Seasonal variation in the epidemiology of sepsis.

OBJECTIVE

We sought to investigate seasonal and regional variability in the epidemiology of sepsis and to identify underlying associations based on geography and seasonal viral infections. Understanding seasonal or regional variations may improve knowledge of sepsis epidemiology and pathophysiology and could affect healthcare planning and resource allocation.

DESIGN

Retrospective cohort study using the National Hospital Discharge Survey to identify cases of sepsis, severe sepsis, influenza, and viral pneumonia using ICD-9-CM codes. Incidence rates are reported as mean cases frequencies per season per 100,000 as calculated by normalization to the 2000 U.S. Census.

SETTING

Acute-care nonfederal U.S. hospitals.

PATIENTS: Patients hospitalized between 1979 and 2003 in acute-care nonfederal U.S. hospitals with a diagnosis of sepsis or viral respiratory infection.

INTERVENTIONS: None.

MEASUREMENTS AND MAIN

RESULTS

The seasonal incidence rate of sepsis increased 16.5% from a low of 41.7 in the fall to a high of 48.6 cases per 100,000 in the winter (p<.05). Similarly, seasonal rates for severe sepsis statistically increased 17.7% from fall to winter at 13.0 and 15.3 cases per 100,000, respectively. The greatest change in sepsis incidence occurred with respiratory sources, increasing 40% during the winter compared with the fall (p<.05). Seasonal variations in viral respiratory infections paralleled changes in sepsis incidence but did not fully account for the changes. The greatest seasonal change in sepsis rates occurred in the Northeast (+30%). Sepsis case-fatality rates were 13% greater in the winter compared with the summer (p<.05) despite similar severity of illness.

CONCLUSIONS

The incidence and mortality of sepsis and severe sepsis are seasonal and consistently highest during the winter, predominantly related to respiratory sepsis. Seasonal changes in sepsis incidence vary according to geographic region. The mechanisms underlying these differences require further investigation.