RESUMEN
OBJECTIVE: Herpes family viruses have been recognized as pathogens for many years in immunosuppressed transplant or human immunodeficiency virus patients, but they have garnered little attention as potential pathogens in the nonimmunosuppressed critically ill. The objective of this study was to define the prevalence of and risk factors for development of herpes family virus infection in chronic critically ill surgical patients. DESIGN: Prospective epidemiologic study. SETTING: A 38-bed surgical intensive care unit in a major university hospital. PATIENTS: Nonimmunosuppressed intensive care unit patients in intensive care unit for >/=5 days. INTERVENTIONS: None; patients received no antiviral treatment during the study. MEASUREMENTS AND MAIN RESULTS: Weekly cultures for cytomegalovirus (CMV) and herpes simplex virus, viral serologies, and T-cell counts were performed. The prevalence (95% confidence interval) of positive respiratory cultures for herpes simplex or CMV was 35% (22-49%); 15% (5-25%) cultured positive for CMV, 23% (11-35%) cultured positive for herpes simplex virus, and one patient's respiratory secretions culturing positive for both CMV and herpes simplex virus. The prevalence of CMV viremia was only 5.8% (1-10%). CMV+ patients had longer hospital admissions, intensive care unit admissions, and periods of ventilator dependence than CMV- patients, despite having comparable severity of illness scores. CMV+ patients also had significantly higher numbers of blood transfusions, prevalence of steroid exposure, and prevalence of hepatic dysfunction, and all were immunoglobulin G positive at the beginning of the study. In contrast, herpes simplex virus-positive patients had lengths of hospital admissions, lengths of intensive care unit admissions, and periods of ventilator dependence comparable with patients without viral infections (p >.05). CONCLUSIONS: There is a significant prevalence (22-49%) of occult active herpes family viruses in chronic critically ill surgical patients. The clinical significance of these viral infections is unknown, although CMV+ patients have significantly higher morbidity rates than CMV- patients. Several factors suggest pathogenicity, but further study is needed to define causality.