The significance of laboratory testing for the German diagnosis-related group system--the systematic evaluation of comorbidities of relevance to case reimbursement and continued development of the DRG Watchdog software.

The launch of the G-DRG system version 2004 made it necessary to update our former studies on the importance of laboratory testing (Clin Lab 2002;48:327-333). Using a systematic search algorithm, we established a total of 2,828 comorbidities, the ICD coding of which has a positive effect on case reimbursement, thus helping to secure hospital revenue. 62% of these comorbidities were found to depend exclusively or predominantly on laboratory testing. On average, one such comorbidity can be said to influence approximately 100 DRGs (range 2 to 226). In order to gain a clearer idea of the practical benefit of such "hits", amounting to several 100,000s, we selected about 5% of them for illustration with a computer program called DRG Watchdog. This program shows just how much additional reimbursement can be achieved for a specific DRG upon the ICD coding of a specific comorbidity. The program is freely available on the Internet at www.trillium.de and enjoys more than 100 search runs per day.

Characterization of antibody responses to endogenous and exogenous antigen in the nonobese diabetic mouse.

It is suggested that a T-helper cell 2 (Th2) shift and Th2 spreading of autoimmunity following immunization with beta-cell antigen causes diabetes protection. To address this, antibody titer and subclass to insulin, glutamic acid decarboxylase (GAD)65, IA-2, and IA-2beta proteins were measured by radiobinding assays in untreated or immunized female nonobese diabetic mice. Untreated nonobese diabetic mice developed autoantibodies to insulin (IAA), but not GAD or IA-2/IA-2beta, and IAA-positive mice had increased diabetes risk (P < 0.001). IAA were IgG1 and IgG2b. In immunized mice, IgG1 and lesser IgG2b insulin antibodies were promoted by subcutaneous injection of insulin plus incomplete Freund's adjuvant, insulin plus Montanide ISA 720, and glucagon plus incomplete Freund's adjuvant, but not by incomplete Freund's adjuvant plus GAD65, IA-2beta, or phenylethanolamine N-methyltransferase, or adjuvant alone. Diabetes incidence was significantly reduced in immunized groups with elevated insulin antibody (IA) responses. Spreading of antibody responses to GAD or IA-2/IA-2beta following immunization was rare, and antibody epitope spreading was only detected in IA-2beta immunized mice. Humoral autoimmunity in nonobese diabetic mice is, therefore, limited to IAA with Th2 subclass phenotype and is associated with increased diabetes risk. This contrasts the diabetes protection provided by immunization protocols that promote this response and suggests that Th2 immunity may not be the principal regulator of beta-cell destruction in autoimmune diabetes.

Detection and documentation of DRG-relevant comorbidities using laboratory tests.

Germany will soon begin per case payment by DRG, and preparations are in progress in most hospitals and insurance companies. The Academic Teaching Hospital Munich-Schwabing in Munich decided to explore coding strategies by considering the impact of diagnoses that could be detected by pathology tests. An Australian database was analysed. We detected "discriminating" diagnoses--that is, diagnoses that could be found in level A or B DRGs, and not in the respective lower severity DRG. After isolating 584 diagnoses, they were rated by a laboratory specialist, to determine whether they could be proved by pathology tests. 187 diagnoses were selected in this way. In the next step, theoretical cases were generated and grouped. 157 diagnoses were found to produce a switch to a higher DRG. The diagnoses, the DRGs and the respective laboratory tests were then arranged in a small MS-Excel program to allow comfortable browsing. The overall success rate of 84% shows that laboratory medicine can contribute to correct coding for DRGs.