Effect of pulmonary rehabilitation on all-cause mortality in patients with chronic respiratory disease: a retrospective cohort study in an Australian teaching hospital.

Abstract

Pulmonary rehabilitation (PR) is widely recommended for short-term benefits in chronic respiratory diseases, yet long-term outcomes remain uncertain. This retrospective cohort study addresses this gap, comparing 20-year mortality rates between PR participants and matched controls, and hypothesizing that the short-term benefits of PR contribute to improved long-term survival. The 20-year mortality of stable chronic respiratory patients who participated in an outpatient PR program was compared with a matched control group based on the type of lung disease. Demographic and clinical variables, and the dates of deaths, were extracted and compared between two groups with two sample t-test and chi-square tests. Kaplan-Meier plots and Cox regression analyses were employed to evaluate survival differences. Between 2000 and 2002, 238 individuals enrolled in a pulmonary rehabilitation (PR) program (58% male, mean age ± SD: 69 ± 8 years, mean FEV1% predicted ± SD: 46 ± 21%). An equal number of people with comparable lung disease were selected as controls (88% COPD, 5% ILD). Controls had lower FEV1% predicted values (mean ± SD: 39 ± 17%, P < 0.001), smoked more (mean ± SD: 48 ± 35 pack-years, P = 0.032), and no differences in age, BMI, sex, and Index of Relative Socio-economic Advantage and Disadvantage (IRSAD). Median (IQR) follow-up time was 68 months (34-123), with 371 (78%) deaths. Univariable (HR = 1.71, p < 0.001) and multivariable (HR = 1.64, p < 0.001) Cox regression found higher mortality risk in controls. Subgroup analysis for COPD replicated these findings (HR = 1.70, P < 0.001). Despite some methodological limitations, our study suggests that clinically stable patients with chronic respiratory disease who undertake PR may have lower mortality than matched controls. Retrospectively registered.