ChatGPT4 outperforms endoscopists for determination of post-colonoscopy re-screening and surveillance recommendations.

BACKGROUND: Large language models (LLM) including ChatGPT4 improve access to artificial intelligence, but their impact on the clinical practice of gastroenterology is undefined. In this study, we aim to compare the accuracy, concordance and reliability of ChatGPT4 colonoscopy recommendations for colorectal cancer re-screening and surveillance to contemporary guidelines and real-world gastroenterology practice. METHODS: History of present illness, colonoscopy data and pathology reports from patients undergoing procedures at two large academic centers were entered into ChatGPT4 and it was queried for next recommended colonoscopy follow-up interval. Using McNemar's test and inter-rater reliability, we compared the recommendations made by ChatGPT4 with the actual surveillance interval provided in the endoscopist's procedure report (gastroenterology practice) and the appropriate USMSTF guidance. The latter was generated for each case by an expert panel using the clinical information and guideline documents as reference. RESULTS: Text input of de-identified data into ChatGPT4 from 505 consecutive patients undergoing colonoscopy between January 1st and April 30th, 2023 elicited a successful follow-up recommendation in 99.2% of the queries. ChatGPT4 recommendations were in closer agreement with the USMSTF Panel (85.7%) than gastroenterology practice recommendations with the USMSTF Panel (75.4%) (P<.001). Of the 14.3% discordant recommendations between ChatGPT4 and USMSTF Panel, recommendations were for later screening in 26 (5.1%) and earlier screening in 44 (8.7%) cases. The inter-rater reliability was good for ChatGPT4 vs. USMSTF Panel (Fleiss κ: 0.786, CI95%: 0.734-0.838, P<.001). CONCLUSIONS: Initial real-world results suggest that ChatGPT4 can accurately define routine colonoscopy screening intervals based on verbatim input of clinical data. LLM have potential for clinical applications, but further training is needed for broad use.

Hepatocellular carcinoma after direct-acting antivirals for hepatitis C is associated with KIR-HLA types predicting weak NK cell-mediated immunity.

BACKGROUND AND AIMS: Second-generation direct-acting antivirals (2G DAA) to cure HCV have led to dramatic clinical improvements. HCV-associated hepatocellular carcinoma (HCC), however, remains common. Impaired immune tumor surveillance may play a role in HCC development. Our cohort evaluated the effects of innate immune types and clinical variables on outcomes including HCC. METHODS: Participants underwent full HLA class I/KIR typing and long-term HCV follow-up. RESULTS: A total of 353 HCV+ participants were followed for a mean of 7 years. Cirrhosis: 25% at baseline, developed in 12% during follow-up. 158 participants received 2G DAA therapy. HCC developed without HCV therapy in 20 subjects, 24 HCC after HCV therapy, and 10 of these after 2G DAA. Two predictors of HCC among 2G DAA-treated patients: cirrhosis (OR, 10.0, p = 0.002) and HLA/KIR profiles predicting weak natural killer (NK) cell-mediated immunity (NK cell complementation groups 6, 9, 11, 12, OR of 5.1, p = 0.02). Without 2G DAA therapy: cirrhosis was the main clinical predictor of HCC (OR, 30.8, p < 0.0001), and weak NK-cell-mediated immunity did not predict HCC. CONCLUSION: Cirrhosis is the main risk state predisposing to HCC, but weak NK-cell-mediated immunity may predispose to post-2G DAA HCC more than intermediate or strong NK-cell-mediated immunity.

Differential Effects of Ascites and Hepatic Encephalopathy on Waitlist Mortality in Liver Transplantation by MELD 3.0.

