# Hospital Cost Analysis.

Reviewer: Richard PonArul, California State University, ChicoHospital Cost Analysis provides a comprehensive treatment of the topic, starting with the necessary theoretical foundations and followed by empirical work with data from hospitals in Queensland, Australia. The book is self-contained in its exposition and should be accessible to anyone with a rudimentary knowledge of economics and statistics.

Estimation of cost per unit of output for hospitals becomes necessary for assessing their operational efficiency as well as for the determination of output-based payments from the government. How exactly should the output of a hospital be defined and measured? The answer to this difficult question will decide the meaningfulness of the cost estimates obtained. In the first part of the book, after a brief exposition of multiproduct production functions and cost functions, the author turns to this important question.

In the field of health, as in other fields, the assessment of outcome is a current topic of debate. Should the desired outcome - improvement in the health of the patients - also be the output to be measured? The author examines whether improvement in the patients' health or the provision of treatment should be chosen as the measure of output and concludes that the latter measure is more appropriate and practical. Then he compares the number of cases treated and the number of patient-days of care as two competing measures of treatment provided and makes a convincing case for number of cases treated as a good output measure for hospitals.

For any hospital, it will be easy to imagine why each case is unique. However, to make the cost estimation feasible, cases must be classified into groups with presumed homogeneity. Two classification schemes are presented: the International Classification of Diseases and Diagnosis-Related Groups. Also presented is the information theory-based case mix index, which is a scalar index of the case mix that measures the deviation of the case mix of a hospital from a benchmark. One of the benchmarks is the case mix for the aggregate of all hospitals.

By the end of the first (theoretical) part of the book, readers should appreciate the compromises that must be made in order to come up with estimates of unit cost per output.

The second part of the book presents the empirical results on cost estimation using data from Queensland hospitals. Most of the work is on public hospitals. First, a multiple regression model is used to study the effect of case mix on costs. Due to multicollinearity problems, the results are not very informative. Two approaches are used to reduce the number of parameters and improve the estimation: data reduction through principal components analysis and construction of a scalar case mix index. The latter method does produce estimates of marginal costs that are plausible. The slope coefficients for the different specifications of the scalar index (dependent variable = average cost per case) are uniformly positive, indicating that deviation from the benchmark leads to higher average cost. It is not clear why this should be the case, since deviation from the benchmark can be toward low cost case types as well as toward high cost case types.

Despite the fact that the results are not strong, the exposition of the methodology is extremely detailed and will be very helpful to researchers who want to extend the work to other data sets. The description of the construction scalar case mix index (information theory-based) is also very meticulous.

The remainder of the second part looks at other factors that may potentially affect the unit cost - scale, capacity utilization, input prices, and teaching activities, for example. One chapter compares public and private hospitals' cost functions. Lack of availability of financial data for private hospitals leaves the results tentative. But, as with the earlier chapters, the methodology is presented in good detail. The last chapter in the second part of the book looks at an interstate comparison of hospital costs (in Queensland and New South Wales).

The third and final section of the book examines the issues of performance appraisal of hospitals and payments to hospitals based on output measurements. Average cost per case is decomposed into average stay (relative stay index) and average cost per day to better understand the source of good/bad performance. In the recent past, input-based payments (fee for service) to hospitals have been replaced by output-based payment schemes. Cost estimation becomes an essential input for such payment schemes. Whether hospitals engage in strategic behavior in response to an output-based payment scheme will be an important consideration in determining the effectiveness of such schemes. The book makes a brief mention of this question, leaving it to the reader to explore further.

In sum, the book comprehensively presents the theory of hospital cost estimation and the empirical results. The style of writing is very lucid, making for easy reading. The background information regarding the hospital system in Australia is added at the right places to enable readers to understand the empirical results in context. This book should be of interest to researchers and policy-makers in the field of health care economics. Its usefulness will also extend to other fields such as education, where output measurement and cost estimation are inherently difficult.

Printer friendly Cite/link Email Feedback | |

Author: | PonArul, Richard |
---|---|

Publication: | Journal of Risk and Insurance |

Article Type: | Book Review |

Date: | Dec 1, 1996 |

Words: | 863 |

Previous Article: | Handbook on Insurance Coverage Disputes. |

Next Article: | Estimating Human Life Values to Assess Life Insurance Potential. |

Topics: |