loading
Skip to main content
  • U.S. Department of Health & Human Services
National Institutes of Health (NIH) - Turning Discovery into Health

Research Methods Resources

Site Menu

  • Home
  • GRT
  • IRGT
  • GRT Sample Size Calculator
  • Glossary
  • References
  • FAQs
  • Feedback

Individually Randomized Group-Treatment Trials (IRGTs)

An individually randomized group-treatment trial (IRGT) is a randomized trial in which participants in one or more study conditions receive at least some of their treatment in groups or through a common change agent. This design is common in surgical trials, where each surgeon operates on multiple patients.14, 44 It is common in psychotherapy trials, where a therapist may treat multiple patients, either in groups or as individuals.3, 11, 53 It is common in a variety of intervention trials addressing health behaviors such as weight loss, smoking cessation, and physical activity, which may include group activities as well as individual activities.31

IRGTs always have a hierarchical structure in the intervention condition. Participants may receive some of their treatment in groups, or they may receive their intervention individually, but through a common change agent, whether in person or through a video or other virtual connection.34 There may or may not be a similar structure in the control condition, depending on the nature of the control condition.

IRGTs can be employed in a wide variety of settings and populations to address a wide variety of research questions. They are an appropriate design when the investigator wants to evaluate an intervention that:

  • Involves at least one component that is delivered in a group format, or
  • It is necessary to use a limited number of intervention delivery staff, or change agents, so that each one interacts with multiple participants, or
  • It is necessary to have participants interact with one another in a virtual environment.

IRGTs randomize individuals to study conditions. If the number is large, confounding is not likely to be a threat to the internal validity of the design. If the number is small, confounding could be a threat, and a priori matching or a priori stratification would be useful strategies to protect against confounding.

The more challenging feature of IRGTs is that participants in at least the intervention condition will interact post-randomization. Even if their groups or virtual networks are constructed using random assignment, those participants will interact with one another directly in their groups or indirectly through their common change agent. This interaction creates the expectation that some level of intraclass correlation (ICC) will develop. The magnitude of the ICC in an IRGT will depend on the type, duration, and intensity of these interactions. That ICC may be negligible at baseline, but it can develop over the course of the trial. With a limited number of groups or change agents, the df available to estimate the ICC, or the component of variance associated with the group or change agent, will be limited. As for GRTs, any analysis that ignores the extra variation (or positive ICC) or the limited df will have a type 1 error rate that is inflated, often badly.4, 5, 33-35, 45-47, 50, 51

These issues are especially challenging in an IRGT because the design may not have the same hierarchical structure in all conditions. In that case, the analytic model must accommodate a heterogeneous variance-covariance structure, allowing for ICC in one condition, but not in the other. For this reason, it is even more important for investigators to rely on an experienced methodologist in developing design and analytic plans for an IRGT.

The recommended solution to these challenges is similar to the solution proposed for GRTs. It is important to employ a priori matching or a priori stratification to balance potential confounders if the number of assignment units is limited, to reflect the hierarchical or partially hierarchical structure of the design in the analytic plan, and to estimate the sample size for the IRGT based on realistic and data-based estimates of the ICC and the other parameters indicated by the analytic plan. Extra variation and limited df always reduce power, so it is essential to consider these factors while the study is being planned, and particularly as part of the sample size estimation.

Questions? You may find answers on the Glossary and FAQs tabs.

The remaining sections provide resources that will be helpful to investigators considering an individually randomized group-treatment trial, or partially clustered design.

NIH Webinars

  • Pragmatic and Group-Randomized Trials in Public Health and Medicine Course
  • Medicine: Mind the Gap Webinar: Design and Analysis of Studies to Evaluate Multilevel Interventions in Public Health and Medicine

CONSORT Statement for IRGTs

  • Boutron I, Moher D, Altman DG, Schulz KF, Ravaud P, CONSORT Group. Extending the CONSORT statement to randomized trials of nonpharmacologic treatment: explanation and elaboration. Annals of Internal Medicine. 2008;148(4):295–309.
    • Note: In particular, see Table 2, Item 7, as well as the discussion of sample size on page 302.

