Aims Meta-analysis and systematic review to determine the long-term prognostic significance of premature ventricular complexes (PVCs) in adults without clinically apparent heart disease.

Methods Relevant studies were searched on MEDLINE and EMBASE. Inclusion criteria: controlled studies on adults without clinically apparent heart disease comparing the prognosis of the presence against the absence of PVCs. Endpoints: all-cause mortality, cardiovascular mortality, sudden cardiac death or development of ischaemic heart disease. OR of endpoints were analysed with random effects model. Relationships between study outcomes and study characteristics were assessed by meta-regression and sensitivity analysis.

Results Eight studies satisfied the inclusion criteria. Meta-analysis shows that in adults without clinically apparent heart disease, PVCs on ECG recording are associated with a pooled OR of 1.72 (95% CI 1.28 to 2.31) of endpoints compared with those without PVCs. However, only one study used echocardiogram or stress test to rule out heart disease. Meta-regression identified mean sample age (p=0.001), diabetes (p=0.005) and hypertension (p=0.005) as predictors of events. Only studies that used 100% male, not 100% female or mixed gender, found increased events.

Conclusions Most studies on PVC prognosis in ‘normal hearts’ did not use advanced tests to rule out structural heart disease. Among these patients, PVCs are associated with a worse cardiovascular outcome if patients are older and have higher cardiovascular risk, suggesting that the poor prognosis studies may have inadvertently included patients with occult structural heart disease, the population in which PVCs are known to confer adverse outcomes.