Trial of Repeated Analgesia with Kangaroo Mother Care (TRAKC Trial)

Pharmacologic agents that have been examined for reducing procedural pain in neonates include topical anesthetics and systemic drugs, such as opioids or acetaminophen. Topical anesthetics, while effective for pain management during circumcision [16‐20], have been shown to be ineffective for heel lance, venipuncture, and insertion of intravenous lines [21‐25]. Systemic drugs, specifically opiates, are highly sensitive to the developmental stage of the infant [26, 27]. Therefore, rapid changes in requirements make effective and safe dosing a challenge. In addition, opiates have significantly slower clearance in neonates [28‐32], and have not necessarily been demonstrated as effective for managing procedural pain [33]. Trials of acetaminophen for procedural pain control have demonstrated that it is not effective, and may even be no better than placebo for heel lance procedure [34, 35]. Thus, pharmacological management for common, repeated, painful procedures in preterm neonates is not an option. Given the frequency of painful procedures in NICUs, the short and long term negative effects of repeated pain exposure in this population, and the ethical imperative to manage this pain, other approaches are required.

Non-pharmacological approaches to pain management are thought to be based on the release of endogenous opiates [36, 37]. Although there are scant data in neonates regarding endogenous descending inhibitory mechanisms, the engagement of mechanisms that release endorphins is well established in adults [36, 37]. Animal studies suggest that the endogenous system is not well developed prior to 32-weeks post-conception, however, there is speculation that it may be developed enough to provide some level of comfort [38, 39]. Various non-pharmacological strategies for pain control have been tested in neonates and show varying degree of efficacy. These strategies can be categorized into sensory stimulation [40, 41] (i.e., positioning/swaddling, vestibular action/rocking, non-nutritive sucking, music), nutritive [42‐63] (i.e., breastfeeding, oral sweet solutions), and maternal interventions (i.e., maternal odor and voice, breastfeeding, KMC). For the purposes of this trial, KMC and 24% oral sucrose (considered standard care) are the interventions of interest.

Research in animals and human infants have shown that intra-oral sucrose solutions have an analgesic effect. Studies using sucrose for pain relief started in the late 1980’s [50‐52] and have since included both term and preterm infants. Sucrose, the oral solution used most frequently, has been examined in two systematic reviews [50, 62], and one meta-analysis [4]. These reviews are in agreement regarding the positive efficacy of oral sucrose for reducing procedural pain in preterm neonates.

Infant response to interventions has been assessed using both behavioral (e.g., cry and facial action) and physiological indicators, as well as composite measures of pain. A recent study demonstrated that although sucrose decreased scores on the Premature Infant Pain Profile (PIPP), there was still nociceptive activity in the somatosensory cortex [63]. Given the dissociation between the affective components of pain [64‐66], that is how much it hurts (somatosensory cortex) vs. how aversive it is (frontal and pre-frontal cortex), it is likely that it is the affective component of the pain response that is being modified by sucrose. The authors of the study thus raised concerns about the analgesic properties of sucrose, and whether it can ameliorate the adverse effects of repeated pain exposure. Such findings reinforce the need for continued study of other non-pharmacological interventions such as KMC alone or in conjunction with sucrose.

The effect of sucrose on pain intensity on consecutive days has been addressed, but needs further investigation. While one study examining the use of sucrose for all painful procedures in the first week of life in the NICU reported poorer neurodevelopmental outcome scores with higher doses of sucrose [6], a secondary analysis found that this was only the case in infants who received more than 10 doses over 24 hours [67]. Concerns about the long-term use of sucrose on neurodevelopmental outcomes have also been raised by Holsti and Grunau [5]. They provide a strong and logical argument that the development of the infant’s dopaminergic system is altered with sucrose administration. Specifically, that sugar stimulates dopamine release, and that this release responds or is a reaction to the concentration of sucrose rather than the volume. In addition to this argument, there is also a relationship between dopamine and attention and motor development – the specific neurodevelopmental outcomes used in the study examining repeated sucrose administration [6]. Holsti and Grunau suggest that dopamine regulation is disturbed by receiving repeated doses of 24% sucrose (the recommended concentration for procedural pain management in neonates), and that this interferes with attention and motor development.

