Our final empirical exercise considers directly the incentives of hospitals to adopt higher levels of technology. Identifying the role that hospital characteristics play in determining the allocation of ambulance patients is in many ways similar to a study of a differentiated goods demand system, in which hospitals compete in the marketplace for patients on the basis of geography and characteristics. However, these two settings also differ in some respects; in particular, while hospitals will presumably have incentives to attract some ambulance patients, a given hospital may want to deter particular types of patients (the uninsured or patients whom are hard-to-treat but do not generate significant income). While these distributional questions are extremely interesting, the present analysis will focus on the sensitivity of the overall patient share to particular hospital investments.
|DEPENDENT VARIABLE = L HOSPITAL MARKET SHARE|
|INDIVIDUAL HOSPITAL CHARACTERISTICS|
|URGENT CARE CENTER||0.6070(0.2911)|
|TRAUMA CENTER LEVEL||0.6626(0.3233)|
|INTENSITY OF RIVAL HOSPITAL COMPETITION|
|# of HOSPITALS||-1.0527(0.1432)|
|AVERAGE URGENT CARE||-0.3947|
|AVERAGE CATH LAB||-0.6712(0.3689)|
|AVERAGE OPENHEART FAC||-0.1427(0.5958)|
|AVERAGE TRAUMA CENTER||-0.6226|
Table 12 presents a results which relate the proportion of a county’s patients in the dataset who are allocated to a given hospital, SHARE, to the characteristics of that hospital as well as the characteristics of other hospitals in the county. First, and not surprisingly, the market share of a given hospital is declining in the total number of hospitals present in a given county. Our more interesting results are derived from our analysis of the specific features of hospitals which seem to impact this market share. In particular, simple measures of the overall “size” of the hospital— the total number of physicians, the total number of hospital beds— are uncorrelated with the hospital market share. In contrast, specific technological investments (such as cardiac catheterization laboratories and the rating of the emergency room) are correlated with the overall market share. Since allocation does appear to respond to technology investment, we conclude that the interaction between the pre-hospital system and technology adoption should be considered in analyses of the incentives for investment by hospitals.
One important caveat to our interpretation of Table 12 is that our results do not necessarily imply that if a given hospital increased its technology, it would increase its market share. If our sample contains some hospitals characterized by higher than average quality, larger number of consumers would use that hospital. The large market share could increase the incentives of the hospital to adopt technology; or, it could be that technology is an integral part of maintaining high overall quality. In either case, a low-quality hospital who adopted sophisticated technology would not necessarily increase its market share.
It is also possible to investigate how the sensitivity of market share to hospital characteristics might depend on the type of pre-hospital emergency response system available in a given county. However, in our preliminary analysis of this dataset, we have not found a robust interaction effect.