Article
Health Care Sciences & Services
Richard Hooper
Summary: A stepped wedge trial is an evaluation method of implementing interventions over a staggered timetable, usually cluster randomized, with various designs and conduct practices. The analysis of this trial assumes a constant intervention effect and the need to adjust for period effects, underlying trends, and clustering of outcomes. The advantage lies in practicality or statistical efficiency.
JOURNAL OF CLINICAL EPIDEMIOLOGY
(2021)
Article
Mathematical & Computational Biology
Phillip Sundin, Catherine M. Crespi
Summary: This article presents a linear mixed model for stepped wedge design (SWD) trials with two or more interventions, including multiarm and factorial designs. The standard errors of the intervention effect coefficients are derived, and power calculation methods are provided. The impact of design features, especially treatment allocations, on power is examined.
STATISTICS IN MEDICINE
(2022)
Article
Clinical Neurology
Fan Li, Rui Wang
Summary: This study provides an update on the design of stepped wedge randomized trials, including new methodological tools, for neurosurgeon scientists. The study highlights the indications for this design and discusses key considerations such as analysis methods and sample size determination.
WORLD NEUROSURGERY
(2022)
Article
Health Care Sciences & Services
Kelsey L. Grantham, Andrew B. Forbes, Richard Hooper, Jessica Kasza
Summary: This article introduces the 'staircase' design, which is a new experimental design method based on the zigzag pattern of steps. It can alleviate the burden on participating clusters, promote trial involvement, and reduce attrition. Although there is a lack of consistent methodology, we provide variance expressions for sample size and power calculations for the staircase designs.
STATISTICAL METHODS IN MEDICAL RESEARCH
(2023)
Article
Mathematical & Computational Biology
Fan Li, Xinyuan Chen, Zizhong Tian, Rui Wang, Patrick J. Heagerty
Summary: This article proposes novel variance formulas for analyzing heterogeneity of treatment effects in stepped wedge designs, applicable to both average treatment effect and subgroup treatment effects. The study also investigates optimal design allocations of clusters to maximize precision for evaluating both average and heterogeneous treatment effects.
STATISTICS IN MEDICINE
(2023)
Article
Health Care Sciences & Services
Clement Ma, Alina Lee, Darren Courtney, David Castle, Wei Wang
Summary: This study investigates the effects of balancing cluster-level characteristics on the estimation of treatment effects in stepped-wedge cluster randomized trials (SWCRTs). It is found that fully-balanced designs show higher efficiency and reduce bias in treatment effect estimation. Therefore, pre-balancing cluster-level characteristics is of significance in SWCRTs.
BMC MEDICAL RESEARCH METHODOLOGY
(2023)
Article
Mathematical & Computational Biology
Dustin J. Rabideau, Rui Wang
Summary: This article proposes a framework for calculating randomization-based p-values and CIs for marginal treatment effects in SWT, using test statistics derived from individual-level generalized linear models. It also explores how study design features impact various SWT analysis methods. Data from the XpertMTB/RIF tuberculosis trial are reanalyzed to illustrate the method and compare it to alternatives.
STATISTICS IN MEDICINE
(2021)
Review
Economics
Thomas Lung, Lei Si, Richard Hooper, Gian Luca Di Tanna
Summary: This study summarized 69 studies where 34 did not report any statistical methods for economic evaluation, while only 16 studies reported appropriate methods using mixed/multilevel models. The use of appropriate statistical methods is crucial for conducting health economics analysis in stepped wedge trials.
Article
Health Care Sciences & Services
J. A. Thompson, K. Hemming, A. Forbes, K. Fielding, R. Hayes
Summary: Generalised estimating equations with the sandwich standard-error estimator are a promising method for stepped wedge cluster randomised trials, but may have inflated type-one errors with a small number of clusters. A large simulation study comparing bias-corrected standard errors from different methods found that Fay and Graubard, as well as an approximation to Kauermann and Carroll, showed unbiased results with an independent working correlation matrix and 12 or more clusters.
STATISTICAL METHODS IN MEDICAL RESEARCH
(2021)
Article
Medicine, General & Internal
Grant M. Russell, Katrina Long, Virginia Lewis, Joanne C. Enticott, Nilakshi Gunatillaka, I--Hao Cheng, Geraldine Marsh, Shiva Vasi, Jenny Advocat, Shoko Saito, Hyun Song, Sue Casey, Mitchell Smith, Mark F. Harris
Summary: The study aimed to investigate whether primary care outreach facilitation could enhance the quality of care for general practice patients from refugee backgrounds. The results showed that the intervention led to an increase in the proportion of refugee patients receiving Medicare-billed health assessments, but had limited impact on other aspects of care.
