Volume 145, Issue 5 , Pages 1153-1155, November 2013
published online 23 September 2013.
Philip S. Schoenfeld, Section Editor, John Y. Kao, Section Editor
Wigg AJ, McCormick R, Wundke R, et al. Efficacy of a chronic disease management model for patients with chronic liver failure. Clin Gastroenterol Hepatol 2013;11:850–858.
Cirrhosis affects >5.5 million patients and costs nearly $4 billion annually. Furthermore, there are nearly 30,000 new diagnoses of cirrhosis per year (Hepatology 2002;36:227–242). The prevalence of cirrhosis is increasing owing to an aging population of patients with chronic hepatitis C virus infection and an increasing prevalence of nonalcoholic fatty liver disease. Unfortunately, our current episodic system of health care delivery is suboptimal for patients with chronic diseases, such as cirrhosis (Gastroenterology 2010;139:14–16 e1). Patients with cirrhosis are at high risk of hospitalization, including nearly one third of patients requiring readmission within 1 month (Clin Gastroenterol Hepatol 2013;11:1335–1141). In other chronic diseases, such as congestive heart failure, chronic disease management (CDM) models have significantly improved clinical outcomes, including disease-specific admission rates, all-cause admission rates, and overall survival (Eur J Heart Fail 2005;7:1133–1144). CDM improves outcomes through several components including delivery system design, decision support systems, coordination of care between providers, and self-management support systems (Milbank Q 1996;74:511–544). The authors of this study sought to determine the impact of a CDM intervention on patients with decompensated cirrhosis (Clin Gastroenterol Hepatol 2013;11:850–858).
In this single-center, randomized, controlled, pilot study, the authors enrolled 60 adult patients who had been admitted with any cirrhosis-related decompensation (ascites, encephalopathy, variceal bleeding, spontaneous bacterial peritonitis, hepatocellular carcinoma (HCC), hepatorenal syndrome, alcoholic hepatitis, or sepsis). Patients were randomized 2:1 to CDM (n = 40) or usual care (n = 20), respectively, and then followed for 12 months. The multifaceted CDM intervention was administered by hepatology nursing staff after hospital discharge and continued for the duration of the 12-month trial. The intervention included nurse case management, decision support tools, patient and caretaker education, and nurse-generated patient data sheets for providers. Specifically, delivery system design measures included multidisciplinary care in clinic, home visits by a nurse within 1 week of discharge, and rapid access to care pathways as needed. Nurses provided self-management support, including patient and caregiver education, to optimize care between visits. Finally, decision support tools, such as evidence-based protocols for cirrhosis complications, and clinical information systems, such as patient data sheets, were implemented.
Patients enrolled in the CDM group had a 30% higher attendance rate at planned outpatient appointments compared to the usual care group (incidence rate ratio [IRR], 1.3; 95% confidence interval [CI], 1.1–1.6). Several quality-of-care indicators were higher in the intervention group than the usual care group, including significant differences in rates of HCC screening, vaccination, and liver transplant assessments. However, there was no difference in the primary study endpoint of cirrhosis-related hospital days between the CDM and usual care groups (17.8 vs 11.0 days per person-year, respectively; IRR, 1.6; 95% CI, 0.5–4.8). Similarly, the authors found no difference in cirrhosis-related admission rates (IRR, 2.2; 95% CI, 1.0–4.5), all-cause admission rates (IRR, 1.7; 95% CI, 1.0–3.7), median length of hospitalization (hazard ratio 1.3; 95% CI, 0.9–1.9), or overall survival (hazard ratio, 0.6; 95% CI, 0.3–.5).
CDM programs have been shown to improve clinical outcomes in several chronic diseases. In chronic obstructive pulmonary disease, for instance, CDM models resulted in a significant reduction in hospitalization rate (Arch Intern Med 2007;167:551–561). Similarly, for congestive heart failure, the use of CDM programs has been associated with significant reductions in both hospitalization and mortality rates (Eur J Heart Fail 2005;7:1133–1144). Although the upfront cost of these programs is not trivial, this cost may be offset by a substantial increase in quality-adjusted life-years and perhaps cost-savings with reduced downstream hospitalizations (Heart 2008;94:1601–1606). Cirrhosis is currently responsible for significant morbidity and health care costs and parallels other target populations of prior CDM efforts.