Key References for IRGTs

  • Turner EL, Li F, Gallis JA, Prague M, Murray DM. Review of recent methodological developments in group-randomized trials: part 1—design. American Journal of Public Health. 2017;107(6):907–915. PMC5425852.
  • Turner EL, Prague M, Gallis JA, Li F, Murray DM. Review of recent methodological developments in group-randomized trials: part 2—analysis. American Journal of Public Health. 2017;107(7);1078–1086. PMC5463203.
  • Lai MH, Kwok OM. Estimating standardized effect sizes for two- and three-level partially nested data. Multivariate Behavioral Research. 2016;51(6):740–756.
  • Hedges LV, Citkowicz M. Estimating effect size when there is clustering in one treatment group. Behavioral Research Methods. 2015;47(4):1295–1308.
  • Andridge RR, Shoben AB, Muller KE, Murray DM. Analytic methods for individually randomized group treatment trials and group-randomized trials when subjects belong to multiple groups. Statistics in Medicine. 2014;33(13):2178–2190. PMC4013262.
  • Kahan BC, Morris TP. Assessing potential sources of clustering in individually randomised trials. BMC Medical Research Methodology. 2013;13:58. PMC3643875.
  • Baldwin SA, Bauer DJ, Stice E, Rohde P. Evaluating models for partially clustered designs. Psychological Methods. 2011;16(2):149–165. PMC3987820.
  • Pals SL, Murray DM, Alfano CM, et al. Erratum. American Journal of Public Health. 2008;98(12):2120. PMC2636526.
  • Pals SL, Murray DM, Alfano CM, et al. Individually randomized group treatment trials: a critical appraisal of frequently used design and analytic approaches. American Journal of Public Health. 2008;98(8):1418–1424. PMC2446464.
  • Roberts C, Roberts SA. Design and analysis of clinical trials with clustering effects due to treatment. Clinical Trials. 2005;2(2):152–162.
  • Hoover DR. Clinical trials of behavioural interventions with heterogeneous teaching subgroup effects. Statistics in Medicine. 2002;21(10):1351–1364.

Key References for State of the Practice Reviews for IRGTs

  • Oltean H, Gagnier JJ. Use of clustering analysis in randomized controlled trials in orthopaedic surgery. BMC Medical Research Methodology. 2015;15:17. PMC4359453.
  • Pals SL, Wiegand RE, Murray DM. Ignoring the group in group-level HIV/AIDS intervention trials: a review of reported design and analytic methods. AIDS. 2011;25(7):989–996.
  • Pals SL, Murray DM, Alfano CM, et al. Individually randomized group treatment trials: a critical appraisal of frequently used design and analytic approaches. American Journal of Public Health. 2008;98(8):1418–1424. PMC2446464.
  • Lee KJ, Thompson SG. Clustering by health professional in individually randomised trials. BMJ. 2005;330(7483):142–144. PMC544437.

Key References for Sample Size Estimation for IRGTs

  • Heo M, Litwin AH, Blackstock O, Kim N, Arnsten JH. Sample size determinations for group-based randomized clinical trials with different levels of data hierarchy between experimental and control arms. Statistical Methods in Medical Research. 2017;26(1):399–413. PMC4329103.
  • Moerbeek M, Teerenstra S. Power Analysis of Trials with Multilevel Data. Boca Raton, FL: CRC Press; 2016.
  • Roberts C, Walwyn R. Design and analysis of non-pharmacological treatment trials with multiple therapists per patient. Statistics in Medicine. 2013;32(1):81–98.
  • Candel MJ, Van Breukelen GJ. Varying cluster sizes in trials with clusters in one treatment arm: sample size adjustments when testing treatment effects with linear mixed models. Statistics in Medicine. 2009;28(18):2307–2324.
  • Moerbeek M, Wong WK. Sample size formulae for trials comparing group and individual treatments in a multilevel model. Statistics in Medicine. 2008;27(15):2850–2864.
  • Pals SL, Murray DM, Alfano CM, et al. Erratum. American Journal of Public Health. 2008;98(12):2120. PMC2636526.
  • Pals SL, Murray DM, Alfano CM, et al. Individually randomized group treatment trials: a critical appraisal of frequently used design and analytic approaches. American Journal of Public Health. 2008;98(8):1418–1424. PMC2446464.
  • Roberts C, Roberts SA. Design and analysis of clinical trials with clustering effects due to treatment. Clinical Trials. 2005;2(2):152–162.
This page was last updated/reviewed 6/28/2018

Footer

  • NIH Home
  • En Español
  • Site Map
  • Visitor Information
  • Frequently Asked Questions
  • Web Policies and Notices
  • Freedom of Information Act
  • No Fear Act
  • Office of Inspector General (link is external)
  • USA.gov – Government Made Easy (link is external)

NIH…Turning Discovery Into Health®

National Institutes of Health, 9000 Rockville Pike, Bethesda, Maryland 20892

U.S. Department of Health and Human Services (link is external)

Back to Top