 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
PEDIATRICS Vol. 107 No. 1 January 2001, pp. 105-112
, §,, §,,, §
From the * Department of Pediatrics, University of British
Columbia, Vancouver, British Columbia, Canada; Children's and
Women's Health Centre of British Columbia; and the § Centre for
Community Child Health Research, British Columbia Research Institute
For Children's and Women's Health, Vancouver, British Columbia,
Canada.
 |
ABSTRACT |
|---|
Top
Abstract
Methods
Results
Discussion
Conclusion
References
|
|---|
Background. Management of pain in very low birth weight infants is limited by a lack of empiric knowledge about the multiple determinants of biobehavioral reactivity in infants receiving neonatal intensive care.
Objective. To examine relationship of early neonatal factors and previous medication exposure to subsequent biobehavioral reactivity to acute pain of blood collection.
Design. Prospective cohort study.
Methods. One hundred thirty-six very low birth weight
(
1500 g) infants who underwent heel lance for blood
collection at 32 weeks' postconceptional age formed the study sample,
after excluding those with significant cerebral lesions
(periventricular leukomalacia or cerebral parenchymal infarction
[grade 4 intraventricular hemorrhage]) on cranial ultrasound. Pain
reactions were assessed using the Neonatal Facial Coding System, infant
state, and spectral analysis of change in heart rate variability from
baseline to reaction to invasive stimulation. Factor analysis was used
to provide an empirical basis for deriving summary pain scores, one
factor was primarily behavioral and the other primarily autonomic.
Results. A normal reaction to procedural pain is characterized by facial grimacing and heightened cardiac sympathetic activity. The most significant factors associated with altered behavioral and autonomic pain reactivity at 32 weeks' postconceptional age were a greater number of previous invasive procedures since birth and gestational age (GA) at birth, both of which were related to a dampened response. After controlling for these variables, exogenous steroid exposure made an independent contribution to both the behavioral and autonomic pain scores, also in the direction of dampening the response. Conversely, previous exposure to morphine was associated with "normalized" (ie, increased) rather than diminished responses. In addition, higher mean heart rate at baseline was associated with lower GA at birth and longer time on mechanical ventilation.
Conclusion. Early pain exposure at very low GA may alter the autonomic substrate, resulting in infants who are in a perpetual state of stress. The results of this study suggest that the judicious use of analgesia may ameliorate these effects on later pain reactivity. However, although early morphine exposure may "normalize" subsequent pain reaction, this study did not examine its effects on neurodevelopment. Key words: infant, premature, very low birth weight, pain, morphine, dexamethasone.
Repeated prolonged early pain exposure in premature infants
is a concern because of neurobiologic vulnerability in extrauterine life at early gestational age (GA).1-3 Preterm neonates
seem more sensitive to pain than do more mature infants, because they
show lower tactile thresholds than term infants, with additional
decreases in threshold (windup phenomenon) after exposure to painful
stimuli.4 This altered excitability spreads to multiple
levels of the spinal cord and may cause non-noxious stimuli (handling,
physical examination, and other nursing procedures) to be perceived as
noxious stimuli and stimulate systemic physiologic responses to stress.
This is likely to set up established and long-term responses to tissue injury that outlast the initial noxious stimulus.5
Tactile and nociceptive responses seem to be altered in preterm infants
by immediate and/or prolonged previous experience. In infants of 27 to
36 weeks' GA, Fitzgerald et al6 demonstrated cutaneous
hypersensitivity when flexion reflex was tested in areas of previous
tissue damage. Further, this was reversible with application of topical
analgesia, providing evidence that the pain of the procedure was
causally related to the subsequent sensitization. Such hypersensitivity
is thought to reflect a change in the way afferent nociceptive stimuli
are processed in the spinal cord after early noxious
stimulation.4 In a study examining relationships of
previous invasive procedures to pain reactivity subsequently in the
neonatal period, Johnston and Stevens7 found that
physiologic and facial responses during an invasive procedure at 32 weeks' postconceptional age (PCA) were related to GA at birth and
previous pain experience. Infants handled before blood collection have
shown higher mean heart rate (HR) and facial activity to heel
stick.8 Invasive events in the previous 24 hours were
related to increased facial activity (brow raising) to subsequent
procedures, namely suctioning, chest physical therapy, and diaper
change.9
Concern about potential detrimental effects of pain exposure in the
neonatal intensive care unit (NICU) in recent years has led to
increased use of opioid analgesics, during the acute phase of
mechanical ventilation support of extremely low birth weight infants
and postoperatively. When morphine was administered to infant rats in
the absence of pain, as adults, there was a negative shift in the
dose-response curve.10 However, this did not occur in
adulthood among the rats that received morphine in the presence of pain
as infants, ie, as an analgesic. This suggests that morphine may have a
different effect on the brain in the presence versus absence of
pain.10 Relationships of early analgesia exposure to
subsequent neonatal pain reactivity in human infants have not been
examined to the best of our knowledge.
