I have certain inclusion criteria for my exercise study and am now doing telephone screening to check eligibility of potential participants. When I have someone who has to be excluded because of let's say age or location I am offering to give the treatment (exercise programme) after study completion. This is what I read in most papers and what people recommended me to do. Now, I am asking myself why can't I give them the programme while the study is running? Do you know of any literature I can read to be able to answer this competently?
Would there be an issue of contamination if the excluded participants were added to the study but not counted? Since trials vary, it is important to consider the impact a extra people will have on the study. For a randomized trial, would you put some of the 'excluded people' in the control and others in the treatment group based on the method of randomization? You might want to consider these articles:http://www.ncbi.nlm.nih.gov/pubmed/23451162 and http://apps.who.int/rhl/LANCET_781-785.pdf; however, they mainly address 'intent-to-treat'.
I am going through the articles. However, as you said it depends on the study design. I have collected some other (more practical responses). Some said it is just to not overburden the researcher with so many people that are excluded. The most scientific explanation is similar to yours. My supervisor said its to avoid confounding. The person (excluded) could go to his/her friends who are included and then they talk about that she/he got treatment without a measurement procedure. Something in this direction. I still believe it is usually more practical to do it afterwards.
In cases where exclusion criteria identify individuals who would be harmed, not benefit, or not be able to do the intervention, you would not offer it. In cases where exclusion criteria are used to screen out people as a means of control (e.g., limit to certain age ranges), it is probably not necessary. If you are going to give them the intervention anyway, you could include them in the study and see if the exclusion variables relate to outcomes, e.g., age effects. Often when individuals are randomly assigned to intervention or a control group, the control group is given the intervention at the end of the study, assuming the intervention worked, or at least was not harmful. There can be advantages in doing so as one can see if the control group benefits from the intervention in order to replicate findings from the initial intervention group.
Hi Paul,
yeah. I only offer the intervention to those who can possibly do it without risk. I do not have a control that is doing nothing. I am going to check one independent variable whereas the rest will be the same. In the end of the day it is important in this study to see whether this additional stimuli does anything. Hence, those who need to be excluded (due to age, health and another factor related to the specific independent variable) will only get the main chunk of the intervention without this additional stimuli (that's how I have planned it). Yeah, I might indeed give them the intervention straight away without measuring anything. I however, intend to do it still after the study because of limited resources.
If possible, you might measure the dependent variable/s on the excluded group as an additional control group since they will get the intervention without the extra stimuli. This could be done during the study or afterwards.
- Badges
- Science topic
- Similar topics
- Correction