Improving the Diagnosis and Fibrosis Risk Assessment of Nonalcoholic Fatty Liver Disease in Primary Care Patients with Abnormal Liver Chemistries

摘要 改善原发性非酒精性脂肪肝的诊断和纤维化风险评估 护理肝脏化学异常的患者。 慢性肝病，包括非酒精性脂肪性肝病（NAFLD）、酒精相关性肝病（ALD）， 和病毒性肝炎，往往在早期未被发现，诊断延误直接损害患者。 普遍存在的诊断错误助长了慢性和终末期肝病的攀升，尽管越来越多的 治疗选择的可用性和有效性。 肝化学升高可能提示慢性肝病，以及对这些异常的系统反应 可以导致早期疾病识别和有效治疗的交付。4 - 11目前，对异常 初级保健中的肝脏检查缺乏一致性，并导致诊断错误。11 - 17我们的K23工作发现， 12%的肝功能检查异常的患者缺乏重复评估，只有16%的连续肝功能检查异常的患者缺乏重复评估。 肝功能检查异常及时接受丙型肝炎病毒检测。14，16这些有限且不一致的评估 挑战我们开发将患者水平变量与肝病诊断联系起来的模型的能力。 诊断NAFLD带来了独特的挑战，因为诊断需要阴性酒精暴露史， 全面排除其他肝脏疾病，和/或腹部成像，无法获得或不存在的元素 11，18尽管基于初级保健的风险因素包括肥胖，高血压， 在这种情况下，NAFLD仍然诊断不足。18 - 22我们的K23工作强调了这种情况的严重性。 诊断不足，因为只有31%的患者有肝脂肪变性的放射学证据， NAFLD）慢性肝病（n = 767）曾接受过NAFLD的诊断代码。 除了诊断，在初级保健中的NAFLD管理需要纤维化风险评估，因为 晚期纤维化的发生率是这些患者肝脏相关、心血管和总体死亡率的最佳指标。 25目前的纤维化风险评估开始采用非侵入性风险评分，包括纤维化-4指数（FIB-4） 和NAFLD纤维化评分（NFS），以确定最有可能从肝病治疗中获益的患者。6，18，26 - 28 FIB-4和NFS是在专科和三级护理队列中开发和测试的， 初级保健.当我们将FIB-4和NFS应用于我们有限的初级保健NAFLD队列时，结果显示， 大量的高风险评分，往往是不一致的，并会导致相互矛盾的临床决策 30%的样本。通过振动控制弹性成像测量肝脏硬度的随访测试 可以提高非侵入性检测的准确性，但目前初级保健中还没有这项技术。 在这项提案中，研究人员试图部署一个积极的诊断策略，以改善初级保健 诊断NAFLD（目标1），并评估与高风险NAFLD相关的基于EHR的患者水平临床变量。 通过非侵入性风险评分和振动控制弹性成像确定的晚期纤维化（目的2）。

Abstract Improving the Diagnosis and Fibrosis Risk Assessment of Nonalcoholic Fatty Liver Disease in Primary Care Patients with Abnormal Liver Chemistries. Chronic liver diseases, including nonalcoholic fatty liver disease (NAFLD), alcohol-related liver disease (ALD), and viral hepatitis, often go undetected in their early stages, a diagnostic delay directly harmful to patients.1,2 Pervasive diagnostic error fuels the climbing toll of chronic and end-stage liver disease, despite the increasing availability and efficacy of therapeutic options.3 Liver chemistry elevations may signal chronic liver disease, and systematic responses to these abnormalities can lead to earlier disease recognition and delivery of effective treatment.4-11 Currently, responses to abnormal liver tests in primary care lack consistency and contribute to diagnostic error.11-17 Our K23 work found nearly 12% of patients with abnormal liver tests lacked repeat assessment, and only 16% of patients with consecutive liver test abnormalities received timely viral hepatitis C testing.14,16 These limited and inconsistent evaluations challenge our ability to develop models linking patient-level variables to liver disease diagnoses. Diagnosing NAFLD poses unique challenges, as the diagnosis requires a negative alcohol exposure history, a comprehensive ruling out of other liver conditions, and/or abdominal imaging, elements inaccessible or absent in electronic health records.11,18 Despite primary care-based risk factors including obesity, hypertension, and diabetes, NAFLD remains underdiagnosed in this setting.18-22 Our K23 work highlights the severity of this underdiagnosis, as only 31% of patients with radiographic evidence of hepatic steatosis and no known (non- NAFLD) chronic liver disease (n=767) ever received a diagnosis code for NAFLD. Beyond diagnosis, NAFLD management in primary care requires fibrosis risk assessment because the presence of advanced fibrosis is the best indicator of liver-related, cardiovascular, and overall mortality in these patients.23- 25 Current fibrosis risk assessments begin with non-invasive risk scores, including the Fibrosis-4 index (FIB-4) and the NAFLD Fibrosis score (NFS), to identify patients most likely to benefit from hepatology referral.6,18,26-28 FIB-4 and NFS were developed and tested in specialty and tertiary care cohorts and may perform differently in primary care. When we applied FIB-4 and NFS to our limited primary care NAFLD cohorts, the results revealed an abundance of high-risk scores, were often discrepant, and would have resulted in conflicting clinical decisions for 30% of the sample. Follow-up testing with liver stiffness measurement by vibration-controlled elastography can improve non-invasive test accuracy, but this technology is not currently available in primary care.29,30 In this proposal, the investigators seek to deploy a proactive diagnostic strategy to improve the primary care diagnosis of NAFLD (Aim 1) and evaluate EHR-based, patient-level clinical variables associated with a high-risk for advanced fibrosis identified by non-invasive risk scores and vibration-controlled elastography (Aim 2).