Journal of Perinatology (2007) 27, S4–S11

r 2007 Nature Publishing Group All rights reserved. 0743-8346/07 $30

www.nature.com/jp

ORIGINAL ARTICLE
Pharmacological approaches to the management of pain in the
neonatal intensive care unit
KJS Anand
Department of Pediatrics, University of Arkansas for Medical Sciences, AR, USA

dementia, children with handicaps and infants, as well as anyone

Effective and consistent management of neonatal pain remains a who is incapable of verbal self-report. On average, adult patients
controversial issue. capable of verbal self-expression receive much more pain
Premature infants are repeatedly subjected to painful tests and procedures medication than those who are incapable of verbal expression.2
or suffer painful conditions when they are most vulnerable. With different Similar studies in children undergoing spinal surgery demonstrated
mechanisms transducing various types of pain the practice of ‘one-drug that children with cognitive impairment received significantly less
fits all’ becomes questionable. pain medication compared to children without cognitive
impairment on post-operative days 1 to 3.3
Clinicians must use the latest non-pharmacologic and pharmacologic
To understand the physiologic outcomes of pain, it is helpful to
therapies for effective management of neonatal pain, distress, or agitation.
understand how the body interprets painful stimuli. The
Pharmacologic strategies for dealing with neonatal pain in the neonatal
identification, quantification and evaluation of pain, as well as the
intensive care unit are described. Opioid therapy, once considered the
utilization of these data in treatment protocols, are of paramount
mainstay for neonatal analgesia, may not be as effective as previously
concern to clinicians. Different types of pain are perceived and
thought. Morphine infusions do not alter the neurological outcomes of
registered by many different mechanisms, although with some
preterm neonates and may not be effective against acute pain. Alternative
degree of overlap. Thus, the physiological pain caused by tissue
approaches with methadone, ketamine, or local anesthetics should be
injury is only one type of pain, although most commonly coupled
considered. Clinicians must understand the contextual circumstances
with the term pain. In addition, there is inflammatory pain
underlying pain in individual neonates and tailor therapy accordingly,
associated with inflammation, neuropathic pain resulting from
using the most current evidence related to neonatal pain assessment and
nerve damage or inflammation, or visceral pain caused by
management.
distention, contraction, or inflammation of hollow or solid viscera.
Journal of Perinatology (2007) 27, S4–S11. doi:10.1038/sj.jp.7211712
With the many different pain mechanisms, the practice of
Keywords: infant-newborn; infant-premature; pain; agitation; stress; ‘one-drug fits all’ becomes less viable as a standard approach.
analgesia

What are the sources of infant pain?

Introduction This article examines pain control in the neonatal intensive care
The International Association for the Study of Pain defines pain as unit (NICU) setting, focusing mainly on the pharmacologic
‘an unpleasant sensory and emotional experience associated with management strategies when dealing with acute, established or
actual or potential tissue damage, or described in terms of such prolonged pain. Acute pain may be caused by various diagnostic
damage.’1 They further promulgate that pain is always subjective and therapeutic procedures, minor surgery, or placement of
and takes on meaning based on each person’s experiences, invasive monitoring devices; established pain is commonly
especially in early life. Thus, the definition of pain heavily weighs associated with post-operative pain, inflammatory pain, thermal or
on the patient’s ability to verbalize their experience of pain. chemical burns; whereas prolonged or disease-related pain results
The ability to assess a patient’s pain and respond to it effectively from contractures, nerve injury, or conditions such as necrotizing
depends on his or her ability to describe the pain, specifically the enterocolitis, thrombophlebitis, or osteomyelitis (Table 1).
location and intensity. Unfortunately, a vast number of patients Pain is a multilayered phenomenon often described as
lack the ability to describe their pain. These include adults with producing either primary or secondary hyperalgesia. Primary
hyperalgesia is located in the area of specific tissue injury, whereas,
Correspondence: Dr KJS Anand, Arkansas Children’s Hospital, 800 Marshall Street, Little
Rock, AR 72202, USA. secondary hyperalgesia occurs away from the site of injury or
E-mail: [email protected] inflammation. In addition, if pain is caused by non-noxious
 Management of pain in the NICU
KJS Anand
S5

