Citations

SummaryBackground and AimsMachine learning (ML) can identify the hidden patterns without hypothesis in heterogeneous diseases like acute‐on‐chronic live failure (ACLF). We employed ML to describe and predict yet unknown clusters in ACLF.MethodsClinical data of 1568 patients with ACLF from a tertiary care centre (2015–2023) were subjected to distance‐, density‐ and model‐based clustering algorithms. Final model was selected on best cluster separation, viz. Silhouette width and Dunn's index (for distance‐ or density‐based algorithms) and minimum BIC (for model‐based algorithms). Cluster assignments, patient trajectories and survival were analysed through inferential statistics. Supervised ML models were trained in 70% data that predicted clusters in remaining 30% data followed by an temporal validation.ResultsThe cohort was male‐predominant (87%), aged 44.3 years, with alcohol‐associated hepatitis (62.9%) and survival of 50.5%. Due to poor performance of distance‐ and density‐based algorithms and better explainability, the latent class model (LCM) was selected for exploration. LCM revealed four clusters with distinct trajectories, reversibility and survival (independent of MELD, CLIF‐C ACLF and AARC scores). Cluster1 had patients with none/one organ failure and highest reversibility. Cluster2 had females with viral hepatitis and two organ failures. More‐than‐one acute precipitant, severity, infections, organ failures and irreversibility escalated from clusters 1 to 4. Circulatory and renal failures critically influenced cluster assignments. Incorporating clusters to CLIF‐C ACLF, infection and ACLF definition improved the discriminative accuracy of CLIF‐C‐ACLF by 11%. Extreme gradient boost and decision trees could predict clusters with AUCs of 0.989 (0.979–0.995) and 0.875 (0.865–0.890). MELD, CLIF‐C‐OF, haemoglobin, lactate, CLIF‐C‐ACLF and ALT were critical variables for cluster prediction. Clusters with distinct survival were documented in a temporal validation cohort.ConclusionsML for the first time could identify clusters with distinct phenotypes, trajectories and outcomes in ACLF. Stratification into clusters can address heterogeneity, guide prognosis, recruitment in trials, resource allocation and liver transplant discussions in ACLF.

Abstract Background Outcomes in alcohol‐associated liver disease (ALD) are influenced by several race and ethnic factors, yet its natural history across the continuum of patients in different stages of the disease is unknown. Methods We conducted a retrospective cohort study of U.S. adults from 2011 to 2018, using three nationally representative databases to examine potential disparities in relevant outcomes among racial and ethnic groups. Our analysis included logistic and linear regressions, along with competing risk analysis. Results Black individuals had the highest daily alcohol consumption (12.6 g/day) while Hispanic participants had the largest prevalence of heavy episodic drinking (33.5%). In a multivariable‐adjusted model, Hispanic and Asian participants were independently associated with a higher ALD prevalence compared to Non‐Hispanic White interviewees (OR: 1.4, 95% CI: 1.1–1.8 and OR: 1.5 95% CI:1.1–2.0, respectively), while Blacks participants had a lower ALD prevalence (OR: .7 95% CI: .6–.9), and a lower risk of mortality during hospitalization due to ALD (OR: .83 95% CI: .73–.94). Finally, a multivariate competing‐risk analysis showed that Hispanic ethnicity had a decreased probability of liver transplantation if waitlisted for ALD (SHR: .7, 95% CI: .6–.8) along with female Asian population (HR: .40, 95% CI: .26–.62). Conclusions After accounting for key social and biological health determinants, the Hispanic population showed an increased risk of ALD prevalence, even with lower alcohol consumption. Additionally, Hispanic and Asian female patients had reduced access to liver transplantation compared to other enlisted patients.

Background:Alcohol-associated hepatitis (AH) is associated with significant mortality. Model for End-Stage Liver Disease (MELD) score is used to predict short-term mortality and aid in treatment decisions. MELD is frequently updated in the course of AH. However, once the most updated MELD is known, it is uncertain if previous ones still have prognostic value, which might be relevant for transplant allocation and trial design. We aimed to investigate the predictive performance of updated MELDs in a prospectively collected cohort of patients with AH by the InTeam consortium.Methods:Three hundred seven patients (with 859 MELD values within 60 d of admission) fulfilled the inclusion criteria. The main endpoint was time to death or transplant up to 90 days. We used a joint model approach to assess the predictive value of updated MELDs.Results:Updated MELD measurements had a strong prognostic value for death/transplant (HR: 1.20, 95% CI: 1.14–1.27) (p< 0.0001). Previous MELD values did not add predictive value to the most current MELD. We also showed that MELD at day 28 (MELD28) had a significant predictive value for subsequent mortality/transplant in a landmark analysis (HR: 1.18, 95% CI: 1.12–1.23). We show that the use of an ordinal scale including death, transplant, and MELD28 as a trial outcome could substantially reduce the sample size required to demonstrate short-term benefit of an intervention.Conclusion:We show that updated MELDs during the trajectory of AH predict subsequent mortality or the need for transplant. MELD28 inclusion in an ordinal outcome (together with death or transplant) could increase the efficiency of randomized controlled trials.

