Temporal trends in septicemia in a community hospital.

A prior study of septicemia in our community teaching hospital demonstrated the importance of case mix categories in understanding differences in rates of septicemia seen in hospitals. This study provides a 10 year, in 1982, and a 15 year, in 1987, follow-up of septicemia from the same hospital. A substantial increase in the incidence of septicemia was noted in 1982 and 1987. Underlying illness categories continued to be important predictors of incidence of septicemia and fatality rates. The sites of infections, relative mix of community-acquired and nosocomial infections, and mix of organisms were fairly stable throughout the period. The overall incidence of septicemia increased from 34/10,000 admissions in 1970-1973 to 87/10,000 in 1982 and 103/10,000 in 1987. The major explanations for these increases are: (a) a striking increase in the use of blood cultures from 1 blood culture per 10.4 patients in 1973 to 1 blood culture per 3.3 patients in 1987; (b) an increase in the proportion of patients in the Medicare age group from 13% in 1970-1973 to 24% in 1987, with that age group accounting for 33% of the cases of septicemia in 1970-1973 and 55% in 1982 and 1987; and (c) a modest change in the case mix category of ultimately fatal underlying illness and a probable increase in the acuity of illness for most patients admitted. Parallel changes in frequency of use of blood cultures have occurred at the previously evaluated medical school teaching hospital in the same community. These same explanations are likely to be applicable for the same time period in other acute care hospitals as well.