An Electronic Health Record Model for Predicting Risk of Hepatic Fibrosis in Primary Care Patients.

BACKGROUND: One challenge for primary care providers caring for patients with nonalcoholic fatty liver disease is to identify those at the highest risk for clinically significant liver disease. AIM: To derive a risk stratification tool using variables from structured electronic health record (EHR) data for use in populations which are disproportionately affected with obesity and diabetes. METHODS: We used data from 344 participants who underwent Fibroscan examination to measure liver fat and liver stiffness measurement [LSM]. Using two approaches, multivariable logistic regression and random forest classification, we assessed risk factors for any hepatic fibrosis (LSM > 7 kPa) and significant hepatic fibrosis (> 8 kPa). Possible predictors included data from the EHR for age, gender, diabetes, hypertension, FIB-4, body mass index (BMI), LDL, HDL, and triglycerides. RESULTS: Of 344 patients (56.4% women), 34 had any hepatic fibrosis, and 15 significant hepatic fibrosis. Three variables (BMI, FIB-4, diabetes) were identified from both approaches. When we used variable cut-offs defined by Youden's index, the final model predicting any hepatic fibrosis had an AUC of 0.75 (95% CI 0.67-0.84), NPV of 91.5% and PPV of 40.0%. The final model with variable categories based on standard clinical thresholds (i.e., BMI ≥ 30 kg/m2; FIB-4 ≥ 1.45) had lower discriminatory ability (AUC 0.65), but higher PPV (50.0%) and similar NPV (91.3%). We observed similar findings for predicting significant hepatic fibrosis. CONCLUSIONS: Our results demonstrate that standard thresholds for clinical risk factors/biomarkers may need to be modified for greater discriminatory ability among populations with high prevalence of obesity and diabetes.

Irritable Bowel Syndrome in Veterans with Gulf War Illness Evaluated at VA's War-Related Illness and Injury Study Center.

INTRODUCTION: Excess rates of Gulf War illness (GWI) and irritable bowel syndrome (IBS), two chronic multisymptom illnesses, have long been documented among nearly 700,000 veterans who served in the 1990-1991 Persian Gulf War. We sought to report the prevalence, characteristics, and association of GWI and IBS decades after the war in a clinical cohort of deployed Gulf War veterans (GWVs) who were evaluated at the Department of Veterans Affairs' War Related Illness and Injury Study Center (WRIISC) for unexplained chronic symptoms. MATERIALS AND METHODS: We analyzed data gathered from clinical intake questionnaires of deployed GWVs who were evaluated at WRIISC clinics between 2008 and 2020. We applied Centers for Disease Control (CDC) criteria to determine the prevalence of severe GWI. IBS was identified using Rome IV diagnostic criteria (current IBS) and veterans' self-reported "history of physician-diagnosed IBS." We examined associations between IBS and GWI using bivariate analyses and multivariable logistic regression. RESULTS: Among the N = 578 GWVs evaluated by the WRIISC, severe GWI (71.8%), history of physician-diagnosed IBS (50.3%) and current IBS (42.2%) were all highly prevalent. Nearly half of GWVs with severe GWI met Rome criteria for IBS (45.8%), and over half reported a history of physician-diagnosed IBS (56.1%). In multivariable models, severe GWI was significantly associated both with current IBS (adjusted odds ratio (aOR): 1.68, 95% CI: 1.11, 2.54) and with veteran-reported history of physician-diagnosed IBS (aOR: 2.15, 95% CI: 1.43, 2.23). IBS with diarrhea (IBS-D) was the most common subtype among GWVs with current IBS (61.1%). However, IBS-mixed affected a significantly greater proportion of veterans with severe GWI, compared to veterans who did not have severe GWI (P = .03). CONCLUSIONS: More than 20 years after the Persian Gulf War, our findings indicate a high degree of comorbidity between severe GWI and IBS among deployed GWVs seeking care for unexplained illnesses. Our results suggest GWVs with GWI should be screened for IBS for which evidence-based treatments are available and could potentially reduce symptom burden. Conversely, symptoms of IBS should trigger additional evaluation for non-gastrointestinal symptoms in deployed Gulf War veterans to identify possible GWI and ensure a comprehensive approach to care.

The Impact of Race on Pancreatic Cancer Treatment and Survival in the Nationwide Veterans Affairs Healthcare System.