Table 1 Sources of infant pain dorsal cutaneous flexor reflex reaction to pain as compared to term
infants.13 The afferent limb of the withdrawal reflex is conducted
Acute pain Diagnostic and therapeutic procedures
Minor surgery
by C-fibers, which are pain-conducting fibers. Term infants have a
Suctioning oral/nasal/tracheal lower threshold for this reflex compared to older children or adults
Established pain Postoperative pain because there is a delayed maturation of descending inhibitory
Inflammatory pain fibers.14–16 In practice, if an intravenous line were to be inserted in
Thermal/chemical burn the hand of a 30-week-old baby, there would be less physiological
Prolonged pain Meningitis response as opposed to inserting that same intravenous line in the
Necrotizing enterocolitis baby’s foot. The reasoning behind this is that the descending
Phlebitis inhibitory fibers are growing from the supraspinal brainstem
Osteomyelitis from repeated heelsticks nuclei, only reaching the cervical section of the spinal cord by
30–32 weeks. As they have not reached the lumbar section of the
stimuli, it is labeled as allodynia. For example, tactile stimuli spine by 30 weeks, the pain threshold is higher in the upper
applied to an inflamed joint are converted into painful stimuli extremities and lower in the lower extremities, resulting in
because of the hyperexcitability of nociceptors in the tissue increased sensitivity to pain in the lower extremities.17–19
(peripheral sensitization) and neurons in the spinal cord or A phenomenon known as wind-up occurs when an infant is
supraspinal areas (central sensitization). Repeated noxious stimuli subjected to prolonged pain, resulting in a temporal summation
causing increasing amounts of activation in the pain pathways is and hyperalgesia, which can last for up to 30–60 minutes in
termed as windup or temporal summation.4,5 preterm infants.14,15,16,19
There are specialized nociceptors and nerve fibers in the Despite being unable to express pain, the premature infant
periphery and specialized neurons in the dorsal root ganglia reacts to painful stimuli with activation of the highest level of
responsible for pain transmission.6 The different sensory systems, sensory function in the brain. Bartocci et al.20 evaluated the
such as visual, auditory, or tactile systems, are designed to develop cortical activation responses following tactile and painful stimuli.
normally with incident stimulation that occurs during early life. In Infrared spectroscopy probes were placed on both sides of the
contrast, the pain system is the only sensory system that develops somatosensory cortex and measured for a baseline period. Tactile
normally without any incident stimulation during early life.7 For stimulation produced by an alcohol swab caused an increase in
the neonate, early developmental periods can easily modify pain blood flow to the somatosensory cortex. A venepuncture resulted in
systems, especially from the exposure to frequent and multiple several-fold increases in blood flow to the somatosensory cortex,
painful stimuli. The repetitive stimulation of the pain system may but no changes occurred in blood flow to the occipital cortex,
attenuate peripheral sensitivity or decrease behavioral responses demonstrating that this was not caused by global increases in
resulting in altered development.8–10 cerebral blood flow but a functionally specific response (Figure 1).
The three categories of pain (acute, established, or prolonged) In contrast, hemodynamic responses in older preterm infants were
based on their duration will result in different clinical associated with changes in the contralateral somatosensory cortex
manifestations. Most pain studies in neonates have studied models after painful stimuli from a heelstick (Figure 2). Thus, the
of acute pain, heel sticks or circumcision, two of the most common repeated exposure to painful stimulation may lead to a maturation
invasive procedures performed in neonates. Acute pain results in a of the cortical responses to pain in preterm neonates.
psychophysiological activation, with increased heart rate,
respiratory rate, blood pressure, changes in skin blood flow and
oxygen saturation, vagal tone and palmer sweating. Behavioral Where do we go from here?
changes can also be seen with increases in crying activity, body The scientific proof that neonates feel and react to painful stimuli
movements and facial expressions. When pain becomes established is evident. Pain, unfortunately, is a common occurrence in the
or prolonged, however, there is a dampening of normal behavior NICU setting. The complexity of their inability to verbalize pain
and physiological responses. Heart rate and blood pressure may and inherent variability of pain dependent on its source and
even be below resting rates, also associated with the elimination of duration adds layers of intricacy to appropriate identification
normal physiological variability, like sinus arrhythmia.11,12 (Table 1). A proposed clinical approach (Figure 3) assesses
Clinicians at the bedside need to be keenly aware of the different potential sources considering both continuous and episodic pain,
types of pain to be better able to evaluate and treat infants and prevents unnecessary tests and procedures and develops specific
neonates. protocols for pain management. There are myriad pain
Multiple lines of evidence suggest that infants are more sensitive measurement tools to choose from the Premature Infant Pain
to pain. In evaluating the dorsal cutaneous flexor withdrawal Profile (PIPP), the CRIES (Cry, Requires oxygen, Increased vital
reflex, preterm infants have a third to 50% of the threshold for the signs, Expression, Sleeplessness) tool and the Neonatal Infant Pain

Journal of Perinatology
 Management of pain in the NICU
KJS Anand
S6

Scale are most commonly used. Of these three pain assessment assessment methods. In addition, clinicians must evaluate
tools, only the PIPP adjusts for gestational age. Evidence-based contextual factors, behavioral and physiological indicators.
pain assessment recommends standardized and validated methods Numerous studies have documented the poor compliance in
that are specific for acute versus prolonged pain. For example, a administering analgesia to neonates before painful procedures.
universal scale of 0 to 10 would be beneficial among the different Barker and Rutter21 demonstrated that neonates less than 30 weeks
are more likely to have invasive procedures (Figure 4). Porter
a 7 followed 144 neonates who underwent 7672 pain-inducing
6 procedures, of which 87% were heel sticks and only 3% were
5
preceded by analgesia.22 More recently, Simons et al.,23 followed
[Hb O2] parietal * 151 neonates for the first 14 days of their NICU admission and
[Hb O2] occipital
4
[HbO]2 microM/L

