|
Analgesics for the Treatment of Pain in Children
Charles B. Berde, M.D., Ph.D., and Navil F. Sethna, M.B.,
Ch.B.
Treatment of pain and suffering should be a priority for all clinicians.
Previous reviews1
decried inadequate treatment of pain in infants and children.
Surveys in the 1970s and 1980s2
reported that infants and children were less likely to receive postoperative
analgesics than adults. In that era, some neonates underwent surgery
with minimal anesthesia,3
although this practice received some criticism.4
Studies over the past 15 years suggest that neonates, infants,
and children can receive analgesia and anesthesia safely, with proper
age-related adjustments in clinical practice and dosing. Although
the emphasis in this review is on the pharmacologic management of
pain, several nonpharmacologic approaches, including hypnosis and
related cognitive behavioral approaches, have had good efficacy in
children with acute or chronic pain.5,6
Making the hospital environment a less terrifying place may reduce
anxiety and fear, which can themselves exacerbate pain.7
Conversely, nonpharmacologic approaches should not be used as an
excuse to withhold appropriate analgesics.
Development of Nociception
Recent studies of the developmental neurobiology of pain have
been reviewed elsewhere.8
Such studies indicate that neonates have considerable maturation of
peripheral, spinal, and supraspinal afferent pain transmission by 26
weeks of gestation9;
respond to tissue injury with specific behavior and with autonomic,
hormonal, and metabolic signs of stress and distress10;
and develop descending inhibitory pathways later than afferent
excitatory pathways.11
Several studies12,13
have examined whether untreated pain in neonates has prolonged
behavioral consequences. Infants who were circumcised without
anesthesia as neonates showed increased distress during routine
immunizations at four to six months of age, as compared with
uncircumcised infants or with those who were circumcised as neonates
with the use of a topical local anesthetic.13
These observations are intriguing, although interpretation should be
circumspect, pending replication and longer-term controlled studies.
Among children with newly diagnosed cancer, those who had inadequate
analgesia during a first bone marrow aspiration or lumbar puncture
showed more severe distress during subsequent procedures than those
who received a potent opioid (oral transmucosal fentanyl citrate)
during the first procedure.14
Developmental Issues in Pain Assessment and
Measurement
Children eight or more years of age can generally use
visual-analogue pain scales used by adults, which involve rating the
intensity of pain on a horizontal ruler. For children from three to
eight years old, self-reported measures use either face scales
(series of photographs15
or drawings16,17,18
of faces showing increasing degrees of distress) or color-analogue
scales (rulers with increasing intensity of red color signifying
increasing intensity of pain).19
Good agreement was reported between the results obtained with a
photographic face scale and those obtained with a color-analogue scale
among three-to-seven-year-old children who had undergone surgery.20
Behavioral observational scales are the primary methods of pain
assessment for neonates, infants, and children under four years of
age or for children with developmental disabilities.21
Such scales may score facial expressions,22
limb and trunk motor responses, verbal responses, or combinations of
behavioral and autonomic measures.23
Some of these scales record "distress," which reflects
fear and anxiety as well as pain.24
Behavioral scales may underrepresent the intensity of persistent
pain, as compared with self-reports.20
Physiological indexes of pain are useful during surgery and intensive
care, although they may be nonspecific. For example, tachycardia may
be caused by hypovolemia or hypoxemia, rather than pain. Thus, pain
assessment in neonates, infants, and children under four years of
age and in children with major disabilities remains a challenge.
When clinical signs are unclear, therapeutic trials of comfort
measures, feeding, and analgesics may clarify the sources of
distress.
General Aspects of Developmental Pharmacology
The pharmacokinetics and pharmacodynamics of analgesics change
during development. Age-related trends in several physiological variables
relevant to drug action are summarized in Table 1. Different
hepatic-enzyme systems for drug metabolism mature at different
rates,25
accounting for many of the observed findings.
