AI-based histologic scoring enables automated and reproducible assessment of enrollment criteria and endpoints in NASH clinical trials.

Clinical trials in nonalcoholic steatohepatitis (NASH) require histologic scoring for assessment of inclusion criteria and endpoints. However, guidelines for scoring key features have led to variability in interpretation, impacting clinical trial outcomes. We developed an artificial intelligence (AI)-based measurement (AIM) tool for scoring NASH histology (AIM-NASH). AIM-NASH predictions for NASH Clinical Research Network (CRN) grades of necroinflammation and stages of fibrosis aligned with expert consensus scores and were reproducible. Continuous scores produced by AIM-NASH for key histological features of NASH correlated with mean pathologist scores and with noninvasive biomarkers and strongly predicted patient outcomes. In a retrospective analysis of the ATLAS trial, previously unmet pathological endpoints were met when scored by the AIM-NASH algorithm alone. Overall, these results suggest that AIM-NASH may assist pathologists in histologic review of NASH clinical trials, reducing inter-rater variability on trial outcomes and offering a more sensitive and reproducible measure of patient therapeutic response.

Correction: The effects of patient cost sharing on inpatient utilization, cost, and outcome.

[This corrects the article DOI: 10.1371/journal.pone.0187096.].

The effects of patient cost sharing on inpatient utilization, cost, and outcome.

BACKGROUND: Health insurance and provider payment reforms all over the world beg a key empirical question: what are the potential impacts of patient cost-sharing on health care utilization, cost and outcomes? The unique health insurance system and rich electronic medical record (EMR) data in China provides us a unique opportunity to study this topic. METHODS: Four years (2010 to 2014) of EMR data from one medical center in China were utilized, including 10,858 adult patients with liver diseases. We measured patient cost-sharing using actual reimbursement ratio (RR) which is allowed us to better capture financial incentive than using type of health insurance. A rigorous risk adjustment method was employed with both comorbidities and disease severity measures acting as risk adjustors. Associations between RR and health use, costs and outcome were analyzed by multivariate analyses. RESULTS: After risk adjustment, patients with more generous health insurance coverage (higher RR) were found to have longer hospital stay, higher total cost, higher medication cost, and higher ratio of medication to total cost, as well as higher number and likelihood that specific procedures were performed. CONCLUSION: Our study implied that patient cost-sharing affects health care services use and cost. This reflects how patients and physicians respond to financial incentives in the current healthcare system in China, and the responses could be a joint effect of both demand and supply side moral hazard. In order to contain cost and improve efficiency in the system, reforming provide payment and insurance scheme is urgently needed.

A quantitative metabolomics profiling approach for the noninvasive assessment of liver histology in patients with chronic hepatitis C.

BACKGROUND: High-throughput technologies have the potential to identify non-invasive biomarkers of liver pathology and improve our understanding of basic mechanisms of liver injury and repair. A metabolite profiling approach was employed to determine associations between alterations in serum metabolites and liver histology in patients with chronic hepatitis C virus (HCV) infection. METHODS: Sera from 45 non-diabetic patients with chronic HCV were quantitatively analyzed using (1)H-NMR spectroscopy. A metabolite profile of advanced fibrosis (METAVIR F3-4) was established using orthogonal partial least squares discriminant analysis modeling and validated using seven-fold cross-validation and permutation testing. Bioprofiles of moderate to severe steatosis (≥33 %) and necroinflammation (METAVIR A2-3) were also derived. The classification accuracy of these profiles was determined using areas under the receiver operator curves (AUROCSs) measuring against liver biopsy as the gold standard. RESULTS: In total 63 spectral features were profiled, of which a highly significant subset of 21 metabolites were associated with advanced fibrosis (variable importance score >1 in multivariate modeling; R(2) = 0.673 and Q(2) = 0.285). For the identification of F3-4 fibrosis, the metabolite bioprofile had an AUROC of 0.86 (95 % CI 0.74-0.97). The AUROCs for the bioprofiles for moderate to severe steatosis were 0.87 (95 % CI 0.76-0.97) and for grade A2-3 inflammation were 0.73 (0.57-0.89). CONCLUSION: This proof-of-principle study demonstrates the utility of a metabolomics profiling approach to non-invasively identify biomarkers of liver fibrosis, steatosis and inflammation in patients with chronic HCV. Future cohorts are necessary to validate these findings.