The aim of this study was to examine relationships between infant
characteristics and experiences in the NICU and subsequent pain
reactivity. The hypotheses were: 1) previous pain alters subsequent
behavioral and cardiac autonomic responses to subsequent pain, and 2)
exposure to opioid analgesia during the NICU stay ameliorates the
effect of previous pain on subsequent pain. To attempt to isolate the
effects of previous pain and opioid exposure in a descriptive study, a
number of other infant characteristics were examined, such as GA at
birth, illness severity, and exposure to other pharmacologic agents
during NICU care. Pain response to an acute invasive event, namely a
heel prick blood collection, was studied at 32 weeks' PCA, because
this is the latest time point before many infants are transferred out
of level III NICU care. To the best of our knowledge, this is the first
study to examine effects of early medications on subsequent acute pain response in the neonatal period.
Study Participants
Written informed consent was obtained from the parent(s) or
other legal guardian according to a protocol approved by the Clinical Research Ethics Committee of the University of British Columbia. Because the aim was to examine multiple determinants of pain responses, sample size was determined based on requirements for multiple regression analyses.11 Allowing for 8 participants
per predictor variable and up to 17 predictors, 135 infants were
required. A continuous series of n = 162 infants with
birth weight TABLE 1
METHODS
Top
Abstract
Methods
Results
Discussion
Conclusion
References
1500 g with no major congenital anomaly were recruited
after admission to the level III NICU in British Columbia's
Children's Hospital and were available for testing at 32 weeks' to 32 weeks 6 days' PCA. Conservative criteria for time since last analgesia
or sedation were used, based on the recent report that the clearance
rate of morphine in premature infants is considerably longer than
previously thought.12 Anticholenergic eye drops were
considered as well, because this agent may affect HR. Infants were
excluded from the final dataset for the following reasons: 1 infant had
questionable GA; 5 had incomplete HR data because of technical
difficulties; 1 did not receive a heel lance at 32 weeks' PCA.
Exclusions because of pharmacologic exposure: 3 received morphine
within 72 hours of the test time; 1 received midazolam within 24 hours
of the test time; 1 received acetominophen within 18 hours of the test
time. An additional 12 infants were excluded because of significant
neurological changes revealed on head ultrasound, ie, periventricular
leukomalacia or cerebral parenchymal infarction (grade 4 intraventricular hemorrhage). Sleep/waking state at baseline was
examined, and 2 infants were excluded who were highly aroused. The
remaining n = 136 infants with complete behavioral and
physiologic data comprised the study sample. Subject characteristics
are provided in Table 1.
Participant Characteristics (n = 136)
Measures
Infant Sleep/Waking State Infant state was coded as defined by Als.13 Sleep/waking state was coded from 1 to 7 as follows: 1 = deep sleep; 2 = light sleep; 3 = drowsy; 4 = quiet awake; 5 = active awake; 6 = highly aroused, agitated, upset, and/or crying; and 7 = prolonged respiratory pause >8 seconds. Preterm infants at times seem to go into a transitory state of collapse or withdrawal, characterized by muscular flacidity and prolonged pause in breathing, which is captured in state 7. Because the lance event of the blood collection procedure is too brief to assess state, the squeeze phase was the period used, corresponding to the time used to measure the heart rate variability (HRV) responses.
Facial Activity The 10 facial actions of the Neonatal Facial Coding System (NFCS)14,15 were coded at bedside16: brow lowering, eyes squeezed shut, deepening of the naso-labial furrow, open lips, vertical mouth stretch, horizontal mouth stretch, taut tongue (cupping of the tongue), chin quiver (high-frequency vibration of the chin), lip purse (tightening the muscles around the lips to form "oo"), and tongue protrusion (tongue pushed forward). Each face action was coded as 1/0 (occurred/did not occur) during each event of blood collection. The frequency distribution of individual NFCS face actions was examined. Frequency of <5% during the heel lance and squeeze phases of blood collection was used as the cutoff for inclusion. Lip purse was excluded, because <2% of infants displayed this face action. Tongue protrusion seems to be a stress indicator in preterm infants, unlike term born infants.16 Tongue protrusion occurred most frequently during the invasive phase (25% of infants during squeeze); thus, it was included in the score. The 9 face actions were summed to provide a total facial activity score for each event (baseline, contact, swab, lance, squeeze, and recovery), with a possible range of 0 to 9.