|
Neonates have reduced clearance (normalized to body weight) of many drugs, as
compared with infants, children, and adults, largely because of the
incomplete maturation of their hepatic-enzyme systems. In contrast,
children two to six years of age have greater weight-normalized clearance than
adults for many drugs. Higher rates of drug metabolism by cytochrome
P-450 in children than in adults are ascribed to a larger liver mass
per kilogram of body weight, rather than to age-related changes in
intrinsic enzyme catalytic rates.26
More rapid drug clearance in children than in adults may mean that
more frequent drug dosing is required. For example, a
sustained-release oral morphine formulation used twice daily in
adults requires thrice-daily dosing in children.27
Genetic variability in drug metabolism can either enhance or diminish
the analgesic effects of drugs in different persons. For example,
genetic absence of cytochrome P-450 subtype 2D6, which converts
codeine to morphine, may render codeine ineffective as an analgesic.28
Renal blood flow, glomerular filtration, and tubular secretion
increase in the first weeks of life, approaching adult values by
8 to 12 months. Renal drug clearance may be particularly decreased
in preterm neonates.29
There are age-related differences in body composition. The
fraction of body weight due to water is greater in neonates than in
older children. In neonates, a larger fraction of body mass consists
of highly perfused tissues, including brain, heart, and viscera, and
a lower fraction consists of muscle and fat. Neonates have lower
plasma concentrations of proteins that bind drugs, including 1-acid
glycoprotein and albumin. For drugs with a high degree of protein
binding, the lower plasma protein concentrations in neonates may
lead to an increased fraction of free (unbound) drug and thus to
increased drug effect or increased toxicity.
Age-related changes in protein binding of drugs and in brain lipid
content may alter drug partitioning and cerebrospinal fluid–blood or
brain–blood concentration ratios, independently of changes in the
permeability of the blood–brain barrier. Drug entry into the central
nervous system depends not only on passive permeation, but also on
specific carriers for either uptake or exclusion, such as
P-glycoproteins.30
Children make up a comparatively small market for pharmaceutical
companies, which have historically been reluctant to conduct pediatric
clinical trials.31
Pediatric trials are important for defining how infants and children
respond to medications and for identifying age-specific toxic
effects. A series of federal laws and policies issued over the past
seven years to encourage pediatric trials culminated in the
"final rule," issued in 2000 and still, at this writing,
in effect. Pediatric trials are mandated for all new drugs that, on
review by the Food and Drug Administration, are determined to have
potential clinical value for a sufficient number of newborns,
infants, children, or adolescents.
Most drugs are packaged primarily for adult use, and dose
calculations or serial dilutions may produce medication errors.
Common patterns of pediatric drug errors32
include milligram–microgram errors, decimal-point errors, confusion
between daily dose and fractional dose (e.g., 100 mg per kilogram
per day divided every six hours vs. 100 mg per kilogram per dose
every six hours), and dilution errors.
Acetaminophen, Aspirin, and Nonsteroidal
Antiinflammatory Drugs
Pediatric use of aspirin has declined since the 1970s, after reports
of its association with Reye's hepatic encephalopathy.33
Aspirin remains useful for rheumatologic conditions and for inhibition
of platelet adhesiveness. A comparison of aspirin with ibuprofen in
childhood arthritis found that both were equally effective, but that
there was better compliance and fewer adverse reactions with
ibuprofen.34
The recommended aspirin dosage is 10 to 15 mg per kilogram every
four hours by mouth. Therapeutic plasma aspirin concentrations for
fever control are 15 to 20 mg per deciliter. Dosage guidelines for
aspirin, acetaminophen, and the nonsteroidal antiinflammatory drugs
(NSAIDs) ibuprofen and naproxen are summarized in Table 2.
|
Acetaminophen (paracetamol) has supplanted aspirin as the most widely
used antipyretic and mild analgesic for children. The plasma
concentrations effective for fever control and analgesia35
are 10 to 20 µg per milliliter. The recommended oral dosage is
10 to 15 mg per kilogram every four hours for children. Rectal administration
produces delayed and variable uptake; single doses of 35 to 45 mg
per kilogram generally produce therapeutic plasma concentrations,36
with prolonged clearance. Subsequent rectal doses should be smaller
(e.g., 20 mg per kilogram), and the interval between doses should be
extended to at least six to eight hours.36,37
Single rectal doses of 20 mg per kilogram produced safe plasma
concentrations in preterm neonates.38
Daily cumulative acetaminophen doses by the oral or rectal route
should not exceed 100 mg per kilogram per day for children, 75
mg per kilogram for infants, 60 mg per kilogram for term and preterm
neonates beyond 32 weeks of postconceptional age, and 40 mg per
kilogram for preterm neonates from 28 to 32 weeks of
postconceptional age. An appropriate rectal regimen for a preterm
neonate 30 weeks of postconceptional age would be 20 mg per kilogram
every 12 hours. Excessive dosing has produced hepatic failure in
both infants and children.