3
A proposed clinical approach…..
2 *
Identify the potential sources of pain
1 PAIN ASSESSMENT

Invasive Procedures
0 Continuous Pain (e.g., IV / CVL, intubation,
(e.g., Post-Operative, chest tube insertion,
Inflammatory) suctioning, etc.)
–1

–2
Baseline Tactile Venipuncture • Prevent unnecessary tests and procedures
• NICU/PICU–Specific Pain Assessment Tools
b 5.5 • Hospital–Specific Polices And Standards
5.0
Environmental
4.5 Pharmacologic Agents
Agents
4.0
[Hb O2] left Behavioral
3.5 Agents
[Hb O2] right
*
MicroMol/L

3.0 PAIN RE-ASSESSMENT

2.5
Figure 3 A proposed clinical approach for assessment and
2.0
management of pain in neonates.
1.5

1.0
* 500
0.5
*
0.0

–0.5
Baseline Tactile Venipuncture
400
Figure 1 Acute pain is associated with increased blood flow
(representing neuronal activation) to the bilateral somatosensory cortex
(lower panel). This is functionally specific because pain activates the
Number of procedures

contralateral somatosensory cortex but not the occipital cortex. Bartocci

300
M et al. PAIN 2006; 122: 109–117. Used with permission.

200

100

0
Figure 2 Hemodynamic responses in the somatosensory cortex of a 22 26 30 34 38 42
ventilated 25 week premature infant, demonstrating the evoked changes Gestation (weeks)
in total hemoglobin in the contralateral (black) and ipsilateral (grey) Figure 4 Total number of invasive procedures experienced by preterm
cortex following a heelstick at t ¼ 20 s. Slater R, Cantarella A, Gallella neonates during their stay in the Neonatal ICU. Archives of Disease in
S, et al. Cortical Pain Responses in Human Infants. Childhood, 1995, vol. 72, pp. F47–F48, reproduced with permission for
J Neuroscience 2006; 26(14): 3662–3666. Used with permission. the BMJ Publishing Group.

Journal of Perinatology
 Management of pain in the NICU
KJS Anand
S7

documented almost 20,000 pain-producing procedures, of which blood draws. However, the following discussion will focus only on
31% were repeat attempts when the first procedure was not the pharmacological management with systemic analgesics, most
successful. This averaged 196 procedures per neonate, or between notably treatment with opioids and other options including
12 and 15 procedures per day. Although physicians and nurses ketamine and midazolam (Table 2).27,28
rated these procedures in the range of moderate to severe pain
using a visual analog scale, less than 35% of these procedures were
associated with some form of analgesia. Opioids
Quality improvement projects have evaluated pain management The mainstay of systemic analgesia for moderate to severe pain is
in several NICUs. One such study found that if pain was assessed by the use of opioid therapy. Opioids provide both sedation and
clinicians with prescriptive authority, such as physicians or nurse analgesia, have a wide therapeutic window, and decrease
practitioners, neonates had a fourfold increase in the odds ratio for hemodynamic and metabolic stress responses, but they do not
post-operative analgesia, as opposed to pain assessment by provide amnesia. Benzodiazepines are a better choice when
nurses.24 A collaborative approach by all clinicians in the NICU to amnesia is required.29 Morphine and fentanyl are the most
manage pain may better serve these patients. For physicians and commonly used opioids in the NICU population.30
nurse practitioners, pain assessment should be part of their routine
physical exam.24 Incorporating parental assessment including Morphine
baseline behavior may be especially valuable. Parents know their Morphine is the most commonly used opiate used for analgesia,
infants and children on a patterned and intuitive level, this but wide variability occurs in dosing and clinical usage.31 Some
knowledge should be utilized when assessing pain in these studies suggest that babies require higher plasma concentrations
patients.25 than older children or adults to receive pain relief,32 although
other studies disagree.33–36 In the past, morphine was thought to
be a pure opioid agonist, but this theory was disproven long ago.37
Morphine has a ceiling effect; after a certain therapeutic level has
Pharmacologic management been reached, higher doses will produce more adverse effects rather
Both non-pharmacological and pharmacological approaches play a than increased analgesia. Clinical experience suggests that a
role in the management of pain in the neonate and infant. The ceiling effect may be reached by using doses up to 0.5 mg/kg.
use of swaddling, sucrose, pacifiers and decreased environmental Morphine has a slow onset of analgesia owing to lower lipid
stimuli have shown limited therapeutic efficacy in treating mild to solubility, especially in premature infants.29 Its mean onset of
moderate painful stimuli. Local and systemic analgesia are often action is 5 min and the peak effect is at 15 min. Morphine is
the treatment of choice for invasive procedures. Local analgesia, metabolized in the liver into two active compounds, morphine-3-
such as various topical agents or lidocaine infiltration, can be glucuronide (M3G) and morphine-6-glucuronide (M6G). M3G is
effective in managing procedural pain, although not as effective as an opioid antagonist, which at high levels can cause seizures;
opioid therapy.26 Most topical agents also need to be applied whereas M6G is a potent analgesic with a half-life greater than
30 min before a procedure and are not effective against heel-stick morphine. In the maturation of hepatic glucuronidation,

