Measuring resilience by cognitive diagnosis models and its prediction of 6-month quality of life in Be Resilient to Breast Cancer (BRBC)

ObjectiveThe application of advanced Cognitive Diagnosis Models (CDMs) in the Patient Reported Outcome (PRO) is limited due to its complex statistics. This study was designed to measure resilience using CDMs and its prediction of 6-month Quality of Life (QoL) in breast cancer. MethodsA total of 492 patients were longitudinally enrolled from Be Resilient to Breast Cancer (BRBC) and administered with 10-item Resilience Scale Specific to Cancer (RS-SC-10) and Functional Assessment of Cancer Therapy-Breast (FACT-B). Generalized Deterministic Input, Noisy And Gate (G-DINA) was performed to measure cognitive diagnostic probabilities (CDPs) of resilience. Integrated Discrimination Improvement (IDI) and Net Reclassification Improvement (NRI) were utilized to estimate the incremental prediction value of cognitive diagnostic probabilities over total score. ResultsCDPs of resilience improved prediction of 6-month QoL above conventional total score. AUC increased from 82.6-88.8% to 95.2-96.5% in four cohorts (all P < 0.001). The NRI ranged from 15.13 to 54.01% and IDI ranged from 24.69 to 47.55% (all P < 0.001). ConclusionCDPs of resilience contribute to a more accurate prediction of 6-month QoL above conventional total score. CDMs could help optimize Patient Reported Outcomes (PROs) measurement in breast cancer.

Automated summary

This study aimed to measure resilience using advanced Cognitive Diagnosis Models (CDMs) and predict the 6-month Quality of Life (QoL) in breast cancer patients. The results showed that cognitive diagnostic probabilities (CDPs) improved the accuracy of predicting 6-month QoL compared to conventional total score. CDMs can help optimize Patient Reported Outcomes (PROs) measurement in breast cancer.