NSAIDs are widely used for children. Systematic reviews have found
few differences among NSAIDs for analgesia in adults and little
advantage of injected over oral administration.39
Pharmacokinetic studies of several NSAIDs in children found
weight-normalized clearance and volumes of distribution greater than
those in adults, but similar elimination half-lives.40
Adverse gastrointestinal or renal events from short-term use of
either ibuprofen or acetaminophen appear to be quite rare in
children.41
Some studies comparing acetaminophen and NSAIDs have found no
difference in analgesic effectiveness,42
whereas others have found better analgesia with NSAIDs.43
NSAIDs may increase the risk of bleeding after tonsillectomy.44
NSAIDs provide good postoperative analgesia and result in lower
opioid requirements than in control groups not receiving NSAIDs.45
Selective cyclooxygenase-2 (COX-2) inhibitors46
have been designed to retain the analgesic and antiinflammatory
effects of NSAIDs while reducing the risk of gastric irritation and
bleeding. There are few published studies of the pediatric use of
selective COX-2 inhibitors,47
except for nimesulide,48
which is not available in the United States. Additional large-scale
studies are needed to evaluate efficacy and cost–benefit and
risk–benefit issues.
Opioids
The indications for opioids include postoperative pain, pain due
to sickle cell disease, and pain due to cancer. The suggested dosage
guidelines for subjects who have never received opioids are
presented in Table
3. As with adults, the risk of addiction (compulsive
drug-seeking behavior) appears low among children receiving opioids
for pain. Over the past 15 years, opioids in infants and children
have received intensive study.
|
Pharmacokinetics of Opioids in Neonates, Infants, and Children
The weight-normalized clearance of several opioids is diminished
in neonates and reaches mature values over the first two to six
months of life.50,51,52,53
The elimination half-life of morphine, in a pooled analysis,
averaged 9 hours in preterm neonates, 6.5 hours in term neonates,
and 2 hours in older infants and children.2
The active metabolites of morphine are excreted by the kidneys and
can accumulate in neonates because renal function is not yet mature.
Delayed renal clearance of morphine metabolites may contribute to
the analgesic, respiratory depressant, and rarely, convulsant
effects of morphine in the neonate. Fentanyl clearance may be
impaired during and after abdominal surgery in neonates.50
Opioid Pharmacodynamics and Clinical
Outcomes in Neonates, Infants, and Children
The respiratory-reflex responses to airway obstruction,
hypercapnia, and hypoxemia are immature at birth and mature
gradually over the first two to three months of life in both preterm54
and term55
neonates. Neonates and infants with chronic lung disease56
have impaired ventilatory reflexes, which might increase their risk
of opioid-induced respiratory depression. Case series and outcome
studies of children not undergoing intubation suggest a higher
frequency of opioid-induced respiratory depression among neonates
than among infants over six months of age or older children.57,58,59
However, morphine infusions during the postoperative period in
intubated neonates are associated with low behavioral pain scores
and good hemodynamic stability.60
In infants three to six months of age, the analgesic effects of
morphine or fentanyl are similar to, and the ventilatory depression
is no greater than, that seen in adults with similar plasma
concentrations of morphine61,62
or fentanyl.63
Some of the cited studies assessed ventilatory drive in infants
breathing through endotracheal tubes; studies conducted in intubated
infants may underestimate the risk of airway obstruction or
hypoventilation in non-intubated infants.
Continuous opioid infusions during the postoperative period have
been used extensively in older infants and children, with generally
good efficacy and safety,64
although with a substantial incidence of peripheral side effects.65
Starting rates of morphine infusion ranged from 0.01 mg per kilogram
per hour in infants under 6 months of age66
to 0.025 to 0.04 mg per kilogram per hour in children over 12 months
of age. In neonates, the weight-scaled rates of opioid infusion
should be lower, and the repeated weight-scaled intermittent doses
should be smaller, less frequent, or both, than in infants and
children.
