Fifteen RCTs (n=996) were included in the review.
Quality assessment.
The quality of each of the included studies was low. None of the included studies stated that the allocation of treatment was concealed, only one study stated that intention-to-treat analysis was used, and most studies did not use standardised cointerventions. However, most of the studies employed a complete follow-up and used objective outcome criteria.
Propofol compared with midazolam (9 RCTs, n=690).
The time taken to recover from sedation was faster in patients treated with propofol than with midazolam (7 RCTs; data not given). Five of the studies demonstrated that propofol was a more effective sedative than midazolam, while two found equal effectiveness (data not provided).
Propofol compared with alfentanil (1 RCT, n=44).
The time taken to awaken was 5, 11 and 18 minutes, respectively, for patients receiving low, moderate and high doses of alfentanil; the time taken to breath spontaneously was 8, 22 and 27 minutes, respectively. No data were provided for those patients receiving propofol.
Propofol compared with papaveretum (1 RCT, n=27).
No difference was found in the number of patients who were able to follow commands or spontaneously breath prior to stopping the infusions (data not provided).
Isoflurane compared with midazolam (2 RCTs, n=120).
The proportions of patients adequately sedated following treatment with isoflurane were 70.7 and 86%, compared with 67.4 and 64% of patients receiving midazolam. However, this difference was only significant in the latter study (P=0.0005). Both studies found that the time to extubation was significantly shorter for patients receiving isoflurane than those receiving midazolam (data not provided).
Lorazepam compared with midazolam (2 RCTs, n=115).
One study (n=20) found that there was no significant difference in the time to reach adequate sedation, or the return to baseline mental status, following sedation with lorazepam or midazolam (data not provided). The second study (n=95) found that there was no significant difference in the quality of sedation between lorazepam and midazolam. However, less lorazepam was required to achieve the same level of sedation attained through treatment with midazolam (data not provided).