Topical anesthetics have been reasonably well researched in this population primarily related to several assumed benefits including noninvasive method of administration, lack of systemic effects, and potential for effectiveness. In an early review paper, Taddio and colleagues [71] evaluated the use of lidocaine–prilocaine cream (EMLA^®, Astra Pharma) compared to placebo in treating pain from heel lance, venipuncture, arterial puncture, lumbar puncture, percutaneous venous catheter placement, and circumcision in preterm and term infants. Nine randomized controlled trials (RCTs) were included. Unfortunately, for the most commonly performed procedure in the NICU, heel lance for blood procurement, EMLA was not shown to be beneficial. Similarly, in two later studies examining the effect of tetracaine 4 % gel (Ametop^®, Smith & Nephew), on the pain of heel lance in both preterm and term newborns, no reduction in pain scores or duration of crying was noted between the groups [72, 73]. It has been postulated that variation in perfusion and skin thickness of an infant’s heel may contribute to this ineffectiveness [74].

Similarly, the use of topical anesthetic has not been shown to be effective in diminishing pain associated with the insertion of intravenous lines or peripherally inserted central catheters (PICC) [75–77]. Some evidence was provided for the use of EMLA in relieving pain during venipuncture; however, results remain inconclusive. There have been five clinical trials examining the effect of topical anesthetic (EMLA and tetracaine 4 %) for pain associated with venipuncture [76, 78–81]. Results show that the application of local anesthetic could decrease the duration of cry but increase the procedure time [79], as well as being dose (0.5 vs. 1 ml) [78] and application time dependent (30 vs. 60 min) [76, 80].

There are no known contradictions to using preemptive local or topical anesthetics for lumbar puncture in neonates, and their use has been associated with increased success in obtaining cerebral spinal fluid (CFS) [82] and potential benefits related to a reduction in pain score and physiological stability [83]. In a randomized trial comparing the effect of lidocaine–prilocaine (EMLA) (1 g over 60–90 min) compared to placebo for 60 infants undergoing a lumbar puncture, infant in the intervention group had lower mean HR at needle insertion (P = 0.001) and needle withdrawal (P < 0.001) and lower total behavioral score again at insertion (P < 0.004) and needle withdrawal (P < 0.001) [84].

Currently, the most widely utilized local anesthetic for injection is lidocaine hydrochloride 1 %. It is effective as an adjuvant pain relieving strategy for lumbar puncture, chest tube insertion, and circumcision [85–89].

The use of EMLA to relieve pain caused by a frequently performed procedure in neonates, circumcision, has been shown to be more effective than placebo, as indicated by changes in physiological and behavioral pain indicators [86], and these findings were similar to a later Cochrane systematic review [71]. In another Cochrane review regarding pain relief for circumcision that included 35 trials involving 1,997 full-term and preterm infants, when compared to placebo, dorsal penile nerve block (DPNB), EMLA, and sweet taste all reduced pain response [90]. Of the six trials (n = 190) specifically examining EMLA compared to placebo, infants receiving EMLA demonstrated significantly lower facial action scores, decreased time crying, and lower heart rate. However, when EMLA and sweet taste were compared with DPNB, crying and elevation in heart rate were lowest in the DPNB group. Despite the large number of trials, small sample sizes, lack of blinding, large variations in practice, and little use of age appropriate validated pain tools limited the author’s ability to make concise recommendations. The authors concluded that topical anesthetic in conjunction with DPNB as well as other pain relieving strategies could be safely implemented as part of routine practice related to circumcision.

Acetaminophen is one of the most commonly used analgesics for both mild ongoing pain and intermittent medical procedures [91]. Interestingly, despite its widespread use, there is limited evidence regarding its efficacy related to procedural pain alleviation in newborns [92–95]. Even at very high oral doses (40 mg/kg), it did not diminish the pain associated with heel lance [96]. The widespread use of prophylactic acetaminophen prior to immunization has been recently refuted, although its administration for local pain or swelling postinjections is still supported [97, 98].

The efficacy of intravenous acetaminophen has been better studied, and it appears to be beneficial for the relief of postoperative pain and act as an opioid sparing agent [99–101]. Its use for intermittent procedure pain has not been reported.

