Skip to main content

Table 1 Secondary analyses in meta-analyses reporting only birthweight mean difference outcome (N = 21)

From: Dichotomisation of a continuous outcome and effect on meta-analyses: illustration of the distributional approach using the outcome birthweight

Meta-analysis    Published Distributional estimates for low birthweight  
Number of studies Pooled sample size Mean difference (g) (95% CI) P-value Distributional RR (95% CI) P-value Comments
Abou El Senoun 2010 [13] 1 55 −170 (−558, 218) 0.39 1.13 (0.85, 1.51) 0.39  
Alfirevic 2010 [14] 7 3,887 28 (−10, 66) 0.15 0.96 (0.89, 1.04) 0.28  
Alfirevic 2010 [15] 2 5,914 −18 (−42, 7) 0.16 1.08 (0.97, 1.19) 0.15  
Begley 2010 [16] 2 3,207 −77 (−109, −45) < 0.01 1.49 (1.27, 1.77) < 0.01  
Bevilacqua 2010 [17] 7 5,372 −83 (−124, −42) < 0.01 1.04 (1.02, 1.08) < 0.01  
Blanco 2011 [18] 2 786 113 (−45, 271) 0.16 0.74 (0.48, 1.16) 0.19  
Buchanan 2010 [19] 7 692 −12 (−91, 67) 0.76 1.00 (0.97, 1.03) 0.87  
Coleman 2010 [20] 3 614 158 (−53, 370) 0.14 0.65 (0.36, 1.18) 0.16  
Crowther 2011 [21] 9 5,626 −76 (−118, −34) < 0.01 1.02 (1.01,1.03) 0.01  
Dhulkotia 2010 [22] 6 1,388 24 (−36, 83) 0.44 0.93 (0.77, 1.13) 0.48  
Gebreselassie 2011 [23] 17 6,208 39 (−7, 85) 0.09 0.88 (0.76, 1.02) 0.08  
Imdad 2011 [24] 13 4,189 60 (33, 87) < 0.01 0.79 (0.71, 0.87) < 0.01  
Lassi 2010 [25] 2 1,050 11 (−39, 62) 0.66 0.98 (0.84, 1.14) 0.78  
Mackeen 2011 [26] 2 117 159 (−44, 361) 0.13 0.97 (0.93, 1.02) 0.19  
Mak 2010 [27] 4 251 8.33 (−143, 159) 0.91 0.99 (0.62, 1.58) 0.96 4/6 primary studies were accessible
Middleton 2010 [28] 2 159 −3 (−180, 175) 0.98 0.99 (0.53, 1.90) 0.99  
Nabhan 2011 [29] 1 125 −100 (−364, 164) 0.46 1.18 (0.78, 1.80) 0.46  
Quinlivan 2011 [30] 4 537 8.5 (−85, 102) 0.86 0.97 (0.67, 1.41) 0.88  
Rumbold 2011 [31] 5 7,497 6.1 (−17, 29) 0.61 0.99 (0.91, 1.08) 0.84  
Stampalija 2010 [32] 1 3,133 −34 (−69, 0.63) 0.05 1.15 (0.99, 1.33) 0.05  
Vazquez 2011a[33] 1 128 −461 (−608, −314) < 0.01 4.91 (2.88, 8.37) < 0.01 Data from meta-analysis where outcome was analysed as continuous for one intervention and as binary for another.
  1. aMeta-analysis where outcome was analysed as continuous for one intervention and as binary for another; CI: confidence interval; RR: risk ratio.