Dataset Information

Assessing potentially time-dependent treatment effect from clinical trials and observational studies for survival data, with applications to the Women's Health Initiative combined hormone therapy trial.

ABSTRACT: For risk and benefit assessment in clinical trials and observational studies with time-to-event data, the Cox model has usually been the model of choice. When the hazards are possibly non-proportional, a piece-wise Cox model over a partition of the time axis may be considered. Here, we propose to analyze clinical trials or observational studies with time-to-event data using a certain semiparametric model. The model allows for a time-dependent treatment effect. It includes the important proportional hazards model as a sub-model and can accommodate various patterns of time-dependence of the hazard ratio. After estimation of the model parameters using a pseudo-likelihood approach, simultaneous confidence intervals for the hazard ratio function are established using a Monte Carlo method to assess the time-varying pattern of the treatment effect. To assess the overall treatment effect, estimated average hazard ratio and its confidence intervals are also obtained. The proposed methods are applied to data from the Women's Health Initiative. To compare the Women's Health Initiative clinical trial and observational study, we use the propensity score in building the regression model. Compared with the piece-wise Cox model, the proposed model yields a better model fit and does not require partitioning of the time axis.

SUBMITTER: Yang S

PROVIDER: S-EPMC4393788 | biostudies-literature | 2015 May

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Assessing potentially time-dependent treatment effect from clinical trials and observational studies for survival data, with applications to the Women's Health Initiative combined hormone therapy trial.

Yang Song S Prentice Ross L RL

Statistics in medicine 20150217 11

For risk and benefit assessment in clinical trials and observational studies with time-to-event data, the Cox model has usually been the model of choice. When the hazards are possibly non-proportional, a piece-wise Cox model over a partition of the time axis may be considered. Here, we propose to analyze clinical trials or observational studies with time-to-event data using a certain semiparametric model. The model allows for a time-dependent treatment effect. It includes the important proportio ...[more]

PMID: 25689356

Similar Datasets

Project description:ObjectiveThe Women's Health Initiative hormone trials identified a 44% increase in ischemic stroke risk with combination estrogen plus progestin and a 39% increase with estrogen alone. We undertook a case-control biomarker study to elucidate underlying mechanisms, and to potentially identify women who would be at lower or higher risk for stroke with postmenopausal hormone therapy (HT).DesignThe hormone trials were randomized, double-blind, and placebo controlled.SettingThe Women's Health Initiative trials were conducted at 40 clinical centers in the United States.ParticipantsThe trials enrolled 27,347 postmenopausal women, aged 50-79 y.InterventionsWe randomized 16,608 women with intact uterus to conjugated estrogens 0.625 mg with medroxyprogesterone acetate 2.5 mg daily or placebo, and 10,739 women with prior hysterectomy to conjugated estrogens 0.625 mg daily or placebo.Outcome measuresStroke was ascertained during 5.6 y of follow-up in the estrogen plus progestin trial and 6.8 y of follow-up in the estrogen alone trial.ResultsNo baseline clinical characteristics, including gene polymorphisms, identified women for whom the stroke risk from HT was higher. Paradoxically, women with higher baseline levels of some stroke-associated biomarkers had a lower risk of stroke when assigned to estrogen plus progestin compared to placebo. For example, those with higher IL-6 were not at increased stroke risk when assigned to estrogen plus progestin (odds ratio 1.28) but were when assigned to placebo (odds ratio 3.47; p for difference = 0.02). Similar findings occurred for high baseline PAP, leukocyte count, and D-dimer. However, only an interaction of D-dimer during follow-up interaction with HT and stroke was marginally significant (p = 0.03).ConclusionsBiomarkers did not identify women at higher stroke risk with postmenopausal HT. Some biomarkers appeared to identify women at lower stroke risk with estrogen plus progestin, but these findings may be due to chance.

Project description:BackgroundCase-control studies have reported that exogenous estrogen use is associated with increased risk of skin cancer. The effects of menopausal hormone therapy on incidence of nonmelanoma skin cancer and melanoma were evaluated in post hoc analyses of the Women's Health Initiative randomized placebo-controlled hormone therapy trials of combined estrogen plus progestin (E + P) and estrogen only (E-alone).MethodsPostmenopausal women aged 50-79 years were randomly assigned to conjugated equine estrogen (0.625 mg/d) plus medroxyprogesterone acetate (2.5 mg/d) or placebo in the E + P trial if they had an intact uterus (N = 16,608) or to conjugated equine estrogen alone or placebo in the E-alone trial if they had a hysterectomy (N = 10,739); the mean follow-up was 5.6 and 7.1 years, respectively. Incident nonmelanoma skin cancers (n = 980 [E + P trial]; n = 820 [E-alone trial]) and melanomas (n = 57 [E + P trial]; n =38 [E-alone trial]) were ascertained by self-report. Incident cases of cutaneous malignant melanoma were confirmed by physician review of medical records. Incidences of nonmelanoma skin cancer and melanoma were compared between the two randomization groups within each trial using hazard ratios (HRs), with corresponding 95% confidence intervals (CIs) and Wald statistic P values from Cox proportional hazards models. All statistical tests were two-sided.ResultsRates of incident nonmelanoma skin cancer and melanoma were similar between the active hormone (combined analysis of E + P and E-alone) and placebo groups (nonmelanoma skin cancer: HR = 0.98, 95% CI = 0.89 to 1.07; melanoma: HR = 0.92, 95% CI = 0.61 to 1.37). Results were similar for the E + P and E-alone trials when analyzed individually.ConclusionsMenopausal hormone therapy did not affect overall incidence of nonmelanoma skin cancer or melanoma. These findings do not support a role of menopausal estrogen, with or without progestin, in the development of skin cancer in postmenopausal women.

Dataset Information

Assessing potentially time-dependent treatment effect from clinical trials and observational studies for survival data, with applications to the Women's Health Initiative combined hormone therapy trial.

Publications

Assessing potentially time-dependent treatment effect from clinical trials and observational studies for survival data, with applications to the Women's Health Initiative combined hormone therapy trial.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets