Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

OBJECTIVES: The introduction of an internal market in the British National Health Service (NHS) has highlighted the importance of developing appropriate, valid and timely measures of hospital activity, both for the purposes of specifying and monitoring contracts and for evaluating the success of the NHS reforms in general. This paper compares the validity of five case mix methods (Diagnosis Related Groups (DRGs); Healthcare Resource Groups (HRGs) versions 1 and 2; specialty classification; a simple age categorization) in predicting resource use. METHODS: Two data sets were used to compare different case mix methods. A 3% random sample (n approximately equal to 300,000) of the 1992/3 Hospital Episodes Statistics was used to test their ability to predict variation in length of stay, and a second set of individually costed patient episodes from two hospitals (n approximately equal to 40,000) was used to test their ability to explain cost variation. Analysis of variance models were used to assess the fit of each of the case mix systems to test data and a simple significance test of differences in mean squared error between models was applied. RESULTS: All case mix methods performed poorly on untrimmed data. When lengths of stay greater than 29 days were excluded, version 2 of HRGs explained 31% of total variance in length of stay and 25% of cost variation. DRGs explained less variance but performed better than HRGs version 1. For a typical hospital patient population consisting of a range of specialties, the difference in explanatory power between HRGs V2 and DRGs was statistically significant at the 5% level for sample sizes of approximately 2000 or greater. For individual specialties, the minimum sample size required for the difference between the groupers to be significant ranged from around 300 to over 2000. CONCLUSIONS: The locally developed HRGs version 2 system appears to offer superior performance in terms of resource homogeneity to other currently available approaches. It is also more adaptable and cheaper than imported alternatives and has been formally endorsed by the UK medical Royal Colleges.


Journal article


J Health Serv Res Policy

Publication Date





10 - 19


Analysis of Variance, Data Interpretation, Statistical, Diagnosis-Related Groups, Episode of Care, Evaluation Studies as Topic, Health Care Reform, Health Expenditures, Health Resources, Health Services Research, Hospitals, Humans, Length of Stay, Medicine, Specialization, State Medicine, United Kingdom, Utilization Review