- Email: [email protected]
Bedside Assessment of Heel Lance Pain in the Hospitalized Infant
PDFlib PLOP: PDF Linearization, Optimization, Protection Page inserted by evaluation version www.pdflib.com – [email protected]
CLINICAL STUDIES
Bedside Assessment of Heel Lance Pain in the Hospitalized Infant Denise Harrison, RN, RM, MN, Cheryl Evans, RN, RM, BN, Linda Johnston, RN, PHD, Peter Loughnan, MB, BS, FRACP
Objective: To evaluate a method of pain assessment to be used for hospitalized infants requiring blood test by a heel lance procedure. Design: Observational study evaluating pain measurement and interrater reliability of pain measurement in hospitalized infants. Setting: A Level III neonatal unit and a cardiac surgical unit at a major pediatric teaching hospital. Participants: 20 infants whose gestational age ranged from 28 weeks to full-term. Main Outcome Measures: Observations included behavioral measurements (facial expressions, body movements, and crying characteristics) and the physiologic measure of heart rate. Interrater reliability and the feasibility of using the procedural pain assessment method at the bedside were considered. Results: The three behavioral measurements and heart rate were responsive to the heel lance. Interrater reliability was high for facial expressions and crying scores but was low for body movements. Conclusions: Modifications have been made to the method of procedural pain assessment to be used in a subsequent study. The modified method is expected to be a reliable measure of procedural pain caused by a heel lance and can easily be used at the bedside during the course of further research. JOGNN, 31, 551–557; 2002. DOI: 10.1177/088421702237738 Keywords: Neonatal pain—Pain assessment— Pain scoring—Procedural pain Accepted: October 2001 Procedures that cause pain in infants are common occurrences in neonatal nurseries. Recommendations are being made in the current literature to utiSeptember/October 2002
lize interventions to reduce procedural pain. To adequately assess the effectiveness of pain reduction interventions, nurses caring for infants must be able to carry out valid and reliable pain assessments. However, because there is no gold standard of pain measurement in infants, there is no single parameter by which to accurately assess pain. Thus, pain assessment in this age group is a difficult task. This difficulty is a salient reason for the ongoing discussion in the literature concerning the most appropriate and reliable method of assessing pain.
Objective The aim of the current study was to test a method of pain assessment that could subsequently be used to measure the effects of a pain reduction strategy for infants undergoing a heel lance procedure. For the purposes of this study, the pain assessment method had to be applicable to hospitalized infants and suitable for bedside application. The method also needed to be appropriate for the assessment of the short-term acute procedural pain of a heel lance, versus pain associated with surgery or chronic pain.
Background Assessment of the complex, subjective nature of pain in the infant must take into account many factors. Pain responses demonstrated by well infants often have both dramatic observable behavioral and physiologic components, but these responses also can occur outside of observable painful situations (Johnston, Stevens, et al., 1999). The responses to pain vary widely among infants, especially in those who are sick or premature and whose behavioral
JOGNN 551
responses may be less robust or even absent, yet their physiologic responses may still be dramatic (Johnston, Stevens, et al., 1999). Infants’ behavioral responses to pain include facial expressions, body movements, and crying characteristics. Infants’ facial expressions relating to pain have been studied in detail (Grunau & Craig, 1987) and have been cited as the most specific indicator of pain resulting from noxious stimuli (Stevens, Johnston, Petryshen, & Taddio, 1996). The Neonatal Facial Coding System (NFCS), comprising nine facial expressions, was described by Grunau and Craig (1987). The NFCS has been shown to detect changes in facial expression in response to a heel lance procedure in infants of all gestational ages, even very premature infants (Grunau, Oberlander, Holsti, & Whitfield, 1998; Johnston, Stevens, Yang, & Horton, 1995), although with less frequency and vigor than in more mature infants (Craig, Whitfield, Grunau, Linton, & Hadjistavropoulos, 1993; Johnston, Stevens, et al., 1999). A limited subset of four of the original nine facial expressions from the NFCS has been reported to occur most commonly as a result of a heel lance procedure. Brow bulge, eye squeeze, nasolabial furrow, and open lips were
B
edside application of a subset of the Neonatal Facial Coding System was shown to be both reliable and clinically feasible.
