Hi Rania,

Have a look at the tools developed by the MOOSE group and the Cochrane collaboration. This should help you in evaluating non randomised studies.

Cochrane Handbook: chapter 13, section 5 - http://handbook.cochrane.org

MOOSE statement: 

Hi Frédérique,

First thank you so much for your help but i have a problem identifying the type of studies in which an intervention is given (by researcher) without control group and without comparing before -after is that considered observational?? As i understand from my readings , in observational studies intervention  (not given by researcher) and i don't know if we said uncontrolled is that the same as non randomized ??? Cochrane has a risk of bias tool for non randomized studies but (with control group) 

if you have a minute have a look on this paper 

Hi Rania,

As mentioned in chapter 13 from the Cochrane Handbook, study designs are nowadays so diverse that it becomes difficult to label them with a single appellation. The best would be instead to evaluate specific design features (like: patient recruitment, treatment allocation, outcome assessment, etc). 

Regarding your questions:

An observational study is indeed observational: no treatment or intervention is provided by the researchers. Instead, features from the participants (such as risk factors for something) are collected and researchers try to establish an association between these factors and a specific outcome. They can start with the risk factors and follow participants over time to see if they develop the outcome (prospective design) or start with the outcome and look in the past if participants had risk factors (retrospective design).

An interventional study is a study where researchers assign participants to an intervention  (treatment, device, strategy, etc). You can have only one group (as it is the case in your beta-carotene study), or 2 (the most common design) or more. Treatment assignment can be randomised (determined by random chance) or not adequatelly randomised (quasi randomised) or determined by other factors. You can start with outcome assessment and determine if participants received the intervention in the past (retrospective design) or start with the intervention and see if participants develop the outcome (prospective design). 

Each specific design feature can be assessed with these tools. Having only 1 study group makes this feature at high risk of bias, but other features in this study might be at lower risk. You can still evaulate each of them using the appropriate tool (MOOSE for observational studies, Chapter 13 for interventional studies). 

Hope it helps

Frédérique

Please go through the article (link), ''the methodological quality assessment tools for preclinical and clinical studies, systematic review and meta-analysis, and clinical practice guideline: a systematic review'' which listed all current evaluation scales required for Systematic Reviews (also have a look on Supplementary Appendix of the article).

Good Luck ...

http://dx.doi.org/10.1111/jebm.12141/epdf

thanks Frédérique

thanks  Sharanbasappa