SYMPOSIUM: INTENSIVE CARE

Sedation and analgesia in background discomfort of ongoing interventions such as the

presence of the endotracheal tube, and allow for the provision of

critically ill children temporary deeper levels of sedation, with appropriate analgesia,
to facilitate intermittent painful procedures. Depth of sedation
must be titrated to allow for the interaction with and mobiliza-
Ramakrishnan Subramaniam
tion of appropriate patients.
Stephen D Playfor
Current practice
Considerable variation in the use of pharmacological agents for
Abstract analgesia and sedation for critically ill children has been
Effective analgesia and sedation in the Paediatric Intensive Care Unit demonstrated. In a recent prospective observational study of 338
(PICU) encompasses the provision of physical comfort and caring for critically ill children in 20 PICUs throughout the UK, a total of 24
the psychological well-being of critically ill children. In the UK the most different sedative and analgesic agents found to be used. Surveys
commonly used sedative and analgesic agents for critically ill children have shown that the most commonly used sedative and analgesic
are midazolam and morphine; consensus clinical practice guidelines for agents for critically ill children within the UK are midazolam and
the provision of sedation and analgesia in critically ill children were pub- morphine, whereas in the USA midazolam and fentanyl are more
lished in 2006 by the UK Paediatric Intensive Care Society. commonly administered. In critically ill adults, the administra-
It is important to treat pain, and in addition to the obvious immediate tion of continuous infusions of sedative agents has been associ-
effects of untreated pain there is increasing evidence that pain experi- ated with prolonged periods of mechanical ventilation and
enced early in life may result in long-term changes in neurosensory func- a routine daily interruption of intravenous sedative agents is now
tion. There are however emerging concerns that sedative and analgesic recommended practice. This practice has led to a reduction in
agents may themselves be associated with developmental neurotoxicity, duration of mechanical ventilation and reduced duration of
particularly amongst neonates. intensive care admission, without any apparent adverse
The extent and impact of psychological morbidity associated with the psychological effects. Whilst it has been proven that increased
use of sedative and analgesic agents are poorly understood. Withdrawal sedative use in the first 24 h of weaning from mechanical
syndrome and delirium probably represent part of a spectrum of adverse ventilation is associated with subsequent failure of extubation in
psychological effects. There appears to be considerable overlap in the infants and children, a routine daily interruption of intravenous
features of psychological morbidity contributed to by pharmacological sedative agents has yet to be thoroughly evaluated in critically ill
agents, and other disease states encountered in the critically ill. The children where the potential adverse effects of discontinuing
most important single factor in reducing avoidable psychological sedative agents include inadvertent self-extubation, adverse
morbidity in survivors of PICU is to minimize the administered doses of cardiovascular effects and possible negative psychological
sedative and analgesic agents. outcomes.
The introduction of clinical guidelines has been associated
Keywords analgesia; critical care; delirium; midazolam; morphine; with a significant reduction in sedative costs per bed day in adult
sedation; withdrawal critical care units. Consensus clinical practice guidelines for the
provision of sedation and analgesia in critically ill children were
published in 2006 by the UK Paediatric Intensive Care Society.
Providing effective analgesia and sedation for critically ill chil-
dren in the Paediatric Intensive Care Unit (PICU) involves caring Analgesia
for both the physical comfort and the psychological well-being of
Analgesic agents
the child. Correctable environmental and physical factors
Commonly used analgesic agents include opioids for the relief of
causing discomfort should be addressed before the commence-
severe pain, nonsteroidal anti-inflammatory drugs (NSAIDs) for
ment of pharmacological agents. Once an adequate level of
moderate to severe pain, and paracetamol for the treatment of
analgesia has been achieved, additional sedative agents may be
mild to moderate pain. Regional techniques have also been used.
required. The aims of sedation are to reduce anxiety and distress
in the child, to facilitate uncomfortable therapeutic and diag-
Regional analgesia: subcutaneous or topically administered
nostic procedures, and to reduce the risk of inadvertent self-
local anaesthetics such as tetracaine, prilocaine and lignocaine
extubation and other forms of so-called ‘treatment interference’.
are used in critically ill children for the short-term relief of
Analgesia and sedation strategies must account for the
painful procedures. Regional techniques such as epidural anal-
gesia may be particularly effective in the post-operative patient,
whilst peripheral nerve blocks are used less frequently in PICU.
Ramakrishnan Subramaniam MD is a Clinical Fellow in the Paediatric
Intensive Care Unit, Royal Manchester Children’s Hospital, Oxford Road Opioids: these produce analgesia via a variety of central and
Manchester M13 9WL, UK. Conflict of interest: none. peripheral opioid receptors, particularly the mu and kappa
receptors. It is believed that interaction at other receptors may be
Stephen D Playfor DM is a Consultant Paediatric Intensivist in the responsible for the adverse effects associated with these agents.
Paediatric Intensive Care Unit, Royal Manchester Children’s Hospital, The pharmacokinetics and pharmacodynamics of systemic
Oxford Road Manchester M13 9WL, UK. Conflict of interest: none. analgesic agents vary with age. Neonates frequently demonstrate

