TABLE 4.

Sensitivity analysis of the treatment effect for the MMSE outcome.

Sensitivity analysis	No. of studies	No. of participants	Statistical method	MD [95% CI]	I2%
All studies	11	823	IV, Random effect	1.90 [1.22, 2.58]	87
All studies	11	823	IV, Fixed effect	2.03 [1.81, 2.25]	87
Low risk or some concerns in overall judgement	10	741	IV, Random effect	1.86 [1.15, 2.56]	88
Low risk in overall judgement	1	30	IV, Random effect	1.56 [-0.45, 3.57]	N/A
Low risk in randomization process	6	389	IV, Random effect	2.01 [1.58, 2.44]	0
Low risk in deviations from the intended interventions	7	425	IV, Random effect	1.45 [0.29, 2.61]	92
Low risk in missing outcome data	10	741	IV, Random effect	1.86 [1.15, 2.56]	88
Low risk in measurement of the outcome	11	823	IV, Random effect	1.90 [1.22, 2.58]	87
Low risk in selection of the reported outcome	2	112	IV, Random effect	0.52 [-0.63, 1.68]	39

Abbreviations: CI, confidence interval; I²: index of heterogeneity; IV, inverse variance; MD, mean difference; N/A: not applicable.