Predictive modeling of response to surveillance outreach invitations can assist in the development of tailored strategies to maximize effectiveness of hepatocellular carcinoma (HCC) surveillance implementation, according to a study published in Clinical Gastroenterology and Hepatology.
There is a need for interventions to increase HCC surveillance in patients with cirrhosis due to a high proportion of late-stage detection. Accurate prediction of patients’ likelihood to respond to interventions could allow for a cost-effective approach to outreach and facilitate targeting more intensive interventions to non-responders.
Therefore, researchers conducted a secondary analysis of a randomized clinical trial that evaluated a mail-based outreach strategy to promote HCC surveillance among 1200 cirrhosis patients at a safety-net health system between December 2014 and March 2017. They developed regularized logistic regression (RLR) and gradient boosting machine (GBM) algorithm models to predict surveillance completion during three screening rounds in a training set (n=960). Model performance was assessed using multiple performance metrics in an independent test set (n=240).
Researchers found that among 1200 patients, surveillance was completed in 41-47% of patients over the three rounds. The RLR and GBM models demonstrated good discriminatory accuracy, with area under receiver operating characteristic (AUROC) curves of 0.67 and 0.66, respectively, in the first surveillance round, which improved to 0.77 by the third surveillance round after incorporating prior screening behavior as a feature.
Additional performance characteristics including the Brier score, the Hosmer-Lemeshow test, and reliability diagrams were also evaluated. The most important variables for the predictive model were prior screening completion status and past primary care contact.
“In summary, we found predictive models can help stratify patients’ likelihood to respond to HCC surveillance invitations, facilitating tailored outreach invitation strategies to maximize effectiveness and cost-effectiveness,” stated the authors.
Disclosure: Several study authors declared affiliations with the pharmaceutical industry. Please see the original reference for a full list of authors’ disclosures.
Singal AG, Chen Y, Sridhar S, et al. Novel application of predictive modeling: a tailored approach to promoting HCC surveillance in patients with cirrhosis. Clin Gastroenterol Hepatol. Published online February 22, 2021. doi: 10.1016/j.cgh.2021.02.038