Hospitalizations for Chronic Liver Disease: Time to Intervene at Multiple Levels

Chronic liver disease (CLD) is an important, and likely underestimated, national and global public health problem. CLD and its underlying risk factors are highly prevalent and mostly clinically silent, but may eventually lead to cirrhosis, which is frequently complicated by considerable morbidity and mortality from gastrointestinal varices, ascites, encephalopathy, and hepatocellular carcinoma. Asrani et al¹ report using data from a large health care system, including 16 hospitals in Texas in which CLD hospitalization rates increased over time (2004–2013) more than those for congestive heart failure or chronic obstructive airway disease, and had longer durations of stay and higher readmissions rate. The use of standardized hospitalization ratio and several adjustments for demographic, insurance, and facility features allowed for fair comparisons among these disorders. The authors made a strong case that it is about time that CLD is given a similar respect, and perhaps more similar systematic interventions as those for heart failure and obstructive airway disease.

The high and increasing inpatient burden of CLD should catch the attention of hospitals, health care systems, and payers who are interested in reducing cost and improving care in the short term. The fact that the inpatient burden of CLD was higher than congestive heart failure and chronic obstructive pulmonary disease, 2 disorders with long track records of disease as well as facility-related quality measures related to hospitalization. For example, the Centers for Medicare & Medicaid Services and Hospital Quality Alliance began publicly reporting 30-day mortality measures for heart failure in 2007 and the Centers for Medicare & Medicaid Services has since expanded the publicly reported outcome measures to include 30-day readmission for heart failure. Similar to hospitalizations, mortality owing to cirrhosis according to death certificates data from the Vital Statistics has been increasing in the United States since 2009.² Furthermore, deaths owing to cirrhosis are expected to triple by 2030.³ These findings should fuel the growth of the emerging literature on the magnitude and determinants of readmission among CLD patients, as well as interventions related to inpatient and postdischarge management algorithms and disease-specific quality metrics, as well as considering facility-related quality measures that are CLD specific.

Apart from these immediate important implications, I highlight the significance of hospitalizations rates as disease burden indictors as well as quality measure of ambulatory care and public health interventions, and determinants of current and future overall burden of CLD.

Hospitalizations as a Measure of Overall Disease Burden?

Disease burden is the impact of a health condition as measured by financial cost, mortality, and morbidity, with the incidence and prevalence being the major determinants of disease burden. CLD is among several chronic health conditions that are ambulatory care-sensitive conditions—that is, when the illness is controlled with effective and timely outpatient care, hospitalizations may be often avoided.⁴ Therefore, although hospitalization rates for CLD measure the narrow but important and expensive domain of inpatient burden, they do not correlate well with the overall disease burden. Hospitalizations reflect the tip of the overall CLD burden, the rest of which depends on the prevalence, incidence, and etiology of CLD in the underlying population. For example, most hepatitis C virus (HCV)-related hospitalizations reflect the maturation of the existing birth cohort of Baby Boomers with untreated or uncured HCV in the face of a stable or even declining overall HCV-related CLD burden. Conversely, hospitalization rates may serve as indicators of access to and quality of ambulatory care for preventing, detecting, and treating CLD before cirrhosis develops. Therefore, the high or increasing CLD hospitalization rates reported by Asarni et al¹ also reflect likely gaps and failures in ambulatory care to prevent CLD and to prevent CLD patients from progressing from uncomplicated to complicated cirrhosis.

The drivers of CLD incidence and prevalence and eventually CLD related hospitalizations and mortality are shown in Table 1. The drivers include the magnitude of existing cases (prevalence) and new ones (incidence) as well as the virulence of the risk factors (relative risk), which together with prevalence translate into the population-attributable fraction. The population-attributable fraction represents the proportion of CLD that can be prevented by removing the underlying risk factor, an endeavor that heavily depends on the existence of efficacious preventive or therapeutic measures and the degree of their successful adoption in the community (effectiveness). HCV and hepatitis B virus (HBV) have low prevalence, but those infected are at a considerable risk of CLD and its complications. We expect that most patients with active HCV or HBV who are linked to care will be treated and cured (HCV) or adequately suppressed (HBV). Studies have shown a considerable reduction but not complete elimination of risk of future decompensation or hepatocellular carcinoma with HCV cure⁵ or HBV suppression. However, the extent and speed of the decrease in HCV-related CLD depends on screening and detecting the HCV- and HBV-infected cohorts, the dissemination of direct- acting antiviral and nucleotide analog treatments, and to a lesser degree on the severity of liver disease at the time of successful treatment, and the clinical course after achieving a sustained Virologie response and whether additional factors (eg, alcohol and or fatty liver] or a small but worrisome recent increase in HCV acquisition will bend the curve of the expected decrease in viral hepatitis-related CLD. For HBV, the combination of HBV vaccination in the United States and other countries should be dealing a major blow to the incidence of HBV.

