Development of Clinical Prediction Rules for Neutropenia in Lung Cancer Patients Receiving Chemotherapy at Maharaj Nakorn Chiang Mai Hospital

Abstract

Objective: To develop clinical prediction rules for chemotherapy-induced neutropenia in lung cancer patients. Methods: A retrospective cohort study was conducted by retrospectively collecting data from medical records and electronic medical records of lung cancer patients recieving care at Maharaj Nakorn Chiang Mai Hospital, Faculty of Medicine, Chiang Mai University between January 1, 2016, and October 31, 2022. There were two groups of subjects. i.e., those with neutropenia and those without. The study analyzed the relationship of various factors and neutropenia, then used the results to develop clinical prediction rules by weighting scores with mixed effect model. The study determined cut-offs of the risk scores using receiver operating characteristic curve and presented the performance of the risk scores in predicting neutropenia. Results: The subjects were 799 patients with a total of 2,617 treatment visits. Neutropenia occurred in 20 visits and did not occur in 2,597 visits. Factors significantly associated with neutropenia before receiving chemotherpay included neutrophil count <3,000 cells/mm³, and being treated with platinum compounds, vinca alkaloids, and topoisomerase II inhibitor chemotherapies. Risk scores developed from significant factors ranged from 0-6.5. Area under the ROC curve was 83.63%. Derived cut-off risk scores were low risk (<2.0), moderate risk (2.0-2.5), and high risk (>2.5), with LHR+ (95%CI) of 0.37 (0.19-0.73), 2.54 (1.55-4.14), and 12.99 (5.79-29.11), respectively. Conclusion: The developed clinical prediction tool for chemotherapy-induced neutropenia in lung cancer patients demonstrated high accuracy and good predictive ability. However, it should be further validated before implemention in other healthcare settings with different contexts.