Frailty and Survival for Diagnoses Feasibly Managed Operatively or Nonoperatively

Abstract

Objective: Compare outcomes, stratified by frailty, of patients with eight common conditions with plausible operative and nonoperative management strategies.

Summary background data: A surgical pause, evaluating potential adverse outcomes among frail patients, improves postoperative outcomes; however, the outcomes among patients opting for nonoperative management are unknown.

Methods: In an observational cohort study across a multi-hospital healthcare system including adults presenting to outpatient surgical clinics (2016-2023) for evaluation of eight conditions feasibly managed operatively or nonoperatively as defined by modified Delphi consensus. In a landmarked analysis, we compared 2-year survival by management strategies across frailty categories (robust, normal, frail, very frail) as defined by the Risk Analysis Index (RAI). Secondarily we compared 365-day hospital free days (HFD-365), postoperative length of stay, and discharge disposition.

Results: Among 49,169 patients (mean±SD age, 60.4±14.6 y; 54.6% female), operative management was associated with lower observed and adjusted mortality (1.3% vs 2.5%; aHR=0.55 [95% CI, 0.47-0.66], P<0.0001) overall and among all frailty categories expect the very frail (8.1% vs 12.1%, P=0.1). Additionally, operative management was associated with fewer HFD-365 again overall which was specifically prominent among the very frail (median 365 [IQR, 358-365] vs 361 days [IQR, 357-363], P<0.0001). Postoperatively, frailty portended more protracted recoveries with greater postoperative lengths of stay (1.7±2.6 vs 1.2±2.1) days, P<0.0001) and fewer discharges home (370 [85.1%] vs 5,087 [91.8%], P<0.0001; odds ratio=2.0 [95%CI 1.5-2.6]).

Conclusions: Considering the protracted postoperative recovery of very frail patients, nonoperative management might be the preferred treatment option for those presenting with these eight clinical conditions.

Keywords: elective; frailty; management.