Background: MELD 3.0 introduces changes to address waitlist disparities for liver transplant (LT) candidates. Ascites and hepatic encephalopathy (HE) are important milestones in the natural history of cirrhosis regardless of the Model for End-Stage Liver Disease (MELD) score. We aim to assess the impact of ascites and HE and its interaction with MELD 3.0 on waitlist mortality. Methods: This is a retrospective study of patients listed for LT in the Organ Procurement and Transplantation Network database from 2016 to 2021. The primary outcome was waitlist mortality (death/delisting for too sick to LT). Ascites/HE were classified as moderate ascites without moderate HE (mAscites), moderate HE without moderate ascites (mHE), both moderate ascites/HE (mBoth), and neither. MELD 3.0 scores were categorized as <20, 20-29, 30-39, and ≥40. Results: Of 39 025 candidates, 29% had mAscites, 3% mHE, and 8% mBoth. One-year waitlist mortality was 30%, 38%, and 47%, respectively, compared with 17% (all P < 0.001) for those with neither. In multivariable Cox regression, the adjusted risk of waitlist mortality associated with mAscites (versus neither) was a hazard ratio (HR) of 1.76 (95% confidence interval [CI], 1.55-2.00) when the MELD 3.0 score was <20, significantly higher than when the MELD 3.0 score was 20-29 (HR 1.40; 95% CI, 1.27-1.54), 30-39 (HR 1.19; 95% CI, 1.04-1.35), and ≥40 (HR 1.14; 95% CI, 0.91-1.43, interaction P < 0.05 for all). A similar pattern was observed by MELD 3.0 for both moderate ascites/HE. Conclusions: The presence of moderate ascites alone, or combined with moderate HE, not only increases the risk of waitlist mortality but also has a differential effect by MELD 3.0, especially at lower MELD scores. Earlier strategies addressing this group and improving treatment plans or access to LT regardless of MELD remain needed.

Association of Renin-Angiotensin System Inhibition With Liver-Related Events and Mortality in Compensated Cirrhosis.

BACKGROUND & AIMS: While renin-angiotensin system inhibition lowers the hepatic venous gradient, the effect on more clinically meaningful endpoints is less studied. We aimed to quantify the relationship between renin-angiotensin system inhibition and liver-related events (LREs) among adults with compensated cirrhosis. METHODS: In this national cohort study using the Optum database, we quantified the association between angiotensin-converting enzyme (ACE) inhibitor or angiotensin-receptor blocker (ARB) use and LREs (hepatocellular carcinoma, liver transplantation, ascites, hepatic encephalopathy, or variceal bleeding) among patients with cirrhosis between 2009 and 2019. Selective beta-blocker (SBB) users served as the comparator group. We used demographic and clinical features to calculate inverse-probability treatment weighting-weighted cumulative incidences, absolute risk differences, and Cox proportional hazard ratios. RESULTS: Among 4214 adults with cirrhosis, 3155 were ACE inhibitor/ARB users and 1059 were SBB users. In inverse probability treatment weighting-weighted analyses, ACE inhibitor/ARB (vs SBB) users had lower 5-year cumulative incidence (30.6% [95% confidence interval (CI), 27.8% to 33.2%] vs 41.3% [95% CI, 34.0% to 47.7%]; absolute risk difference, -10.7% [95% CI, -18.1% to -3.6%]) and lower risk of LREs (adjusted hazard ratio [aHR], 0.69; 95% CI, 0.60 to 0.80). There was a dose-response relationship: compared with SBB use, ACE inhibitor/ARB prescriptions ≥1 defined daily dose (aHR, 0.65; 95% CI, 0.56 to 0.76) were associated with a greater risk reduction compared with <1 defined daily dose (aHR, 0.87; 95% CI, 0.71 to 1.07). Results were robust across sensitivity analyses such as comparing ACE inhibitor/ARB users with nonusers and as-treated analysis. CONCLUSIONS: In this national cohort study, ACE inhibitor/ARB use was associated with significantly lower risk of LREs in patients with compensated cirrhosis. These results provide support for a randomized clinical trial to confirm clinical benefit.

National prevalence estimates for steatotic liver disease and subclassifications using consensus nomenclature.

