Identifying the probable timing and setting of respiratory virus infections.

Lessler, Justin; Brookmeyer, Ron; Reich, Nicholas G; Nelson, Kenrad E; Cummings, Derek A T; Perl, Trish M

OBJECTIVE Show how detailed incubation period estimates can be used to identify and investigate potential healthcare-associated infections and dangerous diseases. METHODS We used the incubation period of 9 respiratory viruses to derive decision rules for distinguishing between community- and hospital-acquired infection. We developed a method, implemented in a simple spreadsheet, that can be used to investigate the exposure history of an individual patient and more specifically to identify the probable time and location of infection. Illustrative examples are used to explain and evaluate this technique. RESULTS If the risks of hospital and community infection are equal, 95% of patients who develop symptoms of adenovirus infection within 5 days of hospital admission will have been infected in the community, as will 95% of patients who develop symptoms within 3 days for human-coronavirus infection, 2.5 days for severe acute respiratory syndrome, 1 day for influenza A, 0.5 day for influenza B, 12 days for measles, 2 days for parainfluenza, 4 days for respiratory syncytial virus infection, and 1.5 days for rhinovirus infection. Sources of infection suggested by analysis of the symptom onset times of individual patients are consistent with those from detailed investigations. CONCLUSIONS This work shows how a detailed understanding of the incubation period can be an effective tool for identifying the source of infection, ultimately ensuring patient safety.