Prior hospitalization experience of DRG outliers versus inliers

The Diagnosis-Related Group (DRG)-based Medicare Prospective Payment System has raised a number of concerns. One major concern has centered on both the definition of outlier patients and how hospitals are to be paid for their care. The epidemiology of outlier patients, however, has received relatively little attention. Using a retrospective database, we constructed a case control study in which the cases were DRG outliers and the control patients were DRG-matched inlier patients. We then examined both the prior hospitalization experience of outlier patients and inlier controls, and the prior outlier experience of the attending physicians of outlier patients and inlier control patients. We demonstrated that DRG-based outlier patients were more likely to have had prior admissions to the hospital as compared with DRG-matched inlier control patients. Moreover, DRG-matched outlier cases were more likely to have had prior outlier admissions as compared to controls--an effect that was more pronounced when only the subset of patients who actually had a prior admission was evaluated. Finally, physicians of the DRG outlier patients were more likely to have had previous outlier patients than physicians of the DRG-matched inlier control patients. In summary, we were able to demonstrate that DRG-based outlier patients have a different prior admission and outlier experience than DRG-matched inlier control patients. However, because we showed the same relationship when we evaluated the outlier experience of physicians of outlier patients as compared to physicians of DRG-matched inlier controls, we were unable to determine if the effect was patient- or physician-based.(ABSTRACT TRUNCATED AT 250 WORDS)