PAEDIATRICS AND CHILD HEALTH 21:4 177 Ó 2010 Elsevier Ltd. All rights reserved.
 SYMPOSIUM: INTENSIVE CARE

reduced clearance of agents because of hepatic enzyme system NSAIDs and paracetamol: NSAIDs provide analgesia through
immaturity, whilst older children may demonstrate higher the non-selective, competitive inhibition of cyclo-oxygenase,
weight-indexed clearance than adults because of their relatively a critical enzyme in the inflammatory cascade. Although the
large liver mass. administration of NSAIDs has been shown to significantly reduce
opioid requirements in adult and paediatric pain after surgery by
Morphine e is the only poorly lipid soluble opioid in common approximately 15e30%, the analgesic benefits of NSAIDs have
use. When administered in a single dose of 0.1 mg/kg intrave- yet to be systematically studied in critically ill children.
nously, its peak analgesic effect occurs after 20 min and its Paracetamol is an analgesic used to treat mild to moderate pain.
duration of action is approximately 4 h. Morphine undergoes When used in combination with opioid agents, paracetamol
extensive hepatic and extra-hepatic glucuronidation and metab- produces a greater analgesic effect than higher doses of opioid
olites are excreted primarily in the urine. The elimination of alone, and has also been shown to have an opioid-sparing effect
morphine from the body is slow and quantitatively different in in adults.
newborns, but equates towards adult values within the first 6
months of life. Morphine may stimulate the release of significant Long-term effects of early pain
amounts of histamine and inhibits compensatory sympathetic In addition to the obvious immediate physiological effects of
responses; hence the vasodilation produced by morphine may untreated pain there is increasing evidence that pain experienced
result in significant hypotension particularly after bolus admin- early in life can result in long-term changes in neurosensory func-
istration. Discontinuation of morphine infusions has been asso- tion. The likely reason for this is that the nervous system is sensitive
ciated with withdrawal phenomena, which may include pupillary to changes levels of neuronal activity. Similarly, increased levels of
dilatation, lacrimation, sweating, goose pimples on the skin, neuronal activity due to pain or injury may lead to changes in the
hypertension, pyrexia, vomiting, abdominal pain, diarrhoea, subsequent pattern of neural connectivity and sensitivity.
muscle and joint pains, and behavioural changes. Any long-term changes in neurosensory function produced as
a result of exposure to pain in early life depend on multiple
Fentanyl e is a synthetic opioid with approximately 100 times factors and mechanisms including the following:
the analgesic potency of morphine. It is highly lipid soluble,  The type, duration and severity of the pain experienced in
which accounts for its rapid onset of action. Fentanyl adminis- early life. It has been demonstrated that ex-preterm infants
tration causes less histamine release than morphine, and there- who have undergone surgery in addition to medical neonatal
fore less hypotension. When given intravenously, fentanyl has intensive care have a greater degree of subsequent sensory