Physiologic Signal Acquisition
Continuous electrocardiographic (ECG) activity was recorded from a single lead of surface ECG (lead II) and digitally sampled at 360 Hz off-line using a specially adapted computer acquisition system and custom physiologic signal processing software.17 R waves were detected from the sampled ECG and used to form a smoothed instantaneous 4-Hz HR time series as described elsewhere.18 Epochs of HR (2.2 minutes each) were selected from: 1) the resting baseline period within 5 minutes before the Lance, 2) a squeeze period starting within 20 seconds after the lance, and 3) a recovery period after the completion of blood collection. The epoch selection criteria were based on quantitative signal stationarity, the presence of a stable behavioral state, and the absence of gross movement artifact. Power spectral estimates of HR were quantified19 using the area (power) of the spectrum in low frequency (LF) HR variability (.04-.15 Hz), which reflects sources mediated by both sympathetic and parasympathetic influences, and high frequency (HF) variability (.15-.80 Hz), which reflects the influence of respiratory activity (respiratory sinus arrhythmia) mediated by parasympathetic influences alone. As well, the ratio of LF and HF power (LF/HF) was calculated, which reflects sympathovagal balance. To examine pain reactivity, change from baseline to squeeze (difference scores) was used for the HRV variables.
Chart Review of Infant Characteristics and NICU Experience
Prospective medical chart review was conducted by a neonatal research nurse, to acquire information at birth (birth weight, GA, Apgar scores), during the NICU stay for each 24-hour period from birth until the test day (eg, illness severity, number and type of invasive procedures, medications in specific categories, days of mechanical ventilation), and at the time of the observed blood collection (eg, respiratory support, type and time of last feeding). Invasive procedures were defined as those involving tissue damage, including but not limited to: heel lance, venipuncture, insertion of arterial and venous lines, and injections. Intrusive procedures, such as endotracheal tube suctioning, were not included. The procedures were summed to provide a single total, without considering the amount of potential pain from diverse types of experience. Severity of illness was measured using the Score of Neonatal Acute Physiology (SNAP II)20,21 on day 3 since birth, and the Neonatal Medical Index,22 which assesses medical risk, using a discrete number of clinically salient items over a defined period (in this case from birth through 32 weeks' PCA at the time of the observed blood collection).
Medication exposure (analgesics, sedatives, antiinflamatories, steroids, and paralytic agents): only drugs that were given to 20% or more of the infants were included in the data analysis. Midazolam was given to 14% of the infants, chloral hydrate to 1.5%, lorazepam to 2.2%, and diazepam to <1%, so these drugs were not considered further. Morphine was given to 66%, dexamethasone to 50%, indomethacin to 36%, fentanyl to 28%, and pancuronium to 23%. The following measures of medication exposure were retained for data analysis: the number of days on which morphine was administered; average mg/kg of morphine per day; morphine exposure from birth to the test day (average daily mg/kg × number of days on morphine); the number of days on which fentanyl was administered; average µg/kg of fentanyl per day; fentanyl exposure from birth to the test day (average daily µg/kg × number of days on fentanyl); and number of days of dexamethasone, indomethacin, and pancuronium.
Procedures
Infants were recruited in the NICU by a research nurse. Observations of the NFCS and infant sleep/wake state were conducted by a trained coder at bedside during routine blood collection, which was performed by a laboratory technician. Data collection methods using the NFCS at bedside have been described previously.16 The behavioral coder was blind to the purpose of the study and to all information about the infants. The coder was trained on videotapes until interrater reliability reached at least .85, using the conservative formula described previously.14 The infant's NICU nurse applied the foot warmer. The research nurse transferred the electrode connections from the clinical bedside monitor to the Respiratrace-Plus monitor (NonInvasive Monitoring Systems, Miami, FL), which was part of the customized study computer setup for HR data acquisition. HR was recorded continuously during 200 seconds of baseline, throughout blood collection, and for a recovery period of 200 seconds after last contact. Spectral analysis of cardiac activity was conducted for 3 phases: 2.2 minutes of baseline, 2.2 minutes starting during squeeze (after the initial reaction to lance), and 2.2 minutes of recovery. NFCS actions and infant state were rated during the last 60 seconds of baseline, at first contact, at swab, at lance, for squeeze, and for the first 60 seconds of recovery. The median time from first contact to last contact for the blood collection procedure was 238 seconds. If an infant required more than one heel lance, responses were used only from the first lance and squeeze phases.
