UBC Theses and Dissertations
Prescribing in the community hospital : method of assessment and evaluation of modification strategies Jewesson, Peter John
The need for the development of a well-controlled methodological system to evaluate the impact of strategies for modifying prescribing behaviour in the community hospital setting is evident. Equally important is the need to design intervention programs which are both practical and effective strategies for modifying prescribing behaviour. To accomplish these objectives a study was conducted to determine the relative impact of three levels of intervention on prescribing behaviour. The interventions included a control, a standard in-hospital non-facilitated persuasive strategy (pharmacy newsletter) and a novel multi-component non-facilitated persuasive strategy (Drug Usage Information (DUI) Program). This third level of intervention was designed to modify prescribing behaviour in a manner which was effective, yet practical to implement. The four-part DUI Program involved: 1) written announcement and hospital endorsement of the program; 2) distribution of an initial report containing topic drug prescribing patterns and recommendations for change; 3) distribution of a follow-up report identifying improvements in topic prescribing patterns and reiterating the change recommendations; and 4) a medical rounds presentation. Over the 2-month period of the DUI Program, each physician received a personalized letter announcing the program, initial and follow-up reports and was notified of the medical rounds presentation. Six drug-related topics involving areas of previously identified suboptimal usage were studied. These topics focused upon optimization of dosage regimens for prophylactic antibiotics, cefazolin, cimetidine and gentamicin and serum drug level monitoring for theophylline and anticonvulsants. Appropriateness of prescribing behaviour for these topics was categorized according to pre-established explicit process criteria. This was accomplished by a blinded review of health records at two similar community hospitals. Employing a duplicated Youden square experimental design, each topic in each hospital was randomly allocated to one of the three intervention levels. Patterns of prescribing were then compared for periods before and after the interventions using analysis of variance and chi-square statistical procedures. Potential interfering factors including test hospital medical and pharmacy publications were monitored. Prescribing patterns were stable for the six month period prior to the interventions. A comparison of the two component 3-month periods did not identify any specific hospital or topic effects. Similarly, when comparing the 3-month periods immediately before and after the interventions, no effects associated with topics or hospitals were evident. However, intervention effects were apparent with the Drug Usage Information Program demonstrating an Improvement in prescribing patterns greater than both the pharmacy newsletter and control. No difference between the pharmacy newsletter intervention and control was evident. Significant improvement in prescribing patterns was observed for three of the four topics addressed by the Drug Usage Information Program. These effects persisted into the second 3-month period, however, some diminution of effect was apparent. No major extraneous influences that might bias the changes in prescribing patterns were identified. The study methodology developed permitted valid assessment of the impact of modification strategies on prescribing behaviour in the community hospital. The results of this study indicated that dissemination of drug information via standard hospital pharmacy newsletter was ineffective as a means of modifying prescribing behaviour in this setting. An alternative method is the novel Drug Usage Information Program. This program was both effective and practical to implement.
Item Citations and Data