Macrolide antibiotics are used as first-line therapy for the empirical outpatient treatment of community-acquired pneumonia (CAP). The recent emergence of macrolide resistant pathogens is a major concern. The objective of the study was to quantify the frequency of macrolide treatment failure in CAP and examine its impact on health care use.
A cohort with CAP treated with macrolides in outpatient clinics from January 2002 to June 2003 was identified using the PharMetrics health insurance claims database. Macrolide treatment failure was defined as receipt of a second antibiotic, different from the first, within 4 weeks of initial macrolide treatment. Endpoints were numbers of hospitalisations, emergency room and office visits during the 1 month period following initial macrolide prescription. We examined ICD-9 CM codes on claim forms for post-treatment hospitalisations and visits to identify those most likely to be related to treatment failure versus other causes. Data analysis involved Poisson regression.
Among 40,751 patients identified, 11.4% received a second antibiotic within 4 weeks. This treatment failure group was older, had more women, and used more medical care prior to the index visit. After adjusting for age, gender, and prior health care use, patients experiencing treatment failure were more likely to be admitted to the hospital or use emergency room and outpatient care following the index visit (see Tables 1Table 1.
Subsequent Hospitalization, Emergency Room Visits and Clinic VisitsDiagnosis Group*Cases with SubsequentHospitalizationEmergency Room VisitClinic VisitTreatment SuccessTreatment FailureTreatment SuccessTreatment FailureTreatment SuccessTreatment FailureNarrow10.60%7.53%1.20%7.61%27.36%53.08%Broad20.67%8.08%1.80%10.14%34.25%72.40%None31.51%9.37%3.69%13.36%37.05%68.84%All41.69%10.42%4.78%17.37%56.52%92.17%
All p < 0.0001). The magnitude of the association was stronger for admissions and visits for care associated with CAP, as indicated by ICD-9 codes, than for services related to other conditions.
By our definition, about 11% of patients with CAP in this population who were treated with macrolides failed therapy within a 4 week window. Macrolide treatment failure was associated with increased health care utilization.
Reported increases in macrolide treatment failure remain important issues when treating patients with CAP. This supports close monitoring of the current practice guidelines for the use of macrolide antiinfective agents. and 2Table 2.
Impact of Macrolide Treatment Failure on Health Care Utilization among Patients with Community-acquired PneumoniaDiagnosis GroupRelative Risk* (95% CI) in Treatment Failure Group, Compared to Those Without Treatment Failure, For Subsequent:HospitalizationEmergency Room VisitClinic VisitNarrow12.0 (9.40, 15.4)6.13 (5.15, 7.31)2.90 (2.76, 3.04)Broad11.1 (8.75, 14.1)5.61 (4.84, 6.49)3.05 (2.93, 3.17)None1.03 (0.90, 1.18)1.18 (1.10, 1.25)1.39 (1.34, 1.44)All5.66 (4.82, 6.64)3.96 (3.59, 4.37)2.27 (2.20, 2.35)*
From multivariate Poisson regression model adjusted for age group, gender, and prior healthcare utilization (inpatient and outpatient office visits)
J.H. Wu, Aventis