SAN DIEGO – Antibiotic prescribing patterns for pediatric community-acquired pneumonia vary substantially across both children’s hospitals and facilities that are not children’s hospitals, a large analysis found.
Specifically, children’s hospitals are far more likely to prescribe in accordance with national guidelines than are other hospitals.
Dr. Alison Tribble
“Moving forward, I think there’s a need for further study to understand these differences, so we can begin to narrow this gap between children’s and non–children’s hospitals,” lead study author Dr. Alison Tribble said at an annual scientific meeting on infectious diseases. “Across the board, we need to continue efforts to improve guideline adherence for all children hospitalized with community-acquired pneumonia.”
In 2012, community-acquired pneumonia (CAP) accounted for 120,000 known pneumonia admissions among children in the United States and about 7% of all pediatric hospitalizations, said Dr. Tribble, a pediatric infectious disease specialist at C.S. Mott Children’s Hospital and the University of Michigan Medical Center, both in Ann Arbor. “We also know that pneumonia accounts for more days of antibiotic therapy than any other indication for admission to U.S. children’s hospitals,” she said.
In 2011, the Infectious Diseases Society of America and Pediatric Infectious Diseases Society released guidelines for pediatric CAP, which recommend a first-line therapy with penicillin, ampicillin, or amoxicillin for most children who are immunized and healthy. “Only in situations where there’s a significant concern for an atypical organism should we be adding coverage for that – even in older children,” Dr. Tribble said. Following the release of the guidelines, she continued, multiple studies have shown that the use of first-line therapy is increasing in children’s hospitals. “However, a substantial proportion of children with pneumonia are admitted to non–children’s hospitals,” she said. “Prior to release of the guidelines, one study showed that use of first-line therapy for pediatric CAP was low in non–children’s hospitals (J Pediatr. 2014 165[3]:585-91), but postguideline CAP therapy in non–children’s hospitals has not yet been evaluated.”
For the current study, Dr. Tribble and her associates set out to evaluate antibiotic prescribing patterns for pediatric CAP in non–children’s hospitals and to compare prescribing patterns between children’s and non–children’s hospitals. They conducted a retrospective cross-sectional study of children aged 1-17 years admitted for CAP in 2013 to 323 hospitals, captured via the Pediatric Health Information System (PHIS) and Premier Perspective databases. PHIS is an administrative database that includes billing data, diagnosis codes, and procedure codes for about 44 freestanding children’s hospitals nationwide, while Premier Perspective encompasses data from 522 hospitals nationwide. The researchers used a validated ICD-9 code-based algorithm to identify patients with CAP and excluded those with complicated pneumonia or complex chronic conditions, those who received intensive care, and those with methicillin-resistant Staphylococcus aureus infection or colonization.
Children’s hospitals were defined as those with pediatric admissions accounting for more than 75% of all admissions. “This was after excluding newborns and admission for childbirth, because many community hospitals will have a birthing center or a NICU, but otherwise would not be considered a children’s hospital,” Dr. Tribble explained. Any other hospital was considered a non–children’s hospital.
Three different outcomes for antibiotic use were examined: those who ever received penicillin, amoxicillin, or ampicillin (guideline therapy); those who ever received a macrolide, fluoroquinolone, or tetracycline (atypical therapy); and those who received anything other than penicillin, amoxicillin, or ampicillin (nonguideline therapy). The standardized probability of exposure to select antibiotics was compared between children’s and non–children’s hospitals, adjusted for age, sex, and insurance provider.
In all, 323 hospitals contributed 15,495 CAP cases. Of the 323 hospitals, 49 were identified as children’s hospitals (44 from the PHIS database and 5 from the Premier database). Dr. Tribble reported results from 9,224 subjects admitted to children’s hospitals and 6,271 subjects admitted to non–children’s hospitals. The demographics between the two groups were similar: The patients’ mean age was 3 years, and 66% were younger than age 5 years.
After adjustment of data, patients admitted to children’s hospitals were found to be more likely to receive guideline therapy, compared with those admitted to non–children’s hospitals (46% vs. 15%, respectively), were less likely to received atypical therapy (36% vs. 51%), and were less likely to receive nonguideline therapy (78% vs. 94%; P less than .001 for all comparisons).
Dr. Tribble acknowledged certain limitations of the study, including the potential for misclassification of children’s hospitals in the Premier database, “although most likely I think we would have failed to identify a children’s hospital, and this would have biased us toward the null and made our difference less significant,” she said. “We are developing an absolute volume classification so we can look at this in another way.” Another limitation is that the study design did not account for the potential of combination therapy, “and you can’t account for change in therapy during hospitalization. Lastly, we compared data across different databases and across different hospital types.”