observed in 99% of infants within 6 seconds of the heel lance (Grunau & Craig, 1987). The authors of the NFCS also extensively studied infants’ body movements occurring as a result of a heel lance. The Infant Body Coding System (IBCS) was developed as a result of videotaping 56 infants during routine heel lances (Craig et al., 1993). Coders viewing the videos scored the presence of motor activities (hand, foot, arm, leg, head, and torso movements) at intervals throughout the procedure. This study showed that the heel lance instigated the most vigorous bodily activity, although there also was a substantial response to the heel swab alone. Body movements were therefore believed to be less specific than facial expressions to the painful event. Infants’ crying characteristics in response to procedural pain also have been studied by several authors. The crying characteristics studied have included both acoustic and temporal measurements. Acoustic studies have analyzed the pitch, fundamental frequency, and amplitude of infants’ cries (Fuller, 1999). Temporal characteristics have been categorized into cry latency, duration of first cry (Ramenghi, Wood, Griffith, & Levene, 1996), number of 552 JOGNN
cry cycles, plus duration of crying times throughout and following painful procedures (Fuller, 1999; Grunau & Craig, 1987; Hadjistavropoulos, Craig, Grunau, & Johnston, 1994; Johnston, Sherrard, et al., 1999). Infants’ cry characteristics can provide important information and can be an obvious and useful measure in studies evaluating different pain reduction strategies. However, crying is not an “accurate” pain measurement on its own and is not unique to pain (Carbajal, Chauvet, Couderc, & Olivier-Martin, 1999). Crying characteristics also may be altered in a sick infant, may not be assessable in an infant with an endotracheal tube in situ, and may vary widely from infant to infant at any given time. There may be a low incidence of crying as a response to pain in premature infants (Johnston, Sherrad, et al., 1999), as shown by Stevens’s study, in which only 47% of premature infants cried during the painful (lance and heel squeeze) stages of a heel lance (Stevens et al., 1996). Physiologic parameters also can be used to measure change after a painful procedure, yet on their own they are not specific measurements of pain. The most common physiologic parameters used to assess infants’ pain include heart rate, respiratory rate, and oxygen saturation. Other physiologic and biochemical parameters sometimes used to assess pain are changes in intracranial pressure, palmar sweating, skin blood flow, blood pressure, glucose levels, blood and/or saliva stress hormone levels, catecholamines, ketones, and urine nitrogenous compounds (Anand & McGrath, 1993; Anand, Sippell, & AynsleyGreen, 1987; Chang, Anderson, & Wood, 1995; Franck & Miaskowski, 1997). In one study, cerebral blood flow, cerebral oxyhemoglobin, and deoxyhemoglobin were used as indicators of pain (Bucher et al., 1995). Although not a response to pain as such, the infant’s sleep or behavioral state before procedures are performed has been shown to influence subsequent behavioral response to pain and therefore is an important factor to note when procedural pain is monitored (Ballantyne, Stevens, McAllister, Dionne, & Jack, 1999; Grunau & Craig, 1987; Prechtl, 1974). Changes in all the parameters previously discussed may occur because of pain, but they may also occur as a result of stress and illness (Stevens & Franck, 1995). After the relevant literature concerning assessment of procedural pain in the infant was reviewed, the parameters chosen to measure procedural pain in the current study included facial expressions, body movements, pitch of cry, duration of cry, and heart rate.
Participants The study was carried out in the neonatal unit and cardiac surgical unit at the authors’ institution. The neonatal unit is a tertiary referral center with 22 medical-surgical beds that provides care for neonates and infants up to 6 months of age. The cardiac surgical unit is one of the Volume 31, Number 5
main referral centers in Australia for infants and children with congenital heart defects. A total of 20 infants undergoing 24 heel lance procedures were studied over a 10-week period between November 1999 and January 2000. Two of the infants were receiving ventilation support. The remaining 18 infants were mainly convalescing after surgery.
