Effectiveness of oral sucrose and simulated rocking on pain response in preterm neonates

Pain 72 (1997) 193–199

Effectiveness of oral sucrose and simulated rocking on pain response in preterm neonates C. Ce´leste Johnston a,*, Robyn L. Stremler a, Bonnie J. Stevens b, Linda J. Horton c a

School of Nursing, McGill University, 3506 University Street, Montreal, Quebec, H3A 2A7 Canada b Faculty of Nursing, University of Toronto, Toronto, Canada c Royal Victoria Hospital, Montreal, Quebec, Canada Received 10 December 1996; revised version received 27 March 1997; accepted 11 April 1997

Abstract Feeding and carrying have been interventions used by caregivers throughout history in relieving distress in infants. Recent studies on the food substance sucrose have elucidated the comforting effect of the taste component of feeding while studies of rocking have examined the comforting effect of the vestibular component of carrying. The purpose of this study was to determine the effectiveness of sucrose and simulated rocking alone and in combination on diminishing pain response in preterm neonates undergoing routine heelstick procedure in the neonatal intensive care unit (NICU). Preterm infants (n = 85) between 25–34 weeks post-conceptual age (PCA) and 2–10 days postnatal age (PNA) were randomly assigned to one of four conditions for routine heelstick procedure. The conditions were (1) sucrose alone: 0.05 ml of 24% sucrose was placed on the anterior surface of the tongue just prior to the lancing of the heel; (2) simulated rocking alone: 15 min prior to and during the heelstick procedure, the infant was swaddled and put on an oscillating mattress; (3) combination of sucrose and simulated rocking; and (4) placebo: 0.05 ml sterile water administered just prior to heelstick. Physiological (heart rate) and behavioural (facial actions) responses from baseline across 90 s following heelstick were scored second-to-second. Facial actions were analysed with repeated measures MANCOVA and heart rate with repeated measures ANCOVA. Behavioural state and gestational age were covariates. The groups that received sucrose alone or in combination with simulated rocking showed less facial actions indicative of pain than the rocking alone or control group. The addition of rocking to the sucrose condition tended to further blunt the facial expression of pain, but this enhancement did not reach a significant level. Heart rate was not decreased by any intervention compared to the control condition. Although the simulated rocking did promote quiet sleep, which has been reported in earlier studies to blunt pain response, there was no difference between simulated rocking and control groups in either facial expressions indicative of pain or heart rate. The implication of these results is that sucrose, but not simulated rocking may be a means of diminishing pain from minor procedures in preterm infants. Further research is needed on the use of sucrose for more than one procedure as well as examining the contact component of natural rocking, as opposed to simulated rocking.  1997 International Association for the Study of Pain. Published by Elsevier Science B.V. Keywords: Sucrose; Rocking; Preterm neonates

1. Introduction Preterm infants spend the first period of their life in neonatal intensive care units (NICU) during which they undergo many minor painful procedures (Barker and Rutter, 1995; Johnston et al., 1995). While opiate therapy had a

* Corresponding author. Tel.: +1 514 3984157; fax: +1 514 3988455; [email protected]

significant blunting effect on the stress of pain from major surgery (Anand and Hickey, 1987; Anand et al., 1987) there is legitimate concern about using opiates for daily painful procedures such as heelstick. Interventions such as feeding and rocking have been used throughout the history of mankind to soothe infants in distress, but they have only recently been considered for scientific study. There is current interest in using oral sucrose for decreasing procedural pain in infants (Stevens et al., 1997). Rocking by caregivers was shown to decrease time to recover from painful procedures

0304-3959/97/$17.00  1997 International Association for the Study of Pain. Published by Elsevier Science B.V. PII S0304-3959 (97 )0 0033-X

