Sorry for my naive question as I am very new to epidemiology research...
I am trying to understand the best hypothetical study design to investigate if there is an association between diseases X and Y. Both X and Y are relatively rare diseases. Although I don't know for sure, X is a systemic disease that could potentially precede Y temporally. There have been no prior studies looking at such a specific association--just a clinical interest since we have incidentally noted a lot of patients with both diseases.
What is the best study design to examine if an association exists between X and Y (above what would be expected compared to the general population)? Since I don't know the temporal relationship for sure, should I just start with a cross-sectional study?
A prospective cohort study would be the best way to study the association in terms of validity. However, because you do not know the temporal direction for sure and because both diseases are rare, you are better off first doing a cross-sectional study or a retrospective case control study.
Is there a national registry that you could use to identify Y patients? If so, you could perhaps find additional data in their records on X diagnosis and then compare the prevalence rate of their X diagnosis to that of the general population (calculate standardized prevalence ratio). Alternatively, you could do a formal case control study in which you select sufficient Y patients (probably from several hospital registries), you select appropriate controls (either hospital or population-based), contact them and ask for diagnosis of X and then calculate the odds ratio. A complication for the latter approach is that X is a "rare exposure" and you are going to have to select a large group of cases and controls because of that.
- Badges
- Science topic
- Similar topics
- Biological Science
- Physiology
- Physiological Processes
- Stress