In a new video, Dr. Greg Simon explains the intraclass correlation coefficient (ICC) with an analogy to a pie eating contest. The ICC is a descriptive statistic that measures the correlations among members of a group, and it is an important tool for cluster-randomized pragmatic trials because this calculation helps determine the sample size needed to detect an effect.
“When we randomize treatments by doctors, clinics, or even whole health systems, we need to think about how things cluster, and the intraclass correlation coefficient is the measure of that clustering. When we think about sample sizes in pragmatic clinical trials, it’s important to understand what an intraclass correlation coefficient actually is.”
For most pragmatic trials, the ICC will be between 0 and 1. If the outcomes in a group are completely correlated (ICC=1), then all participants within the group are likely to have the same outcome. When ICC=1, sampling one participant from the cluster is as informative as sampling the whole cluster, and many clusters will be needed to detect an effect. If there is no correlation among members of the groups (ICC=0), then the available sample size for the study is essentially the number of participants.
Kidney transplantation is the preferred treatment for patients with end-stage renal disease (ESRD), but an insufficient organ supply renders dialysis the only viable treatment option for most patients. Though clinical outcomes among patients receiving dialysis have improved modestly in recent years, annual rates of hospitalization and mortality remain unacceptably high, and quality of life is poor. Poor outcomes are driven primarily by increased risk of cardiovascular disease (CVD), but interventions that improve outcomes in the general population by targeting traditional CVD risk factors have mostly failed in patients with ESRD. Current clinical practice guidelines advocate aggressive treatment of high serum phosphate to near-normal levels using dietary phosphate binders and restrictive diets. The benefits of this approach, however, are unproven, the optimal serum phosphate target remains unknown, and potential harms of aggressive treatment have not been definitively identified.
“The question at hand is something we grapple with on a daily basis in every dialysis facility across the country. Either answer will be important new information that will help us do a better job taking care of patients and hopefully improve their quality of life.”
HiLo is led by Myles Wolf, MD, of Duke University with support from the National Institute of Diabetes and Digestive and Kidney Diseases. Read more about HiLo.
The National Institutes of Health (NIH) Office of Extramural Research has released new clinical trial requirements for grant applications and contract proposals due on or after January 25, 2018. In anticipation of these new requirements, the NIH modified the Application Guide and the Review Criteria to address methodological problems common to many clinical trials. As group- or cluster-randomization designs are increasingly common in both basic and applied research, the new Application Guide includes links to the new Research Methods Resources website, which provides resources for investigators considering these group- or cluster-randomized designs, including lists of NIH webinars, key references, and statements to help investigators prepare sound applications and avoid methodological pitfalls.
The Active Bathing to Eliminate (ABATE) Infection trial (ClinicalTrials.gov #NCT02063867) has completed its intervention phase—the first NIH Health Care Systems Research Collaboratory UH3 Demonstration Project to reach this major milestone. The large-scale, cluster-randomized pragmatic clinical trial (PCT) was designed to assess an approach for reducing multidrug-resistant organisms and hospital-associated infections (HAIs) in nearly 200 non-critical care hospital units affiliated with Hospital Corporation of America (HCA) across the United States.
The ABATE study is led by principal investigator Dr. Susan Huang of the University of California, Irvine, who stated “We are elated to reach the successful completion of the trial thanks to an incredible investigative team at HCA, Harvard Pilgrim Health Care, Rush University, the University of Massachusetts Amherst, and UC Irvine. We look forward to what the trial data will tell us and hope that we can continue to find effective ways to protect patients from infection.”
In the ABATE study, patients hospitalized in non-critical care units were bathed either according to the hospital unit’s usual care procedures (the control group) or bathed with the topical antibacterial agent chlorhexidine (plus nasal administration of the antibiotic mupirocin for those patients who were colonized or infected with, or had a history of methicillin-resistant Staphylococcus aureus [MRSA] [the intervention group]). The study investigators will compare the number of unit-attributable, multidrug-resistant organisms in clinical cultures between the study arms; these organisms include vancomycin-resistant enterococci (VRE), MRSA, and gram-negative bacteria. In addition, the investigators will compare the number of unit-attributable infections in the bloodstream and urinary tract (all pathogens) and Clostridium difficile infections. Cultures were collected at baseline and post intervention and will be assessed to determine whether resistance emerged to decolonization products.
“We are elated to reach the successful completion of the trial thanks to an incredible investigative team at HCA, Harvard Pilgrim Health Care, Rush University, the University of Massachusetts Amherst, and UC Irvine.We look forward to what the trial data will tell us and hope that we can continue to find effective ways to protect patients from infection.”
Healthcare-associated infections caused by common bacteria, including MRSA and VRE, are a leading cause of preventable illness and death in the United States and are associated with upward of $6.5 billion in annual healthcare costs. Although these bacteria normally live on the skin or in the nose, under certain circumstances they can cause serious or even life-threatening infections. Hospitalized patients who are ill or who have weakened immune systems are especially at risk for such infections. Because these pathogens are resistant to many antibiotics, they can be difficult to treat.
In intensive care units (ICUs), reducing the amount of such bacteria (a process referred to as decolonization) by treating patients’ skin with chlorhexidine and their noses with mupirocin ointment has been shown to reduce MRSA infections and all-cause bacteremias. However, relatively little is known about the effects of decolonization in hospital settings outside of critical care units, although this is where the majority of such infections occur. The ABATE trial, in contrast, is testing its bathing and decolonization strategy in adult medical, surgical, oncology, and step-down units (pediatric, psychology, peri-partum, and bone marrow transplantation units were excluded).