BACKGROUND AND AIMS: The multisociety consensus nomenclature has renamed NAFLD to steatotic liver disease (SLD) with various subclassifications. There is a paucity of data regarding how the new nomenclature modifies our understanding of disease prevalence and patient phenotypes. APPROACH AND RESULTS: Using the National Health and Nutrition Examination Survey from January 2017 to March 2020, we included all participants aged 18 years or above with complete vibration-controlled transient elastography measures. SLD and its subclassifications [metabolic dysfunction-associated SLD (MASLD), MASLD + increased alcohol intake (MetALD), alcohol-associated liver disease (ALD), etiology-specific/cryptogenic] were defined according to consensus nomenclature. National SLD prevalence and subclassifications were estimated, and among key subgroups [age, sex, race/ethnicity, advanced liver fibrosis (liver stiffness measurement [LSM] ≥11.7 kPa)]. Among 7367 participants, 2549 had SLD (mean age 51 y, 57.7% male, 63.2% non-Hispanic White). The estimated prevalence of SLD was 34.2% (95% CI 31.9%-36.5%): MASLD 31.3% (29.2%-33.4%), MetALD 2% (1.6%-2.9%), ALD 0.7% (0.5-0.9%), etiology-specific/cryptogenic 0.03% (0.01%-0.08%). In exploratory analyses, participants classified as non-SLD with (vs. without) advanced fibrosis had a higher mean number of metabolic risk factors [2.7 (2.3-3.1) vs. 2.0 (1.9-2.0)] and a higher proportion with average alcohol use ≥20 g/d (women)/≥30 g/d (men) [20.9% (6.2%-51.3%) vs. 7.2% (6.1%-8.4%)]. In another exploratory analysis, increasing quantities of alcohol use remaining below the threshold for MASLD + increased alcohol intake were associated with advanced liver fibrosis in men, but not women. There was 99% overlap in cases of NAFLD and MASLD. CONCLUSIONS: Our findings highlight the utility of the new consensus nomenclature to address deficiencies present with the old nomenclature, and identify areas that require research to further refine classifications of SLD.

Neighborhood-level Social Determinants of Health and Waitlist Mortality for Liver Transplantation: The Liver Outcomes and Equity Index.

BACKGROUND AND AIMS: To examine neighborhood-level disparities in waitlist mortality for adult liver transplantation (LT), we developed novel area-based social determinants of health (SDOH) index using a national transplant database. METHODS: ZIP Codes of individuals listed for or received LT in the Scientific Registry of Transplant Recipients database between June 18, 2013, and May 18, 2019, were linked to 36 American Community Survey (ACS) variables across 5 SDOH domains for index development. A step-wise principal component analysis was used to construct the Liver Outcomes and Equity (LOEq) index. We then examined the association between LOEq quintiles (Q1 = worst and Q5 = best neighborhood SDOH) and waitlist mortality with competing risk regression among listed adults in the study period and acuity circle (AC) era. RESULTS: The final LOEq index consisted of 13 ACS variables. Of 59 298 adults waitlisted for LT, 30% resided in LOEq Q5 compared with only 14% in Q1. Q1 neighborhoods with worse SDOH were disproportionately concentrated in transplant regions with low median Model for End-Stage Liver Disease at transplant (MMAT) and shorter wait times. Five years cumulative incidence of waitlist mortality was 33% in Q1 in high MMAT regions versus 16% in Q5 in low MMAT regions. Despite this allocation advantage, LOEq Q1-Q4 were independently associated with elevated risk of waitlist mortality compared with Q5, with highest increased hazard of waitlist deaths of 19% (95% CI, 11%-26%) in Q1. This disparity persisted in the AC era, with 24% (95% CI, 10%-40%) increased hazard of waitlist deaths for Q1 versus Q5. CONCLUSIONS: Neighborhood SDOH independently predicts waitlist mortality in adult LT.

Clinical Significance of Ascitic Fluid Polymorphonuclear Leukocyte Percentage in Patients With Cirrhosis Without Spontaneous Bacterial Peritonitis.