Neonates receiving opioids should have continuous electronic monitoring,
preferably by pulse oximetry, and they should be observed in a
setting that permits rapid intervention for airway management,
because respiratory-rate monitoring alone may be an inadequate
predictor of impending apnea. Studies have not firmly established
either morphine or fentanyl as the preferred opioid for neonates or
infants.67
Patient-Controlled Analgesia in Children
The safety and efficacy of patient-controlled analgesia for children
as young as six years are supported by the results of controlled
trials.68
The variables for patient-controlled analgesia should be
individualized. Addition of a low-dose basal infusion improves pain
scores and patient satisfaction according to some reports68
but produces more episodic hypoxemia at nighttime according to
others.69
We routinely prescribe basal infusions for children with cancer or
sickle cell disease. Patient-controlled morphine treatment in
children typically starts with a bolus dose of 0.02 mg per kilogram,
a lockout interval of seven minutes, and a four-hour maximum of 0.3
mg per kilogram. If a basal infusion is used, it is typically begun
at 0.01 to 0.015 mg per kilogram per hour.
Nurse-activated patient-controlled analgesia is now widely used
for infants70
as a convenient way to prevent delays in relieving episodic pain.
Activation of the button for patient-controlled analgesia by parents
("parent-controlled analgesia") is widely accepted in
palliative care. However, its use for postoperative pain is
controversial because of the potential for either overdosing or
underdosing subjects who have not received opioids before. If
parent-controlled analgesia is to be considered, we recommend a
formal education program for parents, together with protocols for
close observation by the nursing staff.
Meperidine (pethidine) in low doses is useful to treat
postoperative shivering or rigors after amphotericin infusion, but
it has no particular advantages as an analgesic. Morphine resulted
in better analgesia and no more side effects than meperidine in
a double-blind comparison in which patient-controlled analgesia was
used.71
Meperidine can produce dysphoria and seizures from accumulation of
its metabolite, normeperidine. The clinical usefulness of
hydromorphone appears to be similar to that of morphine; it is
roughly five times as potent as morphine in children.72
Fentanyl provides analgesia with a rapid onset and short duration of
effect for brief, painful procedures. With repeated dosing or with
prolonged infusions, fentanyl becomes longer-acting.73,74
Rapid administration of fentanyl can produce chest-wall rigidity
that responds to naloxone in some cases; in other cases,
neuromuscular blockade and positive-pressure ventilation are
required.75
Two novel formulations of fentanyl may be useful in selected patients.
Oral transmucosal fentanyl permits rapid onset of analgesia for
brief, painful procedures14,76
in hospitalized children in whom intravenous access is not
available. In adults, the analgesic effect of 800 µg of oral
transmucosal fentanyl is roughly equivalent to that of 10 mg of
intravenous morphine.77
Oral transmucosal administration is effective because it bypasses the
efficient first-pass hepatic metabolism of fentanyl that occurs
after enteral absorption.
Transdermal fentanyl provides a consistent analgesic effect for
selected patients, such as children with severe pain due to cancer.78
Transdermal fentanyl has a slow onset and some variability in
absorption, and it is contraindicated as initial treatment for
patients who have not received opioids before. Oral or intravenous
methadone is useful because of its prolonged duration of action.79
However, because of slow and variable clearance, methadone requires
careful assessment and titration to prevent delayed sedation.
Methadone elixir is useful as a long-acting opioid for patients
unable to swallow whole sustained-release opioid tablets.
Agonist–antagonists such as pentazocine80
and drugs such as buprenorphine81
that act at kappa receptors offer no apparent advantages over
mu-agonist opioids.
