I am attaching a link to the criteria for inclusion in the Cochrane system. You will see a section about non-randomized studies (Chapter 13: Including non-randomized studies). There are guidelines for choosing studies and controlling for bias.

I hope this helps

Ariel

http://handbook.cochrane.org/

Combining the reference Ariel suggested above (COCHRANE) with the STROBE checklist specifically for cross sectional studies would be useful.

STROBE checklist would guide how an individual CS study should be reviewed and COCHRANE would guide how these CS studies can be pooled.

The link to STROBE is below.

http://www.strobe-statement.org/index.php?id=available-checklists

• Checklist for cross-sectional studies 

I think the best scale for systematic reviews of sectional studies is the Newcastle Otawa for Sectional Studies. I have put in attach this scale.

Thank you all!

Appreciate your comment

Dear Shiyam,

You can use STROBE checklist to assess the quality of observational studies. However, some researcher criticize this approach as STROBE ensure standard reporting rather than methodological quality.

The best tool, from my opinion is, NOS (NEWCASTLE - OTTAWA QUALITY ASSESSMENT SCALE) which is widely used for observational studies. (Attached below)

URL: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp

Last year, the Cochrane collaboration developped ACROBAT NRSI tool. However, this is used mainly for non-randomized INTERVENTIONAL studies.

BW,

Negida.A

Many Thanks Jennifer de Beyer for detailed answer

Hi all,

I am developing a proposal for my first systematic review and in a trouble.

I want to clarify the difference between 'assessing quality of included studies' and 'assessing the publication bias'. Are they similar or different? Because, I found that the same tool is being used to assess both.

Thank you!

Quality depends on the sample size, whether it is a representative sample etc. If u draw a forest plot the thickness of the box will show the sample size. And you have to assess certain things in the methodology and data analysis. Random selection of study subjects, degree of bias( selection bias, attrition bias, reporting bias... Etc)

Publication bias can be graphically interpreted by a funnel plot. It will show the publication of positive and negative results are biased or not. Researches may publish positive findings only.

The challenge is assessing risk of bias. It would be great if Cochrane's tool to assess risk of bias widely available

Randomized clinical trials (RCT) have specific and unique reporting requirements. This form is used in the Shiyam Sunder Tikmani American Journal of Orthodontics and Dento-facial orthopedics.

How can systematic and meta analysis studies carried out from cross-sectional studies?