INTRODUCTION: Absolute polymorphonuclear leukocyte (PMN) count (PMN-C) ≥250 cells/mm 3 in ascites is the diagnostic hallmark of spontaneous bacterial peritonitis (SBP) and is associated with high morbidity and mortality. However, the clinical significance of ascitic PMN percentage (PMN-%) and PMN-C in the absence of SBP as additional biomarkers for mortality and future incidence of SBP has not been determined. METHODS: This retrospective cohort included adults with cirrhosis undergoing first-recorded paracentesis with initial PMN-C < 250 cells/mm 3 at 2 tertiary medical centers between 2015 and 2020. Patients with prior SBP were excluded. Outcomes were death and SBP development. Cox regression estimated hazard ratios (HRs) for risk of death and SBP development and Akaike information criterion to compare model fit. RESULTS: Three hundred eighty-four adults (73% male, median age 58 years, 67% with alcohol-associated cirrhosis, median PMN-C 14 cells/mm 3 [interquartile range 5-34], and median PMN-% 10% [interquartile range 4-20]) were included in this study. Univariate risk of death increased 10% per 25-unit increase in PMN-C (95% confidence interval 1.01-1.21, P = 0.03) and 19% per 10-unit increase in PMN-% (95% confidence interval 1.06-1.33, P = 0.003) with PMN-% demonstrating better model fit in assessing mortality risk (Akaike information criterion: 1,044 vs 1,048, respectively). In models adjusted for age, chronic hepatitis C virus infection, and Model for End-Stage Liver Disease-Sodium, PMN-% was associated with risk of death (PMN-% 10%-29%, HR 1.17, P = 0.50; PMN-% ≥ 30% group, HR 1.94, P = 0.03; vs PMN-% < 10%) and SBP development (PMN-% 10%-29%, HR 1.68, P = 0.07; PMN-% ≥ 30%, HR 3.48, P < 0.001; vs PMN-% < 10%). DISCUSSION: Our results suggest PMN-% at first paracentesis represents a better biomarker compared with PMN-C for assessing risk of death and future SBP development in patients with PMN-C < 250 cells/mm 3 .

Association of nativity with survival among adults with hepatocellular carcinoma.

INTRODUCTION: Immigrants comprise a considerable proportion of those diagnosed with hepatocellular carcinoma (HCC) in the United States. Nativity or birthplace affects incidence and risk factors for HCC, but little is known about its influence on survival after diagnosis. METHODS: We identified 51 533 adults with HCC with available birthplace in the California Cancer Registry between 1988 and 2017. HCC cases were categorized as foreign born or US born and stratified by mutually exclusive race and ethnicity groups. Primary outcome was all-cause mortality. Race and ethnicity-specific Cox regression propensity score-weighted models evaluated the relationship between nativity and death as well as region of birth among foreign-born patients. RESULTS: A total of 40% of all HCC cases were foreign born, and 92.2%, 45.2%, 9.1%, and 5.8% of Asian/Pacific Islander (API), Hispanic, White, and Black patients were foreign born, respectively. Five-year survival rates were higher in foreign-born patients compared with US-born patients: 12.9% vs 9.6% for White patients, 11.7% vs 9.8% for Hispanic patients, 12.8% vs 8.1% for Black patients, and 16.4% vs 12.4% for API patients. Nativity was associated with survival, with better survival in foreign-born patients: White patients: hazard ratio (HR) = 0.86 (95% confidence interval [CI] = 0.81 to 0.90), Hispanic patients: HR = 0.90 (95% CI = 0.86 to 0.93), Black patients: HR = 0.89 (95% CI = 0.76 to 1.05), and API patients: HR = 0.94 (95% CI = 0.88 to 1.00). Among foreign-born patients, lower mortality was observed in those from Central and South America compared with Mexico for Hispanic patients, East Asia compared with Southeast Asia for API patients, and East Europe and Greater Middle East compared with West/South/North Europe for White patients. CONCLUSION: Foreign-born patients with HCC have better survival than US-born patients. Further investigation into the mechanisms of this survival disparity by nativity is needed.

Effect of dietary branched chain amino acids on liver related mortality: Results from a large cohort of North American patients with advanced HCV infection.