194

C.C. Johnston et al. / Pain 72 (1997) 193–199

in full-term newborns (Campos, 1994). Caregiver rocking in the NICU is not common due to the equipment the infant may need, so that rocker beds or oscillating waterbeds are used to simulate rocking. Simulated rocking has been used with preterm infants to promote state regulation and growth (Gregg et al., 1976; Barnard and Bee, 1983; Clark et al., 1989; Sammon and Darnall, 1994; Groswasser et al., 1995), but not specifically for painful procedures. In the original studies with animals by Blass and colleagues (Blass et al., 1987; Blass and Fitzgerald, 1988; Shide et al., 1989), it was found that several carbohydrates or carbohydrate laden solutions (sucrose, milk, fructose, polycose) reduced pain responses. These analgesic effects were blocked by the administration of the opiate antagonist, naloxone, such that the antinociceptive effect of carbohydrates was thought to be opiate-mediated. In human infants born to methadone-maintained opiate addicted mothers, the comforting effect demonstrated by Blass (Blass and Hoffmeyer, 1991; Blass and Shah, 1995) and Barr (Oberlander et al., 1992; Barr et al., 1994) was not present (Blass and Ciaramitaro, 1994), further supporting the hypothesis that the calming effect is opiate based, since the opiate receptors were presumably not free in the newborns of these mothers. In situations of procedural pain in fullterm infants, sucrose reduced crying in response to heelstick by 50% and, for circumcision, the addition of sucrose to pacifier further decreased crying by 18% above the initial 19% reduction with pacifier in water (Blass and Hoffmeyer, 1991). Haouari et al. (1995) compared the administration of 2 ml of 12.5%, 25%, and 50% sucrose solution 2 min prior to heelstick to full term infants less than 1 week of age (n = 60). Observing the infants 3 min after heelstick, they found that 50% sucrose solution was effective in reducing heart rate and cry duration after 1 min, and 25% sucrose solution was effective by 2 min. Very recently, preterm infants have been the subjects of studies of the analgesic effect of sucrose. Bucher et al. (1995) studied the effect of 50% sucrose (2 ml) in preterm infants and found that it was effective in decreasing heart rate and cry duration, but not cerebral blood flow. Abad et al. (1996) found that 24%, but not 12% sucrose decreased cry duration in preterm infants undergoing venipuncture. All solutions were superior to placebo in decreasing cry duration and pain score. In summary, solutions of sucrose of 24–25% or greater seem to be effective in decreasing pain response to procedures in infants, both fullterm and preterm. Closeness and vestibular motion are the two components of rocking. Korner et al. (1975) determined that vestibular action accounted for the calming effect of rocking. Korner later reported enhanced sleep duration and, more importantly, an increase in the proportion of quiet sleep in preterm infants effected by oscillating waterbeds (Korner, 1988; Korner et al., 1990). Campos (1994) compared rocking by caregiver and pacifier in calming newborns (n = 60) undergoing heelstick and found that both were effective in redu-

cing crying, although the pacifier tended to produce sleep states and rocking, alert states. Barnard and Bee (1983) used horizontal motion of a rocker bed coupled with the sound of heartbeat at fixed intervals during NICU stay for preterms infants less than 34 weeks PCA. This intervention resulted in more quiet state during the hospitalization. No studies were found that reported the use of rocking on pain response in preterm infants. The behavioural state of quiet sleep has been associated with diminished pain response in both full term (Grunau and Craig, 1987) and preterm (Stevens and Johnston, 1994; Stevens et al., 1994) newborns. Quiet sleep can be considered as a cortical mechanism the infant has that can block inputs from the periphery in the same way that distraction can block painful signals from the periphery in older patients. Since rocking was associated in all but one study with increased quiet sleep or state, we hypothesized that rocking should increase quiet sleep state and that this could buffer the perception of pain from heelstick. The purpose of this study was to determine the effectiveness of sucrose alone, simulated rocking alone and a combination of sucrose and simulated rocking on the pain response of preterm infants to routine heelstick.

