Towards evidence based medicine for paediatricians

We’re all well aware of the problems of doing randomised clinical trials in paediatrics but are we as aware of some alternative approaches that have been used? ‘Sequential design’ studies look at comparing a series of treatments against each other, switching to the ‘better’ arm and comparing against the next candidate as time progresses. They need quickly and easily available outcomes and tend to be usable only for short-course treatments but they’ve been estimated to reduce sample sizes by about 25%. ‘Adaptive design’ begins by pitting all the contenders against each other and dropping off the weakest. It can then roll onwards with the final two into a more ‘traditional’ trial design. Again, this needs pretty rapid outcomes and fairly short-course treatments, but reduces numbers to only half of those needed in repeated traditional design trials. ‘Response adapted’ designs are intriguing. These include the withdrawal-enriched design—where everyone gets the treatment to start with, then those who show response are randomised to having placebo or carrying on with therapy. This is tricky with ethics—if you’ve shown it to work can you stop?—and has the risk of carryover effects, and doesn’t work for curative treatments. Another variation is the threestage design, which starts with a randomisation to treatment/placebo. Stage 2 is taking the non-responders to placebo, and randomising those again to placebo/treatment, and stage 3, where only those who have responded to treatment are taken and randomised to placebo/carry-on (like the withdrawal design). This works well with chronic conditions where therapy is not disease modifying, but changes outcomes for a short period. ‘Placebo-phase’ randomisations are appealing for therapies with a good chance of success. These randomise participants to all receive the treatment ... the randomisation is about ‘when’ to get it ... with variable placebo phases before treatment is commenced. This offers the opportunity to treat every child, and still manage to estimate the effect of a therapy. This does rely on (a) you not requiring immediate disease-altering treatment and (b) the disease process not changing massively over time so the ‘later’ ones have more disease to get rid of than the earlier ones. ‘Bayesian approaches’ are the oddest to think about. They start from the principle that we do not have a 50% knowledge about a therapy, but based on adult evidence and possibly prior phase II child studies, we actually believe it’s about 60%–70% (or whatever) effective. The trial analysis then ‘builds’ on this ‘prior knowledge’ to come up with an answer based on the assumptions and the data. This can greatly reduce the amount of people needed in the trial, but can feel very suspect.


QUESTION 1
What is the best sedative to give as premedication for neonatal intubation? Scenario A baby on the neonatal unit requires endotracheal intubation. You plan to give sedation and analgesia to reduce pain and distress, to reduce cardiovascular disturbance and to give better conditions for successful intubation. You are familiar with commonly used premedication regimens using morphine or fentanyl, but wonder if there are better alternatives.

Structured clinical queStion
In neonates requiring intubation (population), is there a better sedative/analgesic (intervention) than the most common premedications (fentanyl or morphine, with atropine and suxamethonium) to reduce patient distress, reduce cardiovascular instability, and either increase success rates or improve conditions for intubation (outcome)?

Search
Cochrane Library: two relevant found, the first withdrawn from the Cochrane Library 1 and the second 2 only examines one study by Ghanta et al. 3 PubMed/Medline and Embase search: (neonat* OR newborn OR new-born) AND (pre-medication OR premedication OR induction OR sedation OR analgesia OR painkiller OR painkiller) AND (intubation).
Abstracts obtained and reviewed, PubMed/Medline search: 574, Embase: 51; 20 additional abstracts reviewed based on references from review articles, 563 excluded as not relevant based on abstracts, 8 excluded as foreign language.

