Kuo, Yong‐Fang

Abstract Aims Readmission is frequent among patients with cirrhosis and is a complex multifactorial process. To examine the association of alcohol use disorder (AUD) and risk of readmission in patients with alcohol-associated cirrhosis. Methods and Results National Readmission Dataset (2016–2017) was used to extract a retrospective cohort of 53,348 patients with primary or secondary discharge diagnosis code of alcohol-associated cirrhosis with their first admission (26,674 patients with vs. propensity matched 26,674 without a primary or secondary discharge diagnosis code of AUD). Readmission within 30-day was lower (43.9 vs. 48%, P < 0.001) among patients identified to have AUD at the time of discharge. In a conditional logistic regression model, a diagnosis of AUD was associated with 15% reduced odds of 30-day readmission, 0.85 (0.83–0.88). Furthermore, the reason for readmission among patients identified vs. not identified to have AUD was less likely to be liver disease complication. The findings remained similar in a matched cohort of patients where the AUD diagnosis at discharge was listed as one of the secondary diagnoses only. Conclusion Although, our study findings suggest that identification of AUD at the time of discharge among patients hospitalized for alcohol-associated cirrhosis reduces the risk of 30-day readmission, unavailable information on patient counseling, referral for mental health specialist and treatment received for AUD limit the causality assessment. Future studies are needed overcoming the inherent limitations of the database to establish the role of identification and treatment of AUD in reducing readmission and liver decompensation in patients with alcohol-associated cirrhosis.

LINKED CONTENTThis article is linked to Singal et al and Barnard & Dave papers. To view these articles, visit https://doi.org/10.1111/apt.16461 and https://doi.org/10.1111/apt.16491

SummaryBackgroundAlcohol use and alcohol‐associated liver disease (ALD) burden are increasing in young individuals.AimTo assess host factors associated with this burden.MethodsNational Health and Nutrition Examination Survey (NHANES), National Inpatient Sample (NIS), and United Network for Organ Sharing (UNOS) databases (2006‐2016) were used to identify individuals with harmful alcohol use, ALD‐related admissions, and ALD‐related LT listings respectively.ResultsOf 15 981 subjects in NHANES database, weighted prevalence of harmful alcohol use was 17.7%, 29.3% in <35 years (G1) versus 16.9% in 35‐64 years (G2) versus 5.1% in ≥65 years (G3). Alcohol use was about 11 and 4.7 folds higher in G1 and G2 versus G3, respectively. Male gender and Hispanic race associated with harmful alcohol use. Of 593 600 ALD admissions (5%, 77%, and 18% in G1‐G3 respectively), acute on chronic liver failure (ACLF) occurred in 7.2%, (7.2 in G2 vs 6.7% in G1 and G3, P < 0.001). After controlling for other variables, ACLF development among ALD hospitalizations was higher by 14% and 10% in G1 and G2 versus G3, respectively. Female gender and Hispanic race were associated with increased ACLF risk by 8% and 17% respectively. Of 20,245 ALD LT listings (3.4%, 84.4%, and 12.2% in G1‐G3 respectively), ACLF occurred in 28% candidates. Risk of severe (grade 2 or 3) ACLF was higher by about 1.7 fold in G1, 1.5 fold in females and 20% in Hispanics.ConclusionYoung age, female gender, and Hispanic race are independently associated with ALD‐related burden and ACLF in the United States. If these findings are validated in prospective studies, strategies will be needed to reduce alcohol use in high risk individuals to reduce burden from ALD.

OBJECTIVES: Alcohol-associated liver disease is increasing, especially hospitalizations with acute on chronic liver failure and need for liver transplant. We examined trends in prevalence, inhospital mortality, and resource utilization associated with AALD and ACLF in the young. METHODS: The National Inpatient Sample (2006–2014) was queried for hospitalizations with a discharge diagnosis of cirrhosis using the International Classification of Diseases, Ninth Edition, codes. ACLF hospitalization was defined as ≥2 organ failures and stratified by age: young (≤35 years) and older (>35 years). RESULTS: Of 447,090 AALD admissions (16,126 in young) between 2006 and 2014, ACLF occurred in 29,599 (6.6%), of which 1,143 (7.1%) were in young. Compared with older, admissions in young had more women (35% vs 29%), were obese (11% vs 7.6%), were Hispanics (29% vs 18%), have alcoholic hepatitis (AH) (41% vs 17%), and have ACLF grades 2 or 3 (34% vs 25%), P < 0.001 for all. Between 2006 and 2014, ACLF in AALD among young increased from 2.8% to 5.2%, with an AH proportion from 24% to 42%, P < 0.0001 for both. Young had more complications requiring ventilation (79% vs 76%) and dialysis (32% vs 28%), P < 0.001 for both. Compared with older, ACLF admission in young had longer hospitalization (12 vs 10 days) with higher hospital charges ($127,915 vs $97,511), P < 0.0001 for both, with 20% reduced inhospital mortality (54%–45%), P < 0.001. DISCUSSION: AALD-related hospitalizations are increasing in young in the United States, mainly because of the increasing frequency of AH. Furthermore, this disease burden in young is increasing with a higher frequency of admissions with more severe ACLF and consumption of hospital resources. Studies are needed to develop preventive strategies to reduce burden related to AALD and ACLF in young.

