Implementing formulary recommendations in primary care : effect on patient outcomes
This research aimed to measure the effect on health outcomes of implementing selected recommendations of the Grampian Joint Drug Formulary in primary care. Antibiotics used in the treatment of uncomplicated lower urinary tract infections (UTIs), ulcer healing agents and peripheral vasodilators were selected for study, thereby reflecting both acute and chronic prescribing. For the UTI study, 12 randomly selected high and low prescribers of trimethoprim, the recommended agent, each agreed to distribute 20 patient questionnaires. Following a period of 18 months and despite repeated contact with the GPs, only 89/480 (19%) questionnaires had been distributed. Patient response was, however, very high with 80 (90%) questionnaires returned. Health outcome measures identified that trimethoprim resulted in no or mild symptoms in 40/45 (91 %) of patients. These findings must be interpreted with caution due to the low level of questionnaire distribution and thus cannot be extrapolated to the total population of patients. In addition, the poor questionnaire distribution did not permit comparison between trimethoprim and non-recommended therapy. One hundred and eighty four patients receiving repeat prescriptions for ulcer healing agents were identified from one general practice. Therapy in 95 patients did not adhere to formulary recommendations. Changes to therapy were considered inappropriate in 11 patients due to factors such as severe depression and a further 8 were deemed unsuitable for participation for non-clinical reasons. The remaining 76 patients were contacted with 19 (25%) refusing to participate. Fifty seven patients were interviewed using the Glasgow Dyspepsia Severity Score and Short Form 36 (SF-36). Changes in health outcomes were measured for 21 patients where a change in therapy had taken place. These results were difficult to interpret due to the diversity of changes recommended and the lack of data relating to those patients not participating. Work involving peripheral vasodilators aimed to determine the effect on health outcomes of cessation of therapy. Forty five patients receiving these agents in 2 practices were identified, although 8 had not requested a prescription in the previous year. Two further patients were excluded from the study due to cancer and old age. The remaining 35 agreed to be interviewed using the Walking Impairment Questionnaire and SF-36. All patients were subsequently instructed to stop therapy for 2 months, although 6 (17%) refused to follow this instruction, one patient was seriously ill thus was excluded and 3 refused to be reinterviewed. Of the remaining 25 patients, no significant differences were observed in the domains studied. Seventeen patients (68%) expressed no desire to restart therapy, generating considerable savings. These results must be interpreted with caution since those not stopping therapy or refusing re-interview are likely to have responded differently to those completing the study. The measurement of health outcomes following formulary implementation deserves further work.