Kangaroo Mother Care (KMC) was implemented in the modern era as an alternative to incubator care to maintain preterm infants’ body temperature and increase survival rates in South America due to short supply of incubators [68]. During this time, it was serendipitously noted that infants in KMC spent more time in quiet sleep state [69, 70]. As quiet sleep state is associated with decreased pain response [71, 72], the idea developed to use KMC to control procedural pain. When initially studied in full term neonates, decreased crying and heart rate acceleration was observed [73]. Subsequent studies in preterm neonates showed decrease in facial action, decrease in heart rate acceleration, and increased oxygen saturation, with a decrease in overall pain scores [74].

In a recent Cochrane review of the literature on skin-to-skin contact for procedural pain in infants conducted by the authors of this protocol [3], 13 studies that met the inclusion criteria showed positive results. Interestingly, two of the studies [75, 76] showed that KMC was more efficacious than sweet taste in reducing infant’s procedural pain. However, to our knowledge, there have been no published studies to date examining the combined efficacy of KMC and sucrose in managing procedural pain in preterm infants.

Mechanism underlying KMC as a comfort strategy There are likely several mechanisms underlying the specific pain-relieving effect of KMC. One hypothesis that could be related to all non-pharmacological strategies is derived from the Gate Control Theory of Melzack and Wall [36]. According to this theory, stimuli travelling ascending pathways may inhibit the nociceptive signals from painful stimulus through various endogenous mechanisms located along the spino-thalamic tract [77]. The stronger these competing stimuli are, including multiple modalities, the more effective they are at blocking the perception of pain. Such theories help explain why multiple modalities, such as KMC, breastfeeding, or sensorial saturation that involves tactile, auditory, and olfactory mechanisms are more effective than single modalities.

The mechanism of the comforting effect of breastfeeding and KMC is likely related to the release of oxytocin [78, 79]. Oxytocin has been referred to as the love hormone due to its role in affiliative behaviors [80‐83]. In an animal model of pain and oxytocin, massage-like stroking induced acute antinociception that was reversed by an oxytocin antagonist [84]. Also, when oxytocin was injected into the periaqeductal gray region, it had an antinociceptive effect.

In summary, 1) procedural pain in the NICU is infrequently and poorly managed, 2) pharmacological interventions present problems with this population, 3) among non-pharmacological interventions both sucrose and KMC are supported by numerous studies as one-time interventions in diminishing pain response, 4) there are no studies of repeated use of KMC and only two studies that have followed infants for repeated use of sucrose over time have had conclusive results, 5) no published studies to date have examined the combination of KMC and sucrose across time.

Given the findings of research to date, this study proposes that KMC alone or in combination with sucrose will continue to reduce pain across the NICU stay more than sucrose alone. This is based on the belief that the touch and olfactory senses invoked from the mother are fundamental to the maternal role of affiliation and comfort through underlying mechanisms of hormone release. It is further proposed that neurodevelopment will be better in infants who have received KMC for minor painful procedures due to the release of norepinephrine during KMC.

Group assignment will be determined using a password protected website, where randomization is in permuted blocks and participants will be stratified by gestational age (less than or greater than/equal to 32 weeks gestational age at birth) in order to ensure equal distribution across groups for all ages.

Group assignment will be acquired by the research nurse from the secure website. The REB approved research nurse is the only individual in this study who has access to the secure site after consent is obtained.

If the infant is randomized to the KMC condition, the research nurse will begin by recording a one-minute sample of the infant in the incubator. The research nurse will then place the mother and infant into the KMC condition, where the diaper-clad infant is help upright, at an angle of approximately 60°, between the mothers breasts, providing maximal skin-to-skin contact between mother and baby. The infant will remain in KMC for at least 15 minutes prior to the painful procedure. Two minutes prior to the procedure the infant will receive ¼ of the recommended (based on NICU protocol) volume of sterile water by syringe onto the tongue. The remainder of the total recommended dose will be given as needed in small increments during the procedure.

If the assigned condition is sucrose alone, half an hour before the procedure the nurse will place the infant in the supine position in an incubator or infant cot and set up the monitoring equipment. All monitoring will take place while the infant is in the incubator or cot. Two minutes prior to the procedure, the baby will receive ¼ of the recommended volume (based on NICU protocol) of 24% oral sucrose solution by syringe onto the tongue. The remainder of the total recommended dose may be given as needed in small increments during the procedure.

