Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Novel Adaptive Trial Designs for Studies with a Composite Endpoint of Morbidity and Mortality

Xu, Zhongying (2023) Novel Adaptive Trial Designs for Studies with a Composite Endpoint of Morbidity and Mortality. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

This is the latest version of this item.

[img] PDF
Restricted to University of Pittsburgh users only until 3 January 2025.

Download (825kB) | Request a Copy


Adaptive designs that allow for prospectively planned modifications during data collection have been widely used in clinical trials in recent years. Comparing to non-adaptive design, adaptive design may detect a true treatment effect with a lower sample size under the same statistical power. In this dissertation, I propose several novel adaptive trial designs for studies with a composite endpoint of morbidity and mortality.
In the first project, I developed a novel Bayesian response adaptive randomization (RAR) design for a composite endpoint of organ support free days (OSFD). I applied this method to design a multicenter, unblinded, phase II or III trial for treating sepsis patients admitted to the intensive care units with the endpoint of OSFD by Day 28. Comparing with other existing designs, non-adaptive or adaptive, the proposed method allocates more patients to the best performing treatment arm(s) and shows higher power.
In the second project, I extended the adaptation method in the first project to address the heterogeneity of treatment effects stemming from patients' baseline characteristics. Four sepsis phenotypes were considered as categorical covariates in this Bayesian covariate-adjusted response adaptive (CARA) design. I also extended win ratio method to the adaptive design and incorporated stratum-specific win ratios into the adaptive randomization (WR-CARA). Simulations showed that both Bayesian CARA and WR-CARA methods resulted in a higher proportion of patients assigned to the best performing treatment arm(s) in each sepsis phenotype when compared with the RAR method. Bayesian CARA had the highest statistical power among those methods compared because it best captures the underlying OSFD distribution. The WR-CARA approach is a good alternative when the underlying distribution is unknown.
Contribution to public health: All adaptive methods we proposed allocate more patients to the superior arm comparing with the existing methods. The two methods in the second project incorporate patients’ baseline characteristics in the design so that heterogenous treatment effects are taken into consideration. This dissertation will promote the development of adaptive designs and innovation for public health and medical research.


Social Networking:
Share |


Item Type: University of Pittsburgh ETD
Status: Unpublished
CreatorsEmailPitt UsernameORCID
Xu, Zhongyingzhx17@pitt.eduzhx17
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairChang, Chung-Chou H.changj@pitt.educhangj
Committee MemberBandos, Andriy I.anb61@pitt.eduanb61
Committee MemberTang, Lulutang@pitt.edulutang
Committee MemberTalisa, Victor B.VIT13@pitt.eduVIT13
Date: 3 January 2023
Date Type: Publication
Defense Date: 24 October 2022
Approval Date: 3 January 2023
Submission Date: 7 November 2022
Access Restriction: 2 year -- Restrict access to University of Pittsburgh for a period of 2 years.
Number of Pages: 80
Institution: University of Pittsburgh
Schools and Programs: School of Public Health > Biostatistics
Degree: PhD - Doctor of Philosophy
Thesis Type: Doctoral Dissertation
Refereed: Yes
Uncontrolled Keywords: Bayesian model, response adaptive design, covariate-adjusted response adaptive randomization, composite endpoint, win ratio
Related URLs:
Date Deposited: 03 Jan 2023 16:53
Last Modified: 03 Jan 2023 16:53

Available Versions of this Item

  • Novel Adaptive Trial Designs for Studies with a Composite Endpoint of Morbidity and Mortality. (deposited 03 Jan 2023 16:53) [Currently Displayed]


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item