Methods Conduct of this study complied with the ethical requirements of the institution. Infants were recruited on the day of study if a blood test by heel lance had been ordered and if they had not received neuromuscular blockade drugs in the last 48 hours. Two of the authors, DH and CE, carried out all assessments simultaneously and independently, and were blinded to each other’s scores. Equipment used in the study was a dictaphone to record crying times and characteristics and an electronic digital clock timer to time the duration of the procedure and to ensure that pain assessments were documented at the correct time points. All infants were connected to a Hewlett Packard Transport Monitor; model M1276A (Hewlett Packard, Now Phillips, Andover, MA) for measurement of heart rate. The following behavioral and physiologic parameters were used to measure pain in this study: 1. A limited subset of the NFCS (brow bulge, eye squeeze, nasolabial furrow, and open lips) (adapted with permission from Grunau & Craig, 1987). Each parameter was scored 1 if present and 0 if absent. 2. A modified version of the IBCS (to allow feasible scoring at the bedside) (Craig et al., 1993). The modified scoring system consisted of (a) limbs— sweeping, waving arms and legs, or kicking legs; and (b) hands/toes—clenched or splayed. Both parameters were scored 1 if present or 0 if absent. 3. Crying characteristics: duration of crying and pitch of cry. Pitch of cry was scored on a 0 to 4 scale, with 4 being scored for a high-pitched cry, 3 for a mediumpitched cry, 2 for a low-pitched cry, and 1 for an altered respiratory pattern compared to baseline. Zero was scored if there was no cry. The “silent cry” facial expression, based on the silent cry description by Sparshott (1996), was included to enable scoring of crying characteristics in the nonverbal infants with endotracheal tubes in situ. The silent cry was scored on a 0 to 4 scale, ranging from no cry to altered respiratory pattern, graded as minimal, moderate, or maximum effort. 4. Heart rate. Observations were taken at baseline, immediately before the infants underwent the heel lance procedures. Baseline observations of behavioral state (Prechtl, 1974), heart rate, facial and body movement scores, and crying were documented. September/October 2002
Upon heel lance, the authors (DH, CE) independently hand-recorded measurements of the infant’s facial expressions and body movements. The heart rate was recorded at the beginning of each time point. The behavioral parameters and heart rate were then recorded again at 30 seconds, 1 minute, and each subsequent minute until the blood collection was completed, as denoted by application of the adhesive bandage. The same parameters were recorded for the 3 minutes following completion of the procedure. For the infants with endotracheal tubes in situ, the effort of silent cry, as determined by facial expressions, was also scored at these time points. During this period, the bedside nurses and parents, if present, were asked to continue with any normally instituted comfort measures. After all the scoring was completed, the crying characteristics were coded and timed from the audiotape. DH and CE independently scored the pitch of cry on a 0 to 4 scale. Duration of the first cry (until a 5-second pause) was then measured from the audiotape. The duration of total crying time during the procedure was calculated as a percentage of the duration of the heel lance procedure. Crying time following the procedure also was expressed as a percentage of the 3-minute observation period.
Data Analysis Microsoft Excel, version 2000 (Microsoft Corporation, Redmond, WA) was used to analyze the descriptive statistics. Lin’s concordance correlation coefficient (Newton, 2000) was used to examine the strength of agreement between the two raters on the three behavioral parameters (Stata 6.0 Statistical Software, Stata Corporation, College Station, TX). The Kruskal-Wallis test was used to examine analysis of variance with regard to sleep states and pain response.
Results Twelve female infants and 8 male infants were included in the study. Four infants were included twice with a minimum of 1 week between each study. Demographic characteristics are outlined in Table 1. In regard to the infants’ behavioral state at baseline, 12 infants were in State 1 (quiet sleep), two infants in State 2 (active sleep), seven infants in State 3 (quiet awake), two infants in State 4 (awake and active) and one infant was crying. The mean time taken for the blood collection procedure (timed from lance to bandage application) was 164 ± 94 (SD) seconds. There was a wide variability in the facial scores recorded, with low scores at baseline and an increase after the heel lance. The scores remained high throughout the procedure but were substantially reduced over the 3 minutes following the procedure (see Figure 1). Scores for body movements were similar to facial scores, with a wide vari-
JOGNN 553
2
Demographic Characteristics Characteristic
Mean (SD)
Range
Gestational age (weeks) Birth weight (kilograms) Age at day of study (days) Total length of stay (days)
36 (3) 2.52 (0.78) 29 (32) 48 (66)
28–41a 0.98–3.5 2–124 7–220
a28–32
weeks: 3 infants. > 32–36 weeks: 6 infants. > 36 weeks: 11
Body Movement Scores
TABLE 1
1
0
Baseline Lance 30 sec. 1 min. 2 min. 3 min. 1 min. 2 min. 3 min. post post post
infants.
Time Points
4
Mean Facial Scores
FIGURE 2
Mean body movement scores obtained during and following the heel lance.
3
Note. The two investigators, DH ■ and CE ■, independently scored two categories of body movements at baseline, during, and for 3 minutes following a routine heel lance procedure. The scoring system was based on a modified version of the Infant Body Coding System.
