Supervised multimodal prehabilitation of older patients with frailty or prefrailty before elective surgery: a multicentre randomised controlled outcome assessor-blinded trial.

Abstract

Frailty is an increasing perioperative risk in our ageing society, and prehabilitation can mitigate this vulnerability. This trial aimed to investigate if a multimodal, individualised prehabilitation programme could maintain independence and avoid further deterioration of long-term care dependency (LCD) in patients undergoing elective surgery. Prehabilitation of elderly frail or pre-frail patients prior to elective surgery (PRAEP-GO) was a multicentre, parallel-group, randomised controlled trial with blinded outcome assessment comparing a 3-week multimodal, individualised prehabilitation programme designed with shared decision making with standard of care in participants aged 70 years or older with frailty or prefrailty (defined using Fried's physical frailty phenotype) who were scheduled for elective surgery across 19 hospitals in Germany. Participants were randomly assigned by a web-based random 1:1 group allocation procedure. Study personnel were not blinded, except for outcome assessors. In the intervention group shared decision making was used to define patient goals. This was followed by a 3-week personalised prehabilitation programme consisting of 30 supervised 30-min sessions and a recommendation for 18 additional unsupervised sessions during this period focused on these goals. Sessions could cover exercise therapy, nutritional therapy, cognitive or psychological support, and other supportive components such as crutch training, medication review, or social service support. The primary outcome was the change in LCD from baseline to 12 months after surgery. LCD was assessed on an ordinal scale from 0 (no dependency) to 5 (highest dependency), analysed primarily as modified intention-to-treat (defined as all randomly assigned patients, except those who withdrew consent immediately after randomisation and those for whom no treatment was initiated and no post-randomisation follow-up data were available). Between June 27, 2020, and July 14, 2023, 3193 patients were assessed for eligibility, with a screening halt from Oct 30, 2020, to March 1, 2021, owing to the COVID-19 pandemic. The last patient was randomly assigned on July 13, 2023, and the final follow-up was completed on Aug 1, 2024. Of the 1382 randomly assigned participants, 1199 had baseline and 12-month LCD (616 in the prehabilitation group and 583 the standard care group) and were included in the modified intention-to-treat analysis. Baseline characteristics were balanced (median age 78 [IQR 74-82] years; 653 (54%) women; 617 (51%) underwent orthopaedic procedures). In accordance with German data-protection law and the trial ethics approval, race and ethnicity data were not collected. A median of 29·5 (IQR 26-30) of 30 sessions were completed. Change in ordinal LCD from baseline to month 12 did not differ significantly between groups (p=0·33), with an absolute between-group difference in change over time of -0·1 (95% CI -0·25 to 0·07; p=0·26). 32 safety events occurred in the prehabilitation group, mostly unrelated or only possibly related to the intervention. The trial was registered at ClinicalTrials.gov (NCT04418271). A 3-week supervised, multimodal, individualised prehabilitation programme, designed with shared decision making for older patients with frailty or prefrailty undergoing elective surgery did not reduce care dependency 12 months after surgery in this trial. These findings indicate that supervised, short-duration prehabilitation should not be implemented as a universal preoperative pathway for these patients. Future programmes and reimbursement policies should prioritise longer, behaviour change-based interventions; strategies to improve uptake and engagement; and the identification of subgroups most likely to benefit. Innovation Fund (Innovationsfond) of the German Federal Joint Committee (G-BA) by Grant 01NVF18024.