Abstract
OBJECTIVE. The purpose of this work was to evaluate the analgesic properties of oral sucrose during routine immunizations in infants at 2 and 4 months of age.
PATIENTS AND METHODS. A prospective, randomized, placebo-controlled clinical trial was conducted at a pediatric ambulatory care clinic. One-hundred healthy term infants scheduled to receive routine immunizations were recruited, randomly stratified into 2- or 4-month study groups, and further randomly assigned to receive 24% oral sucrose and pacifier or the sterile water control solution. The study preparations were administered 2 minutes before the combined diphtheria-tetanus-acellular pertussis, inactivated polio vaccine, and hepatitis B vaccine. Haemophilus influenzae type b vaccine was administered 3 minutes after the combined injection, followed by the pneumococcal conjugate vaccine, 2 minutes after the H influenzae type b injection. The University of Wisconsin Children's Hospital Pain Scale measured serial acute pain responses for the treatment and control groups at baseline and 2, 5, 7, and 9 minutes after solution administration. Repeated-measures analysis of variance examined between-group differences and within-subject variability of treatment effect on overall pain scores.
RESULTS. Two- and 4-month-old infants receiving oral sucrose (n = 38) displayed reductions in pain scores 2 minutes after solution administration compared with 2- and 4-month-old infants in the placebo group (n = 45). Between-group comparisons for the oral sucrose and placebo groups showed lower pain responses at 5, 7, and 9 minutes after solution administration. The oral sucrose and placebo groups demonstrated their highest mean pain score at 7 minutes, with a mean pain score of 3.8 and 4.8, respectively. At 9 minutes, the placebo group had a mean pain score of 2.91 whereas the mean pain score for the oral sucrose group returned to near baseline, reflecting a 78.5% difference in mean pain score (oral sucrose − placebo) relative to the placebo mean.
CONCLUSIONS. Oral sucrose is an effective, easy-to-administer, short-acting analgesic for use during routine immunizations.
The 2006 immunization schedule requires that infants and toddlers receive as many as 24 injections in the first 2 years of life and as many as 5 injections at a single visit.1 Despite proven benefits of immunization, some parents are reluctant to have their children immunized according to the recommended schedule.2 Reluctance to adhere with the recommended immunization schedule may be partially explained by parental perception that their children endure a substantial amount of pain during routine immunizations: almost twice the amount they hypothesize an adult undergoing a similar injection would experience.3 In a telephone survey conducted on a nationally representative sample of 1600 parents, 25% expressed concerns over the number of immunization injections that their child received during infancy.4 In addition, Woodin et al5 found that when a child received multiple injections at a single visit, the primary concern of both physicians and parents was pain.
Infants possess the anatomic and functional capacity for mounting responses to noxious stimuli before birth.6–10 Although the long-term consequences of pain and stress in human infants are unknown, evidence from animal trials propose that the pain and stress suffered early in life permanently alter the central nervous system.11–14 Long-term effects of unmanaged pain in human infants have been shown to include permanent impairment of elements of cognitive development, including learning, memory, and behavior,15 and increased somatization in childhood.16 The plasticity of the developing brain and the changes that occur in response to painful stimuli17 also contribute to altered perceptions of pain later in life.18,19 Early painful experiences affect children's future response to analgesia. Weisman et al20 found that inadequate analgesia in young children during initial procedures diminished the effects of adequate analgesia during subsequent procedures.
