Kathleen F. Kerr Software

Home Research Software Students Assorted Personal Stuff

Evaluating Biomarkers for Prognostic Enrichment of Clinical Trials

BioPET: Biomarker Prognostic Enrichment Tool. This R package evaluates the utility of a biomarker for prognostic enrichment of a clinical trial with a dichotomouse primary endpoint. Biomarkers are evaluated along several dimensions including the trial sample size, calendar time to enroll the trial, and total of patient costs including the cost of running a patient through the trial and the cost of biomarker screening. (Authors: Jeremy Roth, Kathleen Kerr, Kehao Zhu)
BioPETsurv extends the functionality of BioPET to clinical trials with time-to-event endpoints, such as survival endpoint. Trials can have fixed duration or an accural period plus a follow-up period. Per patient costs can be fixed or depend on the length of time a patient is in the trial. (Author: Si Cheng)
BioPET and BioPETsurv webtools allow investigators to explore the utility of prognostic enrichment for their clinical setting. (Authors: Jeremy Roth, Si Cheng)

References: Kerr KF, Roth J, Zhu K, Thiessen-Philbrook H, Meisner A, Wilson FP, Coca S, Parikh C. Evaluating biomarkers for prognostic enrichment of clinical trials Clinical Trials, 2017. PubMed
Cheng S, Kerr KF, Thiessen-Philbrook H, Coca S, Parikh C. A comprehensive framework for evaluating biomarkers for prognostic enrichment of clinical trials with time-to-event endpoints. Submitted.

Evaluating Biomarkers for Risk-Based Treatment Decisions

rmda: Risk Model Decision Analysis. This R package supports Decision Curves and Relative Utility Curves and supports the evaluating of risk models for both "opt in" and "opt out" treatment policies. (Author: Marshall Brown)

Reference: Kerr KF, Brown MD, Zhu K, Janes H. Assessing the Clinical Impact of Risk Prediction Models With Decision Curves: Guidance for Correct Interpretation and Appropriate Use. Journal of Clinical Oncology, 2016. PubMed
Kerr KF, Brown MD, Marsh TL, Janes H. Assessing the Clinical Impact of Risk Models for Opting Out of Treatment. Medical Decision Making, 2019. PubMed

Recalibrating a Risk Model for Risk-Based Treatment Decisions

ClinicalUtilityRecal This R package implements methods for recalibrating a risk model when a primary purpose of the risk model is for risk-based decision-making. (Author: Anu Mishra)

Reference: Mishra A, McClelland RL, Inoue LYT, Kerr KF. Recalibration Methods for Improved Clinical Utility of Risk Scores. Submitted.

Evaluating the Prognostic Capacity of Cancer Staging Systems

R script for the example in the paper (Table 1)
R function for generating examples with exponential survival data.

Reference: Kerr KF, LeBlanc M, Janes H. Comparisons of cancer staging systems should be based on overall performance in the population. Clinical Trials, 2017. PubMed

Developing Biomarker Combinations for a Binary Outcome

maxTPR: Maximizing the TPR for a Specified FPR. This R package estimates a linear combination of biomarkers or other predictors by maximizing a smooth approximation to the estimated true positive rate (TPR; sensitivity) while constraining a smooth approximation to the estimated false positive rate (FPR; 1-specificity). (Author: Allison Meisner)

Reference: Meisner A, Carone M, Pepe M, Kerr KF. Combining Biomarkers by Maximizing the True Positive Rate for a Fixed False Positive Rates. Submitted.

Selecting Biomarker Combinations for an Ordinal Outcome

multiselect: Selecting Combinations of Predictors by Leveraging Multiple AUCs for an Ordered Multilevel Outcome. When an outcome is ordinal but there is interest in discriminating one particular level from others, this package uses information from multiple levels to identify combinations of predictors. (Author: Allison Meisner)

Reference: Meisner A, Parikh CR, Kerr KF. Using ordinal outcomes to construct and select biomarker combinations for single-level prediction. Diagnostic and Prognostic Research, 2018.

Combining Biomarkers to maximize the covariate-adjusted AUC

maxadjAUC: Maximizing the Adjusted AUC. Fit a linear combination of predictors by maximizing a smooth approximation to the covariate-adjusted AUC (area under the ROC curve). (Author: Allison Meisner)

Reference: Meisner A, Parikh CR, Kerr KF. Developing Biomarker Combinations in Multicenter Studies via Direct Maximization and Penalization. Submitted.

Department of Biostatistics | 206-543-1044 | biostat@uw.edu | Terms | Privacy