Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Chloral hydrate or midazolam: which is better for sedating children for painless diagnostic imaging?

Three Part Question

In [children who require sedation for painless diagnostic imaging] is [chloral hydrate better than midazolam] in achieving [effective and safe sedation]?

Clinical Scenario

A 2-year-old child is acutely admitted at night to a district general hospital following a prolonged focal seizure. A CT scan of the head without sedation fails as the child is too active and distressed, and as the on-call paediatric registrar, you are requested by the radiographer to prescribe a sedative. The NICE guideline for sedation in children and young people advises that midazolam or chloral hydrate can be used to sedate for painless procedures. You wonder which of these drugs is the most effective and safe and which has the least side effects.

Search Strategy

Medline (1948 to March week 5 2011) and Embase (1947 to 8 April 2011) were searched via Ovid on 11 April 2011
((chloral adj2 hydrate) AND midazolam) AND ((imaging or neuroimaging) OR ultrasound OR ultrasonog* OR scan* OR (computed ajd2 tomography) OR CT OR (magnetic adj2 resonance adj2 imaging) OR MRI OR echocardio* OR (painless adj2 procedure)).

Search Outcome

A total of 151 abstracts were reviewed and seven relevant articles were found (table 1). The citations in these articles and the Cochrane Library yielded no further relevant results. The relevant articles are summarised in the table

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
D'Agostino et al,
100 children (2/12–8 years) scheduled for a neuroimaging study (CT or MRI). Equally randomised to receive oral chloral hydrate 75 mg/kg (max 2 g) or oral midazolam 0.5 mg/kg (max 10 mg). Double-blinded randomised trial (level 1bCompletion of imaging;RR 2 (chloral hydrate 100% vs midazolam 50%, p<0.05); RR 0.17 (chloral hydrate 9% vs midazolam 55%, p<0.05). No significant difference in duration of sedation. No instances of physiological deterioration occurred in either group. ITT analysis not performed. Study terminated after the first 40 patients due to high sedation failure rate. Small final sample. 1 patient excluded without reason. Allocation concealment method not reported.
McCarver-May et al,
7 term neonates requiring CT and single-photon emission CT. Oral chloral hydrate 75 mg/kg given for the 1st study and intravenous midazolam 0.2 mg/kg for the 2nd study.Single-blinded randomised cross-over trial (level 1b)Induction of sleep/completion of imaging; onset and duration; physiological deteriorationRR 2.3 (7/7 chloral hydrate vs 3/7 midazolam, p=0.04). Chloral hydrate had longer onset (p=0.03) and duration of deep sleep (p=0.03). Both groups had 57% incidence of SaO2 dropping below 90% (p<0.05). 1 patient in midazolam group developed tachycardia. Small sample size. Included neonates only. Only single-blinded. Randomisation method and order of studies not mentioned. Possible carry-over effects from previous drug therapy.
Layangool et al,
264 children (6/12 to 5 years) requiring an ECHO. Equally randomised to 50 mg/kg oral chloral hydrate or 0.3 mg/kg sublingual midazolam. Double-blinded randomised trial (level 1b)Completion of imaging; supplemental dosing; level, onset and duration of sedation; side effects; physiological deteriorationRR of completing the ECHO in the chloral hydrate vs the midazolam group 1.03. Chloral hydrate 10.6%, midazolam 5.3% (p=0.111). Highest level of sedation: 94.7% chloral hydrate, 8.3% midazolam. Chloral hydrate had a longer onset and duration of sedation (p<0.001). 14.4% choral hydrate, 4.5% midazolam group (p=0.006). No significant physiological deterioration occurred. Partially completed ECHOs were included in the successfully completed ECHO group. Side effects were mainly vomiting and paradoxical reaction.