Branched chain amino acids (BCAA) supplementation may reduce the incidence of liver failure and hepatocellular carcinoma in patients with cirrhosis. We aimed to determine whether long-term dietary intake of BCAA is associated with liver-related mortality in a well-characterized cohort of North American patients with advanced fibrosis or compensated cirrhosis. We performed a retrospective cohort study using extended follow-up data from the Hepatitis C Antiviral Long-term Treatment against Cirrhosis (HALT-C) Trial. The analysis included 656 patients who completed two Food Frequency Questionnaires. The primary exposure was BCAA intake measured in grams (g) per 1000 kilocalories (kcal) of energy intake (range 3.0-34.8 g/1000 kcal). During a median follow-up of 5.0 years, the incidence of liver-related death or transplantation was not significantly different among the four quartiles of BCAA intake before and after adjustment of confounders (AHR 1.02, 95% CI 0.81-1.27, P-value for trend = 0.89). There remains no association when BCAA was modeled as a ratio of BCAA to total protein intake or as absolute BCAA intake. Finally, BCAA intake was not associated with the risk of hepatocellular carcinoma, encephalopathy or clinical hepatic decompensation. We concluded that dietary BCAA intake was not associated with liver-related outcomes in HCV-infected patients with advanced fibrosis or compensated cirrhosis. The precise effect of BCAA in patients with liver disease warrants further study.

Patient survey augments detection of harmful alcohol relapse after liver transplant for alcohol-associated cirrhosis.

BACKGROUND: Predicting the risk of alcohol relapse after a liver transplant for alcohol-associated liver disease is critical to guide candidate selection and optimize alcohol use disorder management. We aimed to use patient survey to augment the detection of alcohol relapse and its risk factors and to understand patient perceptions of the importance of alcohol abstinence. METHODS: In this retrospective cohort study, we used a telephone survey and chart review to assess the incidence of post-transplant harmful alcohol relapse, risk factors, and long-term outcomes for patients transplanted for alcohol-associated cirrhosis at our center from 2002 to 2016. RESULTS: Over the median follow-up of 5.9 years, 20.4% relapsed, with 9.3% harmful relapse after median of 4.0 years. The survey response rate was 44.0% (n=110). Of survey responders, 44.3% did not recall discussing alcohol in post-transplant clinics, and 17.6% of relapses were identified by the survey alone. In univariate analysis, shorter pretransplant sobriety (OR: 0.96 per month, p=0.02) and history of pretransplant relapse (OR: 2.99, p=0.02) were associated with post-transplant harmful relapse. After adjusting for these factors, High-risk Alcoholism Relapse score ≥4 predicted harmful relapse (OR: 3.43, p=0.049). A total of 27.3% of patients with both pretransplant relapse and High-risk Alcoholism Relapse score ≥4 relapsed to harmful use compared with 5.2% of those with 1 or neither risk factor (p < 0.001). Harmful relapse was associated with increased graft loss (30.4% vs. 17.4%) and inferior 10-year post-liver transplant survival (61.5% vs. 80.7%). CONCLUSIONS: Incorporating patient survey data allowed the detection of relapses otherwise unreported to clinicians, highlighting the need for novel strategies to detect relapse. Utilizing this augmented data, we identified pretransplant sobriety length, pretransplant relapse, and High-risk Alcoholism Relapse score ≥4 as risk factors that should be evaluated pretransplant to guide candidate selection and peritransplant alcohol use disorder management.

Population-attributable risk of modifiable lifestyle factors to hepatocellular carcinoma: The multi-ethnic cohort.

BACKGROUND AND AIMS: Lifestyle factors are well associated with risk of hepatocellular carcinoma (HCC). However, the impact of reducing adverse lifestyle behaviours on population-level burden of HCC is uncertain. METHODS: We conducted prospective analysis of the population-based multi-ethnic cohort (MEC) with linkage to cancer registries. The association of lifestyle factors (smoking, alcohol, diet quality assessed by alternate Mediterranean diet score, coffee drinking, physical activity and body mass index) with HCC incidence was examined using Cox regression. Population-attributable risk (PAR, %) for the overall, lean and overweight/obese populations was determined. RESULTS: A total of 753 incident cases of HCC were identified in 181,346 participants over median follow-up of 23.1 years. Lifestyle factors associated with elevated HCC risk included former/current smoking, heavy alcohol use, poor diet quality, lower coffee intake and obesity, but not physical activity. The lifestyle factor with highest PAR was lower coffee intake (21.3%; 95% CI: 8.9%-33.0%), followed by current smoking (15.1%; 11.1%-19.0%), obesity (14.5%; 9.2%-19.8%), heavy alcohol use (7.1%; 3.5%-10.6%) and lower diet quality (4.1%; 0.1%-8.1%). The combined PAR of all high-risk lifestyle factors was 51.9% (95% CI: 30.1%-68.6%). A higher combined PAR was observed among lean (65.2%, 26.8%-85.7%) compared to overweight/obese (37.4%, 11.7%-58.3%) participants. Adjusting for viral hepatitis status in a linked MEC-Medicare dataset resulted in similar PAR results. CONCLUSIONS: Modifying lifestyle factors, particularly coffee intake, may have a substantial impact on HCC burden in diverse populations, with greater impact among lean adults. Diet and lifestyle counselling should be incorporated into HCC prevention strategies.