2. Methods 2.1. Sample Infants were recruited from a university-affiliated level III NICU. Infants who were born between 25 and 34 weeks PCA, who were within 10 days PNA, who had no major congenital anomalies, who had not undergone surgery, who had not suffered IVH greater than grade II as determined by ultrasound, and who were not receiving analgesics or sedatives within the past 48 h were eligible for participation in the study. Since parents visited the NICU at differing and unpredictable times, the explanation of the study and consent form were left at the baby’s cot for the parents to read. Staff nurses asked parents only if they had read the explanation and if they had any questions. Any questions were referred to the research team, but any parent who did not sign the consent was not pursued. 2.2. Procedure and interventions The four group protocols were randomly ordered by computer for 28 infants per group as determined by power estimates of 0.8 from data from earlier studies. The protocols were then put into sequentially numbered envelopes corresponding to subject number. On the morning that the infant was to have blood taken for clinical purposes, the research nurse opened the envelope to determine the infant’s study group. Fifteen minutes prior to the heelstick procedure, she prepared the infant and solutions accordingly.

C.C. Johnston et al. / Pain 72 (1997) 193–199

The four interventions were Sucrose (alone), Rocking (alone), Combined (sucrose and rocking) and Placebo. In the Placebo intervention, the baby was handled and repositioned to be side or supine and 0.05 ml of sterile water was drawn into a 1 ml syringe. At the instant just prior to the actual lancing of the heel, the water was ejected onto the infant’s tongue. In the Sucrose intervention, everything was the same except that 24% sucrose replaced the water in the syringe. In the Rocking condition, 15 min prior to the heelstick procedure, the infant was repositioned to be side or supine and was swaddled with a cotton flannel blanket. An air mattress was placed under the mattress in the cot or isolette. This mattress was attached to an unused respirator which pumped air into the bladder of the air mattress at a rate of 12 cycles/min, a rate chosen from Korner et al. (1975) and Korner (1988) which was shown to increase time in quiet sleep. The mattress was flat and was elevated approximately 2 cm at the peak of each cycle. The baby was left undisturbed, swaddled on the oscillating mattress for 15 min. The infant received 0.05 ml of sterile water prior to heel lance. In the group that received Combined interventions, everything was similar to the Rocking group except that the 24% sucrose was in the syringe. The heelstick procedure conducted by the research nurse consisted of 1 min of baseline, 1 min of warming the heel, the actual heel lance itself, and gentle squeezing of the heel until the required amount of blood was obtained, usually 2– 3 min, but with some procedures taking much longer and with some infants having their heel lanced a second time during the procedure, even within the first 2 min. For consistency of data analysis and before any subsequent heel lancing, the time observed from heelstick was 90 s, broken into three 30 s blocks. 2.3. Measures The primary outcome measures were physiological (heart rate and oxygen saturation) and behavioural (facial actions). The heart rate and transcutaneous oxygen saturation data were obtained from a Nellcor 200 pulse oximeter (Pleasanton, CA, USA) connected to a 386 desktop computer using DataLab software (PBO Developments, Montreal, Canada) for analyzing. These physiological parameters were monitored continuously and sampled every second. Phases were marked during the data collection by a research assistant. At times when the signal was lost for more than 5 s, an artifact marker was put into the computer data file via the software and the research assistant began hand recording the heart rate data from the ECG monitor the baby was on as part of his care in the NICU (various types). Oxygen saturation levels during these times was lost. Some infants were not on another monitor and all physiological data was lost for those infants for varying amounts of time. Average heart rate during each phase was used as the heart rate indicator. Facial actions in infants have been shown to be the most sensitive and specific indicator of pain in both fullterm