Over the course of the study, more than a million showers and baths were taken, and all sites have completed the intervention. The next steps for the ABATE investigators are to finish strain collection over the coming weeks, and then clean, validate, and analyze the data over the coming months.
A new article published this week in JAMA describes the cluster randomized trial design. The article is part of JAMA’s Guide to Statistics and Methods series, which publishes explanations of analytic and methodologic approaches used in current research articles to help clinicians better understand the research.
In “Cluster Randomized Trials: Evaluating Treatments Applied to Groups,” Drs. William J. Meurer and Roger J. Lewis define cluster randomization, describe its advantages and limitations, and provide guidance on interpreting cluster randomized trials. The article discusses aspects of a recent cluster randomized trial, the RESTORE trial, as an example.
In RESTORE, pediatric intensive care units were randomized to assess the effects of a nurse-implemented sedation protocol for children with acute respiratory failure on mechanical ventilation. As Meurer and Lewis point out, “interventions that involve training multidisciplinary health care teams are practically difficult to conduct using individual-level randomization, as health care practitioners cannot easily unlearn a new way of taking care of patients.” Cluster randomized designs are therefore often used for this type of research, and it is important for clinicians to be able to understand and evaluate these studies.
Meurer WJ, Lewis RJ. Cluster randomized trials: evaluating treatments applied to groups. JAMA. 2015;313:2068-2069. PMID: 26010636. doi:10.1001/jama.2015.5199.
Dr. Greg Simon and the Suicide Prevention Team have enrolled the first participants in the Pragmatic Trial of Population-Based Programs to Prevent Suicide Attempt. This groundbreaking study was developed by researchers at Group Health Cooperative in Seattle, Washington, Health Partners Medical Group in Minnesota, and Kaiser Permanente of Colorado, in collaboration with patients who have experienced suicidal thoughts or survived suicide attempts themselves.
Over 9 million adults in the United States experience suicidal thoughts, and more than 1 million adults attempt suicide each year. However, patients at risk for suicidal behavior are not routinely identified, and successful interventions for depression and suicide are not routinely implemented. New evidence suggests that patients who report frequent thoughts of death or self-harm on a commonly-used depression questionnaire are at higher risk for suicide attempt and death over the following year.
This study aims to address the significant problem of suicide by identifying patients who are at risk for suicidal behavior and testing two suicide prevention strategies. Patients at participating institutions will complete a standard depression severity questionnaire during routine clinical care, and the results will be stored in their electronic health records (EHR). Investigators will use the responses in the EHR to identify at-risk individuals, and once identified, the patients will be randomly assigned to either usual care or to two treatment programs. The first is a collaborative care-management approach; the second is an online skills training program called “Now Matters Now,” which is designed to help people manage painful emotions and stressful situations.
On October 20, 2014 the Office for Human Research Protections (OHRP) released a draft guidance on how to apply the Department of Health and Human Services (HHS)regulations on protecting human subjects (45 CFR Part 46) who participate in research studies intended to evaluate risks of treatments or procedures commonly used by healthcare professionals and recognized as “standard of care.” In standard-of-care research (or comparative effectiveness research), participants are randomized to receive one of two (or more) treatments that are accepted by medical experts as appropriate treatments for a given disease or condition.
Because treatments assigned to some participants might be different than the treatments they would have been assigned if they were not participating in the study, and the risks associated with one treatment might be different from the risks associated with another treatment, the OHRP recommends that these risks be fully described to potential participants as a part of the informed consent process.
The NIH Collaboratory’s Biostatistics and Study Design Core has released the first in a series of guidance documents focusing on statistical design issues for pragmatic clinical trials. Each of the four guidance documents are intended to help researchers by providing a synthesis of current developments in the field, discuss possible future directions, and, where appropriate, make recommendations for application to pragmatic clinical research.
The guidance documents are available through the Living Textbook and can be accessed on the “Tools for Research” tab or directly here.
Dr. Huang was one 10 winners of CRF Research Achievement Awards, which are presented annually. A key report related to the ABATE project was published in the New England Journal of Medicine in June of 2013.
Additional details of the award are available here.
One of the NIH Collaboratory’s initial Demonstration Projects, the Lumbar Image Reporting with Epidemiology (LIRE) study, has begun randomization in early April. The LIRE trial is designed to test whether insertingadditional epidemiological information into the lumbar spine imaging reports of patients being treated for lower back pain can help both doctors and patients to better understand and interpret the reports. This in turn could help doctors avoid subjecting patients to unnecessary tests and procedures.
LIRE is a cluster randomized trial, which means that instead of randomizing individual patients, whole clinics (one at the Henry Ford Health System in Detroit; one at Group Health Cooperative in Seattle, with more to follow) are randomly assigned to provide either the experimental treatment or the control treatment to patients.
Cluster-randomized trials offer a number of advantages, including the avoidance of certain kinds of bias that can effect the outcome of a study, but they also raise special issues that can require careful consideration.
The principal investigator of the LIRE trial is Dr. Jeffrey Jarvik of the University of Washington. You can read more about the LIRE trial here.