The American Pediatric Society, the Canadian Paediatric Society, and the American Pain Society21,22 recommend the use of sucrose for minor painful procedures in neonates. Sucrose has been examined for its effectiveness in calming distressed term newborns23 and for its analgesic properties in term and preterm infants for pain associated with venipuncture and heel lance.24 Although the analgesic properties of various amounts and concentrations of oral sucrose for procedure-related pain have been tested, the optimal dose of oral sucrose has not been established. Summaries of randomized, controlled trials (RCTs) suggest that a single dose of 0.05 to 2.00 mL of 12% to 50% sucrose delivered by nonnutritive sucking (NNS) via a pacifier for ∼2 minutes before a painful event is safe25 and effective in decreasing physiologic (eg, heart rate, respiratory rate, and oxygen saturation)24,26 and behavioral pain indicators (eg, crying, facial expression, and motor activity),24,27 and in reducing overall composite pain scale scores among hospitalized neonates at 32 to 40 weeks of gestation.24
For healthy infants beyond the neonatal period, immunization is the most common office procedure. Analgesics for this routine intervention must be effective, safe, practical, and easy to use. Currently, pain associated with immunization is managed by acetaminophen,28 ibuprofen,28 or topical anesthetics,29–36 but these treatments may not provide the best relief or are not always practical for use on a routine basis. Few studies have examined the analgesic properties of sucrose beyond the neonatal period. In published studies, variations in the volume and concentration of oral sucrose have limited efforts to establish effective dosing parameters in infants and young children.24 Moreover, the lack of universally accepted criteria for quantifying analgesic responses in infants has led to considerable differences in interpreting findings from studies evaluating the efficacy of oral sucrose. Previous studies have measured proportion, percentage, or duration of crying as indicators of pain.37–41 Research on cry in preterm and term infants has delineated certain properties of cry, such as pitch, intensity, melody, and harmonics, as indicators of pain,42–45 but these characteristics have not been adequately studied in trials of oral sucrose. Although cry duration is of some value in determining stress responses, it alone does not necessarily confirm or refute the presence of pain in infants.43,46,47 Recent evidence suggests that multivariable instruments that include physiologic, behavioral, and contextual indicators yield composite pain scores that are more predictive and valid measures of pain in infants.44,48 To date, few studies have incorporated such measures for evaluating the efficacy of oral sucrose.
We conducted an RCT to evaluate the analgesic properties of oral sucrose during routine immunization in infants at 2 and 4 months of age. Our primary outcome, acute behavioral pain response, was quantified by using a previously validated composite pain scale for preverbal and nonverbal children.49
METHODS
Study Design
A prospective, repeated measures, randomized, placebo-controlled clinical trial examined the analgesic efficacy of oral sucrose during routine immunizations. Institutional review board approval was obtained for the duration of the study (HY03-315).
Sample-Size Calculation
Sample size was calculated on the basis of the numbers of subjects exposed to each treatment condition to yield sufficient statistical power capable of detecting a 20% reduction in the primary outcome, the UWCH pain level scores, between the sucrose intervention groups and placebo control subjects. Institutional review board approval allowed provisions for oversampling (estimated at 20%) to account for participant attrition, disenrollment, or incomplete data sets. The adjusted sample size of 100 infants was equally divided among the 2 treatment and age groups.
Sample and Randomization Assignment
A convenience sample of 100 eligible infants was identified from a consecutive series of patients who visited the university-affiliated ambulatory pediatric clinic for routine immunizations at 2 and 4 months of age. Healthy, developmentally appropriate infants born at 37 to 42 weeks' gestation, with birth weight >2.5 kg, and who presented without evidence of recent illness were recruited for the study. Infants were excluded if they had been fed 30 minutes before immunization, been given acetaminophen the day of the immunizations, had been introduced to solid food, or were not acclimated to a pacifier. Random assignment was determined using a computer-generated random-numbers table. Infants were assigned by progressive random assignment to the stratified 2- or 4-month group and then further randomly assigned into the placebo control (sterile water) or oral sucrose group (Fig 1). Infants participated in the study at 1 point, at either their 2- or 4-month scheduled appointment.
Randomization table.
Study Procedures
The vaccines used in this study included those vaccines used routinely in this practice site: a combined diphtheria, tetanus, acellular pertussis, hepatitis B, and polio vaccine from GlaxoSmithKline (Research Triangle Park, NC), a Haemophilus influenzae type B vaccine from Aventis Pasteur (Swiftwater, PA), and a heptavalent pneumococcal conjugate vaccine from Wyeth (Philadelphia, PA). A 1-inch, 25-gauge disposable needle was used for all of the infants.
All of the study personnel, except for the research assistant who prepared the study solutions, were blinded to the treatment conditions and assignment throughout the duration of the study. Only the statisticians and data monitoring committee had access to the unblinded data, but none had any contact with study participants. The oral sucrose solution consisted of a 24% disaccharide solution manufactured by Children's Medical Ventures (Norwell, MA). A weight-based dose of sucrose (0.6 mL/kg) was calculated using the average birth weight of a term infant (3.4 kg). This dose is consistent with the 2-mL dose of oral sucrose for term infants recommended by the Cochrane Database for Systematic Reviews.24 The placebo control solution was the same weight-based volume of sterile water. Both solutions were clear, nonodorous, similar in physical characteristics (texture), and indistinguishable. NNS was provided in this study with a pacifier introduced into the infant's mouth without the presentation of breast milk or formula. NNS was encouraged by gentle rhythmic stimulation of the pacifier by the parent or clinic nurse.
