C. V. Berwick, S. A. Roberts

Alder Hey Children's Hospital, UK


Craniofacial patients undergo major remodelling surgery such as total vault reconstruction (TVR) and fronto-orbital advancement reconstruction (FOAR) surgery. A database of all patients undergoing these procedures is maintained with the aim of measuring and improving outcomes including vomiting rate, time to drinking and eating, pain and length of stay. Data from this has been used to implement a successful standardised enhanced recovery package of post-operative measures including regular analgesia and antiemetic prophylaxis.


Patients undergoing major craniofacial surgery have high levels of post-operative vomiting, delaying eating and drinking, however there is considerable variation between patients. Post-operative vomiting rates may be influenced by anaesthetic techniques, such as the use of dexmedetomidine1 as the alpha-2 agonist of choice used as part of a multimodal analgesic technique2.

Strategy for change

The craniofacial database was reviewed to identify variations between practice and whether these had any impact on vomiting rates or times to eating and drinking. 151 patients underwent TVR or FOAR over a 4 year period (2020-2023) and survived to discharge. Intra-operative drug dosing, vomiting rates and times to drinking and eating were analysed. Two core variations in anaesthetic technique were identified, use of intra-operative morphine (107 patients mean morphine dose 90mcg/kg, 47 patients no morphine) and choice of alpha-2 agonist (103 patients given clonidine [mean dose 1.57mcg/kg], 47 patients had dexmedetomidine [mean dose 1.44mcg/kg]).

Measure of improvement

Improvements in outcomes and experience were measured in terms of time to drinking and eating and vomiting rates. All groups had similar vomiting rates on post-operative day 0 (POD0: theatre to midnight) at 0.65-0.67. On POD1, vomiting rates were 0.7 in the clonidine patients vs 0.31 in dexmedetomidine patients, while intra-operative morphine had a lesser variation in rates (0.62 [morphine] vs 0.49 [no morphine]). Time to first drink was 06h07m in clonidine patients vs 02h15m in dexmedetomidine patients. Patients who received morphine drank at 05h17m vs 04h02m for those who received no morphine. Time to first meal was 20h24m in clonidine patients vs 13h51m in dexmedetomidine patients, and 19h28m in morphine patients vs. 15h47m in those who received no morphine.

Lessons learnt

Analysis of patient outcome data is vital to identifying the impact of variations in practice, and maintaining a robust database facilitates this. Time to first meal suggests that patients who receive dexmedetomidine drink earlier and have breakfast on POD1, while those who receive clonidine drink later and do not eat until lunchtime on POD1. In toddlers aged 1-2, eating after waking on POD1 is a measure of recovery and comfort.

Message for others

Use of dexmedetomidine rather than clonidine as an alpha-2 agonist appears to be associated with lower vomiting rates and reductions in length of time to first drink and meal.


  1. Reddy SK, Jones JJ, Gordish-Dressman H, Pestieau SR. Dexmedetomidine as an Opioid-Sparing Agent in Pediatric Craniofacial Surgery. 2020 7:68 Children.
  2. Kattail D, Macmillan A, Musavi L, Pedreira R, Faateh M, Cho R, Lopez J, Dorafshar AH. Pain Management for Nonsyndromic Craniosyntosis: Adequate Analgesia in a Pediatric Cohort?. 2018 29:5. The Journal of Craniofacial Surgery.
Scroll to top