Table 3

Sensitivity analyses for the effect of hand washing with soap on diarrhoeal disease according to different quality appraisal tools

Pooled OR (95% CI)% Change in pooled ORIncluded studies§
Main meta-analysis0.60 (0.47 to 0.77)1–12
EPHPPT* 0.68 (0.47 to 0.98)+13.332, 3, 4
EPHPPT 0.59 (0.41 to 0.86)−1.672, 3, 4, 8
CASP0.64 (0.49 to 0.85)+6.671a, b, 2, 3, 4, 6, 7, 10, 11
NOS0.69 (0.42 to 1.14)+15.001b, 3, 4, 7, 11
Liverpool0.68 (0.47 to 0.98)+13.332, 3, 4
GATE* 0.57 (0.40 to 0.81)−0.052, 3, 4, 8, 10
GATE 0.57 (0.40 to 0.81)−0.052, 3, 4, 8, 10
ROBNo study includedNo study includedNo study included
  • * Inclusion of studies scoring 0 and above according to coding method in this study.

  • Inclusion of moderate and strong studies according to EPHPPT qualitative summary score.

  • Inclusion of studies scoring + or ++ according to GATE qualitative summary score for internal and external validity.

  • § See table 2 for an overview of included studies.

  • CASP, Critical Appraisal Skills Programme; EPHPPT, Effective Public Health Practice Project tool; GATE, Graphical Appraisal Tool for Epidemiological Studies; NOS, Newcastle–Ottawa Scale; ROB, Cochrane Collaboration Risk of Bias tool.