OBJECTIVES: Among patients with pancreatic cancer, studies show racial disparities at multiple steps of the cancer care pathway. Access to healthcare is a frequently cited cause of these disparities. It remains unclear if racial disparities exist in an integrated, equal access public system such as the Veterans Affairs healthcare system. METHODS: We identified all patients diagnosed with pancreatic adenocarcinoma in the national Veterans Affairs Central Cancer Registry from January 2010 to December 2018. We examined the independent association between race and 3 endpoints: stage at diagnosis, receipt of treatment, and survival while adjusting for sociodemographic factors and medical comorbidities. RESULTS: We identified 8529 patients with pancreatic adenocarcinoma, of whom 79.5% were White and 20.5% were Black. Black patients were 19% more likely to have late-stage disease and 25% less likely to undergo surgical resection. Black patients had 13% higher mortality risk compared with White patients after adjusting for sociodemographic characteristics and medical comorbidities. This difference in mortality was no longer statistically significant after additionally adjusting for cancer stage and receipt of potentially curative treatment. CONCLUSIONS: Equal access to healthcare might have reduced but failed to eliminate disparities. Dedicated efforts are needed to understand reasons underlying these disparities in an attempt to close these persistent gaps.

The Predictive Performance of Contemporary Guideline Recommendations for Helicobacter pylori Testing in a United States Population.

BACKGROUND: The Houston Consensus Conference and American College of Gastroenterology (ACG) have recommended Helicobacter pylori screening in United States populations with specific risk factors. However, the performance of these guidelines in clinical practice is not known. METHODS: We identified consecutive patients undergoing upper endoscopy with gastric biopsies for any indication in a safety-net hospital in Houston, TX during January 2015-December 2016. We tested the association between the presence of H pylori (histopathology, stool antigen, urea breath test, immunoglobulin G serology, or prior treatment) and H pylori risk factors using logistic regression models, reported as odds ratios and 95% confidence intervals (CIs). We evaluated the area under the receiver operating characteristic (AUROC) curve for predictive ability of individual risk factors identified by the Houston Consensus Conference and ACG. RESULTS: Of 942 patients, the prevalence of H pylori infection was 51.5%. The risk factors with the highest predictive performance included first-generation immigrant (AUROC, 0.59) and Hispanic or black race/ethnicity (AUROC, 0.57), whereas the remaining 7 risk factors/statements had low predictive value. A model that combined first-generation immigrant status, black or Hispanic race/ethnicity, dyspepsia, and reflux had higher predictive ability for H pylori infection (AUROC, 0.64; 95% CI, 0.61-0.68) than any individual risk factor. CONCLUSIONS: In this contemporary U.S. cohort, the performance of individual risk factors identified by the Houston Consensus Conference and ACG was generally low for predicting H pylori infection except for black or Hispanic race/ethnicity and first-generation immigrant status. A risk prediction model combining several risk factors had improved diagnostic performance and should be validated in future studies.

Using Natural Language Processing to Automatically Identify Dysplasia in Pathology Reports for Patients With Barrett's Esophagus.

BACKGROUND & AIMS: Identifying dysplasia of Barrett's esophagus (BE) in the electronic medical record (EMR) requires manual abstraction of unstructured data. Natural language processing (NLP) creates structure to unstructured free text. We aimed to develop and validate an NLP algorithm to identify dysplasia in BE patients on histopathology reports with varying report formats in a large integrated EMR system. METHODS: We randomly selected 600 pathology reports for NLP development and 400 reports for validation from patients with suspected BE in the national Veterans Affairs databases. BE and dysplasia were verified by manual review of the pathology reports. We used NLP software (Clinical Language Annotation, Modeling, and Processing Toolkit; Melax Tech, Houston, TX) to develop an algorithm to identify dysplasia using findings. The algorithm performance characteristics were calculated as recall, precision, accuracy, and F-measure. RESULTS: In the development set of 600 patients, 457 patients had confirmed BE (60 with dysplasia). The NLP identified dysplasia with 98.0% accuracy, 91.7% recall, and 93.2% precision, with an F-measure of 92.4%. All 7 patients with confirmed high-grade dysplasia were classified by the algorithm as having dysplasia. Among the 400 patients in the validation cohort, 230 had confirmed BE (39 with dysplasia). Compared with manual review, the NLP algorithm identified dysplasia with 98.7% accuracy, 92.3% recall, and 100.0% precision, with an F-measure of 96.0%. CONCLUSIONS: NLP yielded a high degree of sensitivity and accuracy for identifying dysplasia from diverse types of pathology reports for patients with BE. The application of this algorithm would facilitate research and clinical care in an EMR system with text reports in large data repositories.