Antivirals and antibiotics for influenza in the United States, 1995-2002

Purpose: To measure the rates of antiviral and antibiotic prescribing for patients diagnosed with influenza in the United States. Methods: We performed a retrospective analysis of visits to ambulatory clinics and emergency departments in the National Ambulatory Medical Care Survey (NAMCS) and the National Hospital Ambulatory Medical Care Survey (NHAMCS) with a diagnosis of influenza that occurred in seven influenza seasons between 1 October 1995 and 31 May 2002 (n = 1216). Results: There were an estimated 22 million visits (95%CI, 17-26 million visits) with a diagnosis of influenza to community ambulatory clinics (88% of visits), hospital ambulatory clinics (3%) and emergency departments (9%) in the United States between the 1995-1996 and the 2001-2002 influenza seasons, inclusive. The sample was 63% adults, 44% male and 84% white. Physicians prescribed antivirals in 19% of visits and antibiotics not associated with an antibiotic-appropriate diagnosis in 26% of visits. In multivariable modeling, independent predictors of antiviral prescribing were adult age (OR, 2.1; 95%CI, 1.1-4.0) and Medicare insurance (OR, 0.1 compared to private insurance; 95%CI, 0.0-0.6). Antiviral prescribing was marginally associated with influenza season (OR, 1.2 per influenza season; 95%CI, 1.0-1.4). Independent predictors of antibiotic prescribing were influenza season (OR, 0.8 per influenza season; 95%CI, 0.7-0.9), male sex (OR, 0.6; 95%CI, 0.4-0.9), adult age (OR, 2.3; 95%CI, 1.2-4.2) and emergency department visits (OR, 0.5 compared to community ambulatory visits; 95%CI, 0.3-0.8). Conclusions: Physicians prescribed antiviral medications to 19% of patients they diagnosed with influenza; the proportion that would have been clinically appropriate is unknown. In contrast, physicians prescribed apparently inappropriate antibiotics to 26% of these same patients, a rate that, encouragingly, decreased over time.