195

(Grunau et al., 1990), 2 and 4 month old infants (Johnston and Strada, 1986) and preterm (Johnston et al., 1995) infants. This measure has been frequently used as an outcome in recent intervention studies and has appeared to reflect therapeutic changes (Scott et al., 1994; Taddio et al., 1997). Crying, although an indicator of pain in fullterm or older infants (Johnston and Strada, 1986; Johnston and O’Shaughnessy, 1988; Fuller, 1990, 1991) has not been a useful indicator in very-low-birth weight preterm infants as they are often intubated or cannot spare the energy required to cry (Stevens et al., 1994). Continuous videotape recordings were made with a small Elmco 102 camera focused on the baby’s face. The camera fed into a Samsung VT1261 video recorder. Phases were marked with a tone generator which also fed into the recorder. The upper facial actions were later coded second-to-second in real time in the laboratory according to the neonatal facial coding system (NFCS) (Grunau and Craig, 1987, 1990). In this system, each facial action is coded as present or absent each second, and the proportion of time that the facial action is seen is the score. The videotapes were coded independently by two raters trained on the NFCS. Ten of the sessions were coded by both raters and the agreement ranged from 88% to 95%. Behavioural state of the infant was coded from the videotapes according to Prechtl’s definition (Prechtl, 1974; Prechtl and Beintema, 1997) during the last 30 s of the baseline. The infants were coded as being in quiet sleep, active sleep, quiet awake, or active awake. The research assistant who coded the state was trained to 95% agreement with the trainer (R.S.). All sessions were coded by this research assistant over a 4 day period. The medical chart was reviewed to determine the severity of illness according to the score for neonatal acute physiology – perinatal extension (SNAP-PE) (Richardson et al., 1993) over the past 24 h. This score takes into account birth factors (GA, Apgar score, birth weight) cardiovascular, respiratory, hematological, neurological, renal, gastrointestinal and metabolic variables and is associated with morbidity and mortality. Number of procedures, medications, and feedings were also obtained from the medical chart. 2.4. Analyses Data were entered into SPSS 7.0 (Chicago, IL, USA) and all analyses were conducted on SPSS. Descriptive and associative analyses were first conducted in order to determine appropriate analyses for group differences. Repeated measures multivariate analysis of covariance (RM-MANCOVA) was used for the three facial actions across baseline and three 30 s blocks following heelstick. Repeated measures analysis of covariance (RM-ANCOVA) was used for heart rate from baseline. Gestational age at birth and behavioural state at baseline were covariates for both RMMANCOVA and RM-ANCOVA.

196

C.C. Johnston et al. / Pain 72 (1997) 193–199

3. Results 3.1. Sample characteristics Of the 101 consents left at the bedside, 90 were signed. Random assignment was made to one of four groups. Two equipment failures and three blocked views of faces resulted in a final total sample of 85. The original study design called for 28 infants/group based on anticipated effect size. When every group had at least half the required sample, a preliminary analysis was conducted for a report to the funding agency. The clearly significant results and large effect size (0.97) for some interventions and clearly non-significant effect for other interventions caused us to stop enrolment. The consequence of stopping before all originally planned subjects were assigned, was that the group sizes differed (see Table 1). The sample thus consisted of 85 infants, 42 males and 43 females, who ranged in PCA from 25 weeks 4 days to 34 weeks 2 days, with an average of 31 weeks. Their PNA ranged from 2 to 10 days, with an average of 5 days. Birth weight ranged from 565 to 2810 g, with a mean of 1655 g. Apgar scores at 5 min after birth ranged from 5 to 9, with an average of 8.2. Severity of illness scores ranged from 0 to 15, with a mean of 5.1. The number of invasive procedures ranged from 1 to 35, with an average of 12. There were no significant differences between any combination of groups on any of these variables. 3.2. Relationship among variables The three facial actions were moderately correlated at all points in time (bivariate Pearson’s r’s range from 0.53 to 0.87) and were thus analyzed multivariately (Tabachnick and Fidell, 1989). Heart rate correlations with the facial actions were lower, ranging from 0.11 to 0.23 and heart rate was therefore analyzed separately. Gestational age and birth weight were highly correlated (Pearson’s r’s .0.85) and also correlated with the outcomes (e.g. Pearson’s r between heart rate at stick and gestational age at

