| 摘要 |
<p>The expected charges for treating a patient are estimated by the use of linear regression techniques wherein variables and coefficients of estimate models (23, 35, 43, 45) are built from historic patient data of all episode types (21) and which include secondary and collateral illnesses that greatly affect the cost of treating a patient for the primary illness. The evaluation of collateral illness is also made possible since the processing techniques are not simply averaging data for the primary illness as has been done in the past. This allows physicians to understand the magnitude and types of resource usage that are appropriate for particular illnesses. The present invention allows better management of many aspects of the healthcare system by significantly reducing a component of the difference between the estimate and actual charges for treating a particular illness of a patient. This then allows the remaining difference to be attributed to differences in the way that health service providers treat each illness of a patient. Estimates are made for the expected cost (29) of providing specific episodes of care.</p> |