2
1
0 Baseline Lance 30 sec. 1 min. 2 min. 3 min. 1 min. 2 min. 3 min. post post post Time Points
FIGURE 1
Mean facial scores during and following the heel lance. Note. The two investigators, DH ■ and CE ■, independently scored facial expressions at baseline, during, and for 3 minutes following a routine heel lance procedure. Scores were based on the limited subset of the Neonatal Facial Coding System.
ability in response to the procedure. As illustrated in Figure 2, most infants had low body movement scores at baseline but increasingly higher scores after the lance and throughout the heel squeeze phase of the procedure. There was a reduction in body movement scores over the 3-minute period after the procedure. The crying duration differed widely among the infants. The duration of the first cry varied, from 6 infants who had no initial first cry to 13 infants crying for the duration of the procedure. Three infants continued to cry for the full 3-minute period after the procedure. The mean percentage of time spent crying was 76% (SD = 40%) during the procedure and 31% (SD = 40%) for the 3 minutes after the procedure. As illustrated in Figure 3, mean crying scores (effort of silent cry for the 2 infants with endotracheal tubes) increased from baseline and remained constant throughout the heel lance procedure, then reduced to a mean score of 0.4 by 3 minutes after the procedure. All of the infants showed an increase in their heart rates from the baseline period, in response to the heel lance (see Figure 4). The infants’ maximum mean increase 554 JOGNN
in heart rate of 18% from baseline was seen at 2 minutes into the heel lance procedure.
Interrater Agreement Interrater reliability studies examine the extent to which two or more people agree in their observations of an event (Hicks, 1996). For pain measures to be reliable, high agreement of pain ratings should be established between different people observing and scoring the same situation at the same time (Stevens, Johnston, & Grunau, 1995). Lin’s concordance correlation coefficient (Newton, 2000) was calculated to assess the strength of agreement between the two raters on the three behavioral parameters, with a score of 1 showing perfect agreement and a score of zero showing no agreement. The two authors’ (DH, CE) mean pain scores for each infant scored during and for the 3 minutes following the procedure were examined for interrater agreement. The concordance correlation coefficient estimated on the four facial expressions scored during the heel lance procedure was ρc = .95, with a 95% confidence interval of 0.90 to 0.99; for the 3 minutes following the procedure it was ρc = .98, (0.97 to 0.99). The correlation of the scores of body movements was relatively low at all time points, with a concordance correlation coefficient of ρc = .66 and a 95% confidence interval of 0.42 to 0.89 during the procedure, and ρc = .69 (0.47 to 0.90) for the 3 minutes following the procedure. There was high interrater agreement on the pitch of cry, as evidenced by a high concordance correlation coefficient and narrow confidence intervals of ρc = .97 (0.95 to 0.99) during the procedure, and ρc = .98 (0.97 to 0.99) for the 3 minutes following the procedure. Volume 31, Number 5
4
40% Percentage change
35%
Cry Scores
3
2
1
30% 25% 20% 15% 10% 5% 0%
0 Baseline Lance 30 sec.
1 min.
2 min.
3 min.
1 min. post
2 min. post
3 min. post
Baseline Lance 30 sec. 1 min. 2 min. 3 min. 1 min. post
Time Points
FIGURE 3
Mean cry scores obtained at baseline, during, and for the 3 minutes following the heel lance. Note. Mean cry scores obtained at baseline, during, and for the 3 minutes following the heel lance. The two investigators, DH ■ and CE ■, independently scored crying, on a 0 to 4 scale during the routine heel lance procedure. Scores were based on pitch of cry or effort of silent cry.
Discussion The purpose of this study was to examine whether facial expressions, generalized body movements, crying characteristics, and heart rate could be used as measurements of procedural pain in sick hospitalized infants and whether acceptable interrater reliability could be established using these parameters. The bedside application of the subset of four NFCS expressions (Grunau & Craig, 1987) to score procedural pain was shown to be clinically feasible for this small group of sick hospitalized infants. The change in scores (elevation from the baseline to the lance and during the procedure, followed by reduced scores afterward) suggests that the measures were able to detect change occurring as a result of a tissue-injuring event. This is consistent with the results of other studies that have used the NFCS at the bedside (Grunau et al., 1998; Ramenghi et al., 1996; Rushforth, Griffiths, Thorpe, & Levene, 1995; Rushforth & Levene, 1994). The four distinct facial expressions were easily and rapidly identified as being either present or absent at all the observed time points. With respect to the infants’ generalized body movements, based on the IBCS (Craig et al., 1993), low scores were documented at baseline, with the scores increasing on lance and throughout the procedure, followed by a reduction in scores in the 3-minute recovery period. This suggests that body movements in this context are associated with the painful event. The duration of the three measurements of crying time showed variability between infants. This is consistent with other results, as reported by Franck and Miaskowski (1997) in their review of neonatal responses to painful stimuli. September/October 2002
2 min. 3 min. post post
Time points
FIGURE 4
Mean percentage heart rate change from baseline. Note. Infants’ heart rates were recorded at baseline, upon heel lance, at 30 seconds, 1 minute, then each minute until 3 minutes following the procedure. The mean percentage heart rate changes, with standard deviations, are shown.