Acute‐on‐chronic liver failure (ACLF) is characterized by multiple organ failure (OF) with high short‐term mortality. There is lack of population‐based data on trends on etiology specific ACLF related burden. National Inpatient Sample (2006‐2014) was queried using ICD‐09 codes for admissions with cirrhosis and ACLF (≥2 extrahepatic OF). Of 1,928,764 admissions for cirrhosis between 2006 and 2014, 112,174 (5.9%) had ACLF (4.5%, 1.2%, and 0.2% with ACLF 1, 2, and 3, respectively). The brain was the most common OF in 11.9%, followed by respiratory failure in 7.7%, cardiac failure in 6.3%, and renal failure in 5.6%. ACLF increased by 24% between 2006 and 2014 with a 63% increase in 179,104 patients with nonalcoholic steatohepatitis (NASH) cirrhosis (3.5% to 5.7%); a 28% increase in patients with 429,306 alcoholic cirrhosis (5.6% to 7.2%); a 25% increase in patients with 1,091,053 with other etiologies (5.2% to 6.5%); and no significant change in 229,301 patients with viral hepatitis (VH) (4.0% to 4.1%). In‐hospital mortality was higher among ACLF patients compared with patients without ACLF (44% versus 4.7%; P < 0.0001). Each NASH‐related ACLF patient compared with other etiologies had a longer mean length of stay (14 versus 12 days), was associated with higher median total charges (US $151,196 versus US $134,597), and had more frequent use of dialysis (45% versus 36%) and longterm care (32% versus 26%; P < 0.0001 for all). Results remained similar in a subgroup analysis after including half of admissions with cryptogenic cirrhosis as NASH. In conclusion, NASH cirrhosis is the most rapidly growing indication for ACLF‐related hospitalization and use of hospital resources. In the setting of improved treatment options for chronic hepatitis, the health care burden of chronic viral‐related liver disease remains stable. Population‐based strategies are needed to reduce the health care burden of cirrhosis, particularly related to NASH.

BackgroundData on liver transplant (LT) outcomes using deceased donors with heavy drinking (HD) (>2 drinks per day) are scanty.MethodsUsing the United Network for Organ Sharing database (2002-2014), we examined outcomes after LT in adults comparing deceased HD donors with non-HD (ND) donors.ResultsOf 56 182 first LTs performed in the United States for 10 common indications using deceased donors, 47 882 with available information on alcohol use were analyzed. Of these 47 882 LT recipients, 7298 (15%) were from HD donors, with similar proportion over time (2002-2014, Armitage trend testP= 0.75) and for recipient liver disease etiology (χ2P= 0.42). Proportion of liver organ used for LT was lower for HD donors compared with ND donors (63% vs 78%;P< 0.001). Five-year outcomes on first LT comparing 7166 HD donors and 21 498 ND donors matched based on propensity score were similar for liver graft (73.7% vs 73.7%, log rankP= 0.98) and patient survival (77.6% vs 77.0%,P= 0.36). On Cox regression analysis, history of HD in deceased donors did not affect liver graft 1.02 (0.97-1.08) or patient survival 1.03 (0.97-1.09).ConclusionsAmong LT recipients using select liver grafts, history of HD in deceased donors does not impact outcomes after LT.

Background and Aim: Data are scanty on allocating simultaneous liver kidney (SLK) based on model for end-stage disease (MELD) score. Diabetes mellitus (DM) and hypertension (HTN) are frequent in cirrhosis patients. We analyzed transplant recipients with DM and/or HTN to compare MELD-based outcomes of SLK to liver transplantation alone (LTA). Materials and Methods: Of 13,584 first deceased donor liver transplantation among patients with DM and/or HTN (1530 or 11.2% SLK), MELD score predicted SLK [1.02 (1.01-1.03)]. SLK was beneficial for 5-year patient survival at MELD score ≥43 (78.6% vs. 62.6%, P=0.017), but not at MELD score <29 (74.8% vs. 76.2%, P=0.63). Among 11,405 recipients (976 SLK) at MELD score <29, SLK (n=816) was beneficial compared with 706 LTA [75% vs. 64%, P<0.001; 0.71 (0.55-0.91)] at serum creatinine (SC) ≥2 but not at SC<2 [73% vs. 76%, P=0.32; 0.85 (0.60-1.2)]. Among patients with MELD score 29 to 42, SLK (n=484) and LTA (n=1403) had similar survival [69% vs. 69%, P=0.58; 0.9 (0.7-1.5)]. Among patients with MELD score ≥43, SLK (n=70) was associated with 35% improved patient survival at 5 years compared with 222 LTA [0.65 (0.46-0.93)]. Conclusions: Among patients with DM and/or HTN, SLK is useful at: (a) MELD score <29 and SC≥2 and (b) MELD score ≥43. Prospective studies are needed to confirm these findings as basis to optimize use of SLK.