Generation and Performance of Outcome Indicators in Liver Cirrhosis: The Value Based Medicine in Hepatology Study (V.B.M.H.)

Liver Cirrhosis (LC) is responsible for high morbidity, mortality and raising costs. Current guidelines set the standard of care for management of cirrhosis in clinical practice, but explicit outcome indicators (OIs) are lacking. If available, OIs could guide clinical care and decision making, so that efforts and resources can be properly allocated. Aim of our study was to generate and test a set of health care OIs for compensated (CC) and decompensated (DC) LC. This study is part of a larger effort (the V.B.M.H. study) to generate OIs for several liver diseases. An expert panel of hepatologists identified a set of OIs for LC according to experience and scientific evidence (as of 2010), and used a modified Delphi method to rate them through a RAND 9-point agreement scale. A final list of 7 indicators with median rating ≥7 and with disagreement index ≤1 was selected. Three OIs were designed for CC and 4 for DC. In the second phase of the study, the selected OIs were tested in clinical practice through a prospective multicenter observational study, involving three tertiary centers in Lombardy, Italy. A web-based EMR was used to collect data. 1732 LC patients were enrolled in 18 months: 984 CC (57%) and 748 DC (43%). 91% of these LC patients had at least two consultations in a 13 months median follow-up. The annual rate of decompensation in CC (OI#1) was 12%, decompensation being more frequent in HCV-related CC. The annual incidence of 1st variceal bleeding (VB) for low-risk varices was 2% and was null for high risk varices, indicating effective primary profilaxis (OI#2). Annual incidence of HCC was 4%, while the percentage of CC patients diagnosed at early stage HCC (BCLC 0/A) was 74%, reflecting the accuracy of oncologic surveillance (OI#3). In DC patients, survival at 1 year after the first episode of decompensation (ascites in 78% of cases) was 99% for CPT A, 87% for CPT B and 73% for CPT C (OI#4). 3% of DC patients had an episode of VB, with survival of 88% after 6 weeks from the VB episode (OI#5) and with recurrence rate of 27% (OI#6). 19 out of 748 DC patients (3%) had spontaneous bacterial peritonitis, with 79% survival after 6 weeks from the episode (OI#7). In conclusion, the selected OIs performed well in monitoring the rate of decompensation in CC and the accuracy of surveillance for HCC; in DC, OIs were able to capture survival and the efficacy of management of major complications. This study represents the first attempt to identify and test a set of value-based OIs for LC, and provides a reference tool for healthcare policy makers to improve quality of care in patients affected by LC