论文部分内容阅读
目的建立一个非酒精性脂肪肝(NAFLD)筛查模型并对其进行验证。方法对“山东多中心健康管理纵向观察队列”中8 993名接受健康体检且无过量饮酒个体随机抽取80%进行建模,并做组内评价,剩余20%作组外评价。通过逐步Logistic回归,建立NAFLD筛查模型,并进行评价及验证。结果多因素Logistic回归分析表明,性别、体质量指数、高血压、血脂异常、谷草转氨酶/谷丙转氨酶(AST/ALT)和血糖(FBG)进入了模型,构建了NAFLD筛查模型并可通过计算得到脂肪肝指数(fatty liver index,FLI)。筛查模型的鉴别能力采用受试者工作特征曲线下面积(AUC)进行评价(组内:0.859,95%CI:0.851~0.867;组外:0.853,95%CI:0.835~0.869)。当FLI≤1.25时排除疾病,组内和组外的阴性预测值分别为93.1%和93.3%,FLI≥2.25诊断为NAFLD,两组的阳性预测值分别为74.6%和72.7%。结论 FLI是一个简单有效的筛查工具,可以用于高危人群的筛查,具有一定的实用价值。
Objective To establish a non-alcoholic fatty liver disease (NAFLD) screening model and verify it. Methods Eighty-three (3) randomly selected 80% of 8 993 healthy subjects who did not drink alcohol in the “Multi-center Health Management Longitudinal Observation Cohort of Shandong” were modeled and evaluated in-group and the remaining 20% were evaluated outside the group. Through stepwise logistic regression, a screening model of NAFLD was established and evaluated and verified. Results Multivariate Logistic regression analysis showed that gender, body mass index, hypertension, dyslipidemia, AST / ALT and FBG entered the model, and a screening model of NAFLD was established and can be calculated Obtained fatty liver index (fatty liver index, FLI). The discriminatory power of the screening model was assessed using the area under the receiver operating characteristic curve (AUC) (0.859, 95% CI: 0.851-0.867; 0.853, 95% CI: 0.835-0.869, outside the group). The disease was excluded when FLI ≤ 1.25. The negative predictive values were 93.1% and 93.3% in the group and in the group, respectively. The positive predictive value of FLI≥2.25 was 74.6% and 72.7% in both groups. Conclusion FLI is a simple and effective screening tool, which can be used for screening high-risk groups and has certain practical value.