MEDICAL JOURNAL OF AUSTRALIA
(2021)
Article
Biology
Lara Maleyeff, Fan Li, Sebastien Haneuse, Rui Wang
Summary: This article proposes a new model formulation for assessing treatment effect heterogeneity over exposure time in stepped-wedge CRTs, using random effects to represent varying treatment effects by exposure time, leading to more precise estimation of average and exposure-time-specific treatment effects. It also develops a permutation test for the variance component of the heterogeneous treatment effect parameters.
Article
Medicine, General & Internal
Christopher W. Jones, Amanda Adams, Benjamin S. Misemer, Mark A. Weaver, Sara Schroter, Hayat Khan, Benyamin Margolis, David L. Schriger, Timothy F. Platts-Mills
Summary: The objective of this study was to test whether providing relevant clinical trial registry information to peer reviewers would decrease discrepancies between registered and published trial outcomes. The results showed that the tested intervention did not increase agreement between prospectively registered and published trial outcomes. Therefore, other approaches are needed to improve the quality of outcome reporting in clinical trials.
Article
Health Care Sciences & Services
Fan Li, James P. Hughes, Karla Hemming, Monica Taljaard, Edward R. Melnick, Patrick J. Heagerty
Summary: The stepped wedge cluster randomized design is gaining attention in pragmatic clinical trials and implementation science research. The design features clusters crossing over from control to intervention conditions in a unidirectional manner on a staggered schedule, which may confound intervention effects over time. This design first appeared in the Gambia hepatitis study in the 1980s, but the statistical model for its design and analysis was formally introduced in 2007. Various mixed-effects model extensions have since been proposed for the design and analysis of these trials.
STATISTICAL METHODS IN MEDICAL RESEARCH
(2021)
Review
Health Care Sciences & Services
Agnes Caille, Monica Taljaard, Floriane Le Vilain Abraham, Alexis Le Moigne, Andrew J. Copas, Florence Tubach, Agnes Dechartres
Summary: This study aims to explore challenges in recruitment and intervention implementation in recent stepped-wedge cluster randomized trials (SW-CRTs). The results show that approximately 33% of the trials faced recruitment challenges and 44% faced implementation challenges. Most trials with recruitment or implementation challenges made modifications to their design, primarily adjusting the trial duration.
JOURNAL OF CLINICAL EPIDEMIOLOGY
(2022)
Article
Public, Environmental & Occupational Health
Xuan Zhang, Chongya Dong, Nana Wang, Chunpong Chan, Chung Tai Lau, Juan Wang, Jiangxia Miao, Chen Yao, Youping Li, Aiping Lyu, David Moher, Zhaoxiang Bian
Summary: Due to the frequent inadequate reporting of key details in multicenter clinical trials, the development of guidelines to improve reporting quality is crucial. This study uses the CONSORT and SPIRIT checklists as tools to assist in designing and reporting high-quality clinical research.
FRONTIERS IN PUBLIC HEALTH
(2023)
Article
Biology
Kendra Davis-Plourde, Monica Taljaard, Fan Li
Summary: This article presents a method for calculating sample size in SW-CRTs with subclusters, which can properly differentiate between-period and within-period correlation coefficients. It uses an extended block exchangeable correlation matrix to characterize the dependencies of outcomes within clusters and derives a closed-form sample size expression for Gaussian outcomes. For non-Gaussian outcomes, it proposes a generic sample size algorithm based on linearization.
Article
Biology
Siyun Yang, Mirjam Moerbeek, Monica Taljaard, Fan Li
Summary: Pragmatic trials in healthcare interventions often use cluster randomization, but methods for determining sample size and power for continuous coprimary endpoints are lacking. We propose a method based on multivariate linear mixed models to address this gap and demonstrate its effectiveness through simulation studies.
Article
Geriatrics & Gerontology
Diana Koszycki, Monica Taljaard, Cary Kogan, Jacques Bradwejn, David Grimes
Summary: This study evaluated the feasibility and effectiveness of interpersonal psychotherapy (IPT) for Parkinson's disease (PD) patients with depression. The results showed high acceptance and tolerance of IPT treatment, with improvements in depression symptoms.
