Bacteremic pneumococcal pneumonia in one American City: a 20-year longitudinal study, 1978-1997

A surveillance of bacteremic pneumococcal pneumonia was conducted in Huntington, West Virginia, from 1978 to 1997 to investigate case-fatality rates, incidence of disease, capsular types, and antibiotic usage. Our study population comprised consecutive inpatients admitted to the hospitals in Huntington, West Virginia, and included 45 children younger than 15 years and 328 adults. All blood isolates were serotyped by capsular swelling procedures; clinical characteristics, treatment, and outcome for all patients were abstracted from hospital charts. The overall case-fatality rate was 20.3%, with most deaths occurring among adults older than 50 years. Case-fatality rates peaked at 37.7% among patients 80 years of age and older. Only 1 of 45 (2.2%) children died. Case-fatality rates declined in each successive 5-year period, from 30.2% in 1978-1982 to 15.6% in 1993-1997. In that same period, incidence rates increased severalfold among children younger than 4 years to 44.5 cases per 100,000 population and among adults 70 years and 80 years of age and older to 38.5 and 76.2 cases per 100,000, respectively. Of the 34 serotypes isolated, 10 accounted for two thirds of the cases of pneumonia: 1, 4, 9, 14, 3, 6, 12, 5, 23, and 19 (in rank order). Chronic renal disease and arteriosclerotic heart disease increased the risk of death. Treatment regimens that included a macrolide and a penicillin or cephalosporin resulted in the lowest case-fatality rate in adults older than 50 years: 6% in 1993-1997. In conclusion, as bacteremic pneumococcal pneumonia evolved over time, the case-fatality rate decreased, its incidence increased, predominant capsular types changed, and treatment regimens that included a macrolide resulted in the lowest fatality rates.