Validation of the Model of End-Stage Liver Disease for Liver Transplant Allocation in Alberta: Implications for Future Directions in Canada.

Background. Since 2002, the Model of End-Stage Liver Disease (MELD) has been used for allocation of liver transplants (LT) in the USA. In Canada, livers were allocated by the CanWAIT algorithm. The aim of this study was to compare the abilities of MELD, Child-Pugh (CP), and CanWAIT status to predict 3-month and 1-year mortality before LT in Canadian patients and to describe the use of MELD in Canada. Methods. Validation of MELD was performed in 320 patients listed for LT in Alberta (1998-2002). In October 2014, a survey of MELD use by Canadian LT centers was conducted. Results. Within 1 year of listing, 47 patients were removed from the waiting list (29 deaths, 18 too ill for LT). Using logistic regression, the MELD and CP were better than the CanWAIT at predicting 3-month (AUROC: 0.79, 0.78, and 0.59; p = 0.0002) and 1-year waitlist mortality (AUROC: 0.70, 0.70, and 0.55; p = 0.0023). Beginning in 2004, MELD began to be adopted by Canadian LT programs but its use was not standardized. Conclusions. Compared with the CanWAIT system, the MELD score was significantly better at predicting LT waitlist mortality. MELD-sodium (MELD-Na) has now been adopted for LT allocation in Canada.

Ulcerative colitis-associated hospitalization costs: a population-based study.

BACKGROUND: Hospitalization costs for ulcerative colitis (UC) following the introduction of infliximab have not been evaluated. OBJECTIVE: To study predictors of costs for UC patients who were hospitalized for a flare or colectomy. METHODS: Population-based surveillance identified adults (≥18 years of age) admitted to hospital for UC flare or colectomy between 2001 and 2009 in the Calgary Health Zone (Alberta). Medical charts were reviewed and patients stratified into three admission types: responsive to inpatient medical therapy (n=307); emergent colectomy (n=227); and elective colectomy (n=208). The annual median cost with interquartile range (IQR) was calculated. Linear regression determined the effect of admission type on hospital charges after adjusting for age, sex, smoking, comorbidities, disease extent, medication use (eg, infliximab) and year. The adjusted cost increase was presented as the percent increase with 95% CIs. Joinpoint analysis assessed for an inflection point in hospital cost after the introduction of infliximab. RESULTS: Median hospitalization cost for UC flare, emergent colectomy and elective colectomy, respectively, were: $5,499 (IQR $3,374 to $8,904), $23,698 (IQR $17,981 to $32,385) and $14,316 (IQR $11,932 to $18,331). Adjusted hospitalization costs increased approximately 6.0% annually (95% CI 4.5% to 7.5%). Adjusted costs were higher for patients who underwent an elective colectomy (percent increase cost 179.8% [95% CI 151.6% to 211.1%]) or an emergent colectomy (percent increase cost 211.1% [95% CI 183.2% to 241.6%]) than medically responsive patients. Infliximab in hospital was an independent predictor of increased costs (percent increase cost 69.5% [95% CI 49.2% to 92.5%]). No inflection points were identified. CONCLUSION: Hospitalization costs for UC increased due to colectomy and infliximab.

Elastography Assessment of Liver Fibrosis: Society of Radiologists in Ultrasound Consensus Conference Statement.

The Society of Radiologists in Ultrasound convened a panel of specialists from radiology, hepatology, pathology, and basic science and physics to arrive at a consensus regarding the use of elastography in the assessment of liver fibrosis in chronic liver disease. The panel met in Denver, Colo, on October 21-22, 2014, and drafted this consensus statement. The recommendations in this statement are based on analysis of current literature and common practice strategies and are thought to represent a reasonable approach to the noninvasive assessment of diffuse liver fibrosis.

Risk factors for mortality in patients with alcoholic hepatitis and assessment of prognostic models: A population-based study.

