Kirstine L Sibilitz, Selina K Berg, Lars H Tang, Signe S Risom, Christian Gluud, Jane Lindschou, Lars Kober, Christian Hassager, Rod S Taylor, Ann‐Dorthe Zwisler

Exercise-based cardiac rehabilitation may benefit heart valve surgery patients. We conducted a systematic review to assess the evidence for the use of exercise-based intervention programmes following heart valve surgery. To assess the benefits and harms of exercise-based cardiac rehabilitation compared with no exercise training intervention, or treatment as usual, in adults following heart valve surgery. We considered programmes including exercise training with or without another intervention (such as a psycho-educational component). We searched: the Cochrane Central Register of Controlled Trials (CENTRAL); the Database of Abstracts of Reviews of Effects (DARE); MEDLINE (Ovid); EMBASE (Ovid); CINAHL (EBSCO); PsycINFO (Ovid); LILACS (Bireme); and Conference Proceedings Citation Index-S (CPCI-S) on Web of Science (Thomson Reuters) on 23 March 2015. We handsearched Web of Science, bibliographies of systematic reviews and trial registers (ClinicalTrials.gov, Controlled-trials.com, and The World Health Organization International Clinical Trials Registry Platform). We included randomised clinical trials that investigated exercise-based interventions compared with no exercise intervention control. The trial participants comprised adults aged 18 years or older who had undergone heart valve surgery for heart valve disease (from any cause) and received either heart valve replacement, or heart valve repair. Two authors independently extracted data. We assessed the risk of systematic errors ('bias') by evaluation of bias risk domains. Clinical and statistical heterogeneity were assessed. Meta-analyses were undertaken using both fixed-effect and random-effects models. We used the GRADE approach to assess the quality of evidence. We sought to assess the risk of random errors with trial sequential analysis. We included two trials from 1987 and 2004 with a total 148 participants who have had heart valve surgery. Both trials had a high risk of bias.There was insufficient evidence at 3 to 6 months follow-up to judge the effect of exercise-based cardiac rehabilitation compared to no exercise on mortality (RR 4.46 (95% confidence interval (CI) 0.22 to 90.78); participants = 104; studies = 1; quality of evidence: very low) and on serious adverse events (RR 1.15 (95% CI 0.37 to 3.62); participants = 148; studies = 2; quality of evidence: very low). Included trials did not report on health-related quality of life (HRQoL), and the secondary outcomes of New York Heart Association class, left ventricular ejection fraction and cost. We did find that, compared with control (no exercise), exercise-based rehabilitation may increase exercise capacity (SMD -0.47, 95% CI -0.81 to -0.13; participants = 140; studies = 2, quality of evidence: moderate). There was insufficient evidence at 12 months follow-up for the return to work outcome (RR 0.55 (95% CI 0.19 to 1.56); participants = 44; studies = 1; quality of evidence: low). Due to limited information, trial sequential analysis could not be performed as planned. Our findings suggest that exercise-based rehabilitation for adults after heart valve surgery, compared with no exercise, may improve exercise capacity. Due to a lack of evidence, we cannot evaluate the impact on other outcomes. Further high-quality randomised clinical trials are needed in order to assess the impact of exercise-based rehabilitation on patient-relevant outcomes, including mortality and quality of life.