The growing burden of prescription drug costs has focused attention on
factors which influence physicians' prescribing decisions. We hypothe
sised that third party coverage of prescription costs would elicit sel
ection of expensive drugs, but that this tendency could be moderated b
y price reminders. In a mailed survey, primary care physicians through
out Ontario (n = 1072) were provided with the clinical scenario of a p
atient with an infectious exacerbation of chronic obstructive pulmonar
y disease, and asked to select diagnostic tests as well as 1 of 6 anti
biotics. Two antibiotics were expensive (ciprofloxacin and cefaclor; a
verage price $Can52.23), and 4 inexpensive [amoxicillin, cotrimoxazole
(trimethoprim/sulfamethoxazole), erythromycin and tetracycline; avera
ge price $Can2.80]. Neither expensive drug is considered first line th
erapy for the condition described. Questionnaires differed in the pres
ence or absence of drug benefit coverage and price information. The re
sponse rate was 71%. With third party cost coverage and prices shown,
18% of respondents selected an expensive antibiotic. This increased to
38% when the prices were omitted [odds ratio 2.72; 95% confidence int
erval (CI) 1.61, 4.60; p < 0.001], and decreased to 8% when the patien
t was said to have no drug benefits coverage (odds ratio 0.40; 95% CI
0.19,0.84; p < 0.01). On a control questionnaire in which neither cost
coverage nor prices were specified, and in which respondents were ask
ed to indicate the most appropriate antibiotic, 37% selected an expens
ive drug (relative to the version giving both prices and cost coverage
: odds ratio 2.70; 95% CI 1.62, 4.53; p < 0.001). We conclude that ina
ppropriate selection of expensive antibiotics is increased when the pa
tient has full drug benefit coverage but reduced when physicians are r
eminded of drug prices. These results support assessment of a price-or
iented educational intervention in an effort to reduce inappropriately
expensive prescribing.