Equipotency tables are useful for conversion from one opioid to
another or for conversion from one route of administration to
another. Studies of opioid-tolerant adults with cancer showed that,
when treatment was being changed from one opioid to another, the
analgesic and respiratory depressant effects of the second opioid
appeared much stronger than those predicted by conversion ratios
that had been derived from studies of subjects who had not received
opioids before. This phenomenon, known as "incomplete cross-tolerance,"
is especially pronounced when the second opioid is methadone,82
and it is probably due to the fact that the d-isomer of methadone can act as an
antagonist at the N-methyl-d-aspartate
subclass of glutamate receptors.83,84
Mild respiratory depression can be managed by repeatedly awakening
the patient, encouraging deep breathing, and withholding further doses
until the effects subside. In urgent situations, assisted ventilation
or naloxone (10 to 20 µg per kilogram) may be needed. The use of
naloxone in opioid-tolerant patients carries a risk of producing
withdrawal reactions; the hemodynamic consequences can be especially
severe in patients with cardiac disease. If the circumstances are
not too urgent, incremental dosing with naloxone (e.g., 2 µg per
kilogram every 30 seconds until the respiratory rate and tidal
volume increase) may reverse excessive opioid effects without
evoking severe pain or withdrawal reactions. If naloxone is
administered, close observation is recommended and repeated doses
may be needed.
Nonrespiratory side effects of opioids, including nausea, ileus,
itching, and urinary retention, are common among infants and children65
and may cause considerable distress. Many opioid side effects can be
ameliorated by drug therapy directed at the side effect (e.g.,
antiemetics to treat nausea and vomiting, antihistamines to treat
itching, and laxatives to treat constipation).
Local Anesthetics
Local anesthetics are now widely used in children. Their safety
is quite acceptable, although excessive plasma concentrations can
produce seizures and cardiac depression. The amino-amides (e.g.,
lidocaine and bupivacaine) have a narrower therapeutic index for
neonates than for children or adults because of decreased metabolic
clearance,85,86
with resultant drug accumulation during infusions86;
decreased plasma concentrations of 1-acid
glycoprotein, leading to higher concentrations of unbound local
anesthetic; and hard-to-recognize warning signs of impending toxic
effects in preverbal neonates and infants. The maximal recommended
doses of lidocaine are 4 mg per kilogram without epinephrine and 5
mg per kilogram with epinephrine in neonates and 5 to 7 mg per kilogram
in children. The maximal recommended doses of bupivacaine, with or
without epinephrine, are 2 mg per kilogram in neonates and 2.5 mg
per kilogram in children.85
Topical formulations are useful for needle procedures. For repair
of lacerations, combinations of tetracaine with epinephrine (adrenaline)
and cocaine, known as TAC, are widely used.87
Cocaine-free preparations are equally effective.88
Several formulations provide analgesia for intact skin and are
effective for needle procedures, including a cream containing both
lidocaine and prilocaine (EMLA, AstraZeneca),89
and tetracaine gel (Ametop, Smith and Nephew [not yet available in
the United States]).90
EMLA is safe and more effective than placebo for circumcision of
newborns,91
although it is less effective than ring block.92
Regional anesthesia is commonly administered, with excellent efficacy
and safety, to anesthetized children for postoperative analgesia,93
peripheral-nerve block,94
and epidural analgesia.95,96,97
Epidural analgesia is effective even in premature and term neonates.97
Epidural analgesia in neonates and infants requires specific expertise
on the part of physicians and nurses and close observation, as well
as modifications in technique and drug selection.98
Ropivacaine99
and levobupivacaine100
are two new local anesthetics that are attractive because they
involve less potential cardiac risk than bupivacaine in the event of
overdose.101
Clonidine is an attractive adjuvant to epidural local anesthetics,
because it prolongs or intensifies analgesia and also produces less
nausea, ileus, itching, urinary retention, and respiratory depression
than opioids.102
Children undergoing outpatient surgery frequently report high
pain scores, partly because of parental reluctance to administer analgesics.103
Although peripheral-nerve blocks and caudal blocks provide good
analgesia, the duration of analgesia is generally less than eight
hours. Thus, parents should be encouraged to administer analgesics
before pain is severe.
General Anesthesia for Neonates and Infants
General anesthesia has become much safer for neonates and infants
over the past 30 years, and the risk of cardiac arrest or death during
general anesthesia in infants has decreased by a factor of more than
20.104,105
Even the most critically ill neonates can tolerate anesthesia for
major surgery.106
Autonomic and hormonal–metabolic stress responses in neonates are
blunted to varying degrees by high-dose opioid anesthesia,107
epidural local anesthetics,108
and inhalational anesthetics.109
Treatment of Pain Due to Cancer
Pain in children with cancer may be caused by tumor progression;
by consequences of treatment, such as mucositis; or by needle procedures,
including bone marrow aspiration. For needle procedures, both
pharmacologic approaches (topical and infiltration anesthesia, conscious
sedation, and general anesthesia) and nonpharmacologic approaches5
(hypnosis and cognitive–behavioral programs) are efficacious. The
optimal combination of pharmacologic and nonpharmacologic approaches
should be individualized.
