KMDATA

KMDATA provides reconstructed patient-level time-to-event datasets from 153 Phase III oncology clinical trials (breast, lung, prostate, colorectal) published 2014–2016 to support trial design and statistical evaluation.

Key Features:

• Patient-Level Data Reconstruction: Digitization of Kaplan–Meier (KM) curves using Digitizelt to extract time-to-event outcomes including overall survival (OS) and progression-free survival (PFS) and recover individual-level event and censoring information.
• Comprehensive Trial Metadata: Inclusion of detailed study-level information for each trial identified through the curation criteria.
• Support for Clinical Trial Design: Provision of empirical patient-level time-to-event datasets to inform disease- and treatment-class-specific trial design choices.
• Statistical Analysis Enhancement: Enabling estimation of operating characteristics for statistical designs and methods, including power and type I error rate calculations.

Scientific Applications:

• Clinical Trial Design: Use reconstructed patient-level data to model expected event trajectories and tailor trial sample size and endpoint selection.
• Statistical Method Evaluation: Assess operating characteristics of candidate statistical designs and procedures using empirical time-to-event data.
• Outcome Modeling: Improve modeling of OS and PFS distributions for simulation studies and comparative analyses across trials.

Methodology:

A systematic PubMed search identified relevant Phase III trials (2014–2016), selected trials were curated, and KM curves were digitized using Digitizelt to estimate patient-level time-to-event outcomes with censoring information.