The infants’ physiologic response to the heel lance procedure was sustained, as indicated by an increase in heart rate that failed to fully return to baseline levels during the 3 minutes following completion of the heel lance procedure. A mean heart rate increase of 9.6% from baseline was still evident 3 minutes after the procedure. This observation is consistent with the sustained physiologic changes reported in other studies following a heel lance (Craig et al., 1993; Stevens & Johnston, 1994). The behavioral states of infants have been shown to affect their response to painful procedures (Ballantyne et al., 1999; Grunau & Craig, 1987). However, in this study no
P
ain assessment methods validated through research are important to facilitate the use of pain reduction strategies.
correlation was found between sleep states and behavioral responses or heart rate. This lack of observed effect may have been due to the small numbers of infants in the sample. As illustrated, there was variability in the three behavioral parameters measured. Not all infants responded to the heel lance by grimacing, gross body movements, or crying, yet all infants responded physiologically, with heart rate alterations. This observation further demonstrates a need to combine both behavioral and physiologic measures when assessing pain, especially in the group of premature and sick infants who may be less able to
JOGNN 555
mount a strong behavioral response (Franck & Miaskowski, 1997). Interrater agreement on facial scores demonstrated high correlation. This result supports other studies in which high interrater agreement was demonstrated at the bedside (Grunau et al., 1998; Rushforth & Levene,
T
he relatively low interrater agreement for infants’ body movements contrasts with results of the original study evaluating the Infant Body Coding System.
1994). Unlike the scoring of distinct facial expressions, the scoring of the infants’ body movements was more difficult, and there was relatively low agreement between the two scorers. The scorers had varying interpretations regarding degrees of body movements. This was in contrast to the original study that tested the IBCS, in which interobserver reliability was 0.83, as assessed by the proportion of agreement on actions obtained by the two coders relative to the total number of actions they coded (Craig et al., 1993). The IBCS study, however, was conducted using video recordings of the event, with coders having the opportunity to closely examine preselected segments of the videotapes. This may well have contributed to the higher interrater agreement reported in their study. In the current study, high interrater agreement was obtained between scorers on the pitch of cry. However, as noted in the literature review, the acoustic properties of cry in hospitalized infants may be altered as a result of disease processes. Thus, an infant’s cry may not be a reliable or valid measure of degree of response to pain. Based on the results of this pilot study, modifications have been made to the pain assessment method. Pain assessment parameters to be used in a subsequent trial will include the same four NFCS expressions, in addition to measurement of the duration of first cry and the total crying time during and following the heel lance. Because there was poor correlation between body movement scores during and following the procedure, this behavioral parameter will be omitted. As previously discussed, pitch of cry may have poor validity as a measurement of pain in the sick hospitalized infant and therefore will also be omitted from the subsequent score. However, crying times will be retained, although this parameter cannot be measured in intubated infants. Measurement of the infants’ heart rate will be retained. Oxygen saturation, which has been measured in several studies of infants’ responses to pain (Craig et al., 1993; Johnston et al., 556 JOGNN
1995; Stevens & Johnston, 1994; Stevens, Johnston, & Horton, 1993), also will be included in the subsequent study.
Limitations This study included a relatively small number of infants. They had a wide range of postgestational ages and hospitalization experiences that may have added to the normal variability of their responses to pain. The age of the infant has been shown to affect both physiologic and behavioral responses to pain (Craig et al., 1993; Johnston et al., 1995), and hospitalization experiences may alter behavioral responses to pain in infants who have been exposed to prolonged periods of neonatal intensive care (Anand, Coskun, Thrivikraman, Nemeroff, & Plotsky, 1999).
Clinical Implications Pain assessment is an important step toward the recognition of pain and the use of routinely planned pain reduction strategies. The bedside utility of the modified NFCS was established in a diverse patient population and, when combined with other behavioral parameters plus the measurement of heart rate changes, was an effective measure of infants’ responses to the noxious stimulus of a heel lance. A combination of behavioral and physiologic parameters can be utilized at the bedside to capture hospitalized infants’ divergent responses to pain resulting from an acute noxious stimulus. Pain assessment methods studied and validated through research are important to enable and encourage measurement of effectiveness of pain reduction strategies and should be a routine component of nursing assessment.
