Computerised prescribing: assessing the impact on prescription repeats and on generic substitution of some commonly used antibiotics.

OBJECTIVES: To assess the impact of two interventions on computer-generated prescriptions for antibiotics--(i) an educational intervention to reduce automatic computerised ordering of repeat antibiotic prescriptions, and (ii) a legislative change prohibiting the "no brand substitution" box being checked as a default setting in prescribing software--and to compare these findings with those of a similar survey we conducted in 2000. DESIGN AND
SETTING: Prospective audit of consecutive prescriptions for four antibiotics (amoxycillin, amoxycillin/clavulanate, roxithromycin, and cefaclor) commonly prescribed for upper respiratory tract infections in community pharmacies in New South Wales and Queensland between 1 November 2008 and 31 January 2009. PRIMARY OUTCOME: rate of repeat prescription ordering on computer-generated versus handwritten prescriptions. Secondary outcome: rate of checking of the "no brand substitution" box on computer-generated versus handwritten prescriptions.
RESULTS: Data were collected on 2807 prescriptions presented to 51 pharmacies (50 in NSW, one in Queensland), of which 2354 were computer-generated. Repeats were ordered on 1633 computer-generated prescriptions (69%) compared with 183 handwritten prescriptions (40%). These proportions were identical to those found in 2000, although the rates of computer prescribing were much higher in this study (84% v 54%). This difference in repeat prescribing was statistically significant (odds ratio adjusted for clustering at pharmacy level, 2.87; 95% CI, 2.32-3.55). Twenty-three (1%) of the computer-generated prescriptions had the "no brand substitution" box checked compared with 3 (0.7%) of the handwritten prescriptions (27% and 1%, respectively, in our previous survey).
CONCLUSIONS: The legislative change which disallowed having the "no brand substitution" box checked as a default setting in prescribing software had a dramatic impact on the checking of the "no brand substitution" box. In contrast, there was no sustained effect of educating prescribers about software default settings relating to repeat prescribing of antibiotics. Other actions are required if unnecessary repeat prescriptions for some medicines, such as antibiotics, are to be reduced.