Sample size requirements and intra-cluster correlations for stepped wedge cluster randomised trials in intensive care
Thomas Hughes-Gooding1, Diva Baggio2, Edward Litton3,4
1Intensive Care Unit, Wellington Hospital, Wellington, New Zealand.
Objective:
To estimate key statistical parameters and provide practical guidance for planning stepped wedge cluster randomised trials in Australian and New Zealand intensive care units (ICUs).
Design:
Cross-sectional retrospective observational study using routinely collected ICU data.
Setting:
Adult public hospital ICUs contributing to the Australian and New Zealand Intensive Care Society Adult Patient Database between 2010 and 2023.
Participants:
All adult ICU admissions to 132 ICUs. Subgroups included unplanned admissions and admissions involving invasive mechanical ventilation or vasopressor use.
Main Outcome Measures:
In-hospital mortality during the index hospitalisation within 90 days of ICU admission. Intra-cluster correlation coefficients (ICCs) and cluster auto-correlations (CACs) were estimated using exchangeable, block-exchangeable, and discrete time decay models using a cross-sectional design.
Results:
Among 1,291,849 eligible ICU admissions, observed mortality ranged from 10.3% (all ICU admissions) to 23.0% (non-elective invasively ventilated patients in Mega-ROX ICUs). ICCs ranged from 0.008 to 0.022 and CACs from 0.83 to 1.00, with block-exchangeable or discrete time decay models most often providing the best fit. In a worked example, a 50-ICU stepped wedge trial with 10 steps (11 two-month periods) enrolling 45 unplanned ventilated patients per ICU per period (total ≈24,750 patients) would have 81.6% power to detect an absolute mortality reduction of 2.7%.
Conclusions:
Stepped wedge cluster randomised trials are feasible for evaluating ICU-wide interventions when routine data are available. The ICC and CAC estimates presented here provide Australian and New Zealand-specific parameters for future trial planning and demonstrate the potential of this design for pragmatic large-scale ICU research.
Related Concept Videos
01:20Cluster Sampling Method
To choose a cluster sample, divide the population into clusters (groups) and then randomly select some of the clusters. All the members from these clusters are in the cluster sample. For example, if you randomly sample four departments from your...
01:23Sampling Plans
Random sampling is a method where each member of the population has an equal chance of being selected for the sample. It involves selecting individuals randomly, often using random number generators or lottery-type methods. For example, when analyzing the properties of a...
01:13Randomized Experiments
Simple randomization
Simple...
01:23Types of Biopharmaceutical Studies: Controlled and Non-Controlled Approaches
Non-controlled studies, commonly employed for initial exploration, lack a control group, rendering them susceptible to biases and external influences. In contrast,...
01:20Study Designs in Epidemiology
Observational studies are those where the researcher does not intervene but rather observes natural variations. They include cross-sectional, cohort, and...
01:15Study Design in Statistics
Does aspirin reduce the risk of heart attacks? Is one brand of fertilizer more effective at growing roses than another? Is fatigue as dangerous to a driver as the influence of alcohol? Questions like these are answered using randomized experiments with proper...