Acknowledgments This study was funded, in part, by the Nursing Research Management Group at the Royal Children’s Hospital, Melbourne. REFERENCES Anand, K. J., Coskun, V., Thrivikraman, K. V., Nemeroff, C. B., & Plotsky, P. M. (1999). Long-term behavioral effects of repetitive pain in neonatal rat pups. Physiology & Behavior, 66(4), 627-637. Anand, K. J., & McGrath, P. (Eds.). (1993). Pain in neonates (Vol. 5). Amsterdam: Elsevier. Anand, K. J., Sippell, W. G., & Aynsley-Green, A. (1987). Randomised trial of fentanyl anaesthesia in preterm babies undergoing surgery: Effects on the stress response. Lancet, 1(8524), 62-66. Ballantyne, M., Stevens, B., McAllister, M., Dionne, K., & Jack, A. (1999). Validation of the Premature Infant Pain Profile in the clinical setting. Clinical Journal of Pain, 15(4), 297303.
Volume 31, Number 5
Bucher, H. U., Moser, T., von Siebenthal, K., Keel, M., Wolf, M., & Duc, G. (1995). Sucrose reduces pain reaction to heel lancing in preterm infants: A placebo-controlled, randomized and masked study. Pediatric Research, 38(3), 332335. Carbajal, R., Chauvet, X., Couderc, S., & Olivier-Martin, M. (1999). Randomised trial of analgesic effects of sucrose, glucose, and pacifiers in term neonates. British Medical Journal, 319(7222), 1393-1397. Chang, H., Anderson, G. C., & Wood, C. E. (1995). Feasible and valid saliva collection for cortisol in transitional newborn infants. Nursing Research, 44(2), 117-119. Craig, K. D., Whitfield, M. F., Grunau, R. V., Linton, J., & Hadjistavropoulos, H. D. (1993). Pain in the preterm neonate: Behavioural and physiological indices. Pain, 52(3), 287299. Franck, L. S., & Miaskowski, C. (1997). Measurement of neonatal responses to painful stimuli: A research review. Journal of Pain & Symptom Management, 14(6), 343-378. Fuller, B. F. (1999). Acoustic discrimination of three types of infant cries. Nursing Research, 40(3), 156-160. Grunau, R. E., & Craig, K. D. (1987). Pain expression in neonates: Facial action and cry. Pain, 28, 395-410. Grunau, R. E., Oberlander, T., Holsti, L., & Whitfield, M. F. (1998). Bedside application of the Neonatal Facial Coding System in pain assessment of premature neonates. Pain, 76(3), 277-286. Hadjistavropoulos, H. D., Craig, K. D., Grunau, R. E., & Johnston, C. (1994). Judging pain in newborns: Facial and cry determinants. Journal of Pediatric Psychology, 19(4), 485-491. Hicks, C. (1996). Undertaking midwifery research. A basic guide to design and analysis. New York: Churchill Livingstone. Johnston, C. C., Sherrard, A., Stevens, B., Franck, L., Stremler, R., & Jack, A. (1999). Do cry features reflect pain intensity in preterm neonates? A preliminary study. Biology of the Neonate, 76(2), 120-124. Johnston, C. C., Stevens, B. J., Franck, L. S., Jack, A., Stremler, R., & Platt, R. (1999). Factors explaining lack of response to heel stick in preterm newborns. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 28, 587-594. Johnston, C. C., Stevens, B. J., Yang, F., & Horton, L. (1995). Differential response to pain by very premature neonates. Pain, 61(3), 471-479. Newton, H. J. (Ed.). (2000). Stata technical bulletin (Vol. 43). Stata Press, College Station, TX: Stata Corporation. Prechtl, H. F. R. (1974). The behavioural states of the newborn infant (a review). Brain Research, 76, 185-212. Ramenghi, L. A., Wood, C. M., Griffith, G. C., & Levene, M. I. (1996). Reduction of pain response in premature infants using intraoral sucrose. Archives of Disease in Childhood, 74, F126-F128.
