Background
In earlier work1,2, we have demonstrated that Pharmacy Performance Indicators can be used to predict pharmacy workforce using multiple linear regression analysis. In order to develop further this evidence-led approach to workforce planning, we have investigated the possibility of using additional NHS performance indicators as predictors of pharmacy establishment. These NHS trust level indicators are used as indicators of health outcomes and are publicly accessible through the Department of Health website.

Method
Health outcome data was extracted from the Department of Health website3. This data was merged with the existing data provided by senior pharmacy managers in London, which constitutes standardised, summative data on pharmacy performance indicators (PIs) from individual trusts. Pharmacy establishment, expressed as whole time equivalents (WTEs), was designated as the dependent variable and the remainder of the pharmacy performance indicators and health outcome measures indicators were considered as independent variables. Data was audited. Data was analyzed using SPSS v12 and both stepwise and enter methods were used together with model diagnostics.

Results
We were able to match data for the time period 1999/2000 for 37 trusts. A robust model was developed, with a coefficient of determination R2=0.917 (F =88.551 p<0.0001) illustrated by figure 1. The residuals were independent (Durbin Watson =1.886). Collinearity diagnostics were favourable. Staff establishment, expressed as WTE, could be described by inpatient drug expenditure, clinical pharmacy ward visits, occupied bed days and rate of emergency readmission to hospital (Table 1.). For some hospitals, the residual plot revealed a large difference between actual and predicted WTE based on this model.
Fig.1

Conclusion
The present study indicates that there are relationships between WTEs and selected pharmacy performance indicators and trust level health outcome indicators. Multivariate techniques appear to be of value in modelling pharmaceutical services, and will provide evidence for planning and skill mix decisions.
Acknowledgement
The authors would like to thank London senior pharmacy managers for permission to analyse the performance indicator data.

References
1. Sharott P, Webb DG, Bates I. Multiple regression as a procedure to interrogate pharmacy performance indicators. 8th Health Services Research and Pharmacy Practice Conference, Leeds University 2002. (ISBN 0 9538505-3-6. The Royal Pharmaceutical Society), p12.
2. Borja-Lopetegi A, Webb DG, Bates I, Sharott P. Validation of regression techniques to predict pharmacy workforce in NHS trusts. 9th Health Services Research and Pharmacy Practice Conference, Queen's University, Belfast 2003. (ISBN 0 9538505-4-4, Royal Pharmaceutical Society, London), p4.
3. http://www.doh.gov.uk/nhsperformanceindicators/hlpi2000/arealist_t.html