Invasive pneumococcal infection in a community hospital, 1993 to 1995. Characteristics of resistant strains
J. B. Orenstein
Department of Pediatrics and Emergency, George Washington University, Washington, DC, USA.
OBJECTIVES: To review all cases of invasive Streptococcus pneumoniae
infection at a single institution and to identify factors that may allow
distinguishing between penicillin-resistant pneumococcal (PRP) and
penicillin-sensitive pneumococcal strains at presentation for emergency
care. PATIENTS AND METHODS: Consecutive patient series of all children with
positive blood and cerebrospinal fluid cultures for S pneumoniae from
January 1993 to April 1995 at a general community hospital. RESULTS:
Fifty-three patients with invasive S pneumoniae infections were identified;
21% of the infections were resistant to penicillin, 9% were resistant to
multiple antibiotics, and 4% were highly resistant to penicillin (minimum
inhibitory concentration, > 2 mg/L). At admission, a diagnosis of
meningitis or sepsis was made in 14 patients; of these, 8 cerebrospinal
fluid cultures yielded S pneumoniae, and 5 were PRP (P < .001).
Significant differences between sensitive and resistant strains were not
found for sex, age, race, ill household contacts, or physician type. Five
of 11 children with PRP strains were either receiving antibiotics
concurrently or in the prior 30 days compared with 5 of 42 children with
penicillin-sensitive pneumococcal strains (P = .002). Children with PRP
strains had a lower mean white blood cell count (14.9 x 10(9)/L vs 22.8 x
10(9)/L, P = .008) than children with penicillin-sensitive pneumococcal
strains, owing to a lower mean absolute neutrophil count (7250 vs 12700, P
= .006). Children with lower white blood cell and absolute neutrophil
counts did not, however, differ in other objective measures. Regression
analysis showed that the combination of current or prior antibiotic use and
an absolute neutrophil count predicted 63% of the resistant strains and 92%
of the sensitive strains (P = .001). CONCLUSIONS: A high proportion of S
pneumoniae infections were caused by strains that were resistant to
penicillin. We discuss factors that are associated with resistance in this
population. The PRP strains were associated with recent antibiotic therapy
and a low absolute neutrophil count. Further surveillance is warranted by
these findings.
