What is a Stepped Wedge Design?
The stepped wedge design is a sub-type of a randomized controlled trial, where individuals or groups are studied at different times. At the end of the trial, all participants will have received the intervention being studied, although the order in which they receive it is randomized. After clusters have been randomly assigned an order, they are assigned to a step. This creates a treatment graph in the shape of a stepped wedge:
Reasons for Using a Stepped Wedge
A stepped wedge is often used when participants cannot be studied at the same time for logistical or financial reasons. Another common reason for using this design is that you want to see the effects of time on the study. For example, if you are studying the effectiveness of treated mosquito netting on malaria, you may want to study the netting during different seasons.
Individuals or groups act as their own control group, so this creates two major benefits:
- Cost efficiency: fewer units are needed for the study.
- Ethically sound practice: when you have a treatment that appears to be very promising, it may make more sense to use a stepped wedge instead of a traditional experimental/control setup where one group would not be given treatment.
- Contamination can be an issue with stepped wedge design as participants may interact with those who are waiting for treatment.
- Blinding is practically impossible because participants and practitioners will most likely be aware of the switch from control period to treatment period. It is, however, possible to blind treatment assessors so that they don’t know the status of the participants when assessing outcomes. Blinding assessors will also minimize information (observation) bias.
- Stepped wedge designs are fairly new, which means sound analysis techniques are still in development. One meta-analysis of recent stepped wedge studies (Barker, 2016) concluded that some studies have too few clusters — resulting in potentially biased and inefficient intervention effect estimates. Taljaard et. al (2016) note the “substantial risks” with small numbers of clusters, including increased type I and type II errors, limited generalizability, and fewer data analysis options.
- Multiple data collection points are required (one for each cell in the above image), so this design is best suited for routinely collected data.
Stepped Wedge vs. Crossover Design
A crossover design is where patients are assigned all treatments, and the results are measured over time. With a crossover design the treatments can happen in any order. For example, one group might be given treatment A and then treatment B, while a second group gets B than A. The stepped wedge, however, has the crossover in one direction only.
Barker, D. et. al. (2016) Step. wedge cluster randomised trials: a review of the statistical methodology used and available. BMC Med Res Methodol. 16: 69.
Taljaard, M. et. al (2016). Substantial risks associated with few clusters in cluster randomized and stepped wedge designs. In Clinical Trials Vol 13, Issue 4. SAGE Publications.
If you prefer an online interactive environment to learn R and statistics, this free R Tutorial by Datacamp is a great way to get started. If you're are somewhat comfortable with R and are interested in going deeper into Statistics, try this Statistics with R track.Comments are now closed for this post. Need help or want to post a correction? Please post a comment on our Facebook page and I'll do my best to help!