Table 2 Analgesic medications.27,28

Agent Class Neonatal dosing

Morphine Opioid IM, IV, SQ – initial dose 0.05-mg/kg every 4–6 h; maximum 0.1 mg/kg/dose
IV continuous infusion – 0.001–0.03 mg/kg/h
Fentanyl Opioid IV slow push – 0.3–2 mcg/kg/dose
IV continuous infusion – 0.3–5 mcg/kg/h
Methadone Opioid IV–0.05–0.2 mg/kg/dose every 12–24 h (for neonatal abstinence syndrome)
Ketamine NMDA antagonist IV slow push – 0.5–2 mg/kg/dose
IV continuous infusion – 0.5–1 mg/kg/h
IM, SQ – 1–2 mg/kg/dose
Oral – 5–8 mg/kg/dose
Midazolam Benzodiazepine IV continuous infusion – (<32 weeks): 0.03 mg/kg/h or 0.5 mcg/kg/min; (>32 weeks): 0.06 mg/kg/h or 1 mcg/kg/min
IV – 0.05 – 0.15 mg/kg over at least 5 min, repeat Q2–4 h as needed
Intranasal or Sublingual–0.2 mg/kg/dose (use injectable form)
IV, intravenous; IM, intramuscular; SQ, Subcutaneous; mg, milligrams; kg, kilograms; mcg, micrograms; Q, every; h, hour; min, minutes.

Journal of Perinatology
 Management of pain in the NICU
KJS Anand
S8

UDP-glucuronyl transferase, preterm infants produce mostly M3G, control group. Of the original 4254 neonates screened, a total of
which is why after three or four days of therapy the infant develops 898 neonates were enrolled in the trial. Poor neurologic outcomes
a degree of tolerance as a result of continued treatment.38,39 were defined as death at less than 28 days before NICU discharge,
Morphine causes some histamine release and may lead to severe intraventricular hemorrhage (IVH, grade III or IV), or
several side effects as a result.29 For those infants on fluid periventricular leukomalacia (PVL). At the end of the study, the
restriction for patent ductus arteriosus (PDA) and those on incidence of the primary outcome was similar between the two
diuretics, morphine will produce some degree of hypotension. groups (morphine, 27%; placebo, 26%; P ¼ 0.58). In sub-group
Caution should be used when prescribing morphine in patients analyses, there was no change in the stratified gestational age
with asthma or bronchopulmonary dysplasia, because of the risk of groups of 23 to 26 weeks (morphine, 40%; placebo, 42%; P ¼ 0.70)
histamine-induced bronchospasm. Compared to fentanyl, or 30 to 32 weeks (morphine, 10%; placebo 15%; P ¼ 0.31).
morphine has decreased risks of causing dependence or There was, however, a trend towards increases in poor
withdrawal40 and greater effects on gastrointestinal (GI) motility neurologic outcomes in the 27 to 29 week neonates receiving
and biliary spasm.29 morphine (23 versus 15%; P ¼ 0.053). Morphine was administered
There is accumulating evidence that morphine may not be as a loading dose (100 mcg/kg) followed by continuous
effective in the treatment of acute pain. Franck et al.41 revealed infusions based on gestation age (23 to 26 weeks, 10 mcg/kg/h;
that in the post-operative period for preterm infants, a dose of 27 to 29 weeks, 20 mcg/kg/h; 30 to 32 weeks, 30 mcg/kg/h).
morphine did not produce any change in plasma norepinephrine, These dosages were based on single-dose pharmacokinetics, and
vagal tone, or flexor withdrawal. Simons demonstrated that were not based on the infusion pharmacokinetics of morphine.
morphine did not have any analgesic effects associated with These were large doses of morphine, infusions of 3–5 mcg/kg/h
tracheal suctioning.42 More recently, in a blinded study of may be effective, particularly in the younger gestational age
morphine versus placebo, there was no change in the response to groups.46
