 12-month ODI response was examined. The effect of loss to follow-up on 12-month PROM responder results was evaluated with modelled 12-month PROM and with observed 3-month PROM in a group with last ODI response at 3-months. The response rate to 6-month electronic-only collection was evaluated in patients who advocated to also receive electronic questionnaires. Proportionally, 70.8% (1832/2588) were 12-month ODI responders, 19.4% (501/2588) provided a last ODI response at 3-months, and 9.9% (255/2588) were ODI non-responders. The PROM collection method linked to the highest 12-month ODI response rate was two attempts each by electronic and postal questionnaire. Important perioperative factors for lower 12-month ODI response were younger age, smoker, ASA grade, physical comorbidities, male, and anxiety/depression. Twelve-month responder PROM results were unchanged (≤ 1% of metric) by the addition of missing values from modelled PROM or last ODI response at 3-months. The 6-month ODI response rate to electronic-only questionnaire was 48% (467/969). Combined electronic and postal PROM collection achieved a 78% response rate. Populations at risk of underrepresentation in spinal register 12-month results were identified. Twelve-month PROM results are not significantly affected by 29% missing data.