Although opioids continue to be the mainstay in the neonatal intensive care unit (NICU) for the treatment of ongoing painful conditions such as necrotizing enterocolitis, operative procedures, and postoperative care, their use for more common single procedures performed in the NICU has been less promising [91]. Systemic administered drugs, specifically opioids, are highly sensitive to development [102, 103] and have significantly slower clearance in neonates [104–108]. Morphine and fentanyl are the predominate opioids used in hospitalized newborns with morphine being the most studied.

There have been conflicting reports regarding the efficacy and safety of intermittent and continuous intravenous infusions morphine for routine medical procedures and the stress associated with mechanical ventilation. Morphine does not appear to be beneficial for some of the commonly performed procedures in the NICU such as tracheal suctioning [109] or heel lance [110]. Validated pain scores were not significantly different for 42 preterm infants, mean GA at birth of 27 weeks, randomized to receive a loading dose, and continuous infusion of morphine or placebo during heel lance over three time points [110]. Conversely, in an earlier study conducted by Anand [111], procedural pain (endotracheal suctioning) response was found to be much lower in the infants receiving morphine compared to placebo. Similarly in a much larger trial, pain scores in response to endotracheal suctioning were lower with morphine [112]. However, the incidence of longer duration of mechanical ventilation, hypotension, and severe intraventricular hemorrhage was higher in infants receiving more frequent intermittent doses of morphine regardless of assigned group.

In a systematic review of 13 RCTs examining the effectiveness of opioid analgesia in reducing the pain experienced from mechanical ventilation, the authors concluded that there was insufficient data to support the routine use of opioids in mechanically ventilated newborns [113]. The broad range of opioid dose and variation in type of analgesia in the trials also contributed to the findings. Of note, pain scores were significantly lower in four of the trials, and the authors did recommend that opioids should be used cautiously and in combination with well-validated pain scoring measures to evaluate their effectiveness. The authors also reported a higher incidence of hypotension and poorer neurodevelopmental outcome associated with midazolam compared to morphine. Therefore, if sedation is required, morphine appears to be a safer choice than midazolam.

There do appear to be some acutely painful conditions that warrant the use of morphine. Intravenous morphine was found to be more advantageous than topical application of tetracaine for the management of pain associated with insertion of a central venous catheter in neonates [75]. Remifentanil, a fast-acting opioid, has also been found to be analgesic for the insertion of a PICC. When compared to placebo, a 0.03 mcg/kg infusion of remifentanil significantly lowered the pain score of very preterm neonates undergoing insertion of a PICC. Mean pain scores [NIPS and PIPP] at skin preparation T1 and needle insertion T2 were significantly different to baseline T0 and recovery T3. No improvement was noted with respect to the number of attempts needed to successfully perform the procedure [114].

There is increasing consensus that opioids with rapid onset in combination with anticholinergics and muscle relaxants should be used for all infants undergoing elective intubation [115, 116]. In a review of nine trials, Shah [117] reported that the use of premedication was associated with a reduction in physiological pain indicators and intubation times. The most common and preferred agents reported were fentanyl, atropine, and rocuronium, although differences in medication and dosages were common across sites [118]. Morphine’s slower onset of peak effect could contribute to its lack of efficacy [119]. Results from studies examining two synthetic agents, alfentanil and remifentanil, are promising [120, 121]. Ongoing research to determine the optimal dosage, administration route, and combination of medications as well as the long-term neurodevelopmental effects are warranted [118].

Sucking movements start in uterus around 12–14 weeks postconceptional age (PCA) [125], the sucking reflex develops around the 17th week postconception, and regular sucking activity is found in fetuses of 27–28 weeks PCA. NNS is stimulated in neonates by placing a pacifier in the infant’s mouth.

Based on its effect in reducing fussing and crying in preterm infants in neonatal care [126], several studies have looked at NNS, comparing it to other interventions or no intervention in order to determine whether it decreases the responses of neonates to painful procedures. NNS has been found to reduce the increase in heart rate [127]; reduce crying time in term and preterm neonates [127–132], even in those who are intubated and ventilated [133]; and reduce pain scores [129, 131, 134–139].

Compared to swaddling, NNS after heel lance interrupted crying earlier (23.2 s vs. 58.7) and promoted a faster decline in heart rate although infants spent more time in an alert state (59 % of the time vs. 22 %, P < 0.01) [140]. Compared to rocking, heart rate was also significantly reduced by NNS, but these infants slept more in the rocking group [128]. Compared to sucrose, glucose, and sucrose with pacifier, the median pain scores of term infants who received NNS for venipuncture was significantly lower (2) than that of infants who received sweet solutions (5). Although the scores with sucrose and pacifier (1) were the lowest, they were not significantly different from NNS alone (P = 0.06) [136].

Adding other interventions to NNS appears to be beneficial. In a crossover trial, pacifier alone was compared to pacifier plus music therapy, music therapy alone, and no intervention during and after heel lance [139]. All three interventions improved the pain response, compared to no intervention, but NNS combined with music was associated with the lowest NIPS scores and the highest transcutaneous oxygen saturation (TcPaO₂) levels, while music therapy alone produced the lowest heart rate. Regarding the synergistic effect of simultaneously using NNS and sweet solutions, while adding a pacifier to sucrose or glucose seems to enhance the effect of sweet solutions used alone [130, 131], adding sucrose or glucose to a pacifier seems to provide no additional benefit than using a pacifier alone [134–138].

Including pacifier with glucose in a combined intervention named sensorial saturation (SS) that includes, besides taste, sight, touch, voice, and smell has shown to be more efficacious than glucose with pacifier in reducing the pain scores of term newborns during heel lance [137].

The capacity of infants to distinguish between flavors, namely, sucrose, quinine, and corn oil has been demonstrated [141], and the calming effects of sweet taste have been known for a long time. Animal studies reinforce the evidence from studies in human infants that sucrose, glucose, and fructose but not lactose have a calming and pain-reducing effect, increasing the latency to withdraw from a heated surface in rat pups [142]. A recently updated Cochrane systematic review including 44 studies concluded that sucrose is efficacious and safe to use in single and repeated heel lances and should be considered for venipuncture since it significantly reduces pain behaviors and composite measures [143]. The authors of this review state that for other procedures such as eye examination for retinopathy of prematurity, bladder catheterization, nasogastric tube insertion, circumcision, and subcutaneous injections, further studies are required due to conflicting evidence and that the use of sucrose in extremely low birth weight and unstable and/or ventilated neonates needs to be addressed.

The recommended dose is a small volume of 0.05–0.5 ml of a 24 % sucrose solution for preterm neonates and 1–2 ml, administered 2 min prior to the painful procedure, for term neonates [143]. Concentrations of sucrose have varied from 12 to 50 %, but a ceiling effect seems to be reached at 25 % [144]. The most common method of administration is via syringe or dropper placing the solution on the anterior surface of the infants’ tongue, but a pacifier dipped in a sucrose solution may also be used and is estimated to deliver approximately 0.1 ml [134]. The repeated use of sucrose for heel lance seems to not reduce its efficacy [144, 145]. Regarding concerns about long-term effects, one study has found a poorer neurobehavioral development in neonates younger than 31 weeks PCA [146]. However, a secondary analysis of the same data showed that increased risk occurred in neonates who had more than ten doses of sucrose in 24 h [147]. A subsequent study has found that infants who had procedural pain consistently managed by sucrose and pacifier in the first 28 days of life had no difference in adverse events or clinical outcomes such as intraventricular hemorrhage compared to infants who received no sucrose [148].

Glucose is another well-studied source of sweet taste. Its efficacy in a volume range of 0.3–2 ml of a 30 % solution has been shown for heel lance [149] and venipuncture [136, 150, 151], both in term [149, 150] and preterm [151] infants, as well as for subcutaneous injections in very preterm neonates (25–32 weeks GA) [152].

Comparisons between similar volumes and concentrations of sucrose and glucose show similar effects in reducing pain scores [126, 153–155], and both sucrose [156] and glucose [157] compare favorably to a topical anesthetic cream for venipuncture in full-term newborns.