The target outcome in this study was reaction to the invasive phase of blood collection. Data recording times differed for physiologic and behavioral data because of the nature of measurement of different aspects of infant reactivity. Analysis of spectral recordings of cardiac activity require periods of relative stability to eliminate artifactual effects, over sufficient time for variability to be expressed (2.2 minutes); thus, broad-band periods are optimal. Infant state is a construct for which sufficient time is needed to rate, usually minimum of 60 seconds. In contrast, facial behavior changes rapidly from moment to moment, thus, brief recording times are optimal to reflect changing events; therefore, facial activity was rated during both the lance and squeeze events.
Data Analysis
Changes across events of blood collection were examined using
repeated-measures analysis of variance for continuous measures and
2 between temporally adjacent events for
categorical variables. Statistically significant analysis of variance
at P < .05 was followed by planned Student's
t tests for paired comparisons, to identify differences
between specific events. With Bonferonni's correction,
P < .01 was the level set for significance for each comparison. Then, to reduce the data, summary pain scores were generated, based on weightings derived from factor analysis. To examine
relationships of previous infant characteristics and NICU experience to
heel lance at 32 weeks' PCA, first associations among the set of
predictor variables were checked for multicollinearity, using Pearson
product moment correlations for continuous variables and Spearman 
for ordinal variables, with P < .01 the level set for
significance. Generally, correlations exceeding .80 are considered problematic.11 An exception to the preference for low
multicollinearity is the presence of suppressor variables, which are
predictors that are highly correlated with one or more of the predictor
variables, but have little or no correlation with the
criterion.11 Therefore, bivariate correlations between the
predictors (background variables) and outcome variables (pain scores)
were examined, before multivariate analyses. Finally, the set of
predictor variables was subjected to separate hierarchical linear
regression analyses, using the summary pain scores as outcomes
measures, with P < .05 the level set for significance.
| |
RESULTS |
|---|
|
Top
Abstract
Methods
Results
Discussion
Conclusion
References
|
|---|
Reactions Across the Blood Collection Procedure
Behavior
NFCS
Facial activity changed significantly across phases of blood
collection (F [5, 131] = 146.49; P < .0001). Facial activity (mean ± standard deviation [SD])
decreased from baseline 1.37 ± 1.66 to contact .91 ± 1.39 (t = 2.9; P < .004); increased to swab
1.49 ± 1.98 (t = 4.09; P < .0001); increased to lance 4.42 ± 2.30 (t = 13.07; P < .0001); and increased to squeeze 5.32 ± 2.36 (t = 4.53; P < .0001); and
decreased to recovery 2.20 ± 2.36 (t = 
12.60;
P < .0001).
State
Most infants were in light sleep during baseline and shifted
significantly from baseline to squeeze
(2 [20] = 59.94; P = .0001), with only 2% of infants in light sleep,
37% drowsy, and 55% highly aroused or agitated. The shift from
squeeze to recovery was also significant
(2 [3] = 98.52; P = .0001), with 30% in light sleep and 51% drowsy.
Physiologic
The distribution of change in mean HR in beats per minute (BPM)
from baseline to squeeze was examined before statistical analysis, to
check whether there were infants whose HR dropped significantly, which
would be problematic for examining mean changes. The largest drop in
mean HR was 2.49, which was inconsequential. Therefore, change in mean
HR could meaningfully be examined across phases. Mean HR increased
significantly across the 3 events (F [1, 135] = 117.34;
P = .0001). Mean HR (mean ± SD) increased from
baseline 160.4 ± 14.9, to squeeze 184.4 ± 16.7 (t = 23.56; P = .0001), and decreased
to recovery 171.4 ± 15.3 (t = 13.14;
P = .0001). Because of skewed distributions,
statistical analysis of the HRV variables of LF power, HF power, and
the ratio LF/HF were conducted after log transformation. LF, HF, and
ratio LF/HF changed significantly across the 3 events (F
[2, 270] = 36.49; P = .0001; F [2, 270] = 11.21; P = .0001; F [2, 270] = 20.86;
P = .0001, respectively). LF, HF, and ratio LF/HF,
respectively, decreased significantly from baseline to squeeze
(t = 6.38; P = .0001;
t = 2.91; P = .0001; t = 6.38; P = .0001) and increased during recovery
(t = 7.03; P = .0001; t = 4.32; P = .0001; t = 4.75;
P = .0001). Together these findings indicated increased
sympathetic control of HR and decreased parasympathetic control.