Summary commentary
It is now accepted that premedication should be used to reduce pain and marked physiological changes caused by awake intubation of neonates. 4 In the UK opioid sedatives are commonly used, most frequently morphine or fentanyl, with atropine and suxamethonium. 5 6 There is some evidence that these regimens increase ease of intubation. [7][8][9] However morphine has a long onset time of 5-15 min, which may lead to delay or inappropriately early laryngoscopy, 10 11 and has prolonged effects on neonates for up to 24 hours; 12 fentanyl is faster acting but can cause chest wall rigidity and laryngospasm, treated with naloxone 13 or muscle relaxant. 4 Table 1 shows randomised controlled trials (RCTs) of other agents whose control arms use common premedication regimens, comparable with practice in the UK, with key outcomes. A summary of evidence, including other studies, is given below.
Thiopentone is a fast-acting barbiturate used for rapid sequence induction in anaesthesia. 14 In the neonatal unit 15 and in surgical theatres, [16][17][18][19] thiopentone gives better conditions than awake intubation, with stable cardiovascular parameters and anterior fontanelle pressures. In combination with glycopyrrolate, suxamethonium and remifentanil, it provides easier intubation with greater cardiovascular stability compared with morphine with atropine; 20 however, this administration is complex, causing errors and overdoses. A single prospective study suggests another barbiturate, methohexital, may also be effective, although with adverse events, such as twitching, hiccoughs and desaturation. 21 Propofol is a rapidly acting anaesthetic induction agent that suppresses airway reflexes 22 and can cause pain on injection. 23 Small RCTs have shown similar efficacy to morphine with no difference in observations, and decreased preparation and intubation times. 3 24 Similarly propofol with remifentanil seems equally effective to midazolam and remifentanil. 25 However examining non-randomised studies, significant numbers of preterm neonates experience hypotension even at low doses, requiring fluid boluses and even inotropes. [26][27][28][29] Intraventricular haemorrhage is a theoretical though yet unproven risk, as is persistent fetal circulation. 30 Propofol therefore may be effective, but further proof of safety is required.
Midazolam, a benzodiazepine, is superior to awake intubation when used in combination with morphine 31 or fentanyl; 32 33 however, one study noted a high rate of desaturations and chest compressions, 34 possibly benzyl alcohol excipients are to blame. 35 Nasal midazolam in the delivery room may also be effective but associated with hypotension and myoclonus. 36 Overall these complications are concerning, and evidence of superiority is awaited.
Remifentanil is a rapid-onset synthetic opioid and provides good intubating conditions and stable observations but less effective analgesia compared with morphine and midazolam. 37 38 However fentanyl with suxamethonium, a more generalisable comparison, provides better conditions than remifentanil alone, possibly due to the muscle relaxant. 39 There is also some limited evidence of efficacy for alfentanil 40 and sufentanil, 41 although no evidence of superiority over fentanyl. Chest wall rigidity has been observed with remifentanil, 42 43 and sufentanil, although use of muscle relaxants may reduce this with all synthetic opioids.
Ketamine is a dissociative anaesthetic and provides analgesia without cardiovascular compromise, compared with awake intubation; 44 intravenous, rectal or nasal routes can be used, therefore useful in the delivery room, although evidence remains limited.
Although anaesthetic gases do not require intravenous access either and are effective compared with awake intubation, 17 45 46 they require an anaesthetic machine, a considerable barrier. Nitrous oxide is simpler to administer and archimedes clinical bottom line Commonly used premedication regimens using fentanyl or morphine seem to work equally well as using alternative sedatives (Grade D).
provides sedation without cardiovascular compromise, but agitation, retching and desaturations have been noted. 47 Long-term outcomes are unknown; in vitro and animal models suggest that morphine, ketamine, propofol, fentanyl and midazolam may have damaging effects on the developing brain, although in human preterm neonates their effects remain unknown. 48 In conclusion the optimum drug for sedation is not known, partly due to heterogeneity of trials and paucity of evidence. More studies of high risk preterms <24 hours old are necessary, alternative routes of administration should be explored, and INSURE (intubation-surfactant-extubation) and minimally invasive surfactant administration also require study. 49 50 Perhaps with greater understanding of the pharmacokinetics and pharmacodynamics of each agent, neonatologists will be able to provide balanced combinations to provide rapid-onset analgesia and optimal safe conditions for intubation.