a relatively short half-time of 30e60 min owing to rapid redis- change. Similarly, it has been demonstrated that the severity
tribution into peripheral compartments. With prolonged admin- of a burn injury experienced in infancy can influence sensi-
istration, there is accumulation within these peripheral tivity to mechanical and thermal stimulation in later life.
compartments causing an increase in the context sensitive half-  Timing of the pain. Pain pathways undergo significant
time and tolerance may rapidly develop. Metabolism occurs structural and functional changes during postnatal develop-
almost exclusively in the liver and clearance is markedly affected ment; therefore the age at which pain is experienced influ-
by hepatic blood flow. Fentanyl has no active metabolites and ences later effects. These changes include alterations in the
does not cross-reactivity in patients with morphine allergy. expression and distribution of neurotransmitters and recep-
tors, changes in the distribution of sensory fibres throughout
Remifentanil e is an ultra-short acting synthetic opioid, the spinal cord, and changes in the balance of excitatory and
which acts as a pure mu-receptor agonist. Its cardiorespiratory inhibitory modulation. Hence, surgical incisions occurring in
effects are similar to those of the other opioids. This drug has an the first week of life have been associated with long-term
exceptionally short half-time of 3 min in all age groups as it is changes in sensory processing, but not if they occur later in
metabolized by plasma and tissue esterases with a very small life.
volume of distribution. The effects of remifentanil dissipate  Later psychosocial and environmental factors. Any assess-
rapidly, even after prolonged infusion, giving it a very short ment of the long-term effects of early pain requires the
context sensitive half-time. Remifentanil has been used to repeated assessment of children over prolonged periods of
provide ongoing analgesia in PICU although prolonged use of this time, which may in itself influence responses to noxious
agent is associated with the rapid development of tolerance and stimuli experienced later in life.
relatively high cost. This agent may have more potential for
procedural analgesia in the critical care setting given its rapid
Sedation
onset and offset times and effective blunting of airway reflexes,
although respiratory and cardiovascular depressant effects Sedative agents
should be anticipated in this setting. Benzodiazepines: these have specific activity at gamma-amino-
When used in 17 children for sedation during painful proce- butyric acid (GABA) receptors, which form part of the major
dures, Bauman and colleagues found an unacceptably high inci- inhibitory system of the central nervous system. The most
dence of life-threatening respiratory depression at subtherapeutic commonly used benzodiazepines for sedation in PICU are mid-
levels of the drug. This feature may make remifentanil more suitable azolam, lorazepam, and diazepam.
for facilitating painful procedures in situations where the airway is
already protected, either in the mechanically ventilated PICU Midazolam e is a sedative agent which also produces ante-
patient or during procedures such as fiberoptic bronchoscopy. grade amnesia and anxiolysis. It is available as a water-soluble