Because most HRV activity was in the LF spectrum, as is usually found
in infancy,23 the change in LF power from baseline to
squeeze was used as the autonomic measure for the remaining analyses.
Composite PAIN Summary Scores
To derive biobehavioral summary scores for each infant for the
invasive phases of blood collection (PAIN), it was necessary to weight
and sum each measure. Because most HRV activity was in the LF spectrum,
the change in LF power (
LF) from baseline to squeeze was used as the
autonomic measure for the composite pain scores. To derive appropriate
weightings for each of the variables, the measures (NFCS lance, NFCS
squeeze, state squeeze, LF power) were subjected to principal
components analysis (PCAnalysis). Two factors with an eigen value >1
emerged, cumulatively accounting for 74% of the variance. The first
factor accounted for 48% of the variance and mainly reflected NFCS
facial activity and state; therefore, it was named PAIN BEHAV. The
second factor accounted for 26% of the variance and mainly reflected
LF; therefore, it was named PAIN HRV. The factor score coefficients
derived from the PCAnalysis were used to weight each variable for each
infant. The sum of the weighted variables resulted in 2 composite PAIN scores for each infant, calculated as follows:
|
|
|
|
Relations Among Predictor and Outcome Measures
Regarding multicollinearity, correlations in the problematic range
were found between GA and days of mechanical ventilation (.86), GA
and number of previous procedures (.82), and days of mechanical
ventilation and global illness severity (.81). However, because the
predictor variables showed only low to moderate correlations with the
outcome variables (Table 2), all the
planned predictors were retained for multivariate analyses.
|
Time since feeding was examined separately, because type of feeding varied (2 infants were not on feeds, 28 were on continuous feeding, and the remainder were on bolus nasogastric tube feedings every 2 hours). For those infants on bolus feeding, time since last feeding was not significantly correlated with any predictor variable or outcome measure and, thus, was not considered further.
Effects of Infant Measures and NICU Experience on Subsequent Pain Reactivity
The set of predictor variables was subjected to separate hierarchical linear regression analyses for each derived PAIN measure (PAIN BEHAV and PAIN HRV). Because the infant characteristics, indicators of illness severity, number of procedures, and surgeries were all significantly correlated with pharmacologic exposure (see Table 3), the predictor variables were entered in 2 blocks. First, (block 1) the set of birth characteristics (birth weight, GA, Apgar at 5 minutes) and factors associated with time in the NICU (illness severity at day 3, overall illness from birth to test day, days of mechanical ventilation, number of pain procedures defined as resulting in tissue damage, number of surgeries) were entered. Then, after controlling for the background variables in block 1, the pharmacologic variables (block 2: days of morphine, total morphine exposure from birth to test day, average mg/kg of morphine per day, days of fentanyl, total fentanyl exposure from birth to test day, average µg/kg of fentanyl per day, days of pancuronium, days of dexamethasone, days of indomethacin) were entered.
|
PAIN BEHAV
The hierarchical regression to predict PAIN BEHAV was significant overall (F [4, 131] = 7.94; P < .0001) and accounted for 20% of the variance. Number of previous pain procedures was the first variable extracted, followed by GA, and early illness severity (SNAP II). After adjusting for the effects of variables related to birth, illness, and previous procedures (block 1), the number of days on dexamethasone made a significant contribution. The partial correlations indicate the independent contribution of each variable, in the final model, and SNAP II illness severity was no longer significant at this point. At the final step, previous procedures, GA, and number of days on dexamethasone each provided significant unique contributions, and all were in the direction of reducing the behavioral pain response. Specifically, higher number of previous pain procedures, GA at birth, and higher number of days on dexamethasone were associated with dampened pain response at 32 weeks (see Table 4). Number of previous pain procedures in relation to facial reactivity (NFCS scores) over the phases of blood collection are displayed graphically in Fig 1. Formation of groups was for visual display only; all data analyses were conducted with previous procedures as a continuous measure.