BACKGROUND: Severe alcoholic hepatitis (AH) is associated with a substantial risk for short-term mortality. OBJECTIVES: To identify prognostic factors and validate well-known prognostic models in a Canadian population of patients hospitalized for AH. METHODS: In the present retrospective study, patients hospitalized for AH in Calgary, Alberta, between January 2008 and August 2012 were included. Stepwise logistic regression models identified independent risk factors for 90-day mortality, and the discrimination of prognostic models (Model for End-stage Liver Disease [MELD] and Maddrey discriminant function [DF]) were examined using areas under the ROC curves. RESULTS: A total of 122 patients with AH were hospitalized during the study period; the median age was 49 years (interquartile range [IQR] 42 to 55 years) and 60% were men. Median MELD score and Maddrey DF on admission were 21 (IQR 18 to 24) and 45 (IQR 26 to 62), respectively. Seventy-three percent of patients received corticosteroids and/or pentoxifylline, and the 90-day mortality was 17%. Independent predictors of mortality included older age, female sex, international normalized ratio, MELD score and Maddrey DF (all P<0.05). For discrimination of 90-day mortality, the areas under the ROC curves of the prognostic models (MELD 0.64; Maddrey DF 0.68) were similar (P>0.05). At optimal cut-offs of ≥22 for MELD score and ≥37 for Maddrey DF, both models excluded death with high certainty (negative predictive values 90% and 96%, respectively). CONCLUSIONS: In patients hospitalized for AH, well-known prognostic models can be used to predict 90-day mortality, particularly to identify patients with a low risk for death.

Burden of disease and cost of chronic hepatitis C infection in Canada.

BACKGROUND: Chronic infection with hepatitis C virus (HCV) is a major cause of cirrhosis, hepatocellular carcinoma and liver transplantation. OBJECTIVE: To estimate the burden of HCV-related disease and costs from a Canadian perspective. METHODS: Using a system dynamic framework, the authors quantified the HCV-infected population, disease progression and costs in Canada between 1950 and 2035. Specifically, 36 hypothetical, age- and sex-defined cohorts were tracked to define HCV prevalence, complications and direct medical costs (excluding the cost of antivirals). Model assumptions and costs were extracted from the literature with an emphasis on Canadian data. No incremental increase in antiviral treatment over current levels was assumed, despite the future availability of potent antivirals. RESULTS: The estimated prevalence of viremic hepatitis C cases peaked in 2003 at 260,000 individuals (uncertainty interval 192,460 to 319,880), reached 251,990 (uncertainty interval 177,890 to 314,800) by 2013 and is expected to decline to 188,190 (uncertainty interval 124,330 to 247,200) in 2035. However, the prevalence of advanced liver disease is increasing. The peak annual number of patients with compensated cirrhosis (n=36,210), decompensated cirrhosis (n=3380), hepatocellular carcinoma (n=2220) and liver-related deaths (n=1880) are expected to occur between 2031 and 2035. During this interval, an estimated 32,460 HCV-infected individuals will die of liver-related causes. Total health care costs associated with HCV (excluding treatment) are expected to increase by 60% from 2013 until the peak in 2032, with the majority attributable to cirrhosis and its complications (81% in 2032 versus 56% in 2013). The lifetime cost for an individual with HCV infection in 2013 was estimated to be $64,694. CONCLUSIONS: Although the prevalence of HCV in Canada is decreasing, cases of advanced liver disease and health care costs continue to rise. These results will facilitate disease forecasting, resource planning and the development of rational management strategies for HCV in Canada.

Decreasing mortality from acute biliary diseases that require endoscopic retrograde cholangiopancreatography: a nationwide cohort study.

