Unlocking better bladder cancer care with AMAZON: adaptive quantile monitoring under progressive censoring

Monitoring systems are essential tools in various scientific fields for identifying changes in lifetime or survival-related phenomena. Statistical monitoring, especially control charts for censored data, is essential in oncology to track remission times and detect disease progression, particularly when patients are lost to follow-up. Therefore, this study presents a novel Adaptive Monitoring framework (control chart) for the Alpha Power Weibull distribution'z (APWD) quantile based on prOgressive type-II (PT-II) ceNsored data, (AMAZON will be used hereafter). AMAZON leverages the Quantile Function (QF) of APWD and an adaptive exponentially weighted moving average (AEWMA) chart, tailored for Bladder Cancer (BC) Care. Maximum likelihood estimators (MLE) for APWD under PT-II censoring form the backbone of AMAZON. A comprehensive simulation study assesses the performance using average run length. Besides an application to BC remission times confirms its ability to detect OOC signals. A rigorous comparative analysis with Weibull-based V-chart and EEP-based np-chart demonstrate that AMAZON significantly outperforms both benchmarks in sensitivity and stability. Application to real BC remission times confirms that AMAZON maintains perfect IC stability and triggers an immediate OOC signal at the first shifted sample, while V-chart suffers from false alarms and EEP based-chart fails to detect the shift entirely. Figure presents the graphical abstract of the proposed methodology.The AMAZON Framework: An Adaptive Quantile Monitoring Scheme for Alpha Power Weibull Data under Progressive Censoring.Highlights A novel AMAZON (Adaptive Monitoring) scheme is introduced for the Quantile Function of the Alpha Power Weibull Distribution.The scheme effectively monitors lifetime data subject to Progressive Type-II (PT-II) Censoring, crucial for clinical trials.Adaptive EWMA charting provides superior sensitivity to both small and large process shifts (ϖQ).Application to Bladder Cancer remission times shows AMAZON significantly outperforms Weibull V-chart and EEP-based np-chart.