birth, 0.93). Gestational age at birth was entered as a covariate. The variable of oxygen saturation was dropped from analysis since as much as half the data were missing after 30 s from heelstick. Within each phase and for each facial action, there were three cases of missing data from different infants. Total sample means were substituted for these missing data. In instances where heart rate data were missing, the case was dropped from the heart rate analysis and this is reflected in the degrees of freedom. The one case that was missing baseline state was dropped from the analyses. 3.3. Interventions The effect of the interventions can be seen in Figs. 1 and 2. Sucrose alone or in combination with simulated rocking was effective in reducing facial expression of pain after the first 30 s, but did not have an effect on heart rate. The difference between sucrose alone and in combination with rocking did not reach a significant level, although the addition of rocking did further decrease the amount of facial expression of pain. The overall multivariate effect of intervention across phases on facial actions was significant (Pillais’ F(6,150) = 2.76537, P , 0.02). Gestational age was a significant covariate (t = 2.88, P , 0.005) and state at baseline approached significance as a covariate (t = −1.71, P = 0.09). Multivariate group effects were significant (Pillais’ F(9,234) = 2.08, P , 0.03). The univariate analyses for each facial action were consistently significant (brow bulge, F(3,78) = 4.56, P , 0.005; eye squeeze, F(3,78) = 4.48, P , 0.01; naso-labial furrow, F(3,78) = 4.38, P , 0.01) and post-hoc contrasts revealed that the significant differences were between the sucrose alone or in combination with rocking versus the rocking alone or placebo after 30 s from heelstick (see Fig. 1). Although heart rate did increase across the phases of the procedure, (F(3,59) = 2.94, P , 0.04), there were no significant differences between any groups at any point in time (F(3,59) = 0.682, P = 0.566), nor was there an inter-

Table 1 Sample characteristics Characteristic

Sucrose alone

Rocking alone

Both sucrose and rocking

Control

n GA birth PNA Apgar @ 5 min SNAP-PE score Weight at birth Sex (% male) Invasive procedures (#) Quiet sleep baseline (%time) Quiet sleep bandaid (%time)

27 31.4 (2.89) 5.6 (2.1) 8.2 (1.7) 7.3 (8.9) 1726 (545) 48 10.7 (5.1) 25 23

24 30.8 (2.7) 5.6 (2.1) 8.3 (.91) 8.6 (9.8) 1562 (526) 50 12.5 (5.7) 42 39

14 32.3 (1.7) 6.0 (2.6) 8.2 (1.1) 8.2 (12.3) 1797 (461) 50 12.9 (6.4) 51 55

20 31.4 (2.9) 5.6 (2.6) 7.9 (1.3) 10.0 (10.1) 1583 (477) 43 12.2 (7.9) 35 30

GA, gestational age at birth in weeks; PNA, post-conceptual age at time of study in days; SNAP-PE score, score of neonatal acute physiology-perinatal extension, reflecting severity of illness.

C.C. Johnston et al. / Pain 72 (1997) 193–199

197

4. Discussion

response and was different between groups and was thus a significant covariate. The rocking conditions increased the amount of time in quiet sleep. This is consistent with some studies on rocking (Barnard and Bee, 1983; Korner et al., 1990), but inconsistent with others, in which rocking promoted awake states (Gregg et al., 1976; Campos, 1994). Of these studies, Korner et al. (1990) and Campos (1994) reported speed of rocking. The speed in Campos’s study was 30 cycles/min, whereas the speed in Korner’s and this study was 12–13 cycles/min. Slower speeds could promote sleep while faster speeds produce wake states, but the effect of speed of rocking on behavioral state has not been explicitly studied. Although infants in the rocking conditions spent more time in quiet sleep prior to the procedure, being in that state was not sufficient to diminish the pain response. Given that the rocking alone was not effective and the sucrose alone was, it was more likely the sucrose in the combined condition that accounted for the effect, although the combination was more effective than sucrose alone. Other studies reported that infants who were in quiet sleep showed less pain response (Grunau and Craig, 1987; Stevens and Johnston, 1994; Stevens et al., 1994). In those reports, the infants were in quiet sleep states naturally, that is, the state was not induced by rocking or other means. The simulated rocking may have artificially induced quiet sleep state but was not an effective enough intervention to maintain the effect during the procedure. In combination with sucrose, rocking not only promotes quiet sleep, but contributes to diminishing pain response. There were several limitations to this study. By stopping the study before completion of final sample, there were unequal cells as well as some unequal distributions of characteristics, such as GA. The RM-ANCOVA is considered robust enough to deal with the unequal cells and the characteristics of GA and behavioural state were accounted for in the analysis. The unequal distribution of GA was by chance. The unequal distribution of behavioural state was consistent with the effect of rocking. Another limitation of

