Ten studies were included (n=2,262): two randomised controlled trials (n=120); three non-randomised prospective studies (n=634); and five non-randomised retrospective studies (n=1,508). Five studies were designated high quality.
Minimal access aortic valve replacement versus conventional aortic valve replacement:
Incidence of atrial fibrillation (10 studies) displayed no statistically significant difference between the groups (odds ratio 0.85, 95% confidence interval: 0.66, 1.11, p=0.24). No evidence of significant statistical heterogeneity was found. The statistical significance of findings did not change when only high-quality studies were included in analysis.
Sensitivity analyses indicated that intraoperative variables did not significantly affect the findings, with the one exception of a statistically significant effect in favour of minimal access aortic valve replacement was found with the exclusion of studies reporting differences in cardiopulmonary bypass time, or if the means were not stated (odds ratio 0.63, 95% confidence interval: 0.47, 0.85). No evidence of significant heterogeneity was found.
Surrogate outcomes (three studies) showed no statistically significant difference between the groups reported in intensive care unit stay, total length of stay, ventilation time, incidence of chest infections or cerebrovascular accident. Findings were inconsistent for other surrogate outcomes.