LiverRisk score: An accurate, cost-effective tool to predict fibrosis, liver-related, and diabetes-related mortality in the general population
Shanghao Liu,Xiaohan Chen,Xuanwei Jiang,Xiaochun Yin,Ginenus Fekadu,Chuan Liu,Yan He,Huihui Chen,Wenjing Ni,Ruiying Wang,Qing-Lei Zeng,Yuping Chen,Ling Yang,Ruihua Shi,Sheng-Hong Ju,Jie Shen,Jingli Gao,Linhua Zhao,Wai-Kit Ming,Victor W Zhong,Gao-Jun Teng,Xiaolong Qi
DOI: https://doi.org/10.1016/j.medj.2024.03.003
2024-06-14
Med
Abstract:Background: Noninvasive and early assessment of liver fibrosis is of great significance and is challenging. We aimed to evaluate the predictive performance and cost-effectiveness of the LiverRisk score for liver fibrosis and liver-related and diabetes-related mortality in the general population. Methods: The general population from the NHANES 2017-March 2020, NHANES 1999-2018, and UK Biobank 2006-2010 were included in the cross-sectional cohort (n = 3,770), along with the NHANES follow-up cohort (n = 25,317) and the UK Biobank follow-up cohort (n = 17,259). The cost-effectiveness analysis was performed using TreeAge Pro software. Liver stiffness measurements ≥10 kPa were defined as compensated advanced chronic liver disease (cACLD). Findings: Compared to conventional scores, the LiverRisk score had significantly better accuracy and calibration in predicting liver fibrosis, with an area under the receiver operating characteristic curve (AUC) of 0.76 (0.72-0.79) for cACLD. According to the updated thresholds of LiverRisk score (6 and 10), we reclassified the population into three groups: low, medium, and high risk. The AUCs of LiverRisk score for predicting liver-related and diabetes-related mortality at 5, 10, and 15 years were all above 0.8, with better performance than the Fibrosis-4 score. Furthermore, compared to the low-risk group, the medium-risk and high-risk groups in the two follow-up cohorts had a significantly higher risk of liver-related and diabetes-related mortality. Finally, the cost-effectiveness analysis showed that the incremental cost-effectiveness ratio for LiverRisk score compared to FIB-4 was USD $18,170 per additional quality-adjusted life-year (QALY) gained, below the willingness-to-pay threshold of $50,000/QALY. Conclusions: The LiverRisk score is an accurate, cost-effective tool to predict liver fibrosis and liver-related and diabetes-related mortality in the general population. Funding: The National Natural Science Foundation of China (nos. 82330060, 92059202, and 92359304); the Key Research and Development Program of Jiangsu Province (BE2023767a); the Fundamental Research Fund of Southeast University (3290002303A2); Changjiang Scholars Talent Cultivation Project of Zhongda Hospital of Southeast University (2023YJXYYRCPY03); and the Research Personnel Cultivation Program of Zhongda Hospital Southeast University (CZXM-GSP-RC125).