BACKGROUND & AIMS: The management of acute biliary diseases often involves endoscopic retrograde cholangiopancreatography (ERCP), but it is not clear whether this technique reduces mortality. We investigated whether mortality from acute biliary diseases that require ERCP has been reduced over time and explored factors associated with mortality. METHODS: We conducted a cohort study using the Nationwide Inpatient Sample (1998-2008). We identified hospitalizations for choledocholithiasis, cholangitis, and acute pancreatitis that involved ERCP. Multivariate analyses were used to determine the effects of time period, patient factors, hospital characteristics, features of the ERCP procedure, and types of cholecystectomies on mortality, length of stay, and costs. RESULTS: From 1998 to 2008 there were 166,438 admissions for acute biliary conditions that met the inclusion criteria, corresponding to more than 800,000 patients nationwide. During this interval, mortality decreased from 1.1% to 0.6% (adjusted odds ratio [aOR], 0.7; 95% confidence interval [CI], 0.6-0.8), diagnostic ERCPs decreased from 28.8% to 10.0%, hospitals performing fewer than 100 ERCPs per year decreased from 38.4% to 26.9%, open cholecystectomies decreased from 12.4% to 5.8%, and unsuccessful ERCPs decreased from 6.3% to 3.2% (P < .0001 for all trends). Unsuccessful ERCP (aOR, 1.7; 95% CI, 1.4-2.2), open cholecystectomy (aOR, 3.4; 95% CI 2.7-4.3), cholangitis (aOR, 1.9; 95% CI, 1.5-2.3), older age, having Medicare health insurance, and comorbidity were associated with increased mortality. CONCLUSIONS: In-hospital mortality from acute biliary conditions requiring ERCP in the United States has decreased over time. Reductions in the rate of unsuccessful ERCPs and open cholecystectomies are associated with this trend.

Predictors of mortality among patients undergoing colectomy for ischemic colitis: A population-based United States study.

BACKGROUND: Ischemic colitis is a potentially life-threatening condition that can require colectomy for management. OBJECTIVE: To assess independent predictors of mortality following colectomy for ischemic colitis using a nationally representative sample of hospitals in the United States. METHODS: The Nationwide Inpatient Sample was used to identify all patients with a primary diagnosis of acute vascular insufficiency of the colon (International Classification of Diseases, Ninth Revision codes 557.0 and 557.9) who underwent a colectomy between 1993 and 2008. Incidence and mortality are described; multivariate logistic regression analysis was performed to determine predictors of mortality. RESULTS: The incidence of colectomy for ischemic colitis was 1.43 cases (95% CI 1.40 cases to 1.47 cases) per 100,000. The incidence of colectomy for ischemic colitis increased by 3.1% per year (95% CI 2.3% to 3.9%) from 1993 to 2003, and stabilized thereafter. The postoperative mortality rate was 21.0% (95% CI 20.2% to 21.8%). After 1997, the mortality rate significantly decreased at an estimated annual rate of 4.5% (95% CI -6.3% to -2.7%). Mortality was associated with older age, 65 to 84 years (OR 5.45 [95% CI 2.91 to 10.22]) versus 18 to 34 years; health insurance, Medicaid (OR 1.69 [95% CI 1.29 to 2.21]) and Medicare (OR 1.33 [95% CI 1.12 to 1.58]) versus private health insurance; and comorbidities such as liver disease (OR 3.54 [95% CI 2.79 to 4.50]). Patients who underwent colonoscopy or sigmoidoscopy (OR 0.78 [95% CI 0.65 to 0.93]) had lower mortality. CONCLUSIONS: Colectomy for ischemic colitis was associated with considerable mortality. The explanation for the stable incidence and decreasing mortality rates observed in the latter part of the present study should be explored in future studies.

A health technology assessment of transient elastography in adult liver disease.

BACKGROUND: An estimated one in 10 Canadians have some form of liver disease. The reference standard for staging and monitoring liver fibrosis is percutaneous liver biopsy--an invasive procedure associated with risks and complications. Transient elastography (TE) represents a noninvasive, ultrasound-based alternative. OBJECTIVE: To assess the efficacy of TE compared with liver biopsy for fibrosis staging in adults with five common types of liver disease: hepatitis B, hepatitis C, nonalcoholic fatty liver disease, cholestatic liver disease and complications post-liver transplantation. METHODS: A systematic review of published and grey literature from 2001 to June 2011 was conducted. Included were observational studies evaluating the accuracy of TE using liver biopsy as the comparator. An economic model was developed to estimate the cost per correct diagnosis gained with liver biopsy compared with TE. Identification of moderate fibrosis (stages 2 to 4) and cirrhosis (stage 4) were considered. RESULTS: Fifty-seven studies were included in the review. The diagnostic accuracy of TE for the five clinical subgroups had sensitivities ranging from 0.67 to 0.92 and specificities ranging from 0.72 to 0.95. Liver biopsy was associated with an additional $1,427 to $7,030 per correct diagnosis gained compared with TE. The model was sensitive to the sensitivity and specificity of TE and the prevalence of fibrosis. CONCLUSIONS: TE is an accurate diagnostic method in patients with moderate fibrosis or cirrhosis. TE is less effective but less expensive than liver biopsy. Systemic implementation of TE should be considered for the noninvasive assessment of liver fibrosis.

