Estimating Hospitalization Costs for Regional Hospitals (Case Study)

A.               Interpret the coefficient of determination (R²) estimated for the nursing cost function.

B.               Describe the economic and statistical significance of each estimated coefficient in the nursing cost function.

C.               Average nursing costs for the eight for‑profit hospitals in the sample are only $318.52 per patient day, or $33.07 per patient day less than the $351.59 average cost experienced by the 32 not‑for‑profit hospitals.  How can this fact be reconciled with the estimated coefficient of -39.156 for the for-profit status variable?

D.               Would such an approach for nursing cost estimation has practical relevance for publicly-funded nursing cost reimbursement systems?

CASE STUDY SOLUTION

A.               Cost estimation results provided indicate that R² = 76.81%, meaning that 76.81 percent of the total variation in nursing costs per patient day can be explained by the five factors studied. For cross-sectional analysis, such a level of cost explanation is often quite adequate for gaining useful insight concerning cost characteristics.

B.               Each individual coefficient estimate is statistically significant at the 99% confidence level, with the exception of the Shots coefficient estimate, which is significant at the 95% level. In terms of economic interpretation, the 72.765 coefficient for the shots variable indicates an average nursing labor cost of roughly $72.76 per shot.  Similarly, IV therapy results in $215.68 in nursing costs per patient day, pulse taking and monitoring costs $36.24, and wound dressing costs $156.04 per unit.  Each of these four services appears to have a clear impact on nursing costs per patient day.  Interestingly, a coefficient of ‑39.156 for the profit‑status variable indicates that, on average, nursing costs per patient day are roughly $39.16 lower in for‑profit than in not‑for‑profit hospitals after accounting for differences in patient mix as captured by the four service categories.  This suggests that the greater efficiency or operating philosophy of for‑profit hospitals may be responsible for a substantial portion of the lower nursing costs these hospitals enjoy.

C.               By considering differences in the nursing services provided, along with the for‑profit or not‑for‑profit status of each hospital, it is possible to learn whether average cost differences are related to differences in patient mix or, perhaps, to other factors, such as efficiency or operating philosophy.  After accounting for the influences associated with variation in assorted output categories, the organization design variable appears relevant.  Before accounting for output differences, for-profit organizations appear to have nursing costs per patient day that are roughly $33.07 less than the average reported by not-for-profit hospitals. After accounting for differences in the patient mix as captured by the four service categories, for-profit organizations appear to have nursing costs per patient day that are roughly $39.16 less than the average reported by not-for-profit hospitals. This is an interesting finding, but the additional analysis would be necessary to determine if this effect is due to the operating advantages of for-profit hospitals or instead due to subtle differences in geographic location, patient mix, and so on.

D.               Despite obvious limitations, such a regression‑based approach can provide useful measures of costs for both private and public decision making. In practice, nursing care cost estimation and cost reimbursement methodologies that reflect the care needs of patients can be based on a manageable number of services.  In fact, Illinois, West Virginia, Ohio, and Maryland have implemented Medicaid nursing home reimbursement systems based on this concept, and several other states have similar case‑mix reimbursement systems under development.