Preterm infants who received 0.05 ml of 24% sucrose showed less behavioral response to heelstick than infants in the control or simulated rocking group. The combination of sucrose and simulated rocking also had a blunting effect on the behavioral response to heelstick. The physiological parameter of heart rate was not affected by either intervention alone or the combination of the two. Other studies of the analgesic effect of sucrose found a difference in heart rate (Bucher et al., 1995; Haouari et al., 1995). In both of these studies, the decrease in heart rate was found with 50% sucrose solution (Bucher et al., 1995) or after 2 min for 25% sucrose (Haouari et al., 1995) and with 2 ml of solution. This current study used 0.05 ml of 24% sucrose and only followed the infants for 90 s, so that, the dose and timing might not have been great enough to show an effect. Behavioral state at baseline was related to behavioral

Fig. 2. Heart rate changes across phases by group. Based on repeated measures analysis of variance. Changes from baseline across phases are significant. There are no significant differences between groups or interaction between group and phase.

Fig. 1. Facial actions across phases by group. Based on repeated measures multivariate analysis of variance. Changes from baseline across other phases are significant over time for all groups, there is a main effect for group, and there is a phase by group interaction effect. Pairwise posthoc analysis (Tukey) show significant differences between sucrose vs. rocking or control after first squeeze phase.

active effect of intervention and phase (F(9,177) = 1.036, P = 0.413) (See Fig. 2). Although gestational age at birth was a significant within-subjects covariate, (t = 6.19, P , 0.001), it was not a significant between subjects covariate (t = 0.631, P = 0.430). State at baseline was not a significant covariate. 3.4. State There were significant between group differences in sleep and wake states both at baseline (F(3,81) = 3.893, P , 0.02) and after the procedure was completed (F(3,78) = 3.586, P , 0.02). In particular, infants in the rocking or combined sucrose and rocking conditions spent 17–31% more time in quiet sleep states than infants in other conditions, who spent more time in active states, either awake or sleep.

198

C.C. Johnston et al. / Pain 72 (1997) 193–199

the study was that the research nurse who actually conducted the heelstick procedure (L.H.) was not naive as to the interventions. Not only was it obvious whether or not the infant was on the rocking bed, the nurse participated in preparing the infants for the conditions. It could be that in some immeasurable way, she influenced the infants’ response. Similarly, in instances where the pulse oximeter signal was lost and heart rate was recorded by hand, the researcher collecting the data (R.S. or C.J.) knew to which group the infant belonged. The research assistants who coded the behavioral data in the laboratory did not know the purpose of the study, the nature of the interventions, nor the infants’ group assignment. Results from this study support the antinociceptive effect of sucrose for procedural pain in preterm neonates. Future studies should examine varying dosages over time. The hypothesis that simulated rocking has a blunting effect on preterm infants’ response to procedural pain was not supported. Although rocking did promote quiet sleep state, it was not effective in decreasing the pain response. Since carrying is a combination of vestibular action and closeness to the mother, future research could be directed at determining if close maternal proximity is effective in decreasing pain response.

Acknowledgements Families who agreed to let their babies participate, as well as the doctors and nurses in the Royal Victoria Hospital NICU are gratefully acknowledged. Research assistants Adam Sherrard, Mina Ladores, and Stefania Forlini are thanked for their diligence and maintenance of excellent standards. Funding for this project was primarily by the Fonds de Recherche en Sante au Quebec (FRSQ940989104) with additional funding from National Institute of Nursing Research of the NIH (NR03916 ROI) and for salary support from Health Canada (C.J.) and the Ontario Ministry of Health (B.S.).