September/October 2002
Rushforth, J. A., Griffiths, G., Thorpe, H., & Levene, M. (1995). Can topical lignocaine reduce behavioural response to heel prick? Archives of Disease in Childhood Fetal & Neonatal Edition, 72(1), F49-F51. Rushforth, J. A., & Levene, M. I. (1994). Behavioural response to pain in healthy neonates. Archives of Disease in Childhood Fetal & Neonatal Edition, 70(3), F174-F176. Sparshott, M. M. (1996). The development of a clinical distress scale for ventilated newborn infants: Identification of pain and distress based on validated behavioural scores. Journal of Neonatal Nursing, 2(2), 5-11. Stevens, B., Johnston, C., Petryshen, P., & Taddio, A. (1996). Premature Infant Pain Profile: Development and initial validation. Clinical Journal of Pain, 12(1), 13-22. Stevens, B. J., & Franck, L. (1995). Special needs of preterm infants in the management of pain and discomfort. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 24, 856-862. Stevens, B. J., & Johnston, C. C. (1994). Physiological responses of premature infants to a painful stimulus. Nursing Research, 43(4), 226-231. Stevens, B. J., Johnston, C. C., & Grunau, R. V. E. (1995). Issues of assessment of pain and discomfort in neonates. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 24, 849-855. Stevens, B. J., Johnston, C. C., & Horton, L. (1993). Multidimensional pain assessment in premature neonates: A pilot study. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 22, 531-541.
Denise Harrison has a joint role as a research assistant and clinical nurse educator at the Royal Children’s Hospital, Melbourne, Victoria, Australia. Cheryl Evans is a clinical nurse specialist in the Neonatal Unit, Royal Children’s Hospital, Melbourne, Victoria, Australia. Linda Johnston is an associate professor of nursing with The Victorian Center for Nursing Practice Research, The University of Melbourne, and the Royal Children’s Hospital, Melbourne, Victoria, Australia. Peter Loughnan is a consultant pediatrician, in the Neonatal Unit, Royal Children’s Hospital, Melbourne, Victoria, Australia. Address for correspondence: Denise Harrison, RN, RM, MN, 7th floor South East Building, Royal Children’s Hospital, Melbourne, 3052 Victoria, Australia. E-mail: [email protected]. unimelb.edu.au.
JOGNN 557
CLINICAL STUDIES
Bedside Assessment of Heel Lance Pain in the Hospitalized Infant Denise Harrison, RN, RM, MN, Cheryl Evans, RN, RM, BN, Linda Johnston, RN, PHD, Peter Loughnan, MB, BS, FRACP
Objective: To evaluate a method of pain assessment to be used for hospitalized infants requiring blood test by a heel lance procedure. Design: Observational study evaluating pain measurement and interrater reliability of pain measurement in hospitalized infants. Setting: A Level III neonatal unit and a cardiac surgical unit at a major pediatric teaching hospital. Participants: 20 infants whose gestational age ranged from 28 weeks to full-term. Main Outcome Measures: Observations included behavioral measurements (facial expressions, body movements, and crying characteristics) and the physiologic measure of heart rate. Interrater reliability and the feasibility of using the procedural pain assessment method at the bedside were considered. Results: The three behavioral measurements and heart rate were responsive to the heel lance. Interrater reliability was high for facial expressions and crying scores but was low for body movements. Conclusions: Modifications have been made to the method of procedural pain assessment to be used in a subsequent study. The modified method is expected to be a reliable measure of procedural pain caused by a heel lance and can easily be used at the bedside during the course of further research. JOGNN, 31, 551–557; 2002. DOI: 10.1177/088421702237738 Keywords: Neonatal pain—Pain assessment— Pain scoring—Procedural pain Accepted: October 2001 Procedures that cause pain in infants are common occurrences in neonatal nurseries. Recommendations are being made in the current literature to utiSeptember/October 2002
lize interventions to reduce procedural pain. To adequately assess the effectiveness of pain reduction interventions, nurses caring for infants must be able to carry out valid and reliable pain assessments. However, because there is no gold standard of pain measurement in infants, there is no single parameter by which to accurately assess pain. Thus, pain assessment in this age group is a difficult task. This difficulty is a salient reason for the ongoing discussion in the literature concerning the most appropriate and reliable method of assessing pain.
Objective The aim of the current study was to test a method of pain assessment that could subsequently be used to measure the effects of a pain reduction strategy for infants undergoing a heel lance procedure. For the purposes of this study, the pain assessment method had to be applicable to hospitalized infants and suitable for bedside application. The method also needed to be appropriate for the assessment of the short-term acute procedural pain of a heel lance, versus pain associated with surgery or chronic pain.