Direct comparison of the FibroScan XL and M probes for assessment of liver fibrosis in obese and nonobese patients.

BACKGROUND: A novel Fibroscan XL probe has recently been introduced and validated for obese patients, and has a diagnostic accuracy comparable with that of the standard M probe. The aim of this study was to analyze and understand the differences between these two probes in nonobese patients, to identify underlying causes for these differences, and to develop a practical algorithm to translate results for the XL probe to those for the M probe. METHODS AND RESULTS: Both probes were directly compared first in copolymer phantoms of varying stiffness (4.8, 11, and 40 kPa) and then in 371 obese and nonobese patients (body mass index, range 17.2-72.4) from German (n = 129) and Canadian (n = 242) centers. Liver stiffness values for both probes correlated better in phantoms than in patients (r = 0.98 versus 0.82, P < 0.001). Significantly more patients could be measured successfully using the XL probe than the M probe (98.4% versus 85.2%, respectively, P < 0.001) while the M probe produced a smaller interquartile range (21% versus 32%). Failure of the M probe to measure liver stiffness was not only observed in patients with a high body mass index and long skin-liver capsule distance but also in some nonobese patients (n = 10) due to quenching of the signal from subcutaneous fat tissue. In contrast with the phantoms, the XL probe consistently produced approximately 20% lower liver stiffness values in humans compared with the M probe. A long skin-liver capsule distance and a high degree of steatosis were responsible for this discordance. Adjustment of cutoff values for the XL probe (<5.5, 5.5-7, 7-10, and >10 kPa for F0, F1-2, F3, and F4 fibrosis, respectively) significantly improved agreement between the two probes from r = 0.655 to 0.679. CONCLUSION: Liver stiffness can be measured in significantly more obese and nonobese patients using the XL probe than the M probe. However, the XL probe is less accurate and adjusted cutoff values are required.

Predicting in-hospital mortality in patients undergoing complex gastrointestinal surgery: determining the optimal risk adjustment method.

OBJECTIVE: To compare the performance of Charlson/Deyo, Elixhauser, Disease Staging, and All Patient Refined Diagnosis-Related Groups (APR-DRGs) algorithms for predicting in-hospital mortality after 3 types of major abdominal surgeries: gastric, hepatic, and pancreatic resections. DESIGN: Cross-sectional nationwide sample. SETTING: Nationwide Inpatient Sample from 2002 to 2007. PATIENTS: Adult patients (≥18 years) hospitalized with a primary or secondary procedure of gastric, hepatic, or pancreatic resection between 2002 and 2007. MAIN OUTCOME MEASURES: Predicting in-hospital mortality using the 4 comorbidity algorithms. Logistic regression analyses were used and C statistics were calculated to assess the performance of the indexes. Risk adjustment methods were then compared. RESULTS: In our study, we identified 46,395 gastric resections, 18,234 hepatic resections, and 15,443 pancreatic resections. Predicted in-hospital mortality rates according to the adjustment methods agreed for 43.8% to 74.6% of patients. In all types of resections, the APR-DRGs and Disease Staging algorithms predicted in-hospital mortality better than the Charlson/Deyo and Elixhauser indexes (P < .001). Compared with the Charlson/Deyo algorithm, the Elixhauser index was of higher accuracy in gastric resections (0.847 vs 0.792), hepatic resections (0.810 vs 0.757), and pancreatic resections (0.811 vs 0.741) (P < .001 for all comparisons). Higher accuracy of the Elixhauser algorithm compared with the Charlson/Deyo algorithm was not affected by diagnosis rank, multiple surgeries, or exclusion of transplant patients. CONCLUSIONS: Different comorbidity algorithms were validated in the surgical setting. The Disease Staging and APR-DRGs algorithms were highly accurate. For commonly used algorithms such as Charlson/Deyo and Elixhauser, the latter showed higher accuracy.