Wrapping young infants in a cloth is part of the traditional way of care in many cultures [158]. An extensive systematic review of 78 studies evaluating the effects of swaddling, four of which examining pain control, concluded that it reduces crying, physiologic distress, and motor activity; increases sleep; and improves neuromuscular development in preterm infants [158]. Regarding pain control during heel lance, neonates over 30 weeks PCA returned to their baseline facial activity, heart rate, and arterial oxygen saturation levels more quickly [158]. A meta-analysis of four unpublished studies conducted in Thailand also supported the efficacy of swaddling, with moderate to large mean effect sizes in full-term babies during heel lance [159].

Containment or facilitated tucking by holding the infant in a side lying position, arms and legs flexed near the trunk [160], also has been shown to reduce behavioral signs of distress of very low birth weight infants during heel lance [161], endotracheal suctioning [162], and pharyngeal suctioning [163]. Facilitated tucking seems to be more efficacious than water or oxycodone in reducing pain scores during pharyngeal suctioning and heel lance in very low birth weight infants and was equivalent to 0.2 ml of 24 % glucose but presents less short-term adverse effects, such as desaturation and/or bradycardia, than oral glucose [164]. In addition, in two studies, facilitated tucking was performed by parents, offering them an opportunity to participate in alleviating their infants’ distress [163, 164].

Rocking has also been a traditional way to calm infants and promote sleep. When compared to the use of a pacifier after heel lance in term infants, while both interventions reduced crying and can therefore be considered efficacious, rocking promoted arousal levels more than pacifiers, which promoted sleep [128]. A more recent trial compared rocking, expressed breast milk, 20 % sucrose, water, NNS, and massage in term, stable neonates [165]. Neonates were rocked by lifting the baby’s head off the cot on the palm of the hand but not the body and making rocking movements in a gentle, rhythmic manner during and up till 2 min after heel lance. Like infants in the NNS group, infants who received rocking cried less and had lower pain scores at 2 and 4 min after the painful procedure, while infants in the sucrose group had a reduced pain score only at 30 sec [165]. Another trial in preterm infants during heel lance compared simulated rocking (infants in supine or side lying position on an oscillating air mattress), sucrose, usual incubator care with no intervention, and a combination of simulated rocking and sucrose [166, 167]. Simulated rocking combined with sucrose decreased facial expression by 40 % and so did sucrose alone, while simulated rocking was no better than incubator care, suggesting that the pain-reducing effect was related to the sucrose administration.

Human fetuses’ capacity of perceiving sound at different frequencies and responding to them develops from 19 weeks of GA to term [167]. Their ability to learn and remember auditory stimuli from the intrauterine environment as early as 22 weeks GA has been put into evidence by conditioning studies [168]. It was demonstrated that infants as young as 3 days preferred their mothers’ voice to the voice of another female [169], and exposure to familiar sounds has been associated to improved physiological stability [170]. The soothing effects of familiar sounds during painful procedures have been evaluated in a few trials. Maternal heart rate, Japanese drum with identical rhythm, and no sound were offered to 131 full-term infants who underwent heel lance [171]. Infants exposed to maternal heart beat had reduced facial response and crying, as well as lower levels of salivary cortisol. Following an identical rationale, 20 preterm infants 32–36 weeks were exposed to recorded and filtered maternal “singsong” voice or to no voice during a heel-lance procedure in a randomized crossover design [172]. No significant differences were found in pain scores between conditions, and the authors conclude that maternal voice alone, without other components of maternal presence, may not be enough to reduce pain response.

The effects of music to reduce pain have also been examined. A recent systematic review of RCTs of music for medical indications in the neonatal period found six studies that looked at painful procedures, three for circumcision and three for heel lance [173]. Only one of the studies on circumcision [174] had high-methodological quality and showed a lower pain score, lower heart rate increase, and higher arterial oxygen saturation levels, while the other two studies found no significant differences between groups. For heel lance, three trials were included but considered to have poor methodological quality [173]. One crossover trial with 27 infants 28 weeks GA or more found a significant decrease in heart rate and pain scores and an improvement in arterial oxygen saturation levels with music but also with music combined with NNS and with NNS alone [139]. Another crossover trial including 14 infants of 29–36 weeks PCA found a significant effect on heart rate, behavioral state, and pain scores only in infants over 31 weeks [175].

