Importance of events per independent variable in proportional hazards regression analysis II. Accuracy and precision of regression estimates

Peduzzi, Peter; Concato, John; Feinstein, Alvan R.; Holford, Theodore R.

doi:10.1016/0895-4356(95)00048-8

Cited by 1,688 publications

(1,174 citation statements)

References 3 publications

Supporting

Mentioning

1,098

Contrasting

Unclassified

Order By: Relevance

“…Assessment of EFS and OS outcomes is therefore descriptive, and p-values are presented as measures of variability and not for inferential purposes. These were pre-planned analysis even if the number of expected events is low (10,11). Differences in EFS and OS between the trastuzumab group and each of the lapatinib-containing groups are described using hazard ratios and 95% confidence intervals with p-values from two-sided stratified log-rank tests, implemented as Wald tests from the Cox models (12).…”

Section: Resultsmentioning

confidence: 99%

Lapatinib with trastuzumab for HER2-positive early breast cancer (NeoALTTO): survival outcomes of a randomised, open-label, multicentre, phase 3 trial and their association with pathological complete response

Azambuja

Holmes

Piccart‐Gebhart

et al. 2014

The Lancet Oncology

398

290

View full text Add to dashboard Cite

Section: Resultsmentioning

confidence: 99%

Lapatinib with trastuzumab for HER2-positive early breast cancer (NeoALTTO): survival outcomes of a randomised, open-label, multicentre, phase 3 trial and their association with pathological complete response

Azambuja

Holmes

Piccart‐Gebhart

et al. 2014

The Lancet Oncology

398

290

View full text Add to dashboard Cite

“…Univariate and multivariate regression models adjusting for confounding factors were not fit given the small number of observed events (recurrent stroke) in the lowest tertile. 25 A total sample size of 180 subjects with a maximum follow-up of 12 months was targeted to achieve 480% power to detect a difference in the time-to-recurrence distributions between pairs of groups defined by tertiles of coated-platelet levels, where the lowest and highest risk groups have a 12-month recurrence rate of 2% and 25%, respectively, using a logrank test and a two-sided 0.0167 a level. The event rate estimates were based on our previously published research data on coated-platelet levels in ischemic stroke.…”

Section: Statistical Analysesmentioning

confidence: 99%

Higher Coated-Platelet Levels are Associated with Stroke Recurrence following Nonlacunar Brain Infarction

Prodan

Stoner

Cowan

et al. 2012

J Cereb Blood Flow Metab

View full text Add to dashboard Cite

Coated-platelets are procoagulant platelets observed upon dual-agonist stimulation with collagen and thrombin. Coated-platelet levels are elevated in patients with nonlacunar (large-vessel) ischemic stroke and decreased in patients with spontaneous intracerebral hemorrhage as compared with controls. The purpose of this study was to investigate a possible relationship between coated-platelet levels and stroke recurrence in patients with nonlacunar ischemic stroke. We assayed coated-platelet levels in 190 consecutive patients with nonlacunar stroke who were followed for up to 12 months; 20 subjects experienced recurrent stroke. Subjects were categorized into tertiles of coated-platelet levels. The distributions of time-to-recurrent stroke were estimated for each tertile using cumulative incidence curves and compared statistically using a log-rank test. The cumulative incidence of recurrent stroke at 12 months differed among the coated-platelet tertiles: 2% for the first tertile (lowest coated-platelet levels), 18% for the second tertile, and 17% for the third tertile (overall log-rank test, P ¼ 0.019). These data suggest that higher levels of coated-platelets, measured shortly after a nonlacunar stroke, are associated with an increased incidence of stroke recurrence. This observation offers an additional tool for identifying patients at highest risk for stroke recurrence following a nonlacunar (large-vessel) infarct.

show abstract

“…Simulation work has suggested that at least 10 events need to be observed for each covariate considered, and anything less will lead to problems, for example, the regression coefficients become biased (Peduzzi et al, 1995). In the ovarian study, there were 550 deaths and 11 covariates for the five prognostic factors, implying 50 events per covariate.…”

Section: Sample Size Considerationsmentioning

confidence: 99%

Survival Analysis Part III: Multivariate data analysis – choosing a model and assessing its adequacy and fit

et al. 2003

View full text Add to dashboard Cite

Importance of events per independent variable in proportional hazards regression analysis II. Accuracy and precision of regression estimates

Cited by 1,688 publications

References 3 publications

Lapatinib with trastuzumab for HER2-positive early breast cancer (NeoALTTO): survival outcomes of a randomised, open-label, multicentre, phase 3 trial and their association with pathological complete response

Lapatinib with trastuzumab for HER2-positive early breast cancer (NeoALTTO): survival outcomes of a randomised, open-label, multicentre, phase 3 trial and their association with pathological complete response

Higher Coated-Platelet Levels are Associated with Stroke Recurrence following Nonlacunar Brain Infarction

Survival Analysis Part III: Multivariate data analysis – choosing a model and assessing its adequacy and fit

Contact Info

Product

Resources

About