Infants participating in the study were not swaddled, cuddled, restrained, or constrained during the administration of the oral sucrose or placebo or during the data collection period. Infants sat supported in their parent's lap facing the clinic nurse. Parents were asked to hold the infants' hands midline to their body to prevent injury to the child. A very small percentage of parents did not wish to support their infants during the immunizations and placed their infants in a supine position on the examining table. Infants randomly assigned to both the oral sucrose NNS group and placebo control NNS group received the study preparation via a syringe onto the surface of the tongue followed immediately by the insertion of a pacifier into the mouth. The pacifier was held in place by the parent or the clinic nurse 2 minutes before, during, and 7 minutes after the initial immunization.
Once all of the immunizations were administered and pain assessments recorded, the study was completed, and no additional study care or monitoring was required. A follow-up telephone interview with the parent or primary caregiver was offered to all of the participants within 24 hours of the office visit to address any subsequent questions or concerns regarding the study.
Primary Outcome
The primary study outcome, acute behavioral pain response, was assessed by using the University of Wisconsin Children's Hospital (UWCH) Pain Scale (Fig 2). 49 This instrument consists of 5 measurement domains, cry, facial expression, behavioral, body movement, and sleep, which have been shown to be reliable and valid indicators of responses to pain. Acceptable psychometric properties are reported for the tool, which include internal consistency reliability (Cronbach's α) of .93 (182 observations) and interrater reliability of r = .92 (58 observations). Cronbach's α for all categories excluding sleep is .87 (154 observations). Correlations of .86 for vocal, .81 for facial, .82 for body movement, .78 for behavioral, and .68 for sleep are reported for the scale with Wong-Baker Faces Scale criterion validity (r = .62; n = 68) and construct validity demonstrated by statistically significant reductions in scores after administration of pain medications during procedures. Each category in the tool was scored numerically along a 6-point ordinal scale from 0 denoting the absence of any response to 5 reflecting the highest degree of the response. The mean value for the 5 indicators provides an overall composite score for the level of observed behavioral pain, with higher scores representing greater pain.
UWCH pain scale for preverbal and nonverbal children (sleep excluded).
The principal investigator, who was experienced in the use of the UWCH Pain Scale ratings and blind to the treatment conditions, obtained the 5 behavioral pain response scores for each study participant. Baseline pain levels were assessed within 5 minutes before administration of the oral sucrose or placebo solutions. Two minutes after administering the study solutions, the combined diphtheria-tetanus-acellular pertussis, inactivated polio vaccine, and hepatitis B vaccines were administered, and a second pain assessment was performed. Exactly 3 minutes after the combined injection, the H influenzae type b vaccine was administered, and the third pain assessment was obtained. Two minutes after this injection, the final injection, pneumococcal conjugate vaccine, was administered and a fourth pain assessment recorded. The fifth and final pain observation occurred 2 minutes after the last pneumococcal conjugate vaccine injection. Infants were dropped from the study if crying was evident before the study intervention, because this limited the delivery of the solution. Infants were also not enrolled if parents withdrew their consent at any time during the study.
Statistical Analyses
Statistical analyses of the behavioral pain response data were conducted by using SAS 9.1.3 (SAS Institute, Inc, Cary, NC). Demographic and other baseline variables were compared between treatment groups to assess the similarity of the groups. Reliability and validity measures were examined for the UWCH Pain Scale in the context of the study outcomes, and the distribution of UWCH scores is reported. Repeated-measures analysis of variance was performed to determine the effects of treatment, age group, and time on pain responses. Unless otherwise noted, P values were adjusted using the Tukey-Kramer procedure for multiple comparisons. A significance criterion of .01 was used for all of the statistical tests.
RESULTS
During the study period (March 2004 to April 2005), 100 eligible infants were consented and enrolled in the study. Data from 83 infants (oral sucrose: n = 38; placebo: n = 45) were analyzed. No statistically significant differences were found between groups with respect to gender (P = .71), gestation age (P = .91), birth weight (P = .85), and type of delivery (P = .33; Table 1). Data from 17 infants were not included in the analysis for the following reasons: infant moved before appointment (n = 1), infant was inconsolable after physical examination (n = 7), parent fed infant before study solution administration (n = 1), mother was on maintenance methadone (n = 1), infant exhibited sympathy crying when sibling received immunization (n = 1), infant refused solution (n = 3), infant choked and refused the solution (n = 1, self-resolved within 10 seconds), mother did not want to wake child to receive solution (n = 1), and infant received immunizations before study solution was administered (n = 1).
