Using Hospital Antibiogram Data To Assess Regional Pneumococcal Resistance to Antibiotics

Cheryl R. Stein, David J. Weber, Meera Kelley


Emerging Infectious Diseases. 2003;9(2) 

In This Article

Abstract and Introduction

Antimicrobial resistance to penicillin and macrolides in Streptococcus pneumoniae has increased in the United States over the past decade. Considerable geographic variation in susceptibility necessitates regional resistance tracking. Traditional active surveillance is labor intensive and costly. We collected antibiogram reports from North Carolina hospitals and assessed pneumococcal susceptibility to multiple agents from 1996 through 2000. Susceptibility in North Carolina was consistently lower than the national average. Aggregating antibiogram data is a feasible and timely method of monitoring regional susceptibility patterns and may also prove beneficial in measuring the effects of interventions to decrease antimicrobial resistance.

Streptococcus pneumoniae is a leading cause of community-acquired illness, resulting in an estimated 3,000 cases of meningitis, 50,000 cases of bacteremia, 500,000 cases of pneumonia, and 7 million cases of otitis media each year in the United States.[1] Even with appropriate antimicrobial therapy, case-fatality rates for high-risk patients can be as high as 40% for bacteremia and 55% for meningitis.[1] Surveillance systems have shown decreasing antimicrobial susceptibility among pneumococci.[2,3,4,5,6,7,8,9,10,11,12,13] A comparison of susceptibility among combined respiratory and invasive isolates from respiratory seasons 1994-1995 and 1999-2000 showed a decrease in both penicillin susceptibility (from 76% to 66%) and erythromycin susceptibility (from 90% to 74%).[8] An examination of only invasive isolates from 1997 and 2000 also showed a decline in isolates' susceptibility to penicillin (from 75% to 73%) and to erythromycin (from 85% to 78%).[2,5]

Many of the current surveillance systems monitoring emerging drug resistance detect susceptibility patterns over large areas, such as an entire country.[2,6,9,13] Results may not be generalizable to all locations within the study area.[1] For example, although the overall proportion of penicillin-nonsusceptible pneumococci within a seven-region, population-based, active surveillance program was 25%, the proportion ranged from 15% in Maryland to 38% in Tennessee.[14] Data are typically gathered from a limited number of medical establishments within a specified region; national surveillance systems may not collect data from every state. Monitoring trends in pneumococcal susceptibility over smaller geographic areas is necessary to aid clinicians in choosing the best drug treatment for empiric therapy.[1] This local information can also help evaluate efforts to decrease resistance rates through judicious antibiotic use.

Chin et al. compared antimicrobial susceptibility results from active and antibiogram surveillance of pneumococcal isolates at 12 hospitals in the Portland, Oregon, area in 1996.[15] Active surveillance was defined as collecting isolates and patient data from participating hospitals and performing susceptibility testing at a centralized laboratory.[15,16] Antibiogram surveillance is quite different. Clinical laboratories assess the antimicrobial susceptibilities of bacterial isolates and summarize all susceptibility results for a specified period on an antibiogram report. Antibiograms conform to the susceptibility testing practices of individual laboratories, include information on both sterile and nonsterile isolates, may include duplicate isolates from a single patient, and lack an epidemiologic characterization of the patient or isolate. The data contained on laboratory-specific antibiograms, however, can be compiled to assess regional susceptibility, monitor trends over time, and assess effects of interventions designed to reduce antibiotic resistance through judicious antibiotic use.

Chin et al. found no statistically significant difference in results obtained by the two methods for any of the four drugs tested at the 12 Portland area hospitals, except for a single drug at a single hospital where erythromycin susceptibility was reported at 97% by active surveillance and 84% from the antibiogram (chi square p=0.01).[15] The cost difference between the active and antibiogram surveillance systems was substantial: $52,000 for active surveillance and $700 for antibiogram surveillance. The authors concluded that although antibiogram surveillance produced less information than active surveillance, "antibiograms provided accurate, community-specific drug-resistant S. pneumoniae data."