|
|
PAIN HRV
The hierarchical regression to predict the PAIN HRV score was significant overall (F [5, 130] = 8.23; P < .0001) and accounted for 21% of the variance. Number of previous pain procedures was the first variable extracted, followed by GA, and days of mechanical ventilation. After adjusting for the effects of the variables in block 1, number of days on dexamethasone and total amount of morphine administered during the NICU stay made significant contributions. The partial correlations gave an indication of the independent contribution of each variable in the final model, and number of days of mechanical ventilation was no longer significant. At the final step, previous procedures, GA, number of days of dexamethasone, and total amount of morphine each provided significant unique contributions. Higher number of previous pain procedures and more days of dexamethasone were associated with decreased pain response, whereas previous total amount of morphine received increased the pain response (see Table 4). Thus, previous morphine exposure seemed to "normalize" subsequent pain response. The number of previous pain procedures in relation to LF power reactivity over the phases of blood collection are displayed graphically in Fig 2.
|
| |
DISCUSSION |
|---|
|
Top
Abstract
Methods
Results
Discussion
Conclusion
References
|
|---|
Acute infant pain reactivity typically comprises considerable behavioral and autonomic reactions.7,24 However, there is wide variability in response, even among infants of the same PCA. In the present study, diminished behavioral and autonomic pain responses were primarily a function of high exposure to previous invasive procedures and GA at birth. In addition, higher number of days of exogenous steroids (dexamethasone) was related to reduced reactivity in both the behavioral and autonomic dimensions of pain expression, after adjusting for number of previous pain procedures, GA at birth, and indicators of illness severity (either SNAP II or days of mechanical ventilation). Conversely, higher previous exposure to morphine was related to "normalized" (ie, increased) cardiac autonomic response. The total amount of morphine exposure to test day made a unique contribution to explaining the variance of the primarily autonomic outcome measure, but not to the measure that was primarily behavioral. Therefore, morphine seemed to ameliorate the effects of early pain, but primarily in the autonomic domain. All the infants in this study had not received analgesics for at least 72 hours before testing.
It must be noted that the total amount of variance accounted for was
only 24%; therefore, more work is needed to explore additional influences on acute pain reactions. One important question is how much
pain experience is required to shift the subsequent response. From
exploration of both behavioral and autonomic response as a function of
the amount of previous pain procedures (see Figs 1 and 2), the group
who received 20 procedures stood out as different from the others.
Thus, exposure to 20 invasive procedures (as defined in this study) may
be enough at this stage of development to convert an infant from a
stimulus-naive responder to a stimulus non-naive responder.
The behavioral results of the present study were consistent with the
previous findings of Johnston and Stevens.7 In both
studies facial activity to an acute pain procedure at 32 weeks' PCA
was related to previous pain experienced, although the earlier study
used videotaped recordings of a subset of 3 facial actions from the
NFCS, and the present study used the total NFCS coded in real time at
bedside. Thus, facial activity to pain seems to be a robust measure, at
least at this age. Further, in the present study the infants were born
at 23 to 32 weeks' GA, whereas in the earlier study the infants were
born at 28 weeks' or 32 weeks' GA. Thus, the finding that higher
amount of pain exposure is associated with decreased behavioral
reaction to subsequent pain is now generalizable to infants born 1500
g, across this wide gestational range. The 2 studies differed, however,
in the prediction of cardiac autonomic reactivity during the invasive phase. In the present study, change in cardiac HRV response from baseline to lance was a function of both previous pain procedures and
GA at birth, whereas in the earlier study mean maximum HR during lance
was associated only with GA at birth. This likely reflects
substantially different measures and methods used to measure cardiac
response.
In the present study, consistent with previous findings,7
resting (baseline) HR was higher in infants born at lower GAs.
Furthermore, higher baseline HR was correlated with lower birth weight,
higher number of days on mechanical ventilation, higher illness
severity (SNAP II on day 3 of life), higher overall illness severity,
higher average dose per day (mg/kg) of morphine, and higher number of days of dexamethasone. These findings suggest that multiple previous experiences in the NICU converge to alter the physiologic substrate of
very low birth weight infants, potentially inducing a state of
autonomic arousal suggestive of perpetual stress. This is consistent with the view that preterm infants in the NICU are in a state of
chronic arousal because of the windup phenomenon (related to previous
and ongoing stressors).5 In other studies, we found
this apparent state of heightened background arousal in infants of
birth weight 800 g not at 4 months,25 but at 8 months26 postterm. The behavioral findings confirmed the
interpretation that pain reactivity is altered in infants as a function
of GA and is further affected by previous pain.
