Dexmedetomidine in prevention and treatment of postoperative and intensive care unit delirium: a systematic review and meta-analysis

Annals of Intensive Care

Table 3 Subgroup and sensitivity analyses for incidence of delirium

Subgroup and sensitivity analyses	Number of data sets	Number of patients	Risk ratio (95% CI) random effect model	I ²
Placebo
Overall	10	2071	0.52 (0.42–0.63)	37%
Studies with no risk of bias according to our risk of bias assessment	5	887	0.66 (0.43–1.03)	20%
Cardiac surgery	2	349	0.53 (0.22–1.24)	1%
Mechanical ventilation	0
Non-cardiac surgery	6	1589	0.53 (0.35–0.80)	61%
Standard sedatives
Overall	15	2463	0.63 (0.46–0.86)	69%
Studies with no risk of bias according to our risk of bias assessment	2	998^a*	0.64 (0.41–0.98)	0%
Cardiac surgery	5	509	0.40 (0.23–0.69)	11%
Mechanical ventilation	7	1536	0.82 (0.60–1.11)	66%
Non-cardiac surgery^b	2	356	0.46 (0.23–0.90)	NA
Trim-and-fill method	20^c	–	0.77 (0.57–1.05)	64%
Opioids
Overall	2	441	0.61 (0.44–0.83)	0%

^aPropofol or midazolam was used in the comparison group
^bOne out of two studies had no events
^cFive hypothetical studies added with by a trim-and-fill function or R package meta