PAEDIATRICS AND CHILD HEALTH 21:4 178 Ó 2010 Elsevier Ltd. All rights reserved.
 SYMPOSIUM: INTENSIVE CARE

acidic preparation, which at plasma pH converts into an union- include hypnotic agents such as chloral hydrate or triclofos
ized form that rapidly crosses the blood brain barrier and acts sodium, and the sedating antihistamines such as promethazine
through GABAA receptors. After a single intravenous bolus or alimemazine.
injection, the time to peak sedation is 5e10 min with a duration The combination of chloral hydrate and promethazine has
of action of 30e120 min. When given by continuous intravenous been shown to be more effective than intravenous midazolam in
infusion however, the duration of action is significantly longer providing maintenance sedation in critically ill children. Chloral
and, after prolonged administration, sedation effects may persist hydrate is rapidly absorbed from the gastrointestinal tract and
for 48 h after discontinuation of the agent. starts to act within 15e60 min. It is converted to the active
Midazolam is metabolized to 1-hydroxymidazolam and 1,4- metabolite trichloroethanol and is metabolized in the liver and
dihydroxymidazolam by cytochrome P450 isoenzyme 3A4 subsequently excreted in the urine and bile. The usual duration
hydroxylation, and then undergoes glucuronidation; it has the of action is around 60e120 min, but this may be prolonged in
shortest elimination half-time of the benzodiazepines. renal or hepatic disease and there is a risk of drug accumulation
Accumulation of active metabolites may produce prolonged with repeated high doses. Gastrointestinal irritation is the most
sedative effects in patients with renal insufficiency, and substrate commonly reported adverse effect, and this is less commonly
competition for cytochrome P450 isoenzyme 3A4 may also occur seen with triclofos sodium.
leading to prolonged sedation after the co-administration of
certain pharmacological agents such as erythromycin. The main Assessment of analgesia and sedation
adverse events of midazolam include the development of toler-
Pain assessment
ance, dependence, and withdrawal after subsequent discontinu-
Pain is a subjective experience and, provided there is no clear
ation. Vasodilatation and hypotension may occur, particularly
reason to doubt them, a patient’s self-report is the single most
after bolus administration in the setting of hypovolaemia. There
reliable indicator of pain. Although pain-related behaviours and
is also evidence of reduced sedative efficacy when midazolam is
the physiological indicators of pain are neither sensitive nor
administered to younger infants.
specific to pain, these features should be routinely documented,
particularly in those who are unable to communicate normally.
Clonidine e is a sedative agent being used with increasing
In neonates, infants, and children under the age of 3 years,
frequency to provide sedation in critically ill children in the UK. It
behavioural observational scales are the main tools available for
is an alpha2-adrenoreceptor agonist which produces sedation
the assessment of pain. These scales frequently rely on facial
without causing respiratory depression, and also produces an
expression, motor responses, and physiological parameters.
anxiolytic effect similar to that of benzodiazepines. Alpha2-
Children aged between 3 and 8 years are generally able to use
adrenoreceptor agonists can reduce the requirement for other
self-reporting techniques such as ‘faces scales’ using either
sedatives and can improve haemodynamic and sympathoadrenal
photographs or drawings of faces, although their application in
stability. They also have some analgesic properties, which may
the critical care environment is frequently difficult and largely
be mediated through the prevention of substance P release.
unvalidated. Competent children over the age of 8 years can
Adverse effects associated with the use of clonidine include
usually use one the validated unidimensional pain tools, such as
bradycardia and hypotension. Withdrawal after prolonged
the verbal rating scale, a visual analogue scale, or a numeric
administration has been associated with hypertension and
rating scale in the same way as adult subjects.
seizures; abrupt discontinuation of clonidine should therefore be
avoided. Sedation assessment
In order to avoid the adverse consequences of both excessive and
Ketamine e is a dissociative anaesthetic agent, structurally inadequate sedation, the level of sedation of critically ill children
similar to phencyclidine, which produces a cataleptic trance-like should be regularly assessed and documented. The use of
state by inducing an electrophysiological dissociation between a formal sedation assessment scale is recommended, wherever
the limbic and thalamoneocortical systems. It is the only possible using a validated scoring system such as the COMFORT
commonly-used agent that produces significant levels of both scale.
sedation and analgesia. Ketamine has a rapid onset of action and The COMFORT scale is a subjective physiological and
a short duration of action owing to its short redistribution half- behavioural scoring system, which requires no disturbance to the
life of 5 min; the elimination half-life is 130 min. Ketamine patient. It measures eight variables: mean arterial blood pres-
produces virtually no respiratory depression, does not blunt sure, heart rate, muscle tone, facial tension, alertness, calmness/
airway reflexes and induces bronchodilation. In addition, func- agitation, respiratory behaviour, and physical movement during
tional residual capacity, minute ventilation and tidal volume are a 2-min period of observation. Doses of sedative agents should be
also unaffected. Ketamine has fewer cardiac side effects than titrated to produce the depth of sedation appropriate for the
other sedative agents primarily because it tends to stimulate clinical situation.
endogenous catecholamine release, leading to increases in heart
rate, blood pressure and cardiac output. Neurophysiological monitors
Objective measurement of sedation using neurophysiological
Enteral sedative agents techniques such as the bispectral index (BIS) or auditory-evoked
It is a common practice in the UK to administer enteral sedative potentials offers the potential to assess sedation without
agents to critically ill children in PICU when this route of depending on patient responsiveness and could therefore be used
administration is available. The most commonly used drugs to differentiate deeper levels of sedation or during the

