Table 2.

Sensitivity analysis of the pooled estimates

	Risk difference (95% CI) with fixed effect assumptions	Risk difference (95% CI) with random effect assumptions
All entered studies
Improper sequence, I2 = 96%	0.11 (0.09 to 0.13)*	0.14 (0.05 to 0.24)*
Presence of abbreviations, I2 = 93%	0.09 (0.08 to 0.11)*	0.14 (0.07 to 0.20)*
No disease-time interval, I2 = 95%	0.27 (0.24 to 0.29)*	0.28 (0.18 to 0.38)*
Multiple causes in one line, I2 = 75%	0.13 (0.11 to 0.15)*	0.14 (0.10 to 0.17)*
Ill-defined UCOD, I2 = 97%	0.06 (0.04 to 0.08)*	0.10 (− 0.03 to 0.22)
After excluding the outliers
Improper sequence, I2 = 64%	0.18 (0.15 to 0.20)*	0.18 (0.14 to 0.23)*
Presence of abbreviations, I2 = 21%	0.16 (0.13 to 0.18)*	0.16 (0.13 to 0.19)*
No disease-time interval, I2 = 0%	0.33 (0.30 to 0.36)*	0.33 (0.30 to 0.36)*
Multiple causes in one line, I2 = 0%	0.15 (0.13 to 0.17)*	0.15 (0.13 to 0.17)*
Ill-defined UCOD, I2 = 70%	0.15 (0.12 to 0.17)*	0.15 (0.10 to 0.20)*

*Statistically significant