Aims. To describe differences between general practice prescribed and
dispensed medications in terms of patient characteristics and category
of drug. Methods. Computerised prescribing records and prescriptions
presented to pharmacies were retrospectively reviewed. All prescriptio
ns generated from the computers of 13 practices over a 12 week period
in 1992 were compared with prescriptions dispensed. Data from the two
sources were matched and unmatched items were analysed to determine wh
ether patient demography or category of drug prescribed influenced the
rate of prescription dispensing. The nine most commonly prescribed dr
ug categories were examined in detail. Results. A total of 49 756 item
s were prescribed to 19 299 people and 43 302 (87.0%) of these were di
spensed. Antibiotics were the most commonly prescribed category of dru
g accounting for 17.6% of nondispensed items. There was no significant
-difference by gender in the proportion of people (9.8%) failing to cl
aim prescribed items, nor in the number of items (13.0%) prescribed bu
t not dispensed. Differences in dispensing rates by community services
card (CSC) status of patients were statistically significant for both
numbers of people failing to uplift their medications and for numbers
of items not dispensed (p < 0.001). There were also significant diffe
rences by high user health card (HUH) eligibility (p < 0.005) and age
(p < 0.001). Conclusions. There is a high rate of nondispensing of med
icines prescribed in general practice. Patients eligible for governmen
t subsidies are more likely than other patients to have their prescrip
tions filled. Current pharmaceutical subsidies may be inadequately tar
geted. There are no significant gender differences, while increasing a
ge is associated with lower nondispensing rates. General practitioners
need to be aware of these factors when making prescribing decisions.