This pilot study is the first prospective, randomized trial assessing the impact of a CDM model in the management of chronic liver disease. In contrast with other chronic diseases in which CDM intervention significantly improves clinical outcomes, this study showed no difference in cirrhosis-related hospital days, all-cause admission rates, or median length of hospitalization. In fact, there was a nonsignificant trend toward worse outcomes in the intervention group, with more liver-related hospital days and higher all-cause admission rates.
The negative findings of this study may be related to intrinsic differences between patients with cirrhosis and those with other chronic diseases. The authors hypothesize that patients with cirrhosis may be more marginalized from health care services than other chronic diseases. It is also possible that acute exacerbations of cirrhosis are unpredictable and unavoidable, even in the setting of high-quality care. For example, hepatic encephalopathy can be a unique and challenging barrier to providing consistent outpatient care to patients with cirrhosis, including sudden exacerbations owing to a host of etiologies, including infection, medication noncompliance, and renal failure.
Alternatively, CDM interventions may be effective in patients with cirrhosis but simply not detected by this study owing to its limitations. First, this was a pilot study and was not powered a priori for all planned analyses (although the authors performed a post hoc power analysis). Furthermore, the small sample size led to imbalances in several measured (and likely unmeasured) confounders between the intervention group and usual care group. There were higher rates of reversible insults, such as spontaneous bacterial peritonitis and alcoholic hepatitis, in the usual care group. Similarly, there was a trend toward higher Model for End-stage Liver Disease scores and comorbidity index in the intervention group.
Furthermore, this study was conducted in a tertiary care center, with easy access to subspecialty care. This is a setting where CDM may be unnecessary, because rates of quality care were high among patients in the usual care group, with >80% of patients receiving HCC surveillance, variceal screening, and spontaneous bacterial peritonitis prophylaxis as needed. Prior studies have demonstrated that subspecialty consultation from a gastroenterologist results in shorter lengths of hospitalization and lower 30-day readmission rates among patients with cirrhosis (Hepatology 2001;34:1089–1095). Therefore, it is possible that CDM interventions may be more beneficial in a community hospital setting where expert consultation is scarcer or in a subgroup of high-risk patients.
Although the authors chose the components of their CDM intervention based on available evidence from other chronic diseases, it is possible that different components are necessary to improve outcomes in patients with cirrhosis. For example, a high proportion of admissions in the intervention group were for elective procedures, such as paracentesis, liver biopsy, or endoscopic procedures. Availability of an outpatient procedure facility could have prevented 48% of hospitalizations in the intervention group and should be included in future CDM intervention trials. Furthermore, several of the components of the CDM intervention in this study, such home visits and patient education, were directed to outpatient care, although the primary study outcomes were all inpatient based. Future studies are needed to characterize reasons for readmission and prolonged hospitalization among patients with cirrhosis to help develop an effective intervention strategy.
Interestingly, there was no difference in the primary and secondary outcomes of the study despite an improvement in several quality measures, such as HCC surveillance, vaccination, and liver transplant evaluation. Although process measures are more sensitive to differences in quality of care, outcomes are often of greater interest. It is possible that improvements in process measures would translate to improved outcomes with longer follow-up; however, it is also possible that there is a disconnect between these quality metrics and the outcomes of interest. Although a recent study from HALT-C helped to establish a link between HCC surveillance process measures and downstream outcomes (Am J Gastroenterol 2013;108:425–432), similar studies are needed for other cirrhosis quality metrics.
Despite not finding a difference in outcomes with CDM, this is an important proof-of-principle pilot study with several strengths. The study raises several interesting questions regarding the benefits of CDM among patients with decompensated cirrhosis. Future studies are needed to determine the optimal components of a CDM program as well as the most appropriate quality measures. Larger studies with longer follow-up are then needed to determine the effectiveness and cost-effectiveness of CDM interventions in improving outcomes among this patient population.
© 2013 AGA Institute. Published by Elsevier Inc. All rights reserved.