References Abad, F., Diaz, N.M., Domenech, E., Robayana, M. and Rico, J., Oral sweet solution reduces pain-related behavior in preterm infants, Acta Pediatrica, 85 (1996) 854–858. Anand, K.J.S. and Hickey, P.R., Randomized trial of high-dose sufentanil anesthesia in neonates undergoing cardiac surgery: effects on metabolic stress response, Anaesthiology, 67 (1987) 502A. Anand, K.J.S., Sippell, W.G. and Aynsley-Green, A., Randomised trial of fentanyl anaesthesia in preterm babies undergoing surgery: effects on the stress response, Lancet, 31 (1987) 243–247. Barker, D.P. and Rutter, N., Exposure to invasive procedures in neonatal intensive care unit admissions, Arch. Dis. Child. – Fetal Neonatal Edn., 72 (1995) F47–F48. Barnard, K.E. and Bee, H.L., The impact of temporally patterned stimulation on the development of preterm infants, Child Dev., 54 (1983) 1156–1167.

Barr, R.G., Quek, V.S.H., Cousineau, D., Oberlander, T.F., Brian, J.A. and Young, S.N., Effects of intraoral sucrose on crying, mouthing and handmouth contact in newborn and six-week old infants, Dev. Med. Child Neurol., 36 (1994) 608–618. Blass, E.M. and Ciaramitaro, V., A new look at some old mechanisms in human newborns: taste and tactile determinants of state, affect, and action, Monogr. Soc. Res. Child Dev., 59 (1994) 1–80. Blass, E.M. and Hoffmeyer, L.B., Sucrose as an analgesic for newborn infants, Pediatrics, 87 (1991) 215–218. Blass, E.M. and Shah, A., Pain-reducing properties of sucrose in human newborns, Chem. Senses, 20 (1995) 29–35. Blass, E., Fitzgerald, E. and Kehoe, P., Interactions between sucrose, pain and isolation distress, Pharmacol. Biochem. Behav., 26 (1987) 483– 489. Blass, E.M. and Fitzgerald, E., Milk-induced analgesia and comforting in 10-day-old rats: opioid mediation, Pharmacol. Biochem. Behav., 29 (1988) 9–13. Bucher, H.U., Moser, T., Von Siebenthal, K., Keel, M., Wolf, M. and Duc, G., Sucrose reduces pain reaction to heel lancing in preterm infants: a placebo-controlled randomized and masked study, Paediatric Res., 38 (1995) 332–335. Campos, R.G., Rocking and pacifiers: two comforting interventions for heelstick pain, Res. Nurs. Health, 17 (1994) 321–331. Clark, D.L., Cordero, L., Goss, K.C. and Manos, D., Effects of rocking on neuromuscular development in the premature, Biol. Neonate, 56 (1989) 306–314. Fuller, B.F., Potential acoustic measures of infant pain and arousal. In: D.C. Tyler and E.J. Krane (Eds.), Advances in Pain Therapy and Research, Vol. 15, Pediatric Pain, Raven Press, New York, 1990, pp. 137–145. Fuller, B.F., Acoustic discrimination of three types of infant cries, Nurs. Res., 40 (1991) 156–160. Gregg, C.L., Haffner, M.E. and Korner, A.F., The relative efficacy of vestibular-proprioceptive stimulation and the upright position in enhancing visual pursuit in neonates, Child Dev., 47 (1976) 309–314. Groswasser, J., Sottiaux, M., Rebuffat, E., Simon, T., Vandeweyer, M., Kelmanson, I., Blum, D. and Kahn, A., Reduction in obstructive breathing events during body rocking: a controlled polygraphic study in preterm and fullterm infants, Pediatrics, 96 (1995) 64–68. Grunau, R.V.E. and Craig, K.D., Pain expression in neonates: facial action and cry, Pain, 28 (1987) 395–410. Grunau, R.V.E. and Craig, K.D., Facial activity as a measure of neonatal pain expression. In: D.C. Tyler and E.J. Krane (Eds.), Advances in Pain Therapy and Research, Vol. 15, Pediatric Pain, Raven Press, New York, 1990, pp. 147–156. Grunau, R.V.E,. Johnston, C.C. and Craig, K.D., Neonatal facial and cry responses to invasive and non-invasive procedures, Pain, 42 (1990) 295–305. Haouari, N., Wood, C., Griffiths, G. and Levene, M., The analgesic effect of sucrose in full term infants: a randomised controlled trial, BMJ, 310 (1995) 1498–1500. Johnston, C.C., Collinge, J.M., Henderson, S. and Anand, K.J.S., A cross sectional survey of pain and analgesia in Canadian neonatal intensive care units, Clin. J. Pain, under review. Johnston, C.C. and Strada, M.E., Acute pain responses in infants: a multidimensional description, Pain, 24 (1986) 373–382. Johnston, C.C. and O’Shaughnessy, D., Acoustical attributes of pain cries: distinguishing features. In: R. Dubner, G.F. Gebbart and M.R. Bond (Eds.), Advances in Pain Research and Therapy, Vol. 3, Proc. of Vth World Congress on Pain, Elsevier, Amsterdam, 1988, pp. 336– 340. Johnston, C.C., Stevens, B.J., Yang, F. and Horton, L., Differential response to pain by very premature neonates. Pain, 61 (1995) 471– 479. Korner, A.F., Kraemer, H.C., Haffner, M.E. and Cosper, L.M., Effects of waterbed flotation on premature infants: a pilot study, Pediatrics, 56 (1975) 361.

