A logistic dose-ranging method for phase I clinical investigations trials

This paper describes an alternative to the continual reassessment method (CRM) for phase I trials. The logistic dose ranging strategy (LDRS) uses logistic regression and a dose allocation scheme similar to the CRM. It can easily be implemented from any logistic regression program. The LDRS can be a stand alone dose allocation scheme or it can be incorporated into standard three on a dose strategies to indicate when escalation can proceed more rapidly. Finally, the effect of covariates such as age or comorbid conditions on the toxicity expected for the dose selected for a phase II trial can be examined.