The present study addressed only factors associated with altered pain reactivity, for which morphine seemed to be an ameliorating factor. However, this tells us nothing about possible effects of morphine exposure positively or negatively on neurodevelopmental outcomes. Recently differential effects of pharmacologic management of support for mechanical ventilation on neurologic outcomes were reported. Poorer neurologic outcome (neonatal death, intraventricular hemorrhage grade III or IV, or cystic periventricular leukomalacia) was found in infants randomized to receive midazolam (32%), compared with 4% in the morphine group and 24% in the placebo control group.27 In another study, no significant difference was found on psychometric assessment at 5 to 6 years of age, between prematurely born children who were randomly assigned to receive morphine or no morphine to facilitate mechanical ventilation in the NICU.28 However, there was a trend toward better performance in the morphine group in all measures.
Although dexamethasone is beneficial in the short-term for reducing chronic lung disease,29 later negative effects on neurodevelopment in human infants have been reported in 2 randomized, placebo-controlled trials. Among recipients of dexamethasone, there was a higher percentage of major intracranial abnormalities at age 1 year,30 and higher incidence of neuromotor dysfunction among boys and girls, and reduced growth among boys at age 2 years.31 One possible explanation for the effect of dexamethasone on subsequent pain reactivity and other long-term behaviors may be determined by its detrimental effects on the developing hippocampus.32,33
| |
CONCLUSION |
|---|
|
Top
Abstract
Methods
Results
Discussion
Conclusion
References
|
|---|
Multiple determinants of pain expression while undergoing neonatal intensive medical care are starting to emerge. Pain and medication exposure during the early neonatal period in vulnerable immature infants seem to be important factors relating to subsequent reactivity to acute pain and may induce altered arousal to stress in general. Further research is needed to delineate the contributing factors more fully and to identify additional ones, because in this study we accounted for only approximately one quarter of the variance in pain response at 32 weeks' PCA. Currently, substantial variations exist in the routine clinical use of analgesia and sedation in NICU care, and little is known about optimal dosage for opioids in premature neonates. Furthermore, animal research suggests that morphine may have a different effect on the brain in the presence than in the absence of pain.10 Thus, increased knowledge about biobehavioral reactivity of vulnerable infants at very low GA is very important to provide an evidence-based foundation for medical and nursing management of pain early in their NICU stay. To improve pain management in very low birth weight infants receiving neonatal intensive care, a concerted research effort is needed to address multiple questions about mechanisms of altered pain response during early development and about the potential for altered neurobehavioral and neurodevelopmental functioning in the longer term.
| |
ACKNOWLEDGMENTS |
|---|
This project was supported by Grant 96-0023 from the British Columbia Medical Services Foundation.
The participation of the staff and families of the Special Care Nursery at British Columbia's Children's Hospital was very much appreciated.
| |
FOOTNOTES |
|---|
Received for publication Nov 1, 1999; accepted Apr 3, 2000.
Reprint requests to (R.E.G.) Centre for Community Child Health Research, Room L408, 4480 Oak St, Vancouver, British Columbia, V6H 3V4 Canada. E-mail: rgrunau{at}cw.bc.ca
| |
ABBREVIATIONS |
|---|
GA, gestational age;
PCA, postconceptional age;
HR, heart rate;
NICU, neonatal intensive care unit;
HRV, heart rate
variability;
NFCS, Neonatal Facial Coding System;
ECG, electrocardiographic;
LF, low frequency;
HF, high frequency;
SNAP II, Score of Neonatal Acute Physiology;
SD, standard deviation;
BPM, beats
per minute;
LF, change in low frequency.
