Rethinking the assessment of risk of bias due to selective reporting: a cross-sectional study

Table 4 Number of reviews which acknowledged that the synthesis of an outcome was missing data that were not/partially reported

Location of statement that data was missing from the synthesis	Type of review outcome			Chi-squared test P value^b
	Any outcome, number (%) of reviews^a	Primary outcomes, number (%) of reviews	Secondary outcomes, number (%) of reviews	Chi-squared test P value^b
Main text	24/79 (30)	22/52 (42)	2/27 (7)	0.0014
Abstract	11/63 (17)	9/51 (18)	2/12 (17)	0.9358
Summary of Findings table	9/47 (19)	7/35 (20)	2/12 (17)	0.8001

^aThe denominators reflect the number of reviews for which the assessment was possible. For example, only 63 abstracts were assessed because the outcome that was not/partially reported in some studies was only described in the abstract of 63 reviews
^bDifference in the proportion of reviews with acknowledgement that data were missing from the synthesis of an outcome when the outcome was considered primary versus when the outcome was considered secondary by the review authors

ISSN: 2046-4053