PAEDIATRICS AND CHILD HEALTH 21:4 179 Ó 2010 Elsevier Ltd. All rights reserved.
 SYMPOSIUM: INTENSIVE CARE

administration of neuromuscular blocking agents. BIS analyses For certain sedative and analgesic agents the incidence of
the electroencephalogram (EEG) and uses a digital scale from withdrawal syndrome is related to the total dose received; Katz
0 (isoelectric EEG) to 100 (completely awake). Use of the BIS has demonstrated that a total dose of >4 mg/kg during a PICU
score has been found to reliably differentiate between inadequate admission was associated with a 100% incidence of withdrawal
and adequate levels of sedation, but appears to be relatively syndrome. This may also explain why Jenkins and colleagues
insensitive in distinguishing between adequate and overly deep found that routine tapering of sedative and analgesic
sedation. Comparisons between BIS and COMFORT scale agents, which would increase the total dose received, was
measurements during a prolonged critical care admission are not associated with an increase in the incidence of withdrawal
well correlated, and BIS scores may vary between patients at the syndrome.
same subjective level of sedation, particularly at the deeper levels Although tolerance, physical dependence, and eventual
of sedation. BIS also has technical limitations in PICU as a result withdrawal syndrome can be anticipated when patients have
of polypharmacy and electrical interference. Currently there is been given high doses or prolonged infusions of opioids and
insufficient evidence to support the routine use of BIS monitoring sedative agents, the exact cellular mechanisms responsible for
in PICU. the development of withdrawal remain unclear.

Psychological morbidity Delirium

The extent and impact of psychological morbidity associated Within adult Critical Care Units diagnostic tools for the assess-
with the use of sedative and analgesic agents are poorly under- ment of delirium, such as the Confusion Assessment Method for
stood. Within PICU researchers have focused on withdrawal the ICU (CAM-ICU), are in common use. The incidence of
syndrome, whilst in adult critical care units delirium is much delirium is known to increase amongst older patients and is
more commonly diagnosed. Whilst separate diagnostic tools are associated with persistent psychological morbidity and outcome;
available to facilitate the diagnosis of withdrawal syndrome and the number of days of ICU delirium has associated with higher
delirium in critically ill children, reference to the International 1-year mortality after adjustment for relevant covariates amongst
Classification of Diseases (ICD-10), suggests that the two diag- an older ICU population.
noses represent part of a spectrum: Features of withdrawal More recently, Schieveld has demonstrated a significant
(agitation, anxiety, hallucination and seizures, for example) may burden of delirium amongst the PICU population, again with an
be termed delirium if accompanied by ‘concurrent disturbances increasing incidence with age: a paediatric neuropsychiatrist
of consciousness and attention, perception, thinking, memory, diagnosing delirium in 4% of PICU patients aged 3e6 years
psychomotor behaviour, emotion, and the sleep-wake schedule’. increasing to 20% in the 15e18-year-old group. Currently there
There appears to be considerable overlap in the features of are no validated instruments, equivalent to the CAMeICU, for
psychological morbidity contributed to by pharmacological the diagnosis of delirium amongst PICU patients.
agents, and other disease states encountered in the critically ill
such as sepsis and metabolic derangement. The most severe of Developmental neurotoxicity of sedative and analgesic
these cases generating a nonspecific organic cerebral syndrome agents
termed delirium. There are concerns that the administration of sedative and
analgesic agents may be associated with developmental neuro-
Withdrawal syndrome toxicity, particularly amongst neonates. A growing number of
A withdrawal syndrome may occur after the discontinuation or animal studies have demonstrated widespread structural damage
reduction in the administered dose of sedative and analgesic of the developing brain and long-lasting neurocognitive abnor-
agents, particularly the benzodiazepines and opioids. Features of malities after exposure to sedatives such as ketamine, midazolam
withdrawal syndrome usually occur within a few hours of and chloral hydrate. These studies suggest a dose and exposure-
stopping the offending drug. These features commonly include: time dependence of neuronal cell death.
(i) central nervous system manifestations (such as agitation, Significant reductions in the level of neuronal activity, medi-
anxiety, seizures, hallucinations, and psychosis), and (ii) auto- ated through general anaesthesia have been shown to trigger
nomic features (such as vomiting, tachycardia, hypertension, apoptosis, and preterm infants with patent ductus arteriosus
and fever). Until recently assessment and management of have been noted to have a poorer neurodevelopmental outcome
withdrawal syndrome has been hampered by the lack of a vali- if managed surgically rather than conservatively.
dated PICU assessment system with many clinicians and It has also been suggested that prolonged administration of
researchers using modified neonatal abstinence syndrome midazolam is associated with poorer neurodevelopmental
scoring systems. outcomes in mechanically ventilated neonates. Roze and
More recently, Ista and colleagues in Holland have con- colleagues however found, in a large cohort study 1572 prema-
structed the Sophia Observation withdrawal Symptoms-scale ture infants who received mechanical ventilation, that there was
(SOS) by attempting to identify the underlying empirical struc- no association between sedative or analgesic agents and outcome
ture of co-occurrences of withdrawal symptoms that PICU after correction for gestational age and other neonatal factors.
physicians and nurses considered to be of relevance in critically
ill children. Agitation, anxiety, inconsolable crying, increased Strategies to reduce psychological morbidity
muscle tension, tremors, tachycardia and sweating were The most important factor in reducing the avoidable incidence of
considered relevant by 85e95% of the experts, and a final list of psychological morbidity amongst patients in PICU is to reduce
15 features were included in the final instrument. the administration of sedative and analgesic agents as much as