JOURNAL OF GERIATRIC PSYCHIATRY AND NEUROLOGY
(2023)
Article
Clinical Neurology
Kasim E. Abdulaziz, Monica Taljaard, Dar Dowlatshahi, Ian G. Stiell, George A. Wells, Jeffrey J. Perry
Summary: This study aimed to assess the variation in patient management between stroke prevention clinics (SPCs) in Canada and the impact of COVID-19 on patient care. The survey found that during the COVID-19 pandemic, many SPCs experienced longer wait times and an increase in the number of patients without completed bloodwork prior to appointments. However, some SPCs provided virtual care during the pandemic and plan to continue using it post-pandemic.
CANADIAN JOURNAL OF NEUROLOGICAL SCIENCES
(2023)
Article
Health Care Sciences & Services
Laura Kudrna, Paul Bird, Karla Hemming, Laura Quinn, Kelly Schmidtke, Richard Lilford
Summary: This study evaluated the impact of a short-term training program on improving the use of SPCs in hospital board papers. The results showed that the training program was effective in increasing the use of SPCs within the organization.
BMJ QUALITY & SAFETY
(2023)
Article
Medicine, Research & Experimental
James Martin, Lee Middleton, Karla Hemming
Summary: Cluster-randomised trials often use restricted randomisation methods, such as stratified- or covariate-constrained randomisation. Minimisation has the potential to balance on more covariates and can be implemented sequentially. However, minimisation may not maintain close to a 1:1 allocation, which is crucial in trials with a small number of allocation units like cluster randomisation.
Article
Geriatrics & Gerontology
Hayden P. Nix, Emily A. Largent, Monica Taljaard, Susan L. Mitchell, Charles Weijer
Summary: This article introduces a framework for identifying vulnerabilities in cluster randomized trials involving people living with dementia in long-term care homes. It discusses the potential wrongs and suggests additional protections that can be implemented to mitigate vulnerability while preserving the scientific validity of the trials.
JOURNAL OF THE AMERICAN GERIATRICS SOCIETY
(2023)
Article
Medicine, Research & Experimental
Evans A. Asamane, Laura Quinn, Samuel I. Watson, Richard J. Lilford, Karla Hemming, Cheick Sidibe, Ryan T. Rego, Sami Bensassi, Youssouf Diarra, Samba Diop, Om Prasad Gautam, Mohammad Sirajul Islam, Louise Jackson, Kate Jolly, Kassoum Kayentao, Ousmane Koita, Buba Manjang, Susan Tebbs, Nicola Gale, Paula Griffiths, Sandy Cairncross, Ousmane Toure, Semira Manaseki-Holland
Summary: This study presents a protocol for a cluster randomised trial to evaluate the effectiveness of a community-based educational intervention in improving food safety and hygiene behaviors and enhancing child nutrition. The trial will assess changes in complementary-food safety and preparation behaviors, as well as food and water contamination and diarrhoea. The study will also examine secondary outcomes such as maternal autonomy, enteric infection, nutrition, child anthropometry, and development scores.
Article
Medicine, General & Internal
Ioannis Gallos, Adam Devall, James Martin, Lee Middleton, Leanne Beeson, Hadiza Galadanci, Fadhlun Alwy Al-beity, Zahida Qureshi, G. Justus Hofmeyr, Neil Moran, Sue Fawcus, Lumaan Sheikh, George Gwako, Alfred Osoti, Ashraf Aswat, Kristie-Marie Mammoliti, Kulandaipalayam N. Sindhu, Marcelina Podesek, Isobelle Horne, Rebecca Timms, Idnan Yunas, Jenipher Okore, Mandisa Singata-Madliki, Edna Arends, Aminu A. Wakili, A. Mwampashi, Sidrah Nausheen, Shah Muhammad, Pallavi Latthe, Cherrie Evans, Shahinoor Akter, Gillian Forbes, David Lissauer, Shireen Meher, Andrew Weeks, Andrew Shennan, Anne Ammerdorffer, Eleanor Williams, Tracy Roberts, Mariana Widmer, Olufemi T. Oladapo, Fabiana Lorencatto, Meghan A. Bohren, Suellen Miller, Fernando Althabe, Metin Gulmezoglu, Jeffrey M. Smith, Karla Hemming, Arri Coomarasamy
Summary: This study aimed to evaluate a multicomponent clinical intervention for early detection and treatment of postpartum hemorrhage. The results showed that the use of a blood-collection drape and treatment bundle reduced the risk of severe postpartum hemorrhage and improved detection and adherence to treatment.
