Purpose: There are extensive data that describe the prescriptive behaviors of physicians (MDs) for upper respiratory tract infections; however, there is a paucity of data on the antibiotic-prescribing patterns of nurse practitioners (NPs). The purpose of this study was to describe and predict factors that are associated with antibiotic prescribing by NPs and MDs for viral upper respiratory infections in the ambulatory setting.
Data sources: The study utilized a cross-sectional retrospective design of data from the National Hospital Ambulatory Medical Care Survey and the National Ambulatory Medical Care Survey between 1997 and 2001. Data were collected on a national probability sample of 506 NP and 13,692 MD visits for patients with nonspecific upper respiratory tract infection, viral pharyngitis, and bronchitis.
Conclusions: Bivariate analysis found no significant differences in antibiotic prescribing for viral upper respiratory tract infections by NPs (50.4%) and MDs (53%). Broad-spectrum antibiotics accounted for 36.6% of the NP antibiotic prescriptions and for 33.2% of the MD antibiotic prescriptions. Multivariate analysis identified several clinical and nonclinical factors that are associated with NP antibiotic prescribing. The strongest positive predictors of NP antibiotic prescribing were black race, Medicaid insurance, Northeast region, and diagnoses of viral pharyngitis and bronchitis. The significant negative predictor was Medicaid insurance status. The strongest positive predictors of MD prescribing were viral pharyngitis, bronchitis, and non-antibiotic prescription.
Implication for practice: The excessive use of antibiotics for upper respiratory infections of viral etiology by both NPs and MDs suggests the continuing need for educational initiatives such as "academic detailing" as well as increasing involvement by both groups of providers in the dissemination of clinical guidelines and system-based quality assurance programs. Also, the lower rate of antibiotic prescribing for viral infections by NPs for patients with Medicaid insurance suggests more appropriate cost-effective care in this population of patients. More study is needed in general on prescribing by NPs for Medicaid patients. Finally, the strong association of nonclinical factors suggests the need for awareness and improvement of prescribing decisions by both NPs and MDs.