Abstract Background & Aims: Sarcopenia is common in patients with liver disease and both sarcopenia and short telomeres are associated with mortality, however their relationship in patients with liver disease remains unknown. Methods: A cohort of 16,072 adults from the National Health and Nutrition Examination Survey from 1999-2006 was analyzed. Liver disease was defined by aminotransferases and classified into etiology-based categories. Sarcopenia was defined by dual-energy x-ray absorptiometry. All analyses were conducted separately on each multiple imputation data set and combined via Rubin’s rules. P-values for group comparisons were calculated by testing logistic regression parameter estimates. Cox proportional hazards regression was used for mortality analysis with mortality data available until 2015. Results:Sarcopenia was present in 9.5% of patients with liver disease. Age, race, income, education, physical inactivity, and certain medical comorbidities were associated with sarcopenia. Patients with liver disease and sarcopenia had significantly shorter telomeres than patients with liver disease without sarcopenia when unadjusted for age. The interaction between telomere length and sarcopenia was significantly associated with all-cause mortality. Conclusions: The implications of telomere length on all-cause mortality in patients with liver disease varied by age and sarcopenia status. Shorter telomeres appear to be more highly associated with increased mortality in older patients without sarcopenia.

Whether centralised or decentralised, healthcare access models provide varying degrees of benefit to customers and to the organisations they serve. Mayo Clinic Health System in Southwest Minnesota, United States, is a community-based practice of 23 clinics and 6 hospitals that has undergone an evolution of its access model. In 2010, a centralised access model was adopted, including the physical location of schedulers, which was expected to offer greater efficiency, flexibility and staff savings. However, in 2016, the region noted concerning trends of low patient and provider satisfaction with scheduling, decreasing patient access and less-than-optimal patient throughput, all subsequently contributing to declining financial performance. Early concerning trends led to further investigation of practice areas and geographical sites that had notable appointment capacity yet low patient access and throughput. Provider productivity rates, calendar management and patient access varied substantially. Performance in scheduling operations was below par because of limited connection with the practice. The combination of practice variation and a centralised access model was causing an increase in scheduling errors, consistent rework/rescheduling, less-than-optimal patient access, low provider calendar fill rates and high levels of dissatisfaction for patients and providers. Our group was tasked with analysing the problems and developing a new scheduling model. This paper describes the resulting model: a hybrid of centralised and decentralised models, which promised the benefits of both and involved transformation in clinical practice operations and access management. Standardisation for provider scheduling and template management was enhanced. Pods of scheduling personnel were embedded in practice areas. The proximity of the pods to physician, nursing and other clinical staff allowed for increased collaboration and communication. The hybrid model improved access metrics for average speed to answer, abandoned call rates, patient access and throughput, financial performance, and patient and provider satisfaction.

Strategic projects in large healthcare organisations are often highly complex and dynamic, and they require a team of skilled business professionals to execute them. Examples of difficult, costly and extended project executions have been shared in the literature. Successful project execution involves many critical facets. One facet that continues to be of marked interest is effective resourcing of strategic projects. This paper focuses on describing a staffing approach used at Mayo Clinic (Rochester, Minnesota) to facilitate successful, agile and sustainable execution, with desired outcomes. In 2016, Management Engineering & Consulting (ME&C) began exploring a collaborative staffing model that partnered management engineers with project managers (PMs) to drive strategic, multisite projects. This ongoing collaboration model was evaluated from multiple perspectives, including project outcomes, timeliness, client and staff satisfaction and opportunities for innovation. Data and feedback on the model were collected with historical reports, surveys and focus groups with stakeholders, leaders, PMs and management engineers. The feedback and data indicated that the collaborative staffing strategy and model have further optimised service delivery, enabled professional staff to work at the highest level of their competency and dynamically blended knowledge, expertise and skills to achieve better and more sustainable overall project outcomes.