Impact of Bacteria Types on the Clinical Outcomes of Spontaneous Bacterial Peritonitis.

BACKGROUND AND AIMS: Cirrhotic patients presenting with spontaneous bacterial peritonitis (SBP) have elevated risk of short-term mortality. While high Model for End-Stage Liver Disease-Sodium score (MELD-Na) and ascites culture yielding multi-drug resistance (MDR) bacteria are well established risk factors for further aggravating mortality, the impact of individual, causative microorganisms and their respective pathogenesis have not been previously investigated. METHODS: This is a retrospective study of 267 cirrhotic patients at two tertiary care hospitals undergoing paracentesis from January 2015 to January 2021 who presented with ascitic PMN count > 250 cells/mm3. The primary outcome was SBP progression defined as death or liver transplantation within 1-month of paracentesis stratified by microorganism type. RESULTS: Of 267 patients with SBP, the ascitic culture yielded causative microorganism in 88 cases [median age 57 years (IQR 52-64)]; 68% male; median MELD-Na 29 (IQR 23-35). The microbes isolated were E. coli (33%), Streptococcus (15%), Klebsiella (13%), Enterococcus (13%), Staphylococcus (9%) and others (18%); 41% were MDR. Cumulative incidence of SBP progression within 1-month was 91% (95% CI 67-100) for Klebsiella, 59% (95% CI 42-76) for E. coli, and 16% (95% CI 4-51) for Streptococcus. After adjusting for MELD-Na and MDR, risk of SBP progression remained elevated for Klebsiella (HR 2.07; 95% CI 0.98-4.24; p-value = 0.06) and decreased for Streptococcus (HR 0.28; 95% CI 0.06-1.21; p-value = 0.09) compared to all other bacteria. CONCLUSION: Our study found Klebsiella-associated SBP had worse clinical outcomes while Streptococcus-associated SBP had the most favorable outcomes after accounting for MDR and MELD-Na. Thus, identification of the causative microorganism is crucial not only for optimizing the treatment but for prognostication.

Healthcare affordability and effects on mortality among adults with liver disease from 2004 to 2018 in the United States.

