Estimating health care treatment costs

This chapter considers the conceptual and many empirical problems associated with estimating health care costs. It does so by considering the estimation of such costs at different levels of aggregation — initially considering health costs incurred at the individual level, including those arising from treatments undergoing clinical trials, and then at the hospital level, the latter being the sector with the largest level of expenditure in most health care systems. The topics covered include the representativeness of sample populations, the importance of skew in the health care cost distribution, issues of censoring, and missing data and their relevance to the estimation process. The consistency of relevant estimators is assessed through discussion of transformation, linkage and general linear methods. Attention then turns to the hospital sector, where lack of knowledge of the underlying production process is a considerable drawback in specifying appropriate cost functions. The amendment of hospital cost functions to particular concerns, including the hospital's control of demand and reactions to demand uncertainty, are outlined. Given the importance of health care cost estimation in helping to risk adjust and predict future health care expenditure levels, as well as help determine reimbursement levels within the hospital sector, more attention needs to be given to this area of research