Please help. I am doing a systematic review of largely observational studies on the safety and effectiveness of treating autoimmune alopecia with Hydroxychloroquine in older adults. I was wondering if a subgroup analysis is feasible in a SR based largely on observational studies. Thank you.
Subgroup analysis is feasible, provided that analysis according to the factors that you want to subgroup is provided in included or some of the included papers. Otherwise you should opt for individual patient data analysis. Rather than subrgrouping you can also consider meta-regression analysis taking in mind the factors that you want to adjust for. This might be more feasible in certain cases.
Maxine, You should check the level of heterogeneity and then plan for sub-group analysis or meta-regression in case of high heterogeneity. Subgroups should be well thought.