Amoebic liver abscess in USA: a population-based study of incidence, temporal trends and mortality.

BACKGROUND: Amoebic liver abscess (ALA) may be associated with significant morbidity and mortality, but nationwide American data is unavailable. Our objective was to describe ALA epidemiology and outcomes in USA from a population-based perspective. METHODS: Patients hospitalized with ALA between 1993 and 2007 were identified using the Nationwide Inpatient Sample. Patient characteristics, interventions and outcomes including mortality were determined. The annual incidence of ALA and temporal trends were determined using the negative binomial regression models. RESULTS: Between 1993 and 2007, 848 hospitalizations for ALA, corresponding to ∼4100 hospitalizations nationwide, were identified. The annual incidence was 1.38 per million population with a 2.4% [95% confidence interval (CI) 0-4.8%; P=0.06] average annual decline during this study. Most patients were hospitalized in western (54%) and southern states (27%), and 48% were Hispanic. Males (incidence rate ratio vs. females: 4.53; 95% CI 4.19-4.90) had the highest incidence rates. Percutaneous and surgical drainage was required in 48 and 7% of patients respectively. Although length of stay [median, 6 days; interquartile range (IQR) 4-10] and hospital charges (US$25,345; IQR US$15,030-42, 275) were substantial, in-hospital mortality was rare (0.8%). Females [odds ratio (OR) 6.12; CI 1.39-26.8], patients ≥ 60 years (OR 13.3; 95% CI 2.5-71.5), and those with ≥ 3 comorbidities (OR 5.80; 95% CI 1.30-25.8), particularly malnutrition, had an increased risk of death. CONCLUSIONS: ALA is rare and the incidence has decreased in USA. Young, Hispanic males in southwestern states are most frequently affected. Mortality caused by ALA is lower than what was reported previously.

Elective resection of colon cancer by high-volume surgeons is associated with decreased morbidity and mortality.

BACKGROUND: The purpose of this study was to determine whether morbidity and mortality in patients undergoing elective resection of colon cancer are associated with surgeon or hospital volume. METHODS: Using the Nationwide Inpatient Sample database, we identified all adult patients who underwent elective resection for colon cancer as their primary procedure between 2003 and 2007. Cases were divided into three groups according to the mean number of resections performed annually by each surgeon: low volume (≤4/year), intermediate volume (5-9/year), or high volume (≥10/year). Annual hospital case-load was also categorized as low volume (≤30/year), intermediate volume (31-60/year), and high volume (≥61/year). Multiple logistic regression models were used to identify differences in morbidity and mortality. RESULTS: A total of 54,000 patients underwent resection of colon cancer by 7,313 surgeons in 1,398 hospitals. After adjusting for important covariates including hospital volume, colon cancer resection by high-volume surgeons was an independent predictor of decreased morbidity (odds ratio [OR], 0.91; 95% CI, 0.85-0.97) and mortality (OR, 0.75; 95% CI, 0.65-0.86). Mortality was lowest among patients operated on by high-volume surgeons in high-volume hospitals (2.2% vs. 3.9%; OR, 0.56; 95% CI, 0.46-0.68). CONCLUSIONS: In patients undergoing elective resection of colon cancer, procedures done by high-volume surgeons are associated with decreased morbidity and mortality.

Gender, renal function, and outcomes on the liver transplant waiting list: assessment of revised MELD including estimated glomerular filtration rate.