heel sticks before and after a loading dose (100 mcg/kg followed by The most prevalent side effects of morphine were hypotension
an infusion of 10 to 30 mcg/kg/h) of morphine on two different and respiratory depression, which resulted in an increased duration
pain scales (PIPP and DAN scales).38 There are some data of ventilation by one day (P ¼ 0.03). Morphine’s effects on GI
suggesting that acute pain leads to an uncoupling of opiate motility increased the time required to reach full volume
receptors in the forebrain.43 In the case of preterm infants, nasogastric feeds (P ¼ 0.04). Logistic regression analyses revealed
immaturity of opiate receptors could be a reason for the lack of that neonatal death was related to lower gestation (OR 0.71,
response. Some studies suggest that infants have a decreased P<0.0001) and greater severity of illness (OR 1.23, P<0.0001), but
production of M6G and an increased production of M3G, thereby not to treatment group (OR 1.16). Severe IVH was related to
resulting in a decrease in the potent analgesia effect and an gestation (OR 0.74, P<0.0001), severity of illness (OR 1.08,
increase in the opioid antagonist effect.44 P ¼ 0.02), lack of antenatal steroids (OR 0.29, P<0.0001), and
However, there are also recent studies that support the use of maternal race (P ¼ 0.04), but not the treatment group (P ¼ 0.22).
morphine for acute pain. Taddio et al.26 demonstrated that PVL was not related to gestation or treatment group; it was only
intravenous morphine was more effective than topical tetracaine predicted by maternal chorioamnionitis (OR 2.05, P ¼ 0.03) when
for the management of pain associated with central line placement all other clinical factors were controlled for in the logistic
in neonates. Angeles et al.45 showed that opioid use in asphyxiated regression model.46
term neonates had a positive effect on the neuroimaging outcomes Overall, from the NEOPAIN trial, continuous morphine infusion
and clinical outcomes at 18 months. did not alter the neurologic outcomes of preterm neonates.
Overall, the routine use of morphine for the management of Gestational age differences may be related to either severity of
pain in preterm neonates is controversial. In addition to analgesia, illness or the effects of increasing morphine infusion rates.
a question arises if morphine therapy may have beneficial effects Intermittent boluses were required more frequently in neonates
on the neurological outcomes of preterm neonates. The NEOPAIN that had developed IVH or PVL before study enrollment.46
trial investigated this issue.
Fentanyl
The (Neurologic Outcomes & Pre-emptive Analgesia In Another commonly used opioid, fentanyl has a rapid onset of
Neonates) Trial action (2 to 3 min), short duration of action (60 min with bolus
The Neurologic Outcomes & Pre-emptive Analgesia In Neonates doses) and minimal hemodynamic effects.47 Prolonged elimination
(NEOPAIN) trial is the largest randomized, placebo-controlled trial occurs with fentanyl infusions,48 resulting from rebound plasma
for any analgesic in the neonatal age group. The trial was designed concentrations. Chest wall rigidity may occur with rapid fentanyl IV
to evaluate poor neurological outcomes of ventilated preterm boluses. Fentanyl is metabolized in the liver to inactive
neonates treated with a morphine infusion compared to a placebo compounds.47 The potency ratio of fentanyl compared to morphine