Auditory stimulation may be administered through different types of sounds, from music to direct or recorded maternal voice, or filtered voice and heart beat that would resemble the sound heard in the womb. Significant changes during the maturation process occurring in the last trimester of pregnancy with implication on the frequencies and levels of intensity that can be perceived by neonates of different GAs pose an important challenge when designing appropriate auditory interventions and methodologically sound studies.

During their prenatal experience, human fetuses are exposed to the numerous compounds of the amniotic fluid, which play an active role in shaping the development of chemosensory sensitivity and preferences [176]. It has been demonstrated that newborns are able to discriminate odors and have head-orientation behavior toward their own amniotic fluid, showing their preference for a familiar versus non familiar odor [177]. In preterm infants, responses elicited by odorants are weak and irregular at 24 weeks PCA but reliable by week 28 [176]. Given the soothing effect of the smell of amniotic fluid in term neonates separated from their mothers following birth [177], the effect of olfactory stimulation for painful procedures has gained increasing interest.

To determine the soothing effect of familiar and unfamiliar odor in full-term infants undergoing a routine heel lance, 44 breast-fed newborns were randomized to 4 groups [178]: (1) infants naturally familiarized with their mother’s milk odor (2), infants previously familiarized with vanilla odor (3), infants not previously exposed to vanilla odor, and (4) infants who received no intervention. Results showed that the neonates in group 1 and 2 who received the odors during and after heel lance showed less distress during the recovery phase compared with the heel lance phase. Furthermore, the infants who were not exposed previously to the vanilla odor and those in the control group showed no difference in grimacing and cry during and after the heel lance. Babies who smelled their mother’s milk exhibited significantly less motor agitation during the heel lance compared with the other groups. Whether familiarization to the odor was obtained through the mother or without the mother did not make a difference as shown in a replication of the previous study, in which the calming effects of familiar odor were visible during the heel-lance phase [179]. In healthy preterm newborns, a familiar odor compared to unfamiliar or no odor also reduced crying and grimacing [180]. A comparison between mother’s milk, non-mother’s milk, and formula milk given to healthy full-term neonates showed that crying, grimacing, and motor activity during heel lance were decreased only by exposure of the infant’s own mother’s milk [180].

Olfactory stimulation has been used in full-term and preterm infants during heel lance as a component of SS, an intervention that combines visual stimulation (looking the baby in the face to attract his attention), auditory stimulation (speaking to the infant gently but firmly), tactile stimulation (massaging the infant’s face and back), and gustatory stimulation (glucose with pacifier) [137, 138, 181]. Within the original concept of SS, olfactory stimulation was provided by letting the infant smell the fragrance of baby oil on the therapist’s hands [138], but a modified version without perfume has also shown to be effective in reducing pain scores of full-term healthy neonates [181]. Moreover, the modified intervention was shown to be more effective on a cry scale [182] than 1 ml of 30 % glucose with pacifier, raising questions regarding the importance of the olfactory component of SS.

The mechanism underlying the comforting effect of intrauterine, maternal, and familiarized smell remains unclear but it has been postulated that it is an opioid-mediated system. This hypothesis derives, on one hand, from knowledge that the taste system and the olfactory system are linked and that the antinociceptive effect of sweet taste is opioid mediated [183]. Conversely, animal studies indicated that the opioid system modulates olfactory learning and odor preferences [184].