NEW ENGLAND JOURNAL OF MEDICINE
(2023)
Article
Mathematical & Computational Biology
Samuel I. Watson, Joshua O. Akinyemi, Karla Hemming
Summary: This article presents and compares methods for deriving P-values and confidence intervals with strong control of the family-wise error rates and coverage in cluster randomized trials with multiple outcomes. Existing methods for P-value corrections and deriving confidence intervals are limited in their application in this context. The study discusses and adapts the methods of Bonferroni, Holm, and Romano and Wolf for cluster randomized trial inference. A novel search procedure using permutation tests is developed for confidence set limits, and a simulation-based study is conducted to compare different methods. The results show that the Romano-Wolf type procedure has nominal error rates and coverage, and is more efficient than other methods in the simulation-based study. Results from the analysis of a real-world trial are also compared.
STATISTICS IN MEDICINE
(2023)
Article
Medicine, General & Internal
Emma Dickinson-Craig, Jargalsaikhan Badarch, Suzanne Bartington, Karla Hemming, Rasiah Thayakaran, Rosie Day, Francis Pope, Bataa Chuluunbaatar, Damdindorj Boldbaatar, Chimedsuren Ochir, David Warburton, Graham Neil Thomas, Semira Manaseki-Holland
Summary: Despite a decade of policy actions, Ulaanbaatar's residents continue to be exposed to extreme levels of air pollution, which is a major public health concern. The Mongolian government implemented a raw coal ban in May 2019 to address this issue. This study aims to assess the effectiveness of the coal ban policy on air quality and maternal and child health outcomes.
Article
Multidisciplinary Sciences
Lailah Alidu, Lena Al-Khudairy, Ila Bharatan, Paul Bird, Niyah Campbell, Graeme Currie, Karla Hemming, Kate Jolly, Laura Kudrna, Richard Lilford, James Martin, Laura Quinn, Kelly Ann Schmidtke, James Yates, Vidanka Vasilevski
Summary: This research aims to evaluate the effectiveness of the intervention 'mental contrasting plus implementation intentions' in a new workplace context. The study will recruit staff from 60 organizations and assess how this intervention can empower and support them in improving their health and wellbeing. This will be the first evaluation of the intervention outside of healthcare settings.
Article
Health Care Sciences & Services
Samuel Watson, Alan Girling, Karla Hemming
Summary: This article presents a review of statistical and computational methods for identifying optimal cluster randomised trial designs. Three broad classes of methods are identified: exact formulae for treatment effect estimator variance, generalised methods for estimating weights for experimental units, and combinatorial optimisation algorithms. The results show that these methods can be applied in various cluster trial scenarios, including optimal allocation and determination of the number of single observations to make in each cluster-period.
STATISTICAL METHODS IN MEDICAL RESEARCH
(2023)
Article
Medicine, Research & Experimental
K. Hemming, M. Taljaard, E. Gkini, J. Bishop
Summary: Justifying sample size for a pilot trial is important, but few trials provide a clear rationale for their chosen sample size. Unlike full-scale trials, pilot trials should focus on feasibility objectives rather than effectiveness objectives. For pilot cluster trials, sample size calculations depend on the number of clusters, cluster sizes, intra-cluster correlation coefficient, and anticipated proportion for the feasibility outcome.
PILOT AND FEASIBILITY STUDIES
(2023)
Article
Medicine, General & Internal
Ioannis Gallos, Adam Devall, James Martin, Lee Middleton, Leanne Beeson, Hadiza Galadanci, Fadhlun Alwy Al-beity, Zahida Qureshi, G. Justus Hofmeyr, Neil Moran, Sue Fawcus, Lumaan Sheikh, George Gwako, Alfred Osoti, Ashraf Aswat, Kristie-Marie Mammoliti, Kulandaipalayam N. Sindhu, Marcelina Podesek, Isobelle Horne, Rebecca Timms, Idnan Yunas, Jenipher Okore, Mandisa Singata-Madliki, Edna Arends, Aminu A. Wakili, Ard Mwampashi, Sidrah Nausheen, Shah Muhammad, Pallavi Latthe, Cherrie Evans, Shahinoor Akter, Gillian Forbes, David Lissauer, Shireen Meher, Andrew Weeks, Andrew Shennan, Anne Ammerdorffer, Eleanor Williams, Tracy Roberts, Mariana Widmer, Olufemi T. Oladapo, Fabiana Lorencatto, Meghan A. Bohren, Suellen Miller, Fernando Althabe, Metin Gulmezoglu, Jeffrey M. Smith, Karla Hemming, Arri Coomarasamy
Summary: Early detection and bundled treatment of postpartum hemorrhage led to a lower risk of severe complications or death compared to usual care among patients undergoing vaginal delivery.
NEW ENGLAND JOURNAL OF MEDICINE
(2023)