PubMed Health⌕ Search
Home/Archive/PubMed

PubMed · 7129743

An empirical study of cluster randomization.

A Donner
View original source ↗Export metadata ↓

Abstract

Increasing attention has been given recently to the methodological issues associated with randomization of clusters rather than individuals in lifestyle intervention trials. These issues are explored through an empirical study of the 'effective sample size' imposed by randomization of three experimental units frequently considered in epidemiological research: the spouse pair, the general practice, and the large geographic area (county). The measurement of within-cluster dependence for a dichotomous outcome variable is also discussed, and a relationship shown between Fleiss's kappa and Cornfield's inflation factor.

Explore related subjects

EpidemiologyHumansRandom AllocationResearchResearch DesignStatistics as TopicJournal ArticleResearch Support, Non-U.S. Gov't

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

A Donner. 1982. An empirical study of cluster randomization.. https://doi.org/10.1093/ije%2F11.3.283

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2026

Incidence patterns and genetic validation of primary glaucoma subtypes among 1 million adults in China and the UK.

BACKGROUND/AIMS: Primary open-angle glaucoma (POAG) and primary angle-closure glaucoma (PACG) are distinct diseases, yet many glaucoma cases in population-based datasets lack subtype specification. We assessed incidence patterns of glaucoma subtypes in China and the UK and used genetic evidence to infer the likely subtype composition of cases recorded as unspecified glaucoma. METHODS: Incident primary glaucoma was identified from linked inpatient records in the prospective China Kadoorie Biobank (CKB; n=512 504) and UK Biobank (UKB; n=492 329) studies. Cohort-specific phenotyping algorithms defined POAG, PACG and unspecified glaucoma. Adjusted incidence rates were estimated by direct standardisation. To support subtype inference, polygenic risk scores (PRSs) were constructed using ancestry-specific genome-wide association studies, including a new East Asian PACG meta-analysis, and tested for association with glaucoma phenotypes using multivariable logistic regression. RESULTS: Over 12 years of follow-up, 1658 primary glaucoma cases were identified in CKB and 7643 in UKB. Most (>68%) cases lacked subtype specification. Incidence increased with age and was twofold higher among women for PACG in both cohorts and for unspecified glaucoma in CKB. In UKB, POAG incidence was fivefold higher among Black than White participants, with a similar but attenuated pattern for unspecified glaucoma. PRS analyses indicated that unspecified glaucoma closely aligned with PACG in CKB but was more heterogeneous in UKB. CONCLUSION: Healthcare-recorded incidence patterns for POAG and PACG were consistent with established demographic risk factors, whereas unspecified glaucoma showed differences in subtype composition between populations. Integrating epidemiological and genetic evidence improves interpretation of glaucoma phenotypes when detailed clinical information is unavailable.

Epidemiology↗
PubMed2026

Neonatal outcomes following antenatal corticosteroid administration before planned caesarean birth at late preterm or early term (36-38 weeks' gestation): a retrospective cohort study.

OBJECTIVE: To investigate the benefits and harms associated with antenatal corticosteroids (ACS) prior to planned caesarean birth at late preterm or early term. DESIGN: Retrospective cohort study. SETTING: South Australia, Australia, 2005-2018. PATIENTS: Women with singleton pregnancies undergoing planned caesarean birth between 36+0 and 38+6 weeks' gestation. METHODS: We used routinely collected electronic maternity and neonatal data. Adjusted risk differences (aRD) with 95% CIs were calculated for neonatal and maternal outcomes. OUTCOME MEASURES: Neonatal outcomes included respiratory distress requiring oxygen, hypoglycaemia requiring glucose, neonatal unit (NNU) admission and prolonged NNU admission (≥48 hours). RESULTS: Among the cohort of 4049 women, the proportion receiving ACS was 54.8%, 48.4% and 5.6% at 36, 37 and 38 completed weeks' gestation, respectively. ACS administration was associated with a significant reduction in respiratory distress across all gestations, with the greatest reduction seen at 36 weeks' gestation (aRD -12.5%; 95% CI -21.4% to -3.6%). At 38 weeks' gestation, ACS administration was associated with an increased risk of neonatal hypoglycaemia (aRD 5.7%; 95% CI 1.1% to 10.2%), NNU admission (aRD 5.4%; 95% CI 0.1% to 10.7%) and prolonged NNU admission (aRD 4.6%; 95% CI 0.2% to 8.9%). CONCLUSIONS: ACS administration prior to late preterm and early term planned caesarean birth was associated with reduced risk of respiratory distress. The net benefit of ACS administration before planned caesarean birth remains unclear; however, there was evidence of potential neonatal harm when birth occurred after 37 weeks' gestation.

Epidemiology↗
PubMed2006

The full bibliography of Ernst Ludwig Wynder.

The chronology of lifetime publications authored and/or co-authored by Dr. Ernst L. Wynder spans a rich assortment of topics relating to the etiology and prevention of chronic diseases. Dr. Wynder's interdisciplinary approach to research in cancer and heart disease embraces the disciplines of epidemiology, biology, pathology, chemistry, biochemistry, tobacco sciences, and nutrition as well as health behavior, and health education. His works reflect his attention to diverse subjects from methodology in scientific endeavors, to mechanisms in carcinogenesis, brain development, alternative medicine, principles of disease prevention, history of medicine, and societal issues, as well as philosophy, and religion. In 1972 Dr. Wynder created the journal Preventive Medicine as an international forum for the exchange of ideas that address the promotion of health and prevention of disease in our time. Many of his ideas were developed and nurtured in the symposia and workshops the Journal has published.

Epidemiology↗