Study Characteristics
Efficacy of sucrose in mediating pain responses after vaccination was assessed and measured using the UWCH Pain Scale. Within each treatment, contrasts were constructed to test for differences in pain response between the 2- and 4-month age groups at 0, 2, 5, 7, and 9 minutes posttreatment administration. Even without adjusting for multiple comparisons, none of these contrasts were found to be statistically significant. Because these contrasts were not significant, additional contrasts were constructed, combining age groups within treatment, to test for differences in pain response among sucrose and sterile water at 2, 5, 7, and 9 minutes after treatment administration. These results were all statistically significant, with P values, adjusted for multiple comparisons, of <.01.
Mean pain response for sucrose and sterile water, difference in pain response for sucrose and sterile water, and 95% confidence intervals (CIs) and associated P values for baseline and 2, 5, 7, and 9 minutes after treatment administration are shown in Table 2. The mean difference in pain scores between the sucrose and sterile water at baseline was −0.02 (P = .926), at 2 minutes was −1.83 (P < .0001), at 5 minutes was −1.34 (P < .001), at 7 minutes was −1.01 (P < .01), and at 9 minutes was −2.16 (P < .001). At 9 minutes, the group receiving sucrose had mean behavioral pain response scores that were significantly lower (P < .001) and had returned closer to baseline than the control group's scores. The time at which the mean behavioral response exceeded a moderate amount (a score of 2–3 on the UWCH Pain Scale) also differed significantly between the 2 groups. In the group receiving sucrose, the behavioral pain response score exceeded 2 (mean: 2.96) at 5 minutes (after the second of 3 injections). In the group receiving sterile water, the behavioral pain response score (mean: 3.02) exceeded 3 at 2 minutes (after the first of 3 injections; Fig 3).
Behavioral pain response scores over time (mean ± SE) for 2-month-old (A) and 4-month-old (B) infants.
Treatment Intervention Means and 95% CIs
Internal consistency reliability (Cronbach's α) measures were obtained for the UWCH Pain Scale for each of the 5 time points from baseline to 9 minutes. At baseline, Cronbach's α was r = .992 (n = 92) and ranged from .911 to .988 for the other 4 time periods (2–9 minutes). Additional analysis examined the duration of behavioral responses over time (Table 2).
DISCUSSION
Administration of 2 mL of a 24% oral sucrose solution 2 minutes before routine immunizations is effective in decreasing maximum immunization pain and shortens the time before returning to a near normal state in infants at 2 and 4 months of age. The heightened behavioral pain responses observed in infants receiving sterile water reflect greater pain intensity compared with the infants who received sucrose. Both groups reached a maximum pain response immediately after the last immunization (at 7 minutes), of 3.8 and 4.8, respectively. The group receiving sucrose returned to near normal 2 minutes later at 0.59, whereas the placebo group remained at 2.91 (after 9 minutes), reflecting a 78.5% difference in mean pain score relative to sterile-water mean pain score. Effect size, measured as the standardized difference between 2 means,50 for each treatment point ranged from −0.025 to −2.191 for the 5 time periods from baseline to 9 minutes (Table 2). The efficacy of this intervention persisted for ≥9 minutes after administration.
Number-needed-to-treat (NNT) analysis provides additional support for the use of oral sucrose before immunization in infants. Within the concept of NNT analysis, values of 4 and 2 are favorable values for determining the effectiveness of this intervention. The NNT with 0.6 mL/kg of 24% sucrose to achieve a behavioral pain score of 0 or 1 at 2 minutes after sucrose administration is 4 (NNT = 1/[0.605 − 0.311] = 3.40). The NNT to observe a behavioral pain score of 0 or 1 at 9 minutes after the administration of sucrose is 2 (NNT = 1/[0.842 − 0.333] = 1.96). Thus, pediatric health care providers would need to treat only a small number of infants to document the efficacy of oral sucrose in reducing pain associated with immunization.
Our sample size of 17 to 25 in each group is within the range of other RCTs that also detected favorable treatment differences between sucrose and placebo using similar outcomes. Haouari et al51 showed a 50% reduction in crying time with heel sticks using 25% sucrose (volume 2 mL) with 15 subjects in each of 3 treatment groups exposed to graded concentrations of sucrose and a placebo group. Blass et al52 had fewer subjects, 10 in each group, testing the efficacy of sucrose by different methods of delivery, syringe versus pacifier. Reis et al41 enrolled 15 to 16 subjects in multiple groups comparing the effects of sucrose with immunization.
