In Journal of the National Cancer Institute
BACKGROUND : The aim of this study is to provide a comprehensive understanding of the current landscape of AI for cancer clinical trial enrolment, and its predictive accuracy in identifying eligible patients for inclusion in such trials.
METHODS : Databases of PubMed, Embase and Cochrane CENTRAL were searched until June 2022. Articles were included if they reported on AI actively being used in the clinical trial enrolment process. Narrative synthesis was conducted amongst all extracted data-accuracy, sensitivity, specificity, positive predictive value, negative predictive value. For studies where the 2x2 contingency table could be calculated or supplied by authors, a meta-analysis to calculate summary statistics was conducted using the hierarchical summary receiver operating characteristics curve model.
RESULTS : Ten articles, reporting on over 50,000 patients in 19 datasets were included. Accuracy, sensitivity and specificity exceeded 80% in all but one dataset. Positive predictive value exceeded 80% in 5 of 17 datasets. Negative predictive value exceeded 80% in all datasets. Summary sensitivity was 90.5% (95% CI: 70.9%-97.4%); summary specificity was 99.3% (95% CI: 81.8%-99.9%).
CONCLUSIONS : AI demonstrated comparable, if not superior performance to manual screening for patient enrolment into cancer clinical trials. As well, AI is highly efficient, requiring less time and human resources to screen patients. AI should be further investigated and implemented for patient recruitment into cancer clinical trials. Future research should validate the use of AI for clinical trials enrolment in less resource-rich regions, and to ensure broad inclusion for generalizability to all genders, ages and ethnicities.
Chow Ronald, Midroni Julie, Kaur Jagdeep, Boldt Gabriel, Liu Geoffrey, Eng Lawson, Liu Fei-Fei, Haibe-Kains Benjamin, Lock Michael, Raman Srinivas
2023-Jan-23
artificial intelligence, cancer, clinical trial, eligibility, enrolment
