Frontiers in health services · 2026
Optimising non-invasive screening for hepatic fibrosis in people living with HIV and intermediate FIB-4 scores.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Ramos-Ruperto Luis, Montes Maria Luisa, Busca-Arenzana Carmen, Manescu Petru, Gilson Richard, Fernandez-Reyes Delmiro, Arenas-Pinto Alejandro
Аннотация
BACKGROUND: Steatotic liver disease is increasingly recognised in people living with HIV. Non-invasive fibrosis screening strategies commonly rely on the FIB-4 score, but individuals with intermediate values (1.3-2.67) fall into a diagnostic grey zone where additional risk stratification may be required. Identifying simple and effective approaches to guide further assessment is particularly important in routine HIV care settings.
METHODS: We analysed two independent cohorts of people living with HIV undergoing transient elastography (TE): a development cohort from London (n = 229) and an external validation cohort from Madrid (n = 188). Among individuals with intermediate FIB-4 scores, we evaluated the performance of established non-invasive scores, including APRI, and several machine learning models for predicting liver stiffness thresholds associated with significant fibrosis (≥7 kPa and ≥8 kPa). Model performance was assessed using sensitivity, specificity, and predictive values.
RESULTS: In both cohorts, the prevalence of significant fibrosis was low. APRI demonstrated consistently high sensitivity and strong negative predictive value for identifying individuals without significant fibrosis. Machine learning models showed modest discrimination and tended to favour negative predictions, reflecting the low prevalence of fibrosis in this screening population. Across models, no machine learning approach demonstrated clear improvement over APRI in identifying individuals at risk of fibrosis.
CONCLUSION: In people living with HIV with intermediate FIB-4 scores, APRI may provide a simple and effective strategy to further stratify fibrosis risk in routine clinical practice. In this real-world screening population, machine learning models did not outperform established non-invasive scores.
METHODS: We analysed two independent cohorts of people living with HIV undergoing transient elastography (TE): a development cohort from London (n = 229) and an external validation cohort from Madrid (n = 188). Among individuals with intermediate FIB-4 scores, we evaluated the performance of established non-invasive scores, including APRI, and several machine learning models for predicting liver stiffness thresholds associated with significant fibrosis (≥7 kPa and ≥8 kPa). Model performance was assessed using sensitivity, specificity, and predictive values.
RESULTS: In both cohorts, the prevalence of significant fibrosis was low. APRI demonstrated consistently high sensitivity and strong negative predictive value for identifying individuals without significant fibrosis. Machine learning models showed modest discrimination and tended to favour negative predictions, reflecting the low prevalence of fibrosis in this screening population. Across models, no machine learning approach demonstrated clear improvement over APRI in identifying individuals at risk of fibrosis.
CONCLUSION: In people living with HIV with intermediate FIB-4 scores, APRI may provide a simple and effective strategy to further stratify fibrosis risk in routine clinical practice. In this real-world screening population, machine learning models did not outperform established non-invasive scores.