The majority of children with advanced cancer can be made
comfortable with titrated oral doses of opioids and appropriate
management of side effects.110
If oral administration is not tolerated, the alternatives include
intravenous, continuous subcutaneous,111
and transdermal78
opioid administration. A retrospective survey of parents'
recollections suggests a need for improved interventions for pain as
well as for a range of other symptoms, especially fatigue and sleep
disturbance, among children with terminal cancer.112
Methylphenidate is useful in antagonizing opioid-induced sedation.
Marked escalation of opioid doses (e.g., by 100 times or more) may
be required, primarily among patients with solid tumors metastatic
to the spine or central nervous system.110
Some of these patients have pain resistant to high-dose opioids but
can be made comfortable and alert by epidural or subarachnoid infusions
of local anesthetics and opioids.113
Management of cancer pain is best approached in the context of
broad-based supportive or palliative care114
programs not limited to pharmacologic interventions.
Pharmacologic Management of Chronic Noncancer
Pain
Chronic pain can be a burden for children and families and can
impair social functioning and school attendance. It is useful to
distinguish between nociceptive pain and neuropathic pain. Nociceptive
pain involves the detection of tissue injury or inflammation by a
normally functioning nervous system. Neuropathic pain persists
because of abnormal excitability in the peripheral or central
nervous system. Neuropathic pain in children is commonly post-traumatic.115
Prolonged pain after amputation is not rare in children.116
Evidence in adults supports the efficacy of tricyclic
antidepressants and several anticonvulsants, especially gabapentin,117
in several conditions involving neuropathic pain.118
Antidepressants and anticonvulsants are commonly used for children with
neuropathic pain, despite a lack of controlled studies. Our
impression is that they can be effective in children, as they are in
adults, although they can have side effects.
Recurrent headaches are common in children.119
Abortive and preventive therapies for migraine in children have been
studied.120
Sumatriptan, a serotonin 1B/1D receptor agonist, appears effective and
safe as an abortive treatment.121
Dihydroergotamine,122
ibuprofen, and acetaminophen123
were more effective than placebo in interrupting episodes of
migraine, and ibuprofen appeared to be more effective than
acetaminophen.123
For prevention of migraine, trials have reported efficacy with
beta-blockers,124
calcium-channel blockers,124
and antidepressants.125,126
Yet other trials found beta-blockers to be no more effective than
placebo127
and less effective than self-hypnosis.127
Children with sickle cell disease who have vaso-occlusive episodes
should receive opioids as needed to relieve pain. Studies emphasize oral
dosing128
of potent opioids and NSAIDs, home treatment,129
and reduced reliance on emergency departments or inpatient admission.
In several other chronic, debilitating conditions in childhood,
there is a restricted role for long-term episodic or regularly scheduled
administration of opioids as a component of a comprehensive pain-management
program.
With knowledge of principles that influence drug dosage, actions,
and interactions, clinicians should generally be able to provide effective
relief of acute pain, pain due to cancer, and several types of
chronic pain in infants and children with a wide margin of safety.
Supported
in part by grants from the National Institute of Child Health and
Human Development (1RO1HD35737, to Dr. Berde), Advocates for
Children's Pain Relief, the Giannini Foundation, and the Anesthesia
Pain Research Endowment Fund.
Source Information
From the Departments of Anesthesia (C.B.B., N.F.S.) and Medicine
(C.B.B.), Children's Hospital; and the Departments of Anesthesia (C.B.B.,
N.F.S.) and Pediatrics (C.B.B.), Harvard Medical School — both in Boston.
Address reprint requests to Dr. Berde at the Pain Treatment
Service, Children's Hospital, 333 Longwood Ave., Rm. 555, Boston, MA 02115.
References
Edward E.
Rylander, M.D.
Diplomat American
Board of Family Practice.
Diplomat American
Board of Palliative Medicine.