Introduction: Cytomegalovirus is the most common viral opportunistic infection in immunocompromised patients. However, recent studies have demonstrated active cytomegalovirus infection in nonimmunosuppressed intensive care unit patients.

Objective: To define the frequency and mortality rate associated with cytomegalovirus infection in nonimmunosuppressed patients in the intensive care unit.

Methods: A systematic review up to October 2008 was performed. Pooled results were analyzed by fixed- and random-effects models. Cochran Q and I2 were performed for heterogeneity.

Results: Thirteen studies (n = 1258) were selected. The overall rate of active cytomegalovirus infection was 17% (95% confidence interval [CI], 11% to 24%; p < .0001; I2 = 86%). When the patients were screened for > or =5 intensive care unit days, the overall rate of infection increased to 21% (95% CI, 15% to 29%; p < .001). The infection rate for studies that used cytomegalovirus DNA/antigen for diagnosis was 20% (95% CI, 13% to 31%; p < .0001) and for studies that used culture was 12% (95% CI, 6% to 22%; p < .0001). The cytomegalovirus rate for patients with unknown serology was 7% (95% CI, 3% to 14%; p < .0001), whereas the rate for patients with positive serology was 31% (95% CI, 22% to 42%; p < .0001). The rate of infection was higher in patients with severe sepsis: 32% (95% CI, 22% to 45%; p < .0001). And in patients with high disease severity: 32% (95% CI, 23% to 42%; p < .0001). The overall mortality rate associated with active cytomegalovirus infection was 1.93 times (95% CI, 1.29 to 2.88; p = .001) as high as that without cytomegalovirus infection.

Conclusions: Active cytomegalovirus infection occurs frequently in nonimmunosuppressed patients in intensive care, especially in those with positive cytomegalovirus serology, intensive care unit stay > or =5 days, severe sepsis, and high disease severity, in whom the rate of cytomegalovirus infection is up to 36%. Mortality rate is significantly doubled with cytomegalovirus, but a cause-effect relationship cannot be established yet. A large prospective cohort study on the patient population identified by our findings is needed to define who is at the highest risk for developing active cytomegalovirus infection and to determine its effects on mortality.