Casemix funding in rural NSW: exploring the effects of isolation and size

The New South Wales Department of Health (NSW Health) wishes to make appropriate use of casemix data as inputs to the determination of funding levels for small rural hospitals. However, other factors such as hospital size and degree of isolation might need to be taken into account. The study reported here involved correlation of actual expenditures with those predicted by use of a casemix model alone, across 105 small public hospitals in the State. We then explored the extent to which the correlation could be increased by the addition of distance and isolation variables. It was found that actual costs were highly correlated with those predicted from the casemix data alone, and that the correlation increased when both the distance and the size variables were introduced. However, contrary to expectations, reduced size was associated with reduced costs, and reduced isolation was associated with increased costs. It was concluded that, while the predicted relationships may be present, they are likely to be relatively weak and are probably being masked by other factors not present in the model. In particular, it seems likely that there are variations in severity within the acute admitted patient category which are not fully explained by the casemix instrument used in this study (the DRG classification). We suggest that other terms be introduced to control for this possibility before any further attempt is made to test whether size and distance factors can be identified which work in the expected direction.