C.C. Johnston et al. / Pain 72 (1997) 193–199 Korner, A.F., Early intervention with preterm infants. In E.D. Hibbs (Ed.), Children and Families: Studies in Prevention and Intervention, International Universities Press, Madison, CT, 1988, pp. 53–62. Korner, A.F., Lane, N.M., Berry, K.L., Rho, J.M. and Brown, Jr., B.W., Sleep enhanced and irritability reduced in preterm infants: differential efficacy of three types of waterbeds, J. Dev. Behav. Pediatr., 11 (1990) 240–246. Oberlander, T.F., Barr, R.G., Young, S.N. and Brian, J.A., Short-term effects of feed composition on sleeping and crying in newborns, Pediatrics, 90 (1992) 733–740. Prechtl, H.F.R., The behavioral states of the newborn infant: a review, Brain Res., 76 (1974) 185–212. Prechtl, H.F.R. and Beintema, D., The neurological examination of the full-term newborn infant, Clin. Dev. Med., 63 (1977). Richardson, D.K., Gray, J.E., McCormick, M.C., Workman, K. and Goldmann, D.A., Score for neonatal acute physiology: a physiologic severity index for neonatal intensive care, Pediatrics, 91 (1993) 617–623. Scott, C.S., Riggs, I.C.W., Grunau, R.V.E. and Craig, K.D., Morphine pharmacodynamics and pain assessment in infants, Proc. of the IIIrd Int. Symp. on Pediatric Pain, 3 (1994) 202.

199

Sammon, M.P. and Darnall, R.A., Entrainment of respiration to rocking in premature infants: coherence analysis, Am. Physiol. Soc., 77 (1994) 1548–1554. Shide, D.J. and Blass, E.M., Opioid-like effects of intraoral infusions of polycose on stress responses in 10-day-old rats, Behav. Neurosci., 103 (1989) 1168–1175. Stevens, B.J. and Johnston, C.C., Physiological responses of premature infants to a painful stimulus, Nurs. Res., 43 (1994) 226–231. Stevens, B.J., Johnston, C.C. and Horton, L., Factors that influence the behavioral pain responses of premature infants, Pain, 59 (1994) 101– 109. Stevens, B.J., Taddio, A. and Ohlsson, A., Efficacy of sucrose for procedural pain in neonates: a systematic review of randomized controlled trials, Acta Paediatr. (1997) in press. Tabachnick, B.G. and Fidell, L.S., Using Multivariate Statistics, 2nd Edn., Harper Collins, New York, 1989. Taddio, A., Stevens, B.J, Craig, K.D., Rastogi, P., BenDavid, S., Shennen, A., Mulligan, P. and Koren, G., Efficacy and safety of lidocaine-prilocaine cream for neonatal circumcision pain management, N. Engl. J. Med., 336 (17) (1997) 1197–1201.