Thirty-four studies (27,186 patients; 133,141 patient years) were included in the review. Twenty-seven studies (21,802 patients) used the universally accredited indexed effective orifice area of 0.85cm2/m2 or more, categorising 44.2% of patients as having prosthesis-patient mismatch. Seven studies diagnosed 34.2% of patients with moderate prosthesis-patient mismatch (>0.65 to >0.85cm2/m2) and 9.8% with severe prosthesis-patient mismatch (<0.65cm2/m2).
Prosthesis-patient mismatch was associated with a statistically significant increase in all-cause mortality when compared with patients without prosthesis-patient mismatch (HR 1.34, 95% CI 1.18 to 1.51;18 studies; Ι²=35%); the increase in cardiac-related mortality was not statistically significant (HR 1.51, 95% CI 0.88 to 2.60; nine studies; Ι²=67%).
Moderate prosthesis-patient mismatch was associated with statistically greater all-cause mortality (HR 1.19, 95% CI 1.07 to 1.33; 10 studies, Ι²=26%) and cardiac-related mortality (HR 1.32, 95% CI 1.02 to 1.71; three studies, Ι²=0%).
Severe prosthesis-patient mismatch was associated with statistically greater all-cause mortality (HR 1.84, 95% CI 1.38 to 2.45; 12 studies; Ι²=79%) and cardiac-related mortality (HR 6.46, 95% CI 2.79 to 14.97; three studies, Ι²=42%).
There was a constant hazard over time for all-cause mortality, whereas cardiac-related showed more variation in hazard ratios over time.
Subgroup analyses indicated statistical heterogeneity between studies using bioprosthetic and mechanical valves on all-cause mortality. Sensitivity analyses were reported.
There was no evidence of publication bias.