For combined sucrose and KMC, the infant will be placed in KMC as described for the KMC alone condition, however, will receive 24% oral sucrose solution as per the NICU protocol described above.

Although most painful procedures will not be video-recorded for the study analysis, the expectation is that the staff caring for infants enrolled in the study will follow the study intervention protocol for all painful procedures, and that any deviations from the study protocol will be charted and included in the analyses. Babies in the KMC group will have reminder signs at the bedside prompting the staff to place them in KMC for procedures, and the study syringes of either 24% sucrose solution or sterile water will be available to staff at the bedside throughout the infant’s stay in the NICU. Staff documentation of all painful procedures will be monitored to ensure compliance with the study protocol.

In order to ensure equipoise, staff may deviate from the protocol to give off-study doses of rescue sucrose if the infant has a PIPP score > 6 (considered to indicate pain) at their discretion. The deviations from protocol will be recorded on a procedure record kept at each infant’s bedside. If the mother is unavailable for a painful procedure, the infant will be swaddled and given the study solution. If the infant has a PIPP score > 6, off-study sucrose may be given by the nurse caring for the patient. All off-study sucrose doses will be recorded in the procedure record and the research nurse will confirm the doses given from the medication administration record.

The primary outcome in this trial will be infant pain intensity measured by the Premature Infant Pain Profile (PIPP) [72, 88]. The PIPP uses seven indicators to calculate a composite pain measure: three behavioral indicators (facial actions of brow bulge, eye squeeze, and nasolabial furrow), 2 physiological indicators (heart rate and oxygen saturation), and 2 contextual indicators (gestational age and behavioral state) [72]. The contextual factors included in the PIPP are particularly important for this study as infants are being followed over time and therefore age changes will be accounted for. In addition, KMC is known to promote quiet sleep state, and this will be accounted for when considering behavioral state as an indicator in the PIPP scoring. The PIPP was developed 13 years ago and has steadily accumulated evidence of reliability and validity [72, 88]. It had been used in over 40 studies to date [88].

This study will utilize a Somte (Compumedics, Melbourne, Australia) system that samples at a rate of 256 Hz to measure physiological data. Heart rate and oxygen saturation are recorded to the Somte system using a transcutaneous probe, and Electrocardiogram (ECG) is recorded from 3 leads. Epochs will be noted with an event timer in the Somte system. Close up video recordings of the infants face will be made using a Sony DCR-SR82 HDD digital video camera in the first baseline in the incubator (Baseline 1), one minute prior to start of procedure (Baseline 2), and during and after the procedure. The neurobehavioral state component of the PIPP score is determined according to Prechtl’s categories of quiet sleep, quiet awake, active sleep, or active awake [89], during the baseline [90]. Gestational age will be taken from the infant chart, and will be based on ultrasound at 16 weeks.

All data will be analyzed in the research laboratory outside the NICU. Physiologic data will be analyzed using Compumedics E-series Profusion PSG II software. Faces will be coded second-to-second on a stop frame system by trained coders naïve to the purpose of the study. Coders will be trained up to 90% agreement on video recordings from previous studies. Every three months, coders will recode randomly selected sessions (one from their earlier coding and one from each of the other coders) in order to determine both inter and intra-rater reliability.

The secondary outcome will be the Neurobehavioral Assessment of the Preterm Infant (NAPI), developed by Korner and colleagues [91‐95]. In particular, the subscales of alertness and orientation and motor development and vigor will be assessed. The NAPI evaluates the relative maturity of functioning of preterm infants, with higher scores reflecting higher maturity, and can differentiate 2 weeks’ post-conceptual age. Assessments at 32 and 36 weeks corrected gestational age and at term equivalence will be preformed on all infants in accordance with their gestational age at birth, as well as at the time of discharge from the NICU. Much of the NAPI examination consists of observational items, and the remainder rates the infant’s response to stimuli. The assessment takes approximately 30 minutes to administer and includes clusters of single-item neurobehavioral dimensions. Test – re-test reliability ranging over 2 consecutive days ranged from 0.59 to 0.90. Original inter-observer reliability ranged from 0.64 to 0.93 [91]. Clinical and construct validity and sensitivity of the NAPI have been established [94, 96].