Wheeler et al,
40 children (<5 years) requiring ECHO. 25 randomised to receive 75 mg/kg chloral hydrate and 15 randomised to 0.5 mg/kg oral midazolam. Single-blinded randomised trial (level 1b)Completion of imaging; supplemental dosing; level, onset and duration of sedation; side effects; physiological deteriorationRR 1.9 (90% chloral hydrate vs 48% midazolam, p=0.001)*. Chloral hydrate 33%, midazolam 52% (p=0.332). Deeper sedation in the chloral hydrate group (p<0.001). No significant difference in onset. Duration shorter in midazolam group (37.4 vs 80.6 min, p=0.0063). No adverse effects. No physiological deterioration. 3 children given chloral hydrate via nasogastric tube as it was refused orally. Only single-blinded.
Roach et al,
703 children (2–4 years) requiring ECHO. 485 given chloral hydrate (mean dose 76 mg/kg) and 48 given midazolam (mean dose 0.8 mg/kg). The rest given other sedatives. Retrospective comparative study (level 3b)Completion of imaging; level, onset and duration of sedation; complicationsRR 1.4 (97% chloral hydrate vs 67% midazolam). The mean onset in the midazolam group was 32 min and duration 40 min vs 30 and 50 min, respectively, in the chloral hydrate group. Complications were minimal in both groups. A year of the study period was excluded due to a change in the sedation policy. Not all data reported, therefore difficult to find relevant results. Wide range of sedation onset and duration times. Possible information recall bias.
326 patients (<5 years) having CT, MRI or radionuclide study. 310 given chloral hydrate and 16 midazolam in average doses of 65.2 and 0.873 mg/kg, respectively. Retrospective comparative study (level 3b)Completion of imaging; adverse eventsRR 4.6 (87.1% chloral hydrate vs 18.75% midazolam, p<0.001). OR 29.3 although quoted as 35.9. No adverse reactions in the midazolam group, 11 (3.6%) in the chloral hydrate group. Inaccurate statistical calculations. Side effects were mainly vomiting and paradoxical reaction. 1 had prolonged sedation of 2 h 10 min and 1 had a transient self-resolving drop in SaO2 from 99% to 92%. Possible information bias.
Malviya et al,
376 children (4/12–7 years) requiring MRI or CT. 336 given chloral hydrate (64±13 mg/kg) and 40 given midazolam (0.15±0.13 mg/kg). Prospective observational study (level 2b)Delayed side effects18% of the chloral hydrate group experienced anxiety vs 8% of the midazolam group (p=not significant). Chloral hydrate was more commonly associated with imbalance and agitation than midazolam. 3 children who had been given chloral hydrate as a sole sedative, returned to the emergency department due to excessive/prolonged sedation. Children who experienced inadequate sedation were older than those whose sedation was adequate (4.8±3.2 vs 2.7±3.4 years, respectively, p=0.03). Possible bias introduced due to excluding patients lost to follow-up. 10% of the chloral hydrate group were also given midazolam. Not all data were reported for each separate sedation subgroup, therefore comparison is difficult. Some outcomes were subjective.