| |
REFERENCES |
|---|
|
Top
Abstract
Methods
Results
Discussion
Conclusion
References
|
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. Carbajal, A. Rousset, C. Danan, S. Coquery, P. Nolent, S. Ducrocq, C. Saizou, A. Lapillonne, M. Granier, P. Durand, et al. Epidemiology and Treatment of Painful Procedures in Neonates in Intensive Care Units JAMA, July 2, 2008; 300(1): 60 - 70. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. J. S. Anand Pain Assessment in Preterm Neonates Pediatrics, March 1, 2007; 119(3): 605 - 607. [Full Text] [PDF]
|
|
|
|
|
|
American Academy of Pediatrics, Committee on Fetus and Newborn and Section on Surg, Canadian Paediatric Society, and Fetus and Newborn Committee Prevention and Management of Pain in the Neonate: An Update Pediatrics, November 1, 2006; 118(5): 2231 - 2241. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K J S Anand and R W Hall Pharmacological therapy for analgesia and sedation in the newborn. Arch. Dis. Child. Fetal Neonatal Ed., November 1, 2006; 91(6): F448 - F453. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Blew and G. Kenny Attention Deficit Hyperactivity Disorder: the current debate and neglected dimensions. J Child Health Care, September 1, 2006; 10(3): 251 - 263. [Abstract] [PDF]
|
|
|
|
 |
|
 A. E. Thompson, S. A. Nazir, M. J. Abbas, and J. Clarke Switching from immediate- to sustained-release psychostimulants in routine treatment of children with attention-deficit hyperactivity disorder Psychiatr. Bull., July 1, 2006; 30(7): 247 - 250. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K.J.S. Anand and R. W. Hall Morphine, Hypotension, and Intraventricular Hemorrhage Pediatrics, January 1, 2006; 117(1): 250 - 252. [Full Text] [PDF]
|
|
|
|
|
|
S H P Simons, M van Dijk, R A van Lingen, D Roofthooft, F Boomsma, J N van den Anker, and D Tibboel Randomised controlled trial evaluating effects of morphine on plasma adrenaline/noradrenaline concentrations in newborns Arch. Dis. Child. Fetal Neonatal Ed., January 1, 2005; 90(1): F36 - F40. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Holsti, R. E. Grunau, T. F. Oberlander, and M. F. Whitfield Specific Newborn Individualized Developmental Care and Assessment Program Movements Are Associated With Acute Pain in Preterm Infants in the Neonatal Intensive Care Unit Pediatrics, July 1, 2004; 114(1): 65 - 72. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. E. Grunau, J. Weinberg, and M. F. Whitfield Neonatal Procedural Pain and Preterm Infant Cortisol Response to Novelty at 8 Months Pediatrics, July 1, 2004; 114(1): e77 - e84. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. H. P. Simons, M. van Dijk, R. A. van Lingen, D. Roofthooft, H. J. Duivenvoorden, N. Jongeneel, C. Bunkers, E. Smink, K. J. S. Anand, J. N. van den Anker, et al. Routine Morphine Infusion in Preterm Newborns Who Received Ventilatory Support: A Randomized Controlled Trial JAMA, November 12, 2003; 290(18): 2419 - 2427. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. H. P. Simons, M. van Dijk, K. S. Anand, D. Roofthooft, R. A. van Lingen, and D. Tibboel Do We Still Hurt Newborn Babies?: A Prospective Study of Procedural Pain and Analgesia in Neonates Arch Pediatr Adolesc Med, November 1, 2003; 157(11): 1058 - 1064. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. C. Johnston, B. Stevens, J. Pinelli, S. Gibbins, F. Filion, A. Jack, S. Steele, K. Boyer, and A. Veilleux Kangaroo Care Is Effective in Diminishing Pain Response in Preterm Neonates Arch Pediatr Adolesc Med, November 1, 2003; 157(11): 1084 - 1088. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. J. Bouwmeester, J. N. van den Anker, W. C. J. Hop, K. J. S. Anand, and D. Tibboel Age- and therapy-related effects on morphine requirements and plasma concentrations of morphine and its metabolites in postoperative infants Br. J. Anaesth., May 1, 2003; 90(5): 642 - 652. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. W. B. Peters, H. M. Koot, J. B. de Boer, J. Passchier, J. M. Bueno-de-Mesquita, F. H. de Jong, H. J. Duivenvoorden, and D. Tibboel Major Surgery Within the First 3 Months of Life and Subsequent Biobehavioral Pain Responses to Immunization at Later Age: A Case Comparison Study Pediatrics, January 1, 2003; 111(1): 129 - 135. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. C. Johnston, F. Filion, L. Snider, A. Majnemer, C. Limperopoulos, C.-D. Walker, A. Veilleux, E. Pelausa, H. Cake, S. Stone, et al. Routine Sucrose Analgesia During the First Week of Life in Neonates Younger Than 31 Weeks' Postconceptional Age Pediatrics, September 1, 2002; 110(3): 523 - 528. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. F. Oberlander, R. E. Grunau, C. Fitzgerald, and M. F. Whitfield Does Parenchymal Brain Injury Affect Biobehavioral Pain Responses in Very Low Birth Weight Infants at 32 Weeks' Postconceptional Age? Pediatrics, September 1, 2002; 110(3): 570 - 576. [Abstract] [Full Text] [PDF]
|
|
