Table 3. Sensitivity analyses of intensive care unit stay and time to extubation.

Outcome	Number of included trials	Dex patients	Controlpatients	SMD	95% CI	P for effect	P for heterogeneity	I2 (%)
Overall trials	28 trials (27 manuscripts)	1,870	1,778
ICU stay	17	1,274	1,150	−0.48	−0.78 to −0.18	0.002	<0.001	91
-Postoperative elective surgery patients	8	373	372	−0.60	−1.05 to −0.15	0.008	<0.001	86
---- Cardiac surgery	6	337	336	−0.57	−1.11 to −0.03	0.04	<0.001	89
---- CABG surgery	4	103	106	−0.76	−1.94 to 0.42	0.2	<0.001	93
---- Non Cardiac surgery	2	36	36	−0.75	−1.23 to −0.27	0.002	0.3	23
-ICU patients (non elective postoperative)	9	901	778	−0.38	−0.82 to 0.06	0.09	<0.001	93
-excluding outlier studies [14], [32], [33]	13	1438	1427	−0.17	−0.29 to −0.05	0.005	0.15	28
- Long term sedation	6	849	726	−0.54	−1.09 to 0.008	0.053	<0.001	95
- Short term sedation	11	425	424	−0.45	−0.81 to −0.09	0.02	<0.001	82
- Daily interruption sedation	5	839	716	−0.38	−0.96 to 0.18	0.18	<0.001	96
- High maintenance dose	7	879	756	−0.31	−0.79 to 0.17	0.2	<0.001	95
- Low maintenance dose(<0.7 µg kg−1 h−1)	10	395	394	−0.62	−1.04 to −0.20	0.004	<0.001	85
- Loading dose	11	281	282	−0.58	−1.03 to −0.13	0.01	<0.001	83
- No loading dose	6	993	868	−0.36	−0.81 to 0.10	0.13	<0.001	96
- Loading dose and high maintenance dose	2	140	40	−0.09	−0.79 to 0.61	0.8	0.12	60
- SENSITIVITY (including only blinded studies)	8	891	768	−0.56	−1.09 to −0.04	0.04	<0.001	95
SENSITIVITY (including only low risk of bias studies)	10	1065	940	−0.44	−0.86 to −0.02	0.04	<0.001	94
SENSITIVITY (removing 1 study at time)	All 95% CIs of SDM<0 and p<0.05
SMALL STUDY EFFECT (fixed model)				−0.34	−0.43 to −0.26	<0.001
Time to extubation	24	1,804	1,674	−0.39	−0.66 to −0.11	0.005	<0.001	93
-Postoperative elective surgery patients	17	954	942	−0.31	−0.52 to −0.09	0.005	<0.001	77
---- Cardiac surgery	10	558	555	−0.42	−0.75 to −0.10	0.01	<0.001	83
---- CABG surgery	7	310	311	−0.59	−1.13 to −0.05	0.03	<0.001	89
---- Non Cardiac surgery (3 studies did not specify the operative setting)	4	76	76	−0.15	−0.47 to 0.17	0.4	0.8	0
-ICU patients (non elective postoperative)	7	850	732	−0.52	−1.25 to 0.21	0.16	<0.001	97
-excluding outlier studies [14], [32], [33]	20	995	993	−0.16	−0.26 to −0.05	0.003	0.04	39
- Long term sedation	6	830	712	−0.65	−1.44 to 0.15	0.11	<0.001	98
- Short term sedation	18	974	962	−0.28	−0.49 to −0.07	0.009	<0.001	76
- Daily interruption sedation	4	785	664	−0.69	−1.70 to 0.32	0.18	<0.001	99
- High maintenance dose	7	859	737	−0.42	−1.13 to 0.30	0.3	<0.001	76
- Low maintenance dose(<0.7 µg kg−1 h−1)	16	843	839	−0.30	−0.53 to −0.07	0.009	<0.001	76
- Loading dose	16	734	731	−0.23	−0.47 to 0.001	0.051	<0.001	75
- No loading dose	7	968	845	−0.60	−1.25 to 0.05	0.07	<0.001	97
- Loading dose and high maintenance dose	3	74	73	−0.08	−0.44 to 0.27	0.3	0.3	11
SENSITIVITY (including only blinded studies)	10	1241	1112	−0.56	−1.06 to −0.05	0.03	<0.001	97
SENSITIVITY (including only low risk of bias studies)	8	1023	899	−0.72	−1.34 to −0.10	0.02	<0.001	97
SENSITIVITY (removing 1 study at time)	All 95% CIs of SDM<0 and p<0.05
SENSITIVITY(Jakob study [8] included as time on mechanical ventilation) *	24	1,804	1,674	−0.38	−0.66 to −0.10	0.007	<0.001	93
SMALL STUDY EFFECT (fixed model)				−0.31	−0.38 to 0.24	<0.001

The overall analyses using weighted mean differences showed a reduction in intensive care unit stay of −0.79 [−1.17 to −0.40] days and a reduction in time to extubation of −2.74 [−3.80 to −1.65] hours in the dexmedetomidine group. It should be noted that the standard mean differences used in this table is not expressed in days or hours.

Dex: dexmedetomidine; SMD: standardized mean difference; CI: confidence interval; P: p-value; CABG: coronary artery bypass grafting; ICU: intensive care unit; NIV: non invasive ventilation.

duration of mechanical ventilation from randomization until patients were free of mechanical ventilation(including noninvasive).