The objective of this proposal is to provide new statistical methods, or modifications and clarifications of existing methods, related to the design and analysis of clinical or laboratory cancer studies conducted at the University of Wisconsin or at other cancer centers. Four research areas have been selected: 1. Methods for data monitoring; 2. Design of Phase 1 trials; 3. Analysis of longitudinal data; and 4. Survival analysis. Methods for data monitoring are useful in the proper early termination of trials with evidence for either benefit or harm or if there is little chance of obtaining significant results, thereby saving patient and financial resources. The previous grant has developed and studied group sequential methods as well as stochastic curtailment. The proposal is to further develop ongoing efforts. While Phase I trials are numerous, few statistical design methods are available and the most popular current design has not been fully evaluated. Given that our Cancer Center does numerous Phase I trials, we are interested in more efficient designs in obtaining the maximum tolerated dose, using sequential methods. Some methodological progress has already been achieved in our research. Many epidemiological and clinical trials make repeated measurements of patients overtime, often observing the rate of change. One such case is a pilot prevention trial in Stage I breast cancer evaluating the effect of tamoxifen on bone density and serum lipids. Existing statistical methods range from simple regression to more complex multivariate models. Further evaluation is needed as to the advantages, e.g., robustness and efficiency, of the complex models relative to simple regression. Survival analysis is a central method in biostatistics, especially for clinical research. While models such as the proportional hazards regression are commonly used, methods for verification of the model and diagnostics are needed. Such verification and the impact of various observations is highly desirable in order to have confidence in the analyses. Existing work by the investigators, using methods analogous to other regression techniques, needs to be further developed and software provided. Finally, issues related to the relative efficiency of the Cox proportional hazards model and alternatives to the KaplanMeier will be pursued.

National Institute of Health (NIH)
National Cancer Institute (NCI)
Research Project (R01)
Project #
Application #
Study Section
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Wisconsin Madison
Schools of Medicine
United States
Zip Code
Chen, Y H Joshua; DeMets, David L; Lan, K K Gordon (2004) Increasing the sample size when the unblinded interim result is promising. Stat Med 23:1023-38
Fleming, T R; DeMets, D L (1996) Surrogate end points in clinical trials: are we being misled? Ann Intern Med 125:605-13
Lindstrom, M J (1995) Self-modelling with random shift and scale parameters and a free-knot spline shape function. Stat Med 14:2009-21
Lee, J W; DeMets, D L (1995) Group sequential comparison of changes: ad-hoc versus more exact method. Biometrics 51:21-30
Lindstrom, M J; Kunugi, K A; Kinsella, T J (1993) Global comparison of radiation and chemotherapy dose-response curves with a test for interaction. Radiat Res 135:269-77
Kim, K; Demets, D L (1992) Sample size determination for group sequential clinical trials with immediate response. Stat Med 11:1391-9
Palta, M; Yao, T J (1991) Analysis of longitudinal data with unmeasured confounders. Biometrics 47:1355-69
Lindstrom, M L; Bates, D M (1990) Nonlinear mixed effects models for repeated measures data. Biometrics 46:673-87
Lan, K K; DeMets, D L (1989) Changing frequency of interim analysis in sequential monitoring. Biometrics 45:1017-20
Storer, B E (1989) Design and analysis of phase I clinical trials. Biometrics 45:925-37

Showing the most recent 10 out of 17 publications