The mother–infant dyad has an innate mutual bond that is key to survival. Infants actively mediate this bond by eliciting distress cues when separated from their mother that in turn heightens a mothers’ instinctive need to protect and comfort their young. Therefore, it is not surprising that researchers returned to this basic human premise to investigate whether maternal presence could diminish the effects of repeated procedural pain exposure during prolonged hospitalization. The first studies followed the oro-gustatory research and focused on the use of breast milk or breastfeeding to attenuate the pain associated with common newborn procedures such as heel lance, venipuncture, and intramuscular injection. In a systematic review of eleven clinical trials, six examining the effectiveness of supplemental breast milk and five examining breastfeeding, breast milk giving orally by syringe was no different than water and significantly less beneficial than sweet taste in both full-term and preterm infants undergoing routine procedural pain from heel lance and venipuncture [185]. In contrast, breastfeeding when compared to placebo was shown to provide analgesia. In addition, when comparing to sweet taste, breastfeeding has been shown to be equivocal [186] and may even be superior [187]. Healthy term neonates (37–42 weeks of gestation at least 60 h old) undergoing heel lance for metabolic screening had lower median PIPP scores in the breastfeeding group (3.0) than those infants receiving 1 ml sucrose solution (8.5). The benefits of glucose and breastfeeding may be cumulative when provided simultaneously [188]. The efficacy of breastfeeding to diminish the painful effects of immunization appears to continue to at least 1 year of age. Consistent findings have been reported in three studies. Breast-fed infants when compared to controls experienced significantly shorter duration of crying, 35.85 vs. 76.24 s, P = 0.001 [189] and 20.0 s (0–120) vs. 150.0 (0–180), P =☻ 0.001 [190] and 125.33 vs. 148.66 [191]. NIPS scores were also significantly reduced when infants were breast-feed, B 3.0 (0–6) vs. 6.0 (0–7), P = 0.001 [190].

Maternal contact is likely to be the mediating factor why breastfeeding when compared to supplemental breast milk alone is effective. During heel lance, infants being held by mother and breast-fed or being held by mother with pacifier cried significantly less (33 and 45 %) compared to being held by non-mother with pacifier (66 %, P < 0.01 and P = 0.03) [192].

SSC between an infant and mother is also referred to as kangaroo mother care (KMC) due to its similarity to marsupial maternal care [193]. During KMC, a diaper-clad infant is held upright, at an angle of approximately 60°, between the mother’s breasts, providing maximal skin-to-skin contact between baby and parent. Full skin contact and maternal presence have been shown to be beneficial for both term and preterm infants. Advantages for the infant are numerous: stable heart and respiratory rates, balanced thermoregulation, decreased apnea and periodic breathing, improved weight gain, accelerated maturation of the autonomic and circadian systems, and analgesia to painful therapeutic procedures [194–197]. KMC was originally implemented as an alternative to the incubator to maintain preterm infants’ body temperature and increase survival rate in South America where incubators were in short supply [193]. During this time, it was serendipitously noted that infants spent more time in quiet sleep state [197, 199]. Since quiet state is associated with decreased pain response [31, 200], the idea developed to use KMC for procedural pain. In addition, it appeared that holding with skin-to-skin contact provided more comfort than holding with clothed body-to-skin contact [201]. The difference in skin-to-skin contact comfort may be related to inborn tactile receptor response and regulation of opiates, oxytocin, beta endorphins, and vagal tone [202, 203].

Initially studied in full-term neonates, 10–15 min of KMC prior to heel lance reduced crying by 82 %, grimacing by 65 %, and elevation in heart rate (8–10 vs. 36–38) compared to infants who stayed in a cot [204]. Later in the first study to examine the effects of KMC in preterm neonates, pain scores (PIPP) [32, 38], as well as the individual components of decreased facial action, heart rate acceleration, and increased arterial oxygen saturation changes, were reported as lower for the neonates who received KMC compared to those remaining in an incubator during heel lance [196]. Following these studies, numerous trials followed.

In a Cochrane review on skin-to-skin contact for procedural pain in infants, 13 studies that meet the inclusion criteria all show positive results [205]. KMC during heel lance significantly reduced pain scores in full- and in preterm neonates as young as 28 weeks GA [196, 206–210], as well as venipuncture [206] and intramuscular injection [211, 212]. KMC during heel lance has also been associated with a shortened duration of crying [213, 214], more robust heart rate variability [215], and better regulated neurobehavioral response assessed by the Newborn Individualized Developmental Care and Assessment Program (NIDCAP) [216]. Of interest, two of those studies [210, 211] showed that KMC was more effective than sweet taste.