Journal of Perinatology
 Management of pain in the NICU
KJS Anand
S9

is different in adults (80 to 100 times) as compared to infants (13 ketamine or midazolam, recognizing that midazolam is a sedative
to 20 times).38 Tolerance develops rapidly, especially with infusions amnesic agent, and not an analgesic.
compared to boluses. In addition, the risk of dependence and Ketamine is commonly used for procedural pain. It produces a
withdrawal is greater as compared to morphine.40 Fentanyl is also dissociative state by blocking N-methyl-D-aspartic acid (NMDA)
available as a transdermal patch. The use of fentanyl patches is receptors, resulting in sedation, analgesia and amnesia. It is the
limited and should be considered only when IV access is not only drug that causes all three effects. It has been shown to be safe
available. The problem arises because the lowest dose fentanyl patch and effective for sedation in pediatric critical care procedures.52 It
provides 12.5 mcg/h, which is a huge dose for preterm infants. increases muscle tone and blood pressure, thus maintaining
In addition, any vasodilation close to the infant’s skin will increase hemodynamic stability. Ketamine does not cause respiratory
the absorption rate of the drug. One advantage is that the gradual depression and maintains respiratory drive. With regard to opioid
‘tail-off’ effect after removing the patch does not occur because tolerance and withdrawal, methadone and ketamine work in a
there is less subcutaneous fat in preterm infants.11 Intravenous similar way; these novel co-therapies are being investigated and
infusion is considered the safest way to administer fentanyl. may prove to be effective. Low-dose ketamine, 0.3 mg/kg/h with a
morphine infusion, has been efficacious in blocking morphine
Methadone tolerance by blocking NMDA receptors.6 Although apoptosis occurs
Methadone is currently being studied as a viable drug to reduce in the brain of newborn rats, this has been reported with doses that
pain in neonates. It is equipotent with morphine, but has are 100 times of those used in clinical practice;53 these data may
additional mechanisms of action that may be advantageous in the not apply to human neonates receiving clinical doses.54
preterm neonate population. In addition to activating specifically As previously mentioned, midazolam is a benzodiazepine drug
the m-opiate receptors as morphine does, it also causes a that produces anxiolysis, sedation, amnesia, and muscle relaxation
desensitization of delta-opiate receptors and antagonism of the but no analgesia. It is often used before induction of anesthesia
NMDA receptors. The action on the delta-opiate receptors reverses and for conscious sedation before diagnostic procedures.
the tolerance that occurs with morphine.49 As an NMDA antagonist, Midazolam is frequently used for infants <32 weeks during
methadone blocks NMDA receptors, producing additive analgesic mechanical ventilation, with initial doses of 0.03 mg/kg/h
effects and delayed development of tolerance.50 Methadone has a (0.05 mcg/kg/min). It activates the inhibitory g-amino butyric
slow onset of action (20 min with IV, 30 to 60 min with oral),47 acid receptors in the brain, brainstem, and spinal cord. Onset
good oral bioavailability (75 to 85%) and prolonged elimination of action ranges from 2 to 6 min, lasting 1 hour after a single IV
half-life (t1/2: children 19 h; neonates 41 h).22 dose. Although the combination of opioids with midazolam is
common practice in many NICUs, extreme caution is suggested
Critical issues with opioids because of the many serious adverse events. Midazolam may cause
Tolerance and withdrawal respiratory depression, hypotension, and bradycardia.47 In addition,
With the widespread use of opioids, critical issues of tolerance and midazolam has been known to cause a decrease in cerebral blood
withdrawal occur. Clinicians are ever vigilant of patients developing flow in preterm neonates.55
opioid tolerance, requiring higher doses to illicit the same
analgesic effect. As higher doses of opioids are administered over
longer periods of time, withdrawal becomes an even greater Conclusions
concern. Some procedural and therapeutic approaches may help The multifaceted issues of pain management can be addressed in
prevent the development of tolerance. Methadone is commonly different ways rather than relying on the ‘one-drug-fits-all’
used owing to its long half-life. Novel combination therapies to approach (Table 3). Exposure to repetitive pain occurs commonly
prevent tolerance include morphine with low-dose ketamine, which in preterm and term neonates. Pain assessment methods are
will block NMDA receptors.6 Morphine can also be combined with currently available specific for acute pain, but newer methods are
ultra-low dose naloxone. With the use of ultra low-dose naloxone, currently being developed to evaluate prolonged pain. An
picomolar concentrations of naloxone at the receptor level only appropriate approach to the management of pain needs to assess
block the opioid receptors coupled with the stimulating G-protein the type of pain the neonate is experiencing. For acute
and do not affect the receptors coupled to the inhibitory G-protein physiological pain, avoiding invasive procedures, utilizing sucrose
that are producing analgesia.51 Randomized trials are now being pacifiers, and topical/local anesthetics can be useful. For post-
designed to look at these novel therapies. operative pain, a short duration (24 to 48 h) of opioid therapy,
positioning, removing drains, and considering adjuvant therapies
Beyond opioids: ketamine and midazolam would be appropriate. For inflammatory pain, caused by meningitis
In addition to the usual opioids, other options should be evaluated. or phlebitis, anti-inflammatory agents should be considered.
Consideration should also be given to short-term infusions of Opioids can be used in this setting if the pain is severe or extensive.

