Two-point improvement in short physical performance battery during inpatient cardiac rehabilitation predicts lower rate of all-cause mortality and hospitalization for older patients with heart-failure

Background Old and frail patients with heart failure (HF) are a burden worldwide. Late-recovery phase (outpatient) cardiac rehabilitation (CR) program improves exercise capacity and prognosis. However, most old and frail patients have difficulty accessing the outpatient CR program due to logistical issues. Therefore, early-recovery phase (inpatient) CR program during acute decompensated HF hospitalization may be important for improving functional status. The Short Physical Performance Battery (SPPB) is a practical test of physical function in old patients. However, the cut-off value for the in-hospital change in SPPB (ΔSPPB) that is associated with prognosis has not been established. Purpose This study aimed to evaluate the minimum clinically important difference (MCID) of the SPPB for prognostic value during the early-recovery phase CR program for old and frail patients with HF. Methods We enrolled hospitalized patients with acute decompensated HF who were old (≥65 years) and frail (SPPB≤9), and performed the early-recovery phase CR program (60 minutes/day). The SPPB was assessed at baseline and discharge, and the in-hospital change (ΔSPPB = discharge − baseline) was calculated. Our main outcome was a composite of all-cause mortality and unplanned all-cause rehospitalization within 6 months after discharge. We evaluated the cut-off value of ΔSPPB for the outcome. Results 137 patients with HF (median age 84 years, 48.2% Female) were enrolled. The mean number of CR program was 6 sessions. The SPPB improved from a median [25-75% interquartile range] of 6 [4–8] to 8 [6–10] (p < 0.05), and ΔSPPB was 1.7 ± 2.2 points. After discharge for 6 months, 4 patients died and 42 patients were hospitalized. The MCID for ΔSPPB was determined at 1.5 points using Anchor methods (receiver operating characteristic [ROC] curve analysis) and Distribution methods (1.5> 0.94-1.16 points) (Figure 1). Because SPPB is an integer unit, we finally defined the MCID cut-off for the composite outcome as ΔSPPB ≥2 points. Divided them into the Improved group (ΔSPPB ≥2) and the Non-Improved group (ΔSPPB <2), Cox hazard analysis for multivariate analysis showed that ΔSPPB ≥2 points was independently associated with the composite outcomes (adjusted hazard ratio; 0.441, 95% confidence interval, 0.224-0.868; p=0.018) compared with the non-improvement group. And, Kaplan–Meier analysis revealed that patients with ΔSPPB ≥2 had significantly lower event-free survival for the composite outcome within 180 days (log-rank [Mantel–Cox] test: χ² = 7.594, p = 0.006) (Figure 2). Conclusions Early-recovery phase CR program increases SPPB, and an improvement of 2 points or more in SPPB may predicts lower rate of all-cause mortality and unplanned all-cause hospitalization for old and frail patients with HF.