Skip to main content

Table 3 Unadjusted and adjusted odds ratio of adhering to quality criteria

From: A systematic review of cluster randomised trials in residential facilities for older people suggests how to improve quality

  Proportion adhering to criteria Unadjusted odds ratio Adjustedg odds ratio
Reported accounting for clustering in analysis (includes secondary reports, 84 reports)
  No statistician co-author 16/28 Reference category  
  Co-author included statistician 44/56 3.2 (1.2,8.5) 3.1 (1.1,8.7)
  Publication prior to 2005 22/33 Reference category  
  Journal CONSORT endorsement lowf 9/14 0.9 (0.2,3.3) 0.8 (0.2,3.3)
  Journal CONSORT endorsement mediumf 16/21 1.6 (0.5,5.5) 1.1 (0.3,4.2)
  Journal CONSORT endorsement highf 12/16 1.5 (0.4,5.8) 1.4 (0.3,6.5)
Reported accounting for clustering in sample size (unique trials, 73 reports)
  No statistician co-author 2/22 Reference category  
  Co-author included statistician 18/51 5.4 (1.1,26.7) 2.7 (0.5,16.2)
  Publication prior to 2005 4/29 Reference category  
  Journal CONSORT endorsement lowf 4/11 3.6 (0.7,18.0) 3.5 (0.6,18.8)
  Journal CONSORT endorsement mediumf 5/18 2.4 (0.5,10.5) 1.7 (0.4,7.7)
  Journal CONSORT endorsement highf 7/15 5.5 (1.3,23.6) 4.7 (1.0,21.8)
Identification/recruitment bias is not possible or unlikely (unique trials, 73 reports)
  No statistician co-author 12/22 Reference category  
  Co-author included statistician 32/51 1.4 (0.5,3.9) 1.3 (0.4,4.0)
  Publication prior to 2005 15/29 Reference category  
  Journal CONSORT endorsement lowf 9/11 4.2 (0.8,22.9) 4.1 (0.8,22.7)
  Journal CONSORT endorsement mediumf 12/18 1.9 (0.5,6.3) 1.8 (0.5,6.4)
  Journal CONSORT endorsement highf 8/15 1.1 (0.3,3.7) 10.2 (0.3,4.5)
  1. f Reference category is publication pre-2005.
  2. g Adjusted for the other variables in the table.