How Pharmaceutical Cost Indicators Are Affected when Adjusted to Hospitals' Clinical Workload: Data Analysis for Greek NHS Hospitals


Datamed SA
University of Athens
3 Athens Medical Society


OBJECTIVES: To examine how hospitals' pharmaceutical cost changes when clinical workload indicators are incorporated in the measurement and provide corrective estimates to the comparative performance assessment of the MoH for Greek NHS hospitals, for 2013.

METHODS: The study includes annual detailed financial and operational data, as recorded by the MoH's BI system,, and analytical data on the type of incidents (based on GR-DRGs) recorded, for 129 Greek NHS hospitals. Several KPIs were examined, but the study focuses on the mean pharmaceutical cost per patient, which was then adjusted to the case mix of the units, by using a hospital clinical weight indicator, estimated as the weighted average of particular incidents, based on GR-DRG data. The results were statistically analysed, in order to group the changes, before and after the indicator's adjustment, per hospital size category and assess hospitals' relative ranking.

RESULTS: The results reveal that there was a significant -but diminishing as the bed number increases- rise in the mean pharmaceutical cost per patient for three categories (25.49%, 14.23%, 11.74% increase for hospitals with less than 100, 101-250 and 251-400 beds, respectively) after the indicator was adjusted to the hospitals' clinical workload. For hospitals with more than 400 beds, the mean pharmaceutical cost per patient decreased by 15.38%, after the adjustment. The results also indicate significant changes in hospital ranking, even within the 4 categories, ranging from a 34 places ranking rise for Onassis Cardiac Surgery Centre, to a 21 places ranking drop for the G.H. of Thessaloniki 'Agios Dimitrios'.

CONCLUSION: Hospitals with more severe than average incidents, (cancer hospitals, cardiac surgery centres, etc) improved their performance and relative ranking, while small health centres and regional hospitals reported worse results after the indicators' adjustment. The study concludes that clinical workload should be incorporated in hospitals' pharmaceutical cost assessment.

“ISPOR 18th Annual European Congress, November, 2015”
 Download the presentation poster