16 Cochran-mantel-haenszel techniques: Applications involving epidemiologic survey data

Hall, Daniel B.; Woolson, Robert F.; Clarke, William R.; Jones, Martha F.

doi:10.1016/s0169-7161(00)18018-6

Cited by 9 publications

(9 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The primary efficacy hypothesis was assessed with the use of the Cochran–Mantel–Haenszel test, with adjustment for the dichotomized baseline NIHSS score, and the weights of the Cochran–Mantel–Haenszel test were applied in the estimation of the risk difference. 25 At both the interim and the final analyses, an unfavorable outcome (defined as a modified Rankin score of >2) was imputed for participants who had missing data for the primary outcome or for whom data on the primary outcome were obtained outside the specified window. For all analyses of predefined secondary outcomes and subgroup and safety analyses, each test was conducted at a two-sided alpha level of 0.01.…”

Section: Methodsmentioning

confidence: 99%

Endovascular Therapy after Intravenous t-PA versus t-PA Alone for Stroke

Palesch²,

et al. 2013

View full text Add to dashboard Cite

BACKGROUND Endovascular therapy is increasingly used after the administration of intravenous tissue plasminogen activator (t-PA) for patients with moderate-to-severe acute ischemic stroke, but whether a combined approach is more effective than intravenous t-PA alone is uncertain. METHODS We randomly assigned eligible patients who had received intravenous t-PA within 3 hours after symptom onset to receive additional endovascular therapy or intravenous t-PA alone, in a 2:1 ratio. The primary outcome measure was a modified Rankin scale score of 2 or less (indicating functional independence) at 90 days (scores range from 0 to 6, with higher scores indicating greater disability). RESULTS The study was stopped early because of futility after 656 participants had undergone randomization (434 patients to endovascular therapy and 222 to intravenous t-PA alone). The proportion of participants with a modified Rankin score of 2 or less at 90 days did not differ significantly according to treatment (40.8% with endovascular therapy and 38.7% with intravenous t-PA; absolute adjusted difference, 1.5 percentage points; 95% confidence interval [CI], −6.1 to 9.1, with adjustment for the National Institutes of Health Stroke Scale [NIHSS] score [8–19, indicating moderately severe stroke, or ≥20, indicating severe stroke]), nor were there significant differences for the predefined subgroups of patients with an NIHSS score of 20 or higher (6.8 percentage points; 95% CI, −4.4 to 18.1) and those with a score of 19 or lower (−1.0 percentage point; 95% CI, −10.8 to 8.8). Findings in the endovascular-therapy and intravenous t-PA groups were similar for mortality at 90 days (19.1% and 21.6%, respectively; P = 0.52) and the proportion of patients with symptomatic intracerebral hemorrhage within 30 hours after initiation of t-PA (6.2% and 5.9%, respectively; P = 0.83). CONCLUSIONS The trial showed similar safety outcomes and no significant difference in functional independence with endovascular therapy after intravenous t-PA, as compared with intravenous t-PA alone. (Funded by the National Institutes of Health and others; ClinicalTrials.gov number, NCT00359424.)

show abstract

Section: Methodsmentioning

confidence: 99%

Endovascular Therapy after Intravenous t-PA versus t-PA Alone for Stroke

Palesch²,

et al. 2013

View full text Add to dashboard Cite

show abstract

“…This paper discusses implementation in R of some newly developed and already well known non-parametric and goodness-of-fit tests and procedures that are utilized in litigations, environmental studies and biostatistics. Besides of the methods discussed in the paper, lawstat contains such procedures as the Brunner-Munzel test, which is also known as the generalized Wilcoxon test (Brunner and Munzel, 2000); the Cochran-Mantel-Haenszel test (Agresti 2002;Gastwirth 1984;Hall, Woolson, Clarke, and Jones 1999); the Lorenz curve, the coefficient of dispersion and the Gini index (Gastwirth 1988;Bonett and Seier 2008), as well as a number of illustrative data sets from law cases, environmental and archeological studies. We plan to constantly update lawstat with new statistical methods and tests.…”