Journal of Perinatology
 Management of pain in the NICU
KJS Anand
S10

Table 3 Pain management approaches 4 Anand KJS. Clinical importance of pain and stress in preterm neonates. Biol
Neonate 1998; 73: 1–9.
Type of pain Pain management approaches
5 Narsinghani U, Anand KJS. Developmental neurobiology of pain in neonatal
Acute pain Avoid invasive procedures rats. Lab Anim 2000; 29: 27–39.
Sucrose 6 Wiesenfeld-Hallin Z. Combined opioid-NMDA antagonist therapies. What
Pacifier advantage do they offer for the control of pain syndromes? Drugs 1998; 55:
Local/topical anesthesia 1–4.
7 Ruda MA, Ling Q-D, Hohmann AG, Peng YB, Tachibana T. Altered
Postoperative pain Short-term opioid infusion (24–48 h) nociceptive neuronal circuits after neonatal peripheral inflammation.
Positioning Science 2000; 289(5479): 628–631.
Remove drains 8 Anand KJS. Pain, plasticity, and premature birth: a prescription for
Consider adjuvant therapies permanent suffering? Nat Med 2000; 6: 971–973.
9 Anand KJS, Grunau RE, Oberlander T. Developmental character and long-
Inflammatory pain Non-steroidal anti-inflammatory agents term consequences of pain in infants and children. Child Adolesc Psychiatr
(non-specific, or specific Cox-1 or Cox-2 inhibitors) Clin N Am 1997; 6: 703–724.
Consider opioids if severe or extensive 10 Peters JW, Schouw R, Anand KJS, van Dijk M, Duivenvoorden HJ, Tibboel D.
Does neonatal surgery lead to increased pain sensitivity in later childhood?
Visceral pain Oxybutynin for bladder pain/spasms Pain 2005; 114: 444–454.
Metoclopramide for intestinal cramping 11 Johnston CC, Stevens BJ. Experience in a neonatal intensive care unit affect
Consider epidural analgesia pain response. Pediatrics 1996; 98: 925–930.
Specific therapies for underlying condition 12 Grunau RE, Oberlander TF, Whitfield MF, Fitzgerald C, Lee SK. Demographic
and therapeutic determinant of pain reactivity in very low birth neonates at
Neuropathic pain Relieve nerve compression 32 weeks’ postconceptional age. Pediatrics 2001; 107: 105–112.
Steroids for reducing nerve inflammation 13 Fitzgerald M, Shaw A, MacIntosh N. The postnatal development of the
Consider nerve blocks or epidurals or spinal anesthesia cutaneous flexor reflex: a comparative study in premature infants and
Specific agents: Mexiletine, memantine, gabapentin, newborn rat pups. Dev Med Child Neurol 1988; 30: 520–526.
phenytoin and other agents (not tested in neonates) 14 Fitzgerald M, Millard C, MacIntosh N. Hyperalgesia in premature infants.
Lancet 1988; 1(8580): 292.
15 Fitzgerald M, Millard C, McIntosh N. Cutaneous hypersensitivity following
peripheral tissue damage in newborn infants and its reversal with topical
A comprehensive knowledge base and familiarity with current anaesthesia. Pain 1989; 39: 31–36.
research related to pain assessment and management are 16 Andrews K, Fitzgerald M. The cutaneous withdrawal reflex in human
neonates: sensitization, receptive fields, and the effects of contralateral
paramount, to help the clinician treating neonates in the NICU.
stimulation. Pain 1994; 56: 95–101.
Future research must define efficacy of specific therapeutic
17 Fitzgerald M, Koltzenburg M. The functional development of descending
approaches, applicability to specific neonatal populations, inhibitory pathways in the dorsolateral funiculus of the newborn rat spinal
combination approaches, and comparative studies between cord. Brain Res 1986; 389: 261–270.
analgesic drugs. 18 Ren K, Blass EM, Zhou Q, Dubner R. Suckling and sucrose ingestion
suppress persistent hyperalgesia and spinal Fos expression after forepaw
inflammation in infant rats. Proc Natl Acad Sci 1997; 94: 1471–1475.
Acknowledgments 19 Fitzgerald M. The development of nociceptive circuits. Nat Rev Neurosci
I gratefully acknowledge grant support from the NICHD (U10 2005; 6: 507–520.
HD500009), the help of Dr Evelyn R. Hermes-DeSantis (Ernest 20 Bartocci M, Bergqvist LL, Lagercrantz H, Anand KJS. Pain activates cortical
Mario School of Pharmacy) and Dr Sohini Sarkar (Robert Wood areas in the preterm newborn brain. Pain 2006; 122: 109–117.
21 Barker DP, Rutter N. Exposure to invasive procedures in neonatal intensive
Johnson University Hospital) in preparing this paper.
care unit admissions. Arch Dis Child 1995; 72: F47–F48.
22 Porter FL, Anand KJS. Epidemiology of pain in neonates. Res Clin Forums
1998; 20: 9–16.
References 23 Simons SHP, van Dijk M, Anand KJS, Roofthooft D, van Lingen RA, Tibboel
1 Merskey H. The definition of pain. Eur J Psychiatry 1991; 6: 153–159. D. Do we still hurt newborn babies? A prospective study of procedural pain
2 Anand KJS, Craig KD. New perspectives on the definition of pain. Pain 1996; and analgesia in neonates. Arch Pediatr Adolesc Med 2003; 157:
67: 3–6. 1058–1064.
3 Malviya S, Voepel-Lewis T, Tait AR, Merkel S, Lauer A, Munro H et al. Pain 24 Taylor BJ, Robbins JM, Gold JI, Logsdon TR, Bird TM, Anand KJS. Assessing
management in children with and without cognitive impairment following postoperative pain in neonates: A multi-center observational study.
spine fusion surgery. Paediatr Anaesth 2001; 11: 453–458. Pediatrics 2006; 118: 992–1000.

Journal of Perinatology
 Management of pain in the NICU
KJS Anand
S11

25 Chambers CT, Ewis GJ, McGrath PJ, Finley GA. Development and preliminary 41 Franck LS, Boyce WT, Gregory GA, Jemerin J, Levine J, Miaskowski C. Plasma
validation of a postoperative pain measure for parents. Pain 1996; 68: norepinephrine levels, vagal tone index, and flexor reflex threshold in
307–313. premature neonates receiving intravenous morphine during the post-
26 Tadio A, Lee C, Yip A, Parvez B, McNamara PJ, Shah V. Intravenous operative period: a pilot study. Clin J Pain 2000; 16: 95–104.
morphine and topical tetracaine for treatment of pain in preterm neonates 42 Simons SHP, van Dijk M, van Lingen RA, Roofthooft D, Duivenvoorden HJ,
undergoing central line placement. JAMA 2006; 295: 793–800. Jongeneel N et al. Routine morphine infusion in preterm newborns who
27 Young TE, Mangum B (eds.). Neofaxs. 19th edition. Raleigh: Acorn received ventilatory support: A randomized controlled trial. JAMA 2003; 290:
Publishing, 2006, 150–167. 2419–2427.
28 Taketomo CK, Hodding JH, Kraus DM (eds.). Pediatric Dosage Handbook. 43 Liu JG, Rovnaghi CR, Garg S, Anand KJS. Opioid receptor desensitization
12th edition. Hudson: Lexi-Comp Inc. 2005. contributes to thermal hyperalgesia in infant rats. Eur J Pharmacol 2004;
29 Hardman JG, Limbird LE (eds) Goodman and Gilman’s The 491: 127–136.
Pharmacological Basis of Therapeuticsth edition. McGaw Hill: New York, 44 Anand KJS, Suresh S. Opioid tolerance in neonates: a state-of-the-art review.
2001. Paediatr Anaesth 2001; 11: 511–521.
30 Johnston CC, Collinge JM, Henderson SJ, Anand KJS. A cross-sectional survey 45 Angeles DM, Wycliffe N, Michelson D, Holshouser BA, Deming DD, Pearce WJ
of pain and pharmacological analgesia in Canadian neonatal intensive care et al. Use of opioids in asphyxiated term neonates: effects on neuroimaging
units. Clin J Pain 1997; 13: 308–312. and clinical outcome. Pediatr Res 2005; 57: 873–878.
31 Kahn DJ, Richardson DK, Gray JE, Bednarek F, Rubin LP, Shah B et al. 46 Anand KJ, Hall RW, Desai N, Shephard B, Bergqvist LL, Young TE et al.
Variation among neonatal intensive care units in narcotic administration. Effects of morphine analgesia in ventilated preterm neonates: primary
Arch Pediatr Adolesc Med 1998; 152: 844–851. outcomes from the NEOPAIN randomised trial. Lancet 2004; 363:
32 Chay PCW, Duffy BJ, Walker JS. Pharmacokinetic-pharmacodynamic 1673–1682.
relationships of morphine in neonates. Clin Pharmacol Ther 1992; 51: 47 Taketomo CK, Hodding JH, Kraus DM (eds). Pediatric dosage handbook.
334–342. 11th edn. LexiComp: Hudson, Ohio, 2004.
33 Farrington EA, McGuinness GA, Johnson GF, Erenberg A, Leff RD. 48 Koehntop DE, Rodman JH, Brundage DM, Hegland MG, Buckley JJ.
Continuous intravenous morphine infusion in postoperative newborn Pharmacokinetics of fentanyl in neonates. Anesth Analg 1986; 65: 227–232.
infants. Am J Perinatol 1993; 10: 84–87. 49 Liu JG, Anand KJS. Protein kinases modulate the cellular adaptations
34 Saarenmaa E, Neuvonen PJ, Rosenberg P, Fellman V. Morphine clearance associated with opioid tolerance and dependence. Brain Res Brain Res Rev
and effects in newborn infants in relation to gestational age. Clin 2001; 38: 1–19.
Pharmacol Ther 2000; 68: 160–166. 50 Chana SK, Anand KJS. Can we use methadone for analgesia in neonates?
35 Scott CS, Riggs KW, Ling EW, Fitzgerald CE, Hill ML, Grunau RV et al. Arch Dis Child Fetal Neonatal Ed 2001; 85: F79–F81.
Morphine pharmacokinetics and pain assessment in premature newborns. 51 Wang HY, Friedman E, Olmstead MC, Burns LH. Ultra-low dose naloxone
J Pediatr 1999; 135: 423–429. suppresses opioid tolerance, dependence and associated changes in mu
36 Bouwmeester NJ, Anderson BJ, Tibboel D, Holford NH. Developmental opioid receptor-G protein coupling and Gbg signaling. Neuroscience 2005;
pharmacokinetics of morphine and its metabolites in neonates, infants and 135: 247–261.
young children. Br J Anaesth 2004; 92: 208–217. 52 Green SM, Denmark TK, Cline J, Roghair C, Abd Allah S, Rothrock SG.
37 Selley DE, Sim LJ, Xiao R, Liu Q, Childers SR. mu-Opioid receptor-stimulated Ketamine sedation for pediatric critical care procedures. Pediatr Emerg
guanosine-50 -O-(gamma-thio)-triphosphate binding in rat thalamus and Care. 2001; 17: 244–248.
cultured cell lines: signal transduction mechanisms underlying agonist 53 Ikonomidou C, Bosch F, Miksa M, Bittigau P, Vockler J, Dikranian K et al.
efficacy. Mol Pharmacol 1997; 51: 87–96. Blockade of NMDA receptors and apoptic neurodegeneration in the
38 Saarenmaa E, Huttunen P, Leppaluoto J, Meretoja O, Fellman V. Advantage developing brain. Science 1999; 283: 70–74.
of fentanyl over morphine in analgesia for ventilated newborn infants after 54 Soriano SG, Anand KJS, Rovnaghi CR, Hickey PR. Of mice and men: Should
birth: a randomized trial. J Pediatr 1999; 134: 144–150. we extrapolate rodent experimental data to the care of human neonates?
39 Bhat R, Abu-Harb M, Charl G, Gulati A. Morpine metabolism in acutely ill Anesthesiology 2005; 102(4): 866–869.
preterm newborn infants. J Pediatr 1992; 120: 795–799. 55 van Alfen-van der Velden AAEM, Hopman JCW, Klaessens JHGM, Feuth T,
40 Franck LS, Vilardi J, Durand D, Powers R. Opioid withdrawal in neonates Sengers RCA, Liem KD. Effects of midazolam and morphine on cerebral
after continuous infusions of morphine or fentanyl during extracorporeal oxygenation and hemodynamics in ventilated premature infants. Biol
membrane oxygenation. Am J Crit Care 1998; 7: 364–369. Neonate 2006; 90: 197–202.

Journal of Perinatology