BACKGROUND & AIMS: Liver disease is associated with substantial morbidity and mortality, likely incurring financial distress (i.e. healthcare affordability and accessibility issues), although long-term national-level data are limited. METHODS: Using the National Health Interview Survey from 2004 to 2018, we categorised adults based on report of liver disease and other chronic conditions linked to mortality data from the National Death Index. We estimated age-adjusted proportions of adults reporting healthcare affordability and accessibility issues. Multivariable logistic regression and Cox regression were used to assess the association of liver disease with financial distress and financial distress with all-cause mortality, respectively. RESULTS: Among adults with liver disease (n = 19,407) vs. those without liver disease (n = 996,352), those with cancer history (n = 37,225), those with emphysema (n = 7,937), and those with coronary artery disease (n = 21,510), the age-adjusted proportion reporting healthcare affordability issues for medical services was 29.9% (95% CI 29.7-30.1%) vs. 18.1% (95% CI 18.0-18.3%), 26.5% (95% CI 26.3-26.7%), 42.2% (95% CI 42.1-42.4%), and 31.6% (31.5-31.8%), respectively, and for medications: 15.5% (95% CI 15.4-15.6%) vs. 8.2% (95% CI 8.1-8.3%), 14.8% (95% CI 14.7-14.9%), 26.1% (95% CI 26.0-26.2%), and 20.6% (95% CI 20.5-20.7%), respectively. In multivariable analysis, liver disease (vs. without liver disease, vs. cancer history, vs. emphysema, and vs. coronary artery disease) was associated with inability to afford medical services (adjusted odds ratio [aOR] 1.84, 95% CI 1.77-1.92; aOR 1.32, 95% CI 1.25-1.40; aOR 0.91, 95% CI 0.84-0.98; and aOR 1.11, 95% CI 1.04-1.19, respectively) and medications (aOR 1.92, 95% CI 1.82-2.03; aOR 1.24, 95% CI 1.14-1.33; aOR 0.81, 95% CI 0.74-0.90; and aOR 0.94, 95% CI 0.86-1.02, respectively), delays in medical care (aOR 1.77, 95% CI 1.69-1.87; aOR 1.14, 95% CI 1.06-1.22; aOR 0.88, 95% CI 0.79-0.97; and aOR 1.05, 95% CI 0.97-1.14, respectively), and not receiving the needed medical care (aOR 1.86, 95% CI 1.76-1.96; aOR 1.16, 95% CI 1.07-1.26; aOR 0.89, 95% CI 0.80-0.99; aOR 1.06, 95% CI 0.96-1.16, respectively). In multivariable analysis, among adults with liver disease, financial distress (vs. without financial distress) was associated with increased all-cause mortality (aHR 1.24, 95% CI 1.01-1.53). CONCLUSIONS: Adults with liver disease face greater financial distress than adults without liver disease and adults with cancer history. Financial distress is associated with increased risk of all-cause mortality among adults with liver disease. Interventions to improve healthcare affordability should be prioritised in this population. IMPACT AND IMPLICATIONS: Adults with liver disease use many medical services, but long-term national studies regarding the financial repercussions and the effects on mortality for such patients are lacking. This study shows that adults with liver disease are more likely to face issues affording medical services and prescription medication, experience delays in medical care, and needing but not obtaining medical care owing to cost, compared with adults without liver disease, adults with cancer history, are equally likely as adults with coronary artery disease, and less likely than adults with emphysema-patients with liver disease who face these issues are at increased risk of death. This study provides the impetus for medical providers and policymakers to prioritise interventions to improve healthcare affordability for adults with liver disease.

Quantifying days at home in patients with cirrhosis: A national cohort study.

BACKGROUND AND AIMS: Days at home (DAH) is a patient-centric metric developed by the Medicare Payment Advisory Commission, capturing annual health care use, including and beyond hospitalizations and mortality. We quantified DAH and assessed factors associated with DAH differences among patients with cirrhosis. APPROACH AND RESULTS: Using a national claims database (Optum) between 2014 and 2018, we calculated DAH (365 minus mortality, inpatient, observation, postacute, and emergency department days). Among 20,776,597 patients, 63,477 had cirrhosis (median age, 66, 52% males, and 63% non-Hispanic White). Age-adjusted mean DAH for cirrhosis was 335.1 days (95% CI: 335.0 to 335.2) vs 360.1 (95% CI: 360.1 to 360.1) without cirrhosis. In mixed-effects linear regression, adjusted for demographic and clinical characteristics, patients with decompensated cirrhosis spent 15.2 days (95% CI: 14.4 to 15.8) in postacute, emergency, and observation settings and 13.8 days (95% CI: 13.5 to 14.0) hospitalized. Hepatic encephalopathy (-29.2 d, 95% CI: -30.4 to -28.0), ascites (-34.6 d, 95% CI: -35.3 to -33.9), and combined ascites and hepatic encephalopathy (-63.8 d, 95% CI: -65.0 to -62.6) were associated with decreased DAH. Variceal bleeding was not associated with a change in DAH (-0.2 d, 95% CI: -1.6 to +1.1). Among hospitalized patients, during the 365 days after index hospitalization, patients with cirrhosis had fewer age-adjusted DAH (272.8 d, 95% CI: 271.5 to 274.1) than congestive heart failure (288.0 d, 95% CI: 287.7 to 288.3) and chronic obstructive pulmonary disease (296.6 d, 95% CI: 296.3 to 297.0). CONCLUSIONS: In this national study, we found that patients with cirrhosis spend as many, if not more, cumulative days receiving postacute, emergency, and observational care, as hospitalized care. Ultimately, up to 2 months of DAH are lost annually with the onset of liver decompensation. DAH may be a useful metric for patients and health systems alike.

