Twenty-six studies (25 papers) were included in the review. Eleven studies were RCTs (965 patients) and 15 were non RCTs (7,368 patients). All RCTs were reported as underpowered to detect adverse events. Five RCTs were reported as having an unclear randomisation method. Three RCTs were reported as not having blinding. Two of the non RCTs were prospective studies and six were retrospective comparative studies. Seven non RCTs were non comparative studies.
Propofol-based sedation (six RCTs): In all cases of propofol sedation, the incidence of major respiratory complications was 0.3%. Minor respiratory events occurred more frequently (up to 24% in one study), particularly in very young children. One RCT that compared propofol with general anaesthesia found that propofol provided an equivalent alternative to general anaesthesia. One RCT found that the addition of fentanyl or midazolam to propofol significantly lowered the propofol dose required, with no increase in adverse events. One RCT found that propofol was associated with significantly higher levels of effectiveness and patient comfort, with fewer adverse events when compared with midazolam and meperidine. One RCT found that ketamine premedication and propofol were significantly more effective in reducing propofol infusion pain than lidocaine-propofol mixture without premedication.
Opioid and benzodiazepine-based sedation (three RCTs): In all of the opioid plus benzodiazepine-based regimens, the incidence of major respiratory complications was 0.2%. One RCT found that use of meperidine and midazolam compared with propofol, with or without premedication midazolam, was associated with significantly higher need for supplemental oxygen, equally high procedural success rate, but it took a significantly longer time to achieve adequate sedation, mean procedure time and recovery time. One RCT reported that fentanyl plus midazolam and meperidine plus midazolam were equally safe and effective. One RCT found that intravenous meperidine plus midazolam and intranasal midazolam plus intravenous meperidine were equally safe, without major adverse events, but children receiving intranasal midazolam premedication had less intense negative behaviour than other groups.
Midazolam alone (one RCT): One RCT reported procedural success rates of approximately 100% with oral and intravenous midazolam but insufficient data on effectiveness. There were no major adverse events.
Premedication (four RCTs): Two RCTs found that premedication with midazolam prior to a propofol or opioid-based sedation regimen significantly improved the ease and comfort of both intravenous catheter placement and separating the child from the parents, and increased the level of patient comfort. One RCT found that intranasal midazolam premedication significantly reduced the intensity of negative behaviours, but not the total incidence of these behaviours. One RCT that compared atropine premedication with placebo in children sedated with other regimens had limited data to assess outcomes.
Details of the findings of non RCTs were reported in the paper.