BACKGROUND & AIMS: The Model for End-Stage Liver Disease (MELD) allocation system for liver transplantation (LT) may present a disadvantage for women by including serum creatinine, which is typically lower in females. Our objectives were to investigate gender disparities in outcomes among LT candidates and to assess a revised MELD, including estimated glomerular filtration rate (eGFR), for predicting waiting list mortality. METHODS: Adults registered for LT between 2002 and 2007 were identified using the UNOS database. We compared components of MELD, MDRD-derived eGFR, and the 3-month probability of LT and death between genders. Discrimination of MELD, MELDNa, and revised models including eGFR for mortality were compared using c-statistics. RESULTS: A total of 40,393 patients (36% female) met the inclusion criteria; 9% died and 24% underwent LT within 3 months of listing. Compared with men, women had lower median serum creatinine (0.9 vs. 1.0 mg/dl), eGFR (72 vs. 83 ml/min/1.73 m(2)), and mean MELD (16.5 vs. 17.2; all p <0.0005), but within most MELD strata, had higher bilirubin and INR. After adjusting for relevant covariates including creatinine and body weight, women were less likely than men to receive a LT (hazard ratio [HR] 0.85; 95% CI 0.79-0.87) and had greater 3-month mortality (HR 1.13; 95% CI 1.05-1.21). Revision of MELD and MELDNa to include eGFR did not improve discrimination for 3-month mortality (c-statistics: MELD 0.896, MELD-eGFR 0.894, MELDNa 0.911, MELDNa-eGFR 0.905). CONCLUSIONS: Women are disadvantaged under MELD potentially due to its inclusion of creatinine. However, since including eGFR in MELD does not improve mortality prediction, alternative refinements are necessary.

Transient elastography for the noninvasive assessment of liver fibrosis: a multicentre Canadian study.

BACKGROUND: Liver stiffness measurement (LSM) using transient elastography (TE) is a promising tool for the noninvasive assessment of hepatic fibrosis. OBJECTIVES: To determine the feasibility and performance of TE in a North American cohort of patients with chronic liver disease. METHODS: LSMs were obtained using TE in 260 patients with chronic hepatitis B or C, or nonalcoholic fatty liver disease from four Canadian hepatology centres. The accuracy of TE compared with liver biopsy for the prediction of significant fibrosis (Metavir fibrosis score of F2 or greater), bridging fibrosis (Metavir fibrosis score of F3 or greater) and cirrhosis (Metavir fibrosis score of F4 ) was assessed using area under ROC curves (AUROCs), and compared with the aspartate aminotransferase-to-platelet ratio index. The influence of alanine aminotransferase (ALT) levels and other factors on liver stiffness was determined using linear regression analyses. RESULTS: failure of TE occurred in 2.7% of patients, while liver biopsies were inadequate for staging in 0.8%. Among the remaining 251 patients, the AUROCs of TE for Metavir fibrosis scores of F2 and F3 or greater, and F4 were 0.74 (95% CI 0.68 to 0.80), 0.89 (95% CI 0.84 to 0.94), and 0.94 (95% CI 0.90 to 0.97), respectively. LSM was more accurate than the aminotransferase-to-platelet ratio index for bridging fibrosis (AUROC 0.78) and cirrhosis (AUROC 0.88), but not significant fibrosis (AUROC 0.76). At a cut-off of 11.1 kPa, the sensitivity, specificity, and positive and negative predictive values for cirrhosis (prevalence 11%) were 96%, 81%, 39% and 99%, respectively. For significant fibrosis (prevalence 53%), a cut-off of 7.7 kPa was 68% sensitive and 69% specific, and had a positive predictive value of 70% and a negative predictive value of 65%. Liver stiffness was independently associated with ALT, body mass index and steatosis. The optimal LSM cut-offs for cirrhosis were 11.1 kPa and 11.5 kPa in patients with ALT levels lower than 100 U/L and 100 U/L or greater, respectively. For fibrosis scores of F2 or greater, these figures were 7.0 kPa and 8.6 kPa, respectively. CONCLUSIONS: the major role of TE is the exclusion of bridging fibrosis and cirrhosis. However, TE cannot replace biopsy for the diagnosis of significant fibrosis. Because liver stiffness may be influenced by significant ALT elevation, body mass index and/or steatosis, tailored liver stiffness cut-offs may be necessary to account for these factors.