2005 HSR&D National Meeting Abstract
3028 — Antibiotic Prescribing in Veterans with Acute Respiratory Infections of Non-Bacterial Origin
Aspinall SL (VA Pittsburgh Healthcare System)
Good CB (VA Pittsburgh Healthcare System)
Metlay JP (Philadelphia VA Medical Center)
Fine MJ (VA Pittsburgh Healthcare System)
To determine patient, provider, and system characteristics associated with prescribing antibiotics for outpatients with nonspecific upper respiratory tract infections (URIs) or acute bronchitis at two academic Veterans Affairs Medical Centers (VAMCs).
We conducted a retrospective cohort study of antibiotic prescribing for outpatients with respiratory infections of likely viral etiology (i.e., nonspecific URIs and acute bronchitis) at the VA Pittsburgh Healthcare System from June 2003 - June 2004 and the Philadelphia VAMC from December 2003 - March 2004. Patients who presented to the emergency department and were diagnosed by the provider with a nonspecific URI or acute bronchitis were eligible for inclusion. Patients diagnosed with a concomitant bacterial infection were excluded (e.g., sinusitis, pneumonia). We reviewed patients’ medical records and interviewed providers to obtain patient and provider factors that were potentially associated with antibiotic prescribing. We also collected the time of day of patient presentation. Stepwise logistic regression was used to assess the factors independently associated with antibiotic prescribing.
At one VAMC, providers prescribed antibiotics for 15.5% (75/483) of the patients diagnosed with nonspecific URIs and 68.5% (87/127) of those with acute bronchitis. At the second VAMC, providers prescribed antibiotics for 30.9% (30/97) of the patients with nonspecific URIs and 97.9% (46/47) of those with acute bronchitis. Using multivariable analysis, presence of 1 or more comorbidities, presence of chronic obstructive pulmonary disease, diagnosis of acute bronchitis, purulent nasal discharge, purulent sputum, fever, shortness of breath, altered breath sounds, hospital site, and presentation on an evening or weekend were all independently associated with antibiotic prescribing.
Although antibiotic prescribing for URIs and acute bronchitis varied substantially between the two sites, they were overused in both VAMCs for these infections with a likely viral etiology. Site and evening/weekend effects, symptoms perceived to indicate a bacterial infection, and underlying pulmonary disease were independently associated with prescribing antibiotics.
Interventions to decrease prescribing in nonspecific URIs and acute bronchitis are warranted. These interventions should address the institutional culture of antibiotic prescribing and physicians’ beliefs that certain signs and symptoms are predictors of bacterial infection and necessitate antibiotics.