Children undergoing painless diagnostic imaging including MRI, CT and echocardiography frequently require sedation in order to reduce anxiety, and to facilitate a successful diagnostic study without motion-degraded imaging results. With the increasing availability of such studies, non-anaesthetic trained paediatricians are frequently being requested to provide mild to moderate sedation. An ideal sedating agent for this indication would be a drug that has a rapid onset, a short duration and predictable effects, and is easily administered and well tolerated. Chloral hydrate and midazolam are both commonly used in paediatric sedation. Chloral hydrate is a hypnotic that can be administered by mouth or by rectum and has been used in children since 1894, although there have been concerns about its unpredictable onset and long duration (Buck). Midazolam is a newer, more expensive benzodiazepine that has appealing pharmacodynamics and can be given by the intravenous, oral, intramuscular, intranasal, sublingual, buccal and rectal routes. NICE recommends the use of chloral hydrate or midazolam as a safe sedative for paediatric use for painless procedures provided that the patient's suitability for sedation has been assessed, the sedation is administered and the effects monitored by adequately trained staff, and there is immediate access to resuscitation equipment(NICE)

Four randomised trials have compared the efficacy and safety of chloral hydrate versus midazolam for use in paediatric sedation for painless imaging (D'Agostino, McCarver-May, Layangool, Wheeler).When assessed for methodological rigour, all of the studies had limitations, and there was some heterogeneity between the studies, particularly in terms of the dosing regimens, children's age ranges and imaging type; however, the results were remarkably consistent across the studies. All found that the chloral hydrate group had a significantly higher rate of successful completion of imaging studies, with RRs of 1.03–2.3. Two of the three studies which monitored the need for supplemental dosing following failure of the initial dose to cause sedation did not find a significant difference between the two groups (Layangool, Wheeler). while one study showed a significantly reduced need for further sedation in the chloral hydrate group (McCarver-May). Two studies assessed the level of sedation and both found that chloral hydrate resulted in a deeper level of sedation (Layangool, Wheeler). the result of one study was significant (Wheeler). but the significance was not mentioned in the other (Layangool). Three studies looked at the onset and duration of sedation; two found that chloral hydrate had a significantly longer onset and duration of sedation (McCarver-May, Layangool). while one study did not find a significant difference in the onset of sedation between the two groups but reported a significantly shorter duration in the midazolam group (Wheeler). The study by D’Agostino et al also compared duration of sedation and although the mean duration was longer in the chloral hydrate group, this was not significant. No significant physiological deterioration was mentioned in any of the four trials, other than one patient becoming tachycardic following administration of midazolam (McCarver-May)Layangool et al reported significantly more minor side effects in the chloral hydrate group, and Wheeler reported poor palatability of chloral hydrate which necessitated administration via a nasogastric tube in three of the patients randomised to receive chloral hydrate.

Two large retrospective studies have also compared the safety and effectiveness of chloral hydrate versus midazolam for use in children less than 5 years of age as sedation for painless imaging (Roach, Schmalfuss).These studies also reported chloral hydrate was more effective than midazolam in enabling successful completion of the painless imaging studies, and reported minimal adverse events in both group. One prospective observational study by Malviya et al evaluated the delayed side effects after discharge of children receiving sedation for MRI or CT.The study showed a higher rate of anxiety in the children who received chloral hydrate when compared with those who receive midazolam, although the difference was not significant. The study also reported three cases who returned to the emergency department after being discharged following sedation with chloral hydrate, due to excessive or prolonged sedation.

There is limited evidence available to compare the use of chloral hydrate with midazolam in older children. This may be because chloral hydrate is less well tolerated in older children due to larger volumes causing nausea and vomiting (NICE).

Editor Comment

* Percentages estimated from bar chart. ECHO, echocardiogram; ITT, intention-to-treat.

Clinical Bottom Line

When used as sedation for young children, the evidence suggests that chloral hydrate is more effective than midazolam in facilitating completion of painless imaging studies, although it has a longer onset and duration. (Grade B)

Both chloral hydrate and midazolam have a wide margin of safety when used in limited dosage. (Grade B)

Limited evidence suggests that chloral hydrate may have more associated side effects than midazolam, although these are relatively minor. (Grade B)


  1. The National Institute for Health and Clinical Excellence (NICE). Sedation in children and young people (CG112) Sedation for diagnostic and therapeutic procedures in children and young people, 2010 (accessed 6 Nov 2011).
  2. Buck ML. The use of chloral hydrate in infants and children. Pediatric Pharmacotherapy 2005;11:1–4.
  3. D'Agostino J, Terndrup TE. Chloral hydrate versus midazolam for sedation of children for neuroimaging: a randomized clinical trial. Pediatr Emerg Care 2000;16:1–4.
  4. McCarver-May DG, Kang J, Aouthmany M, et al. Comparison of chloral hydrate and midazolam for sedation of neonates for neuroimaging studies. J Pediatr 1996;128:573–6
  5. Layangool T, Sangtawesin C, Kirawittaya T, et al. A comparison of oral chloral hydrate and sublingual midazolam sedation for echocardiogram in children. J Med Assoc Thai 2008;91 Suppl 3:S45–52.
  6. Wheeler DS, Jensen RA, Poss WB. A randomized, blinded comparison of chloral hydrate and midazolam sedation in children undergoing echocardiography. Clin Pediatr (Phila) 2001;40:381–7.
  7. Roach CL, Husain N, Zabinsky J, et al. Moderate sedation for echocardiography of preschoolers. Pediatr Cardiol 2010;31:469–73.
  8. Schmalfuss I. Oral sedation of pediatric patients for noninvasive radiological procedures: chloral hydrate versus midazolam. J Radiol Nurs 2005;24:42–48.
  9. Malviya S, Voepel-Lewis T, Prochaska G, et al. Prolonged recovery and delayed side effects of sedation for diagnostic imaging studies in children. Pediatrics 2000;105:E42.