OBJECTIVES: The governments of many countries are undertaking initiatives to assess the extent to which health care organisations fulfil important objectives of health care, such as health improvement, fair access and efficiency. However, the extent to which these health care organisations can influence these objectives is unclear. The purpose of this study is to examine the potential influence of English National Health Service territorial health authorities on 14 indicators of system performance. METHODS: The study uses performance data relating to approximately 5000 small geographical areas with average populations of 10,000. Multi-level statistical models are used to attribute variation in the indicators to three hierarchical levels--small areas, district health authorities and regional health authorities--after controlling for socio-demographic characteristics. Variations in indicators attributable to district or regional level give an indication of the extent to which health authorities may influence performance. RESULTS: After adjusting for socio-demographic characteristics, the proportion of variation in performance attributable to district health authorities varies from about 8% (for standardised mortality ratios) to about 76% (for waiting time for elective surgery). Variation at the regional level is smaller than at the district level. CONCLUSIONS: There appear to be very large variations between indicators in the extent to which health care organisations can influence health system performance. Choice of performance indicators and the managerial incentive regime based on the indicators should recognise this variability, as it is highly dysfunctional to hold managers accountable for measures of performance that are beyond their control.
OBJECTIVES: The governments of many countries are undertaking initiatives to assess the extent to which health care organisations fulfil important objectives of health care, such as health improvement, fair access and efficiency. However, the extent to which these health care organisations can influence these objectives is unclear. The purpose of this study is to examine the potential influence of English National Health Service territorial health authorities on 14 indicators of system performance. METHODS: The study uses performance data relating to approximately 5000 small geographical areas with average populations of 10,000. Multi-level statistical models are used to attribute variation in the indicators to three hierarchical levels--small areas, district health authorities and regional health authorities--after controlling for socio-demographic characteristics. Variations in indicators attributable to district or regional level give an indication of the extent to which health authorities may influence performance. RESULTS: After adjusting for socio-demographic characteristics, the proportion of variation in performance attributable to district health authorities varies from about 8% (for standardised mortality ratios) to about 76% (for waiting time for elective surgery). Variation at the regional level is smaller than at the district level. CONCLUSIONS: There appear to be very large variations between indicators in the extent to which health care organisations can influence health system performance. Choice of performance indicators and the managerial incentive regime based on the indicators should recognise this variability, as it is highly dysfunctional to hold managers accountable for measures of performance that are beyond their control.