Implantable cardioverter defibrillators (ICDs) prevent sudden cardiac death, but patients need to be counseled about potential harms. We summarized the evidence on adverse events from ICDs with a focus on ICD use for primary prevention of sudden cardiac death.

We searched MEDLINE and Cochrane Central Register of Controlled Trials from 2002 through 2012 for reports of adverse events from ICDs implanted for primary prevention or mixed indications (primary and secondary prevention). Studies had to have ≥500 patients and specify patient numerators and denominators.

Data from 35 independent cohorts reported in 53 articles were included. Reports from one registry provided high quality evidence on adverse events during hospitalization for ICD implantation. Adverse events ranged from 2.8 to 3.6 %. Serious adverse events ranged from 1.2 to 1.4 %. The most frequent serious adverse events were pneumothorax (0.4–0.5 %) and cardiac arrest (0.3 %). The quality of the evidence for long-term adverse events was low. Frequency of adverse events post-hospitalization was variable, as was follow-up: device-related complications <0.1–6.4 % (2–49 months), lead-related complications <0.1–3.9 % (1.5–40 months), infection 0.2–3.7 % (1.5–49 months), and thrombosis 0.2–2.9 % (1.5–49 months). Evidence for inappropriate shock was of moderate quality with 3–21 % of patients experiencing at least one inappropriate shock during 1 to 5 years of follow-up.

The limitation of the evidence reviewed in this study is low quality evidence for adverse events post-hospitalization. Evidence is predominantly from mixed primary and secondary prevention populations.

In-hospital adverse events after ICD implantation are infrequent. The estimates for long-term adverse events are uncertain. Up to one-fifth of patients receive inappropriate shocks.