The uniform dosing used for all of the study infants is a strength of this study. In previous research, the volume and concentration of sucrose varied from study to study. A weight/volume dose of study solution was administered to all of the infants participating in the study, and this dose of oral sucrose significantly reduced the behavioral pain responses of both 2- and 4-month-old infants.
A composite pain scale is considered the most valid and predictive measure of pain in infants.44,48 Although previous studies37–41 measured proportion, percentage, or duration of cry as indicators of pain, our study used a multivariate pain scale (UWCH) that includes physiologic, behavioral, and contextual indices to assess behavioral pain response. High internal consistency reliability was demonstrated for the UWCH Pain Scale using data from 92 participants with a total of 460 pain rating observations, a much larger number compared with the 74 subjects and 182 observations reported in the original work on this pain measure.49
Parents have shown a strong desire to want to hold infants at the time of administration of injections.41 Although there are limited data regarding factors affecting infant responses to pain during immunization, 1 RCT demonstrated no differences in crying duration, facial expressions, and pain levels whether infants were positioned supine or held by a parent.53 The study design controlled for nonnociceptive fiber stimulation by requesting that parents support their infants in their laps facing the clinic nurse or lay their child supine on the examining table. Parents were discouraged from swaddling, cuddling, or restraining their infants during data collection, because these activities could activate nerve fibers that carry nonnociceptive information and decrease or inhibit nociceptive transmission. Although gender and age have not been found to be significant predictors of pain response, parents have exhibited more coping-promoting gestures, such as talking with female infants.54
Although necessary to quantify the duration of sucrose analgesia, the practice of extending routine immunizations over several minutes is not common place and limits the ability to generalize our findings. Many ambulatory pediatric clinics administer multiple immunizations serially without significant time in between vaccinations or provide them simultaneously during a single visit. In all 4 of the treatment groups, the behavioral pain response scores increased with each immunization. Additional research should determine whether the concentration and weight/volume dose of sucrose administered independent of other interventions are adequate to balance the pain intensity of 3 serial or multiple simultaneous immunizations.
CONCLUSIONS
Efforts to minimize pain associated with immunizations in the first few months of life have implications for promoting parental adherence to immunization schedules and thereby preventing a resurgence of vaccine-preventable diseases. Recent data suggest that exposure to repeated or severe pain in early life can subsequently affect the development of the central nervous system11–14 and may ultimately have long-lasting effects on neurologic function and behaviors.18 This challenges clinicians to identify strategies to reduce pain associated with this important and common procedure.
The ideal preprocedural analgesic for minor invasive procedures would be a cost-effective, inexpensive, short-acting agent with few associated risks.55 The analgesic should be easy to implement and require no additional training for the provider or additional monitoring of the infant. This study observed that a weight-based volume dose calculation of 0.6 mL/kg of a 24 g/1000 mL (24%) sucrose solution administered as a preprocedural analgesic before routine immunizations significantly reduced pain scores after each of the 3 immunizations. Although sucrose did not eliminate pain at any point in time, other pain reduction or comforting measures (acetaminophen, distraction, holding, feeding, etc) used in conjunction with sucrose administration could provide additional comfort for infants.21
Sucrose is inexpensive, short acting, nonsedating, easily administered, noninvasive, and commercially available. Administration of sucrose does not require additional training and does not expose the infant to risks greater than those associated with breast or bottle feeding. The rapid onset and the absence of long-term effects of the analgesia facilitate its use for pain prevention during common procedures in ambulatory practice sites and in hospital settings.
Acknowledgments
This work was supported by American Nurses Foundation grant 2003051. Dr Hatfield is an Aventis Pasteur American Nurses Foundation Scholar and recipient of the 2003 American Nurses Foundation grant.
We thank David Mauger, PhD (Public Health Sciences, Penn State University College of Medicine), for assistance in preparing the randomization table and Denise Drago, MD (Pennsylvania State University College of Medicine), for assistance in recruitment and data collection. We also acknowledge the contributions of the physicians, nurses, and medical charts assistants at University Pediatric Associates at Pennsylvania State Children's Hospital and the parents who gave their consent for this study.
Footnotes
- Accepted July 19, 2007.
- Address correspondence to Linda A. Hatfield, PhD, CNNP, Pennsylvania State University School of Nursing, College of Health and Human Development, 307B HHD East, University Park, PA 16802. E-mail: lal153{at}psu.edu
This trial has been registered at www.clinicaltrials.gov (identifier NCT00150189).
The authors have indicated they have no financial relationships relevant to this article to disclose.
REFERENCES
- Copyright © 2008 by the American Academy of Pediatrics