Objectives: To conduct a systematic review of studies externally validating the ADNEX model for ovarian cancer diagnosis and perform a meta-analysis of its performance. Design: Systematic review, meta-analysis Data sources: Medline, EMBASE, WOS, Scopus, and EuropePMC up to 15/05/2023. Review methods: We included external validation studies of the performance of ADNEX using any study design and any study population comprising patients with an adnexal mass. Two independent reviewers extracted data. Disagreements were resolved through discussion. Reporting quality of the studies was scored using the TRIPOD reporting guideline and methodological conduct and risk of bias using the PROBAST tool. We performed random effects meta-analysis of the AUC, sensitivity and specificity at the 10% risk of malignancy threshold, and Net Benefit and Relative Utility at the 10% risk of malignancy threshold. Results: We included 47 studies (17,007 tumours) with median study sample size 261 (range 24-4905). On average, 61% of TRIPOD items were reported. Handling of missing data, sample size justification, and model calibration were rarely described. 91% of validations were at high risk of bias, mainly due to the unexplained exclusion of incomplete cases, low sample size, or absent calibration assessment. The summary AUC to distinguish benign from malignant tumours in operated patients was 0.93 (95% CI 0.92-0.94, 95% prediction interval 0.85-0.98) for ADNEX with CA125 as a predictor (9202 tumours, 43 centres, 18 countries, 21 studies) and 0.93 (95% CI 0.91-0.94, 95% prediction interval 0.85-0.98) for ADNEX without CA125 (6309 tumours, 31 centres, 13 countries, 12 studies). The estimated probability that the model has clinical utility in a new centre was 95% (with CA125) and 91% (without CA125). When restricting analysis to studies at low risk of bias, summary AUCs were 0.93 (with CA125) and 0.91 (without CA125), and estimated probabilities that the model has clinical utility were 89% (with CA125) and 87% (without CA125). Discussion: ADNEX performed well to distinguish benign from malignant tumours in populations from different countries and settings regardless of whether CA125 was used or not. A key limitation is that calibration was rarely assessed. Review registration: PROSPERO, CRD42022373182
This report addresses, from a machine learning perspective, a multi-class classification problem to predict the first deterioration level of a COVID-19 positive patient at the time of hospital admission. Socio-demographic features, laboratory tests and other measures are taken into account to learn the models. Our output is divided into 4 categories ranging from healthy patients, followed by patients requiring some form of ventilation (divided in 2 categories) and finally patients expected to die. The study is conducted thanks to data provided by "Sociedad Española de Medicina Interna" (SEMI) and "Red de Investigación en Servicios de Salud de Enfermedades Crónicas" (REDISSEC). Results show that logistic regression is the best method for identifying patients with clinical deterioration.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.