PAEDIATRICS AND CHILD HEALTH 21:4 180 Ó 2010 Elsevier Ltd. All rights reserved.
 SYMPOSIUM: INTENSIVE CARE

possible. Clinical guidelines for sedation and analgesia should be Playfor SD. The use of bispectral index monitors in paediatric intensive
used, and doses of pharmacological agents should be titrated care. Crit Care 2005; 59: 25e6.
using pain and sedation scoring systems. The use of lighter levels Playfor S, Jenkins I, Boyles C, et al. United Kingdom Paediatric Intensive
of sedation has been associated with fewer days of mechanical Care Society SedationAnalgesia and Neuromuscular Blockade Working
ventilation, reduced ICU stay and no increase in psychological Group. Consensus guidelines on sedation and analgesia in critically ill
morbidity in critically ill adults. children. Intensive Care Med 2006; 32: 1125e36.
There is little evidence to support the practice of routine Playfor SD, Thomas DA, Choonara I. Sedation and neuromuscular
tapering of sedative and analgesic agents, or the planned blockade in paediatric intensive care; a review of current practice in
substitution of one class of agent for another (drug holidays), in the United Kingdom. Paediatr Anaesth 2003; 13: 147e51.
reducing the subsequent incidence of withdrawal. A Schieveld JN, Leroy PL, van Os J, Nicolai J, Vos GD, Leentjens AF. Pediatric
delirium in critical illness: phenomenology, clinical correlates and
treatment response in 40 cases in the pediatric intensive care unit.
Intensive Care Med 2007; 33: 1033e40.
FURTHER READING
Ambuel B, Hamlett KW, Marx CM, Blumer JL. Assessing distress in
pediatric intensive care environments: the COMFORT scale. J Pediatr
Psychol 1992; 17: 95e109.
Practice points
Finnegan LP, Connaughton Jr JF, Kron RE, Emich JP. Neonatal abstinence
syndrome: assessment and management. Addict Dis 1975; 2: 141e58. C Correctable environmental and physical factors causing
Ista E, van Dijk M, de Hoog M, Tibboel D, Duivenvoorden HJ. Construction
discomfort should be addressed before starting pharmaco-
of the Sophia Observation withdrawal Symptoms-scale (SOS) for
logical agents.
critically ill children. Intensive Care Med 2009; 35: 1075e81. C Consensus clinical practice guidelines for the provision of
Jenkins IA, Playfor SD, Bevan C, Davies G, Wolf AR. Current United
sedation and analgesia in critically ill children are available
Kingdom sedation practice in pediatric intensive care. Paediatr
from the UK Paediatric Intensive Care Society.
Anaesth 2007; 17: 675e83. C Routine assessment of pain and depth of sedation should
Kress JP, Pohlman AS, O’Connor MF, Hall JB. Daily interruption of sedative
form part of standard PICU practice.
infusions in critically ill patients undergoing mechanical ventilation. C Clinicians should be aware of the risk of psychological
N Engl J Med 2000; 342: 1471e7.
morbidity associated with the use of sedative and analgesic
Parkinson L, Hughes J, Gill A, Billingham I, Ratcliffe J, Choonara I. A
agents and should minimize the use of these agents
randomized controlled trial of sedation in the critically ill. Paediatr
Anaesth 1997; 7: 405e10.

PAEDIATRICS AND CHILD HEALTH 21:4 181 Ó 2010 Elsevier Ltd. All rights reserved.