Section: Resultsmentioning

confidence: 99%

lawstat: AnRPackage for Law, Public Policy and Biostatistics

Hui¹,

Gel²,

Gastwirth³

2008

J. Stat. Soft.

View full text Add to dashboard Cite

We present a new R software package lawstat that contains statistical tests and procedures that are utilized in various litigations on securities law, antitrust law, equal employment and discrimination as well as in public policy and biostatistics. Along with the well known tests such as the Bartels test, runs test, tests of homogeneity of several sample proportions, the Brunner-Munzel tests, the Lorenz curve, the Cochran-Mantel-Haenszel test and others, the package contains new distribution-free robust tests for symmetry, robust tests for normality that are more sensitive to heavy-tailed departures, measures of relative variability, Levene-type tests against trends in variances etc. All implemented tests and methods are illustrated by simulations and real-life examples from legal cases, economics and biostatistics. Although the package is called lawstat, it presents implementation and discussion of statistical procedures and tests that are also employed in a variety of other applications, e.g., biostatistics, environmental studies, social sciences and others, in other words, all applications utilizing statistical data analysis. Hence, name of the package should not be considered as a restriction to legal statistics. The package will be useful to applied statisticians and "quantitatively alert practitioners" of other subjects as well as an asset in teaching statistical courses.

show abstract

“…Although all of these techniques are standard in the context of simple random sampling, the stratified sampling design of the study required that adjustments to these methods be utilized. Although these adjustments were available for most methods, it was necessary to develop methods to compute the CMH adjusted rate differences [63,64]. All programming pertinent to implementation and completion of these goals was completed before June 1996, when final data from the interviews were received.…”

Section: Methodsmentioning

confidence: 99%

Methodologic issues in a population-based health survey of Gulf War veterans

Doebbeling

Jones

Hall

et al. 2002

Journal of Clinical Epidemiology

Self Cite

View full text Add to dashboard Cite

This report describes the principal methods used in the development, conduct, and analysis of the research study "Health Assessment of Persian Gulf War Veterans from Iowa" (Iowa Gulf War Study). The methods presented include an outline of the organizational structure, study timeline, hypotheses, outcome definitions, and study design. Adhering to a strict timeline, the study protocol and instruments were developed, and a stratified sample of 3,695 military personnel (76% participation) was located and surveyed by structured telephone interview. The study tracked personnel from all service branches residing nationally and internationally, including those discharged from service. This study required development and implementation of methods appropriate to analysis of data collected in a complex sampling framework and methodological procedures to ensure scientific rigor in a highly public and politicized environment. Statistical analyses were conducted on a priori health outcomes and required development of methods to compute Cochran-Mantel-Haenszel adjusted rate differences. This environment facilitated rapid implementation, critique by scientific and public advisors, a high participation rate, and rapid publication. AbstractThis report describes the principal methods used in the development, conduct, and analysis of the research study "Health Assessment of Persian Gulf War Veterans from Iowa" (Iowa Gulf War Study). The methods presented include an outline of the organizational structure, study timeline, hypotheses, outcome definitions, and study design. Adhering to a strict timeline, the study protocol and instruments were developed, and a stratified sample of 3,695 military personnel (76% participation) was located and surveyed by structured telephone interview. The study tracked personnel from all service branches residing nationally and internationally, including those discharged from service. This study required development and implementation of methods appropriate to analysis of data collected in a complex sampling framework and methodological procedures to ensure scientific rigor in a highly public and politicized environment. Statistical analyses were conducted on a priori health outcomes and required development of methods to compute Cochran-Mantel-Haenszel adjusted rate differences. This environment facilitated rapid implementation, critique by scientific and public advisors, a high participation rate, and rapid publication.

show abstract

16 Cochran-mantel-haenszel techniques: Applications involving epidemiologic survey data

Cited by 9 publications

References 38 publications

Endovascular Therapy after Intravenous t-PA versus t-PA Alone for Stroke

Endovascular Therapy after Intravenous t-PA versus t-PA Alone for Stroke

lawstat: AnRPackage for Law, Public Policy and Biostatistics

Methodologic issues in a population-based health survey of Gulf War veterans

Contact Info

Product

Resources

About