Hospital antibiograms that are based on pooled pediatric and adult data may not present a clear picture of institution-specific resistance patterns and thus lead to suboptimal antibiotic choices. My team at Duke University Hospital System (DUHS) hypothesized that significant susceptibility differences do exist for pediatric-specific isolates and that knowledge of the pediatric-only data could alter empirical antibiotic choices made by our physicians. The results support what has been seen in similar settings and argue strongly for using pediatric-specific resistance information to improve patient care on a large scale.
For office-based primary care clinicians, our findings suggest that the more information you can obtain on specific resistance patterns in your community, the more effective your antibiotic choices will be.
Resistance scenarios
For approximately 1 year, we studied pediatric Escherichia coli resistance patterns for children through age 12 years who were treated for urinary tract infections (UTIs) at the medical center. We compared the data with hospital-wide isolates and found the pediatric isolates were more resistant to ampicillin(Drug information on ampicillin) and trimethoprim(Drug information on trimethoprim)/sulfamethoxazole and less resistant to amoxicillin/clavulanate and ciprofloxacin(Drug information on ciprofloxacin) (P < .0005 for all).
We then asked pediatric attending and resident physicians what antibiotic they would use to treat a UTI in a 3-month-old and 12-year-old based on 3 scenarios—no antibiogram for reference; using the DUMH hospital-wide antibiogram; and using the pediatric-specific antibiogram data from the study. For the infant vignettes, effective antibiotic choices increased from 69% to 82% (P = .06) to 92% (P < .01) across the scenarios. For the adolescent vignettes, effective choices increased from 32% to 57% (P < .01) to 79% (P = .01).
Conversely, we did find in some instances that the more specific pediatric data may have contributed to inappropriate or less appropriate medication choices. An antibiogram presents objective data about the susceptibility results of bacteria to agents in vitro. What cannot be presented is the context or utility of a given agent for a particular patient. In most cases, amikacin(Drug information on amikacin) is reserved for hospitalized patients with nosocomial infections and is only available for intravenous administration. Regardless of susceptibility data, cefuroxime(Drug information on cefuroxime) should not be used to treat meningitis. Microbiology labs do selectively report some information depending on context (source of specimen, age of patient), but ultimately the clinician needs to know what to do with the information. Our study may have captured some trainees early in their clinical experience who still have more to learn.
Access to antibiograms
The availability of pediatric-specific antibiogram programs is extremely variable. Almost all freestanding children's hospitals have only pediatric-specific data because there is only one lab. At children's hospitals within larger hospitals, the data are not usually provided.
As for the office-based physician, it requires a bit more work to obtain community- or setting-based resistance information, but it’s not impossible. Labs at local hospitals should be able to provide susceptibility data for E coli and other common isolates. Alternatively, an ambitious practice can independently track susceptibility within its caseload to learn if certain subpopulations harbor resistant organisms. Of course, the inquiring clinician can contact a local infectious disease specialist to ask about local patterns of resistance.
The bottom line: any knowledge of local trends can help inform subsequent decisions about empiric antibiotics.
Reference
Boggan JC, Navar-Boggan AM, Jhaveri R. Pediatric-specific antimicrobial susceptibility data and empiric antibiotic selection. Pediatrics. 2012;130:e61.
