This study is a secondary analysis of a prospective observational study, which focused on the effects of 3-month postoperative recovery as measured by the quality of recovery-15 in hospital on disability-free survival. This study was approved by the Nara Medical University Institutional Review Board (Kashihara, Nara, Japan; Chairperson, Prof. M Yoshizumi; approval number: 2975; 28 April 2021). The statistical protocol of this secondary analysis was approved on 17 August 2023; Kashihara, Nara, Japan; Chairperson, Prof. M Yoshizumi; approval number: 2975).

Our initial study, which focused on the effects of postoperative recovery as measured by the quality of recovery-15 in hospital on disability-free survival three months later, included a total of 230 patients aged 65 years or older who underwent elective major abdominal surgery with a cancer diagnosis [10]. Among them, patients without atrial fibrillation and cardiac pacemakers were included in this study.

We collected various preoperative patient characteristic data, including co-morbidities, daily medications, and frailty at the perioperative management center where patients underwent medical interviews and were scheduled for surgery. Frailty was assessed using the Fried Frailty Phenotype Questionnaire, including five domains (fatigue, resistance, ambulation, inactivity, and loss of weight) with the total score ranging from 0 to 5 points [11]. Patients with a total score ≥ 3 were classified as frailty patients [11]. In terms of intraoperative data, we collected information on anesthetics used, total administered dose of ephedrine and phenylephrine, total fluid volume, surgical field, postoperative analgesia, surgical duration, and SAS. The SAS, with a total score of 0 (bad) to 10 (excellent), was calculated based on the following three parameters: lowest mean blood pressure (0–3), lowest heart rate (0–4), and blood loss volume (0–4) [5].

Daily oral medications used by the patients were continued except for angiotensin receptor blockers and angiotensin-converting enzyme inhibitors. No pre-surgery medication was administered on the day of surgery. Patients were allowed to have clear water orally up to two hours before entering the operating room. Intraoperative management, including the insertion of an arterial catheter, fluid therapy, and choice of cardiovascular agents, was determined by the attending anesthesiologist. Mean arterial blood pressure values were recorded at 2.5-minute intervals (when blood pressure was measured using oscillometry) or at 1-minute intervals (when an arterial catheter was used).

The primary outcome of this study was the SAS. Secondary outcomes were postoperative severe complications, defined as a Clavien–Dindo classification ≥ 3 [12], and length of postoperative stay.

Continuous data are presented as median [1st quartile, 3rd quartile], and categorical variables are presented as number (%). Univariate analysis was performed using the Mann–Whitney U test or Fisher's exact test as appropriate, to compare the two groups (robust vs. frailty). To assess the effect of SAS on postoperative severe complications and length of postoperative hospital stays, a cut-off value of SAS 7 was determined because of median SAS scores in patients with and without frailty were 7 and 8. Subsequently, the secondary outcomes were compared between patients with SAS ≤ 7 or >7.

Since this study involved a secondary analysis, sample size calculation was not performed. However, as an alternative, we performed a post hoc power analysis using G*power version 3.1 (Faul, Erdfelder, Lang, & Buchner, 2007) with a type I error of 0.05 and effect size of 0.5 (large effect size). With these parameters and the existing number of patients (robust = 165 and frailty = 45), the power was determined to be 0.82 to detect a significant difference. IBM SPSS Statistics (version 25.0; IBM Corp., Armonk, NY) was used to analyze all data, and p-values < 0.05 were considered statistically significant.

One post hoc analysis using nonlinear restricted cubic splines in the regression model was performed to confirm the nonlinearity of SAS for secondary outcomes.