Jp. Metlay et al., NATIONAL TRENDS IN THE USE OF ANTIBIOTICS BY PRIMARY-CARE PHYSICIANS FOR ADULT PATIENTS WITH COUGH, Archives of internal medicine, 158(16), 1998, pp. 1813-1818
Background: Increased antibiotic use for outpatient illnesses has been
identified as an important determinant of the recent rise in antibiot
ic resistance among common respiratory pathogens. Efforts to reduce th
e inappropriate use will need to be evaluated against current trends i
n the outpatient use of antibiotics. Objectives: To examine national t
rends in the use of antibiotics by primary care physicians in the care
of adult patients with cough and identify patient factors that may in
fluence antibiotic use for these patients. Methods: This study was bas
ed on a serial analysis of results from all National Ambulatory Medica
l Care Surveys beginning in 1980 (when therapeutic drug use was first
recorded) to 1994 (the most recent survey year available). These surve
ys are a random sampling of visits to US office-based physicians in 19
80, 1981, 1985, and annually from 1989-1994. Eligible visits included
those by adults presenting to general internists, family practitioners
, or general practitioners with a chief complaint of cough. A total of
3416 visits for cough were identified over the survey years. Survey r
esults were extrapolated, based on sampling weights in each year, to p
roject national rates of antibiotic use for patients with cough. Addit
ional analyses examined the rates of antibiotic use stratified by pati
ent age, race, and clinical diagnosis. Results: Overall, an antibiotic
was prescribed 66% of the time during office visits for patients with
cough: 59% of patient visits in 1980 rising to 70% of visits in 1994
(P = .002 for trend). In every study year, white, non-Hispanic patient
s and patients younger than 65 years were more likely to receive antib
iotics compared with nonwhite patients and patients 65 years or older,
respectively. Conclusions: The rate of antibiotic use by primary care
physicians for patients with cough remained high from 1980 to 1994, a
nd was influenced by nonclinical characteristics of patients.