Intent to Sole Source - Quantifying Prediction Error in Two-time Scale Competing Risks Models

The National Cancer Institute (NCI) intends to award a sole‑source, firm‑fixed‑price purchase order to develop and validate predictive‑accuracy measures for continuous‑time competing‑risks survival models that operate on two time scales (time since diagnosis for cancer mortality and age for other‑cause mortality). The contractor will create methodology, implement reproducible software (in R, SAS, or Stata), produce a technical report, and draft a manuscript for journal publication. The effort spans a 12‑month base period with an optional additional 12‑month extension, and deliverables include methodology documentation, software code, and manuscript drafts. • Develop statistical methodology for prediction error in two‑time‑scale competing‑risks models. • Build reproducible software tools to compute the new error measures. • Produce a technical report and a peer‑review‑ready manuscript summarizing findings. • Base period of 12 months (option for another 12 months). • Sole‑source award to Jason P. Fine; however, other qualified parties may submit capability statements by the deadline.