Dataset Information

A Rank-Based Test for Comparison of Multidimensional Outcomes.

ABSTRACT: For comparison of multiple outcomes commonly encountered in biomedical research, Huang et al. (2005) improved O'Brien's (1984) rank-sum tests through the replacement of the ad hoc variance by the asymptotic variance of the test statistics. The improved tests control the Type I error rate at the desired level and gain power when the differences between the two comparison groups in each outcome variable fall into the same direction. However, they may lose power when the differences are in different directions (e.g., some are positive and some are negative). These tests and the popular Bonferroni correction failed to show important significant difference when applied to compare heart rates from a clinical trial to evaluate the effect of a procedure to remove the cardioprotective solution HTK. We propose an alternative test statistic, taking the maximum of the individual rank-sum statistics, which controls the type I error and maintains satisfactory power regardless of the directions of the differences. Simulation studies show the proposed test to be of higher power than other tests in certain alternative parameter space of interest. Furthermore, when used to analyze the heart rates data the proposed test yields more satisfactory results.

SUBMITTER: Liu A

PROVIDER: S-EPMC3102319 | biostudies-literature | 2010 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A Rank-Based Test for Comparison of Multidimensional Outcomes.

Liu Aiyi A Li Qizhai Q Liu Chunling C Yu Kai K Yu Kai F KF

Journal of the American Statistical Association 20100601 490

For comparison of multiple outcomes commonly encountered in biomedical research, Huang et al. (2005) improved O'Brien's (1984) rank-sum tests through the replacement of the ad hoc variance by the asymptotic variance of the test statistics. The improved tests control the Type I error rate at the desired level and gain power when the differences between the two comparison groups in each outcome variable fall into the same direction. However, they may lose power when the differences are in differen ...[more]

PMID: 21625372

Similar Datasets

Project description:ImportanceLittle is known about the performance of available frailty instruments in estimating patient-relevant outcomes after cardiac surgery.ObjectiveTo examine how well the Johns Hopkins Adjusted Clinical Groups (ACG) frailty indicator, the Hospital Frailty Risk Score (HFRS), and the Preoperative Frailty Index (PFI) estimate long-term patient-centered outcomes after cardiac surgery.Design, setting, and participantsThis retrospective cohort study was conducted in Ontario, Canada, among residents 18 years and older who underwent coronary artery bypass grafting or aortic, mitral or tricuspid valve, or thoracic aorta surgery between October 2008 and March 2017. Long-term care residents, those with discordant surgical encounters, and those receiving dialysis or dependent on a ventilator within 90 days were excluded. Statistical analysis was conducted from July 2021 to January 2022.Main outcomes and measuresThe primary outcome was patient-defined adverse cardiovascular and noncardiovascular events (PACE), defined as the composite of severe stroke, heart failure, long-term care admission, new-onset dialysis, and ventilator dependence. Secondary outcomes included mortality and individual PACE events. The association between frailty and PACE was examined using cause-specific hazard models with death as a competing risk, and the association between frailty and death was examined using Cox models. Areas under the receiver operating characteristic curve (AUROC) were determined over 10 years of follow-up for each frailty instrument.ResultsOf 88 456 patients (22 924 [25.9%] female; mean [SD] age, 66.3 [11.1] years), 14 935 (16.9%) were frail according to ACG criteria, 63 095 (71.3%) according to HFRS, and 76 754 (86.8%) according to PFI. Patients with frailty were more likely to be older, female, and rural residents; to have lower income and multimorbidity; and to undergo urgent surgery. Patients meeting ACG criteria (hazard ratio [HR], 1.66; 95% CI, 1.60-1.71) and those with higher HFRS scores (HR per 1.0-point increment, 1.10; 95% CI, 1.09-1.10) and PFI scores (HR per 0.1-point increment, 1.75; 95% CI, 1.73-1.78) had higher rates of PACE. Similar magnitudes of association were observed for each frailty instrument with death and individual PACE components. The HFRS had the highest AUROC for estimating PACE during the first 2 years and death during the first 4 years, after which the PFI had the highest AUROC.Conclusions and relevanceThese findings could help to tailor the use of frailty instruments by outcome and follow-up duration, thus optimizing preoperative risk stratification, patient-centered decision-making, candidate selection for prehabilitation, and personalized monitoring and health resource planning in patients undergoing cardiac surgery.

Dataset Information

A Rank-Based Test for Comparison of Multidimensional Outcomes.

Publications

A Rank-Based Test for Comparison of Multidimensional Outcomes.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets