Introduction
Coronary heart disease (CHD) is the most common cause of death in UK. The CHD National Service Framework (NSF)¹ sets national standards for the management of high-risk patients with diagnosed CHD. A Medicines Management service delivered by community pharmacists could help to achieve these standards. The Community Pharmacy Medicines Management Project² was a large randomised controlled trial (RCT) to evaluate such a service. The aim of the work reported here was the development and application of a tool to measure appropriateness of treatment of CHD patients, comparing outcomes in patients who received the pharmacy led service (the intervention group) with those receiving standard treatment (controls).

Methods
A cumulative 7-point score (0 (worst case) to 7 (best case)) was designed specifically for the study, based on similar measures³, the NSF criteria, data from the literature and discussions with experts. The score included both clinical (aspirin-related treatment, as well as lipid and blood pressure management) and lifestyle (smoking cessation, physical activity, diet, body mass index (BMI)) components. Data were collected from patient medical records using a standard template, or by patient self-report using questionnaires administered at baseline and follow-up (12 months after intervention). Data had to be recorded within 12 months of the census date. Due to a high rate of missing data at follow-up with BMI components, a sensitivity analysis was applied to exclude this component from the total score. The clinical and lifestyle components were also considered separately. Differences in scores between the intervention and control group at follow-up were adjusted for differences in patient characteristics or outcomes at baseline using multiple regression analysis. Levels of statistical significance were stated at p<0.01. Data were analysed using SPSS v11.5 and STATA v8.

Results
At baseline, the mean (SD) global 7-point score for appropriate CHD treatment was similar in both groups (4.3 (1.1) intervention and 4.2 (1.2) control). At follow-up, there was no statistically significant difference in mean score between groups (4.6 (1.2) and 4.6 (1.1)). In the sensitivity analysis using the 6- point scale without BMI data, the baseline results were 3.4 (1.1) and 3.3 (1.2), respectively. At follow-up, no statistically significant difference was shown (4.2 (1.0) in both groups). When the clinical and lifestyle 3-point sub-scores were considered separately, the increase in global score across time was attributable to a change in the lifestyle component only (baseline 1.3 (0.8) in both groups; follow-up 2.1 (0.7) and 2.0 (0.8)).

Discussion 
A tool to measure appropriateness of treatment for CHD patients has been developed and applied. Sensitivity analyses demonstrated the tool was robust to missing data.

References

1. http://www.nelh.nhs.uk/nsf/chd/nsf/index.htm. (Last visited 03/12/04);
2. 2. http://www.medicinesmanagement.org.uk (Last visited 29/03/04);
3. Campbell N, Ritchie L, Thain J, Deans H, Rawles J, Squair J, et al. Secondary prevention in coronary heart disease: a randomised trial of nurse led clinics in primary care. Heart 1998;80:447-52.