We conducted a study to ascertain Mantoux reaction among diabetic patients. Since we conducted the work in a diabetic clinic, we found it difficult to blind the subjects (from investigators or independent investigators too) as they were from a diabetic clinic, thus all of them being diabetic. To mitigate this, we took steps to standardize all administration and TST reading based on CDC guidelines. We were blinded to other information such as investigative results and other history on comorbities. Being aware that this is some form of inescapable bias, is this still permissible or is the study method flawed?
Hi,
Interesting query.
I understand that you have done an 'interventional case-series' in which you evaluated the magnitude of reaction following Mantoux test in a group of diabetic patients. By the very nature of the study design, it is difficult to 'blind' the research team to the condition of interest - i.e. diabetes.
The risk of measurement or detection bias would have been high if you had two groups - one with diabetes and the other without diabetes and you wanted to compare the reaction of the two groups. The risk would be real if the outcome assessors were aware of the disease status.
In your case, there is no comparison. You are going to just describe the findings in one group of patients. There is a theoretical risk if you are going to compare with the 'control' or 'standard' population of other studies. But that is unavoidable. The best way in that case is to acknowledge as a possible limitation in the 'Discussion' section.
But you highlight the fact that you have tried to 'objectivise' the measurement as much possible by using strict and validated criteria (CDC guidelines). You can also provide the steps taken for quality control of measurement and the inter-observer variability.
I hope this helps...
Best
Jeeva
Dear Jeeva,
Thanks for your answer. Yes, it was a comparative study, and it was almost impossible to blind the investigators. All investigators worked at the clinic and even if we tried hard, there would be a chance they we would know their status anyways. It would be unethical to say we were blinded. So we described this as a limitation in the discussion section and that we chose to mitigate the this bias through various methods. This was the only way to obtain realistic data that reflects findings at a community level.
Rabin
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology