INTRODUCTION — Neonatal discomfort, stress, or pain may be associated with routine patient care (eg, physical examination and diaper changes), frequent invasive procedures (eg, suctioning, phlebotomy, and peripheral intravenous [IV] line placement), or infrequent but even more invasive procedures (eg, chest tube placement and circumcision).

Although it is challenging to detect and measure the intensity of pain in neonates because of their inability to communicate with care providers, parents expect care providers to prevent or protect their infant from experiencing pain if at all possible. Therefore, accurate pain assessment is a necessary part of neonatal pain management to determine therapy should be initiated, as well as for assessing its effectiveness. The need for effective pain management and the assessment of pain in neonates will be reviewed here. Prevention and treatment of neonatal pain are discussed separately. (See "Prevention and treatment of neonatal pain".)

DEFINITIONS — This topic review uses the following terms as defined by the neonatal pain control group of the Newborn Drug Development Initiative [1]:

●Pain – An unpleasant somatic or visceral sensation associated with actual or potential tissue damage

●Stress – A disturbance of the dynamic equilibrium between an infant and his/her environment that results in a physiologic response by the infant

●Stress or pain response – The individual's physiologic response to pain or stress that is characterized primarily by changes in four domains (ie, endocrine-metabolic, autonomic, neurophysiological, and/or behavioral responses)

●Analgesia – Absence or reduction of pain in the presence of stimuli that would normally be painful

●Pain control – Reduction in the intensity, frequency, and/or duration of pain

BACKGROUND — The need to assess and treat neonatal pain has been increasingly appreciated with correction of past misconceptions and better understanding that neonates, including preterm infants, experience pain and the potential effects of inadequate pain control. However, there is no scientific evidence supporting these misconceptions but rather increasing evidence that neonates experience pain and stress.

Historically, pain prevention and control have been underutilized in neonates because of the following misconceptions:

●The pain pathways in neonates are unmyelinated or otherwise immature and cannot transmit painful stimuli to the brain.

●There is no alternative for verbal self-report, which remains the "gold standard" for conveying a subjective experience like pain.

●Pain perception is located only in the cortex, and thalamocortical connections must be fully developed in order to allow pain perception.

●The human infant does not have the psychological context in order to identify any experiences as painful and this does not develop until two years or later.

●Newborn infants are at greater risk for the adverse effects of analgesic or sedative agents, or these drugs have adverse long-term effects on brain development and behavior.

Neonatal sensory pain system — Beginning in the 1980s, accumulating evidence demonstrated that both preterm and term infants experience pain and stress in response to noxious stimuli [2-4]. By the middle of the second trimester, the human fetus has a highly differentiated and functional sensory system [5-7]. The fetal system appears to transmit different sensory modalities (pain, touch, and vibration), which are mediated by different pathways and loci in sensory processing, than in the mature adult system.

Numerous studies have documented neonatal responses to pain, which include autonomic (eg, increases in heart rate, blood pressure), hormonal (eg, cortisol and catecholamine responses), and behavioral changes (eg, facial grimace) [2,8-12]. These responses form the basis of the many pain assessment tools used to evaluate acute pain in the neonate [9,13-16]. (See 'Pain assessment' below.)

Types of neonatal pain — Pain in the neonate was previously classified into three categories [17]:

●Acute or physiological pain – Occurs from skin-breaking procedures or tissue injury caused by diagnostic or therapeutic interventions. Infants admitted to the neonatal intensive care unit (NICU) repeatedly experience acute pain from an average of 12 to 16 invasive procedures each day [18,19].

●Established pain – Occurs following surgery, localized inflammatory conditions (eg, abscess or thrombophlebitis), or birth-related trauma.

●Prolonged or chronic pain – Results from severe diseases such as necrotizing enterocolitis (NEC) or meningitis, or rare conditions such as scalded skin syndrome or Harlequin syndrome (see "Vesicular, pustular, and bullous lesions in the newborn and infant", section on 'Staphylococcal scalded skin syndrome'). However, it has been challenging to define neonatal chronic pain, as perspectives from care providers differ and a scientific rationale for prolonged or chronic pain could not be developed [20,21].

A scientific framework to categorize neonatal pain was developed by the author based on clinical parameters including temporal features, clinical character, secondary effects, and neonatal response patterns [22]. This framework differentiates neonatal pain into several categories, including acute episodic, acute recurrent, prolonged, persistent, and chronic pain. Classification is based on onset, duration, character, presence of primary or secondary hyperalgesia (increased sensitivity to pain), or allodynia (central pain sensitization due to increased response of neurons following normally nonpainful repetitive stimulation), and typical behavioral and physiological responses associated with these types of pain (table 1) [22].

Neonatal brain response to pain — Neuroimaging and neurophysiological studies have reported the following brain responses to painful stimuli in both preterm and term infants [23-28]. However, these measures should only be used in the research setting, as data are limited regarding their accuracy and specificity, and due to the availability and cost of these studies [28].

●In preterm infants, near infrared spectroscopy (NIRS) has demonstrated increased cortical activation in the somatosensory areas of the brain in response to painful stimuli (eg, heel stick or venipuncture) [23,24]. Simultaneous imaging and physiologic testing using NIRS and electroencephalography also confirmed cortical activation with greater temporal and spatial resolution [25].

●In term infants less than seven days old, functional magnetic resonance imaging (fMRI) studies identified brain activation in 18 of the 20 brain regions typically activated in healthy adults following noxious stimulation [26]. There was no activation in the infant amygdala or orbitofrontal cortex. These results demonstrate that sensory and affective components of pain are active in infants and suggest that the infant pain experience closely resembles that of adults.

●Electroencephalography (EEG) has been shown to identify nociceptive brain activity evoked by acute noxious stimuli and sensitive to analgesia [28]. It has low sensitivity and specificity [29]. However, the routine assessment of EEG responses to neonatal pain is not ready for clinical application because of its relatively low sensitivity (64 percent) and specificity (65 percent) and the difficulty of obtaining and interpreting EEG signals from term and preterm neonates [29].

Frequency of painful procedures — Painful procedures are common in neonates, especially in those in the NICU. Analgesic therapy is often not given, despite greater understanding that neonates experience pain [18,19,30].

The prevalence of untreated neonatal pain was best illustrated by a large prospective French study of 430 neonates admitted to tertiary NICUs during a six-week time period beginning in September 2005 [19]. All painful and stressful procedures were recorded for each participant during the first 14 days following NICU admission. The following findings were noted:

●Overall, almost 70,000 first-attempt procedures were recorded, of which 70 percent were painful and 30 percent were stressful. An additional 11,500 subsequent attempts were performed, of which 90 percent were painful and 10 percent were stressful.

●Neonates experienced a median of 115 procedures during the 14-day study period, of which 75 were painful.

●Of the 42,413 painful procedures, specific analgesic therapy was provided in 20.8 percent of patients, which included only nonpharmacologic therapy (18 percent), only pharmacologic therapy (2 percent), and both nonpharmacologic and pharmacologic therapy (0.4 percent). An additional 34 percent of patients were receiving concurrent analgesia or anesthetic therapy for other reasons during the procedure.

●Factors associated with greater use of specific preprocedural analgesia included prematurity, the type of procedure, parental presence, surgery, daytime, and day of procedure after the first day of admission. In contrast, mechanical and noninvasive ventilation and use of concurrent analgesia were associated with lower use of specific preprocedural analgesia.

Subsequent studies looking at changes in practice patterns over the past decade have noted an increased attention to pain assessment in the NICU, a decline in the numbers of painful procedures performed, and increased use of pain management using analgesic drugs and nonpharmacologic approaches [30,31].

Effects of inadequately treated pain — Accumulating data suggest that untreated or inadequately treated neonatal pain may have long-term deleterious effects on pain response and neurodevelopmental outcome.

Altered pain response — Several studies have reported that exposure to repetitive pain in early life may lead to greater risk of developing increased pain sensitivity and/or chronic pain syndromes during their subsequent lifespan [32-41].

For example, infants of diabetic mothers, who were exposed to repeated heel sticks just after birth, exhibited more intense pain responses (facial grimacing and crying) during later venipuncture compared with normal infants [39]. Infants exposed to circumcision pain at birth experienced greater pain at immunization four to six months later [42], whereas those exposed to gastric suctioning at birth evidenced threefold greater odds of developing irritable bowel syndrome (IBS) during adolescence or adulthood [43]. Adolescents born preterm also display higher somatic pain sensitivity than adolescents born at term [44]. These findings and other animal studies substantiate the theory that repeated exposure to neonatal pain leads to permanent changes in pain processing [32].

Neuroanatomical and behavioral changes have also been reported in adult animals that were exposed to neonatal pain [45-47]. In one study, exposure to persistent hind paw pain in newborn rats induced changes in primary afferent neurons and their spinal neuronal circuits in adult animals, both at baseline and following sensory stimulation, compared with control animals [45]. Another study showed that the neuroanatomical and behavioral changes following repetitive inflammatory pain resulted from neuronal excitotoxicity, which was ameliorated by ketamine analgesia [48]. It remains unclear whether such long-term effects will occur in adult humans exposed to prolonged pain as neonates [49-51].

Interventions that reduce neonatal pain/stress also improve clinical outcomes. In postoperative infants, those who received greater amounts of anesthesia and analgesia compared with controls had reduced levels of norepinephrine, epinephrine, glucagon, aldosterone, and cortisol; decreased postoperative morbidity (eg, sepsis, metabolic acidosis, disseminated intravascular coagulation); and a lower mortality rate [8,52].

Neurodevelopmental outcome — Neuroimaging, neuroendocrine, and neurobehavioral studies have also shown the neurodevelopmental impact of repetitive neonatal pain on long-term outcomes [53-57]. Frequency of exposure to neonatal pain-related stress have been correlated with subsequent impairments in cognitive development, altered neurocognitive processing, decreased cortical thickness, and dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis. Thus, it is essential to identify, assess, and manage neonatal pain effectively in order to minimize its impact on the intermediate- and long-term outcomes of preterm or term newborns.

PAIN ASSESSMENT

Overview — A neonatal pain control program that includes routine pain assessment should be established for each healthcare facility that cares for neonates and young infants [58]. Effective neonatal pain assessment is an essential prerequisite for optimal pain management and is based on the following:

●Selection of an appropriately sensitive and accurate clinical pain assessment tool

●Clinical staff training to ensure health care providers can detect neonatal pain using the selected assessment tool

A multicenter observational study found that postoperative pain assessments by clinicians increased the odds of receiving analgesia fourfold as opposed to those not assessed [59], which highlights the importance of pain assessment as a component of routine clinical evaluation.

However, multiple challenges limit the ability of available tools for accurate evaluation (see 'Challenges' below). Given these challenges and concerns, some authors have questioned whether scoring methods that assess pain intensity are even required for neonates [60,61]. They propose instead a "pain detection method," which takes into account the type of the noxious stimulus, the body region being stimulated, and simplifies the pain assessment to identify whether pain is present or not [61]. Although novel, the validity, feasibility, and clinical utility of this approach have not been investigated.

As a result, we continue to monitor for neonatal pain on a routine basis using validated pain assessment tools, which have been the focus of mandatory training for the clinical staff. (See 'Our approach' below.)

Pain assessment tools — Accurate neonatal pain assessment tools are required because of the inability of the infant to self-report. Neonatal pain assessment tools rely on surrogate measures of physiologic and behavioral responses to pain or noxious stimuli:

●Physiologic parameters – Changes in heart rate, respiratory rate, blood pressure, vagal tone, heart rate variability, breathing pattern, oxygen saturation, intracranial pressure, palmar sweating, skin color, or pupillary size. Some studies have used alteration in physiologic electroencephalographic (EEG) or electromyographic (EMG) patterns to assess pain, but these methods are not considered to be valid or reliable, representative of pain perception, or universally available [62,63].

●Behavioral responses – Crying patterns, acoustic features of infant crying, facial expressions, hand and body movements, muscle tone, sleep patterns, behavioral state changes, and consolability. In infants, total facial activity and cluster of specific facial findings (brow bulge, eye squeeze, nasolabial furrow, and open mouth) are associated with acute and postoperative pain [9,13-16,64,65].

Neonatal assessment tools that are used routinely are either unidimensional, meaning they are dependent on either physiologic or behavioral parameters, or multidimensional, including physiologic, behavioral, and contextual parameters (eg, gestational age [GA]) [17,58,66,67].

The scales most commonly used in the neonatal intensive care unit (NICU) for acute pain assessment include the following (table 2) [58]:

●PIPP – Premature Infant Pain Profile [13]

●PIPP-R – Premature Infant Pain Profile-Revised [68]

●N-PASS – Neonatal Pain Agitation and Sedation Scale [69]

●NIPS – Neonatal Infant Pain Scale [70]

●CRIES – Crying, Requires Oxygen Saturation, Increased Vital Signs, Expression, Sleeplessness [15]

●NFCS – Neonatal Facial Coding System [64]

●DAN – Douleur Aiguë Nouveau-né scale [71,72]

●BIPP – Behavioral Infant Pain Profile [73]

●COMFORTneo – Comfort neo scale [74]

Of note, a single assessment tool has not been adopted universally because each tool was developed and validated for selected populations and clinical settings [75]. As an example, several of these scales were initially developed for preterm infants. The PIPP has been revised with simplification of the scoring methods for oxygen saturations, facial expressions, and the baseline behavioral state, while expanding its application for neonates with GA from 25 to 41 weeks [76]. Although initial validation and feasibility of the revised version has been published [68,76], further validation and dissemination are ongoing.

Research efforts to improve the objectivity and accuracy of assessment tools are ongoing. These include using neuroimaging (functional magnetic resonance imaging [fMRI] and near-infrared spectroscopy) and neurophysiologic techniques (amplitude-integrated electroencephalography [aEEG], changes in skin conductance, and heart rate variability) during acute or prolonged pain [23,24,58,77-79].

Challenges — The following challenges limit the ability of available tools for accurate evaluation [77]:

●Interobserver variability and subjectivity – Many signs used in these assessment tools require the subjective evaluation by observers. As a result, there is significant interobserver variability in the evaluation of behavioral responses that can be reduced with multidisciplinary training of the staff [80,81].

In addition, many tools require the observation, mental calculations, and recording of 3 to 10 parameters in real time by the bedside nurse. Often, the nurse performing the painful procedure is also tasked with observing the infant's pain responses at the same time.

●Validity of the assessment tool – Since there is no "gold standard" established for pain in the neonate, the concurrent validity of many assessment tools has been questioned:

•Neuroimaging or neurophysiologic approaches used for research have not reached a level of sensitivity or specificity where they can be accepted as "the gold standard" for testing the accuracy of subjective assessment methods.

•Assessment tools that include multi-modal parameters are often limited by dissociations in the response characteristics of physiologic versus behavioral parameters [82]. These characteristics include the reactivity, responsivity, trigger threshold, onset, or decay of changes in these parameters and affect the scaling properties of the pain score.

•Pain assessment tools generally do not take into account the type of the nociceptive stimulus or the body region where it occurs. For example, very limited data are available on visceral pain or bone pain in newborn infants.

In addition, pain assessment may be limited by the availability of reliable validated tools for selected populations and clinical settings as follows [75]:

●Preterm or critically ill infants – Most pain scales are developed and validated by studies including relatively healthy infants or late preterm infants. Very preterm infants, the group most likely to undergo many painful procedures, consistently demonstrate muted responses to pain measured by these assessment tools [10,64,83-85]. Also, critically ill infants at any GA will have limited vigor or energy to mount a robust response to acute pain.

●Persistent or prolonged pain – Most tools evaluate acute pain and some evaluate postoperative pain, but do not assess persistent or prolonged pain [68]. The definition of prolonged or chronic pain in newborns remains unclear, which has led to challenges for research in this area [22,77]. As a result, tools for the assessment of persistent or prolonged pain in neonates (due to major surgery, osteomyelitis, or necrotizing enterocolitis [NEC]) have not been developed or completely validated [69,74,86,87]. During episodes of persistent pain, neonates may enter a passive state, with limited or no body movements, an expressionless face, reduced variability in heart rate and respiratory rate, and decreased oxygen consumption [58,77]. Thus, assessment tools based on these indicators will not adequately detect and assess the intensity of prolonged neonatal pain [77,83].

The EDIN (Echelle de Douleur et d'Inconfort du Nouveau-né) and COMFORTneo scale were tools developed specifically for assessing prolonged neonatal pain [78,79]. Although they are used widely, these tools have not been extensively validated.

●Mechanically ventilated patients – Most assessment tools were developed for nonventilated infants. However, several have been used in mechanically ventilated infants, including COMFORTneo scale and NFCS [58,75].

●Neurologic impairment including neuromuscular blockade – Responses to pain, including body movement and changes in facial expression, may be decreased or altered in neurologically impaired neonates [70,85] and absent in those who receive paralytic medications.

Choice of assessment tools — As noted above, although several scoring tools have been developed, the use of a single assessment tool to address all the needs for neonates is not advisable as each tool was developed and validated for selected populations and clinical settings. The choice of the pain assessment tool is dependent upon the neonatal population to be assessed, and the different types of pain that need to be evaluated [75].

Three multicenter studies illustrate the wide range of pain assessment tools used in NICUs:

●In the first study, 12 sites evaluated by the 2002 Neonatal Intensive Care Quality Improvement Collaborative used five different assessment tools [88].

●In the second study from the Child Health Accountability Initiative (CHAI), 10 sites used eight different assessment tools [59].

Because of our limited ability to detect and quantify pain in neonates, especially preterm infants, we suggest that judicious pain control measures be used to prevent or reduce pain due to known noxious stimuli [88,89]. (See "Prevention and treatment of neonatal pain".)

Additional indicators of pain — Regardless of which tool is used, we also utilize information from the clinical setting to determine the likelihood of neonatal pain. As a general rule, anything that causes pain in adults or older children will also cause pain in neonates, regardless of their GA. If parents are available, we ask for their opinion: Do they feel that their baby is in pain? In cases in which there is a likelihood of pain or perceived pain, we suggest that pain control measures be administered pre-emptively to prevent or reduce pain due to known noxious stimuli.

Although not universally available, measures of neonatal stress (eg, skin conductance activity) may be indicative of pain. Serum or salivary cortisol levels can provide an indication of the level of stress as well as skin conductance activity [78,79]; however, these tests are generally not clinically useful as results are not available in real-time and may be affected by other factors (such as illness severity).

Staff training — Staff recognition of neonatal pain will determine whether or not neonates receive adequate pain control [59,90]. In a multicenter observational study, the documentation of clinician pain assessment was significantly associated with the use of pharmacologic analgesia after surgery [59]. In contrast, the type of surgery (major or minor) was not associated with the administration of pharmacologic therapy. Thus, each facility that cares for infants should adopt an assessment strategy for the detection and documentation of pain.

Institution of a pain management program increases the awareness of the staff that pain occurs routinely in the NICU [59,81,90] and that its control is an important clinical goal [1,91,92]. The effective use of a clinical assessment tool requires mandatory training of the staff to improve interobserver reliability and to educate the staff on the limitations of the selected tool [77].

Our approach — At our institution, we assess pain every four hours when vital signs are measured, and after each painful or therapeutic intervention. We use the PIPP-R and/or N-PASS for assessment of acute or postoperative pain [13,65] and the N-PASS for the assessment of prolonged pain [69]. We also utilize various contextual factors, and other behavioral or physiological indicators suggesting inadequate analgesia (eg, pupillary dilation). If the parents are available at the bedside, we ask for their opinion: Do they feel that their baby is in pain?

SUMMARY AND RECOMMENDATIONS

●Neonates experience pain from the same interventions or clinical conditions as older children and adults. (See 'Background' above.)

●Untreated or inadequately treated neonatal pain may have immediate and long-term effects including altered pain sensitivity and reactivity, and other adverse health outcomes. (See 'Effects of inadequately treated pain' above.)

●Assessment of neonatal pain is challenging because of the inability of the infant to communicate with care providers. Assessment tools based on contextual factors (eg, gestational age [GA]), and physiologic and behavioral responses to pain have been developed to detect and measure the intensity of neonatal pain (table 2). However, a single assessment tool has not been universally adopted because each tool was developed and validated for selected populations and clinical settings. (See 'Pain assessment tools' above and 'Choice of assessment tools' above.)

●Despite the limitations of currently available tools, each facility that cares for infants should adopt an assessment strategy for the detection of pain. This includes routine assessment by trained health care workers using standardized pain assessment tools that are appropriate for the neonatal population and clinical setting. However, clinicians need to be aware of limitations of these assessment instruments and altered patient responses due to neurologic impairment, prematurity, or neuromuscular blockade. (See 'Choice of assessment tools' above and 'Challenges' above.)

●In our practice, we assess pain every four hours when vital signs are measured, and after each painful or therapeutic intervention using the PIPP-R (Premature Infant Pain Profile-Revised) for evaluation of acute or postoperative pain and the N-PASS (Neonatal Pain Agitation and Sedation Scale) for the prolonged pain. (See 'Our approach' above.)

●Because of the limited ability to detect and quantify neonatal pain, we recommend that pain control measures should be administered to prevent or reduce pain due to known or suspected noxious stimuli (Grade 1C). (See "Prevention and treatment of neonatal pain" and 'Choice of assessment tools' above.)