Table 1.

Drivers of Hospitalizations and Mortality Related to Complications of CLD

	Annual Incidence Trend	Prevalence (Trend)	Relative Risk (Trend)	PAF	Prevention (efficacy/ effectiveness)	Treatment (efficacy/ effectiveness)
HCV	Declined but recent outbreaks	1%–2% (declining)	20 (declining)	20%–30%	++/++	+++/++
HBV	Declining	0.5% (declining)	20 (declining)	5%–10%	+++/++	++/++
Alcoholic liver disease	Rising	5%–10% (stable)	2 (stable/ increasing)	20%–30%	++/+	++/+
NAFLD	Rising	10%–30% (increasing)	1.5–2.0 (increasing)	25%–35%	−/−	+/−

CLD, chronic liver disease; HBV, hepatitis B virus; HCV, hepatitis C virus; NAFLD, nonalcoholic fatty liver disease; PAF, population-attributable fraction.

Increasing rates of alcohol use in the United States will likely result in more alcoholic liver disease, especially in younger age groups and in racial-ethnic minorities.⁶ Heavy alcohol drinking alone or in combination of other risk factors promotes the development and progression of CLD. There has been an increase in the number and proportion of transplants for alcoholic liver disease starting in 2013.⁷

The increase in NAFLD in the United States qualifies its dramatic description of an “epidemic,” with a prevalence of up to one-third of adults in Dallas County.⁸ The progression rates from NAFLD to nonalcoholic steatohepatitis, cirrhosis, or hepatocellular carcinoma are unclear and probably low in the general population, but a frequently quoted 15% to 20% progression from NAFLD to nonalcoholic steatohepatitis does not bode well for the future of CLD-related complications.⁹ Reports from transplant centers indicate an already considerable increase in the proportion of patients listed or transplanted with NAFLD as their main underlying CLD risk factor.¹⁰ Although weight loss improves alanine aminotransferase levels and occasionally decreases hepatic inflammation and fibrosis, there is little to suggest that considerable and sustained weight loss is possible among enough patients to result in a meaningful decrease in the risk of cirrhosis and/or hepatocellular carcinoma. Several medications are being tested for nonalcoholic steatohepatitis treatment, but even if their safety and efficacy are shown, there are major obstacles related to widespread adoption in the near future. The likely role of nonalcoholic, and to a lesser part alcoholic, fatty liver disease and is also evident in the increasing prevalence of CLD among adolescents and young adults reported in the National Health and Nutrition Examination Survey.¹¹ The saving grace here might be that the recency of the NAFLD cohort and the lag period required for cirrhosis and complications allow interventions to potentially change the current projections.^12,13

Although the immediate implications and interventions related to CLD hospitalization are important and should receive the proper attention, the larger goal of prevention, early detection, and early diagnosis and treatment of CLD, thus preventing cirrhosis-related morbidity and mortality, is crucial and unfortunately less likely to receive appropriate attention. The greatest potential for altering decreasing CLD progression and hospitalizations include the following.

• Increased awareness and coordinated approaches, similar to those taken for congestive heart failure and pulmonary diseases.

• Further implementation of electronic health records and integrated health care systems for care management. These conceptually allow for population health management programs for mass screening for risk factors of CLD (eg, NAFLD, HCV, alcohol and HBV], and subsequent triage of positive (and negative) results.

• Addressing detrimental traits in our current “leaky” health care system, including the lack of incentives for insurers and payers to invest in interventions, because they run the risk of paying now and having another health system reap the future benefit.

Without such increased awareness and actions, we observe with great apprehension the anticipated increasing wave of hospitalizations related to CLD.