Multicentered study of patient outcomes after declined for early liver transplantation in severe alcohol-associated hepatitis.

BACKGROUND: Early liver transplantation for alcohol-associated hepatitis is controversial in part because patients may recover, and obviate the need for liver transplantation. METHODS: In this retrospective study among 5 ACCELERATE-AH sites, we randomly sampled patients evaluated and then declined for liver transplantation for alcohol-associated hepatitis. All had Model of End-Stage Liver Disease (MELD) >20 and <6 months of abstinence. Recompensation was defined as MELD <15 without variceal bleeding, ascites, or overt HE requiring treatment. Multilevel mixed effects linear regression was used to calculate probabilities of recompensation; multivariable Cox regression was used for mortality analyses. RESULTS: Among 145 patients [61% men; median abstinence time and MELD-Na was 33 days (interquartile range: 13-70) and 31 (interquartile range: 26-36), respectively], 56% were declined for psychosocial reasons. Probability of 30-day, 90-day, 6-month, and 1-year survival were 76% (95% CI, 68%-82%), 59% (95% CI, 50%-66%), 49% (95% CI, 40%-57%), and 46% (95% CI, 37%-55%), respectively. Probability of 1-year recompensation was low at 10.0% (95% CI, 4.5%-15.4%). Among patients declined because of clinical improvement, 1-year probability of recompensation was 28.0% (95% CI, 5.7%-50.3%). Among survivors, median MELD-Na at 30 days, 90 days, and 1-year were 29 (interquartile range: 22-38), 19 (interquartile range : 14-29), and 11 (interquartile range : 7-17). Increased MELD-Na (adjusted HR: 1.13, p <0.001) and age (adjusted HR: 1.03, p <0.001) were associated with early (≤90 d) death, and only history of failed alcohol rehabilitation (adjusted HR: 1.76, p =0.02) was associated with late death. CONCLUSIONS: Liver recompensation is infrequent among severe alcohol-associated hepatitis patients declined for liver transplantation. Higher MELD-Na and age were associated with short-term mortality, whereas only history of failed alcohol rehabilitation was associated with long-term mortality. The distinction between survival and liver recompensation merits further attention.

Strategies to increase primary care provider capacity for hepatitis C care: The California ECHO-PLUS study.

Expanding capacity to screen and treat those infected with the hepatitis C virus (HCV) is an essential element of the global elimination strategy. We evaluated the hub-and-spoke Project ECHO training versus telementoring models to educate, train and support HCV care by primary care providers in 13 targeted counties in northern California. A novel provider engagement strategy was used. Provider engagement and retention, time to readiness to treat HCV, and knowledge and confidence were the outcomes of interest. 94 participants from 60 unique clinics in the target counties participated in the ECHO-PLUS programme; 39.4% were physicians, 48.9% were advanced practice providers, and 11.7% were nurses. The median (range) participation time was 5 (1-49) hours. Confidence scores (minimum score = 13 and maximum score = 65) increased by a mean of 14.0 (SD:8.2) and 11.4 (SD:12.0) points for the hub-and-spoke and telementoring programmes, respectively (p = .53), with the largest changes in confidence seen in treating patients per guidelines, managing side effects and in serving as a consultant for HCV in their clinic. Among 24 participants with data on time to treatment, median time from beginner to experienced was 8 h (IQR:6-12) for hub-and-spoke and 2 h (IQR:1-2.4) for the telementoring programme (p = .01). A 'boots on the ground' approach to recruiting HCV champions was effective within rural communities. Both tele-ECHO hub-and-spoke and telementoring approaches to training primary care providers yielded increase in knowledge and confidence in HCV care and amplified the number of patients who were screened and treated. Telementoring accelerated the timeline of novice providers being 'ready to treat'.