Background Assessment of the complex, subjective nature of pain in the infant must take into account many factors. Pain responses demonstrated by well infants often have both dramatic observable behavioral and physiologic components, but these responses also can occur outside of observable painful situations (Johnston, Stevens, et al., 1999). The responses to pain vary widely among infants, especially in those who are sick or premature and whose behavioral
JOGNN 551
responses may be less robust or even absent, yet their physiologic responses may still be dramatic (Johnston, Stevens, et al., 1999). Infants’ behavioral responses to pain include facial expressions, body movements, and crying characteristics. Infants’ facial expressions relating to pain have been studied in detail (Grunau & Craig, 1987) and have been cited as the most specific indicator of pain resulting from noxious stimuli (Stevens, Johnston, Petryshen, & Taddio, 1996). The Neonatal Facial Coding System (NFCS), comprising nine facial expressions, was described by Grunau and Craig (1987). The NFCS has been shown to detect changes in facial expression in response to a heel lance procedure in infants of all gestational ages, even very premature infants (Grunau, Oberlander, Holsti, & Whitfield, 1998; Johnston, Stevens, Yang, & Horton, 1995), although with less frequency and vigor than in more mature infants (Craig, Whitfield, Grunau, Linton, & Hadjistavropoulos, 1993; Johnston, Stevens, et al., 1999). A limited subset of four of the original nine facial expressions from the NFCS has been reported to occur most commonly as a result of a heel lance procedure. Brow bulge, eye squeeze, nasolabial furrow, and open lips were
B
edside application of a subset of the Neonatal Facial Coding System was shown to be both reliable and clinically feasible.
observed in 99% of infants within 6 seconds of the heel lance (Grunau & Craig, 1987). The authors of the NFCS also extensively studied infants’ body movements occurring as a result of a heel lance. The Infant Body Coding System (IBCS) was developed as a result of videotaping 56 infants during routine heel lances (Craig et al., 1993). Coders viewing the videos scored the presence of motor activities (hand, foot, arm, leg, head, and torso movements) at intervals throughout the procedure. This study showed that the heel lance instigated the most vigorous bodily activity, although there also was a substantial response to the heel swab alone. Body movements were therefore believed to be less specific than facial expressions to the painful event. Infants’ crying characteristics in response to procedural pain also have been studied by several authors. The crying characteristics studied have included both acoustic and temporal measurements. Acoustic studies have analyzed the pitch, fundamental frequency, and amplitude of infants’ cries (Fuller, 1999). Temporal characteristics have been categorized into cry latency, duration of first cry (Ramenghi, Wood, Griffith, & Levene, 1996), number of 552 JOGNN
cry cycles, plus duration of crying times throughout and following painful procedures (Fuller, 1999; Grunau & Craig, 1987; Hadjistavropoulos, Craig, Grunau, & Johnston, 1994; Johnston, Sherrard, et al., 1999). Infants’ cry characteristics can provide important information and can be an obvious and useful measure in studies evaluating different pain reduction strategies. However, crying is not an “accurate” pain measurement on its own and is not unique to pain (Carbajal, Chauvet, Couderc, & Olivier-Martin, 1999). Crying characteristics also may be altered in a sick infant, may not be assessable in an infant with an endotracheal tube in situ, and may vary widely from infant to infant at any given time. There may be a low incidence of crying as a response to pain in premature infants (Johnston, Sherrad, et al., 1999), as shown by Stevens’s study, in which only 47% of premature infants cried during the painful (lance and heel squeeze) stages of a heel lance (Stevens et al., 1996). Physiologic parameters also can be used to measure change after a painful procedure, yet on their own they are not specific measurements of pain. The most common physiologic parameters used to assess infants’ pain include heart rate, respiratory rate, and oxygen saturation. Other physiologic and biochemical parameters sometimes used to assess pain are changes in intracranial pressure, palmar sweating, skin blood flow, blood pressure, glucose levels, blood and/or saliva stress hormone levels, catecholamines, ketones, and urine nitrogenous compounds (Anand & McGrath, 1993; Anand, Sippell, & AynsleyGreen, 1987; Chang, Anderson, & Wood, 1995; Franck & Miaskowski, 1997). In one study, cerebral blood flow, cerebral oxyhemoglobin, and deoxyhemoglobin were used as indicators of pain (Bucher et al., 1995). Although not a response to pain as such, the infant’s sleep or behavioral state before procedures are performed has been shown to influence subsequent behavioral response to pain and therefore is an important factor to note when procedural pain is monitored (Ballantyne, Stevens, McAllister, Dionne, & Jack, 1999; Grunau & Craig, 1987; Prechtl, 1974). Changes in all the parameters previously discussed may occur because of pain, but they may also occur as a result of stress and illness (Stevens & Franck, 1995). After the relevant literature concerning assessment of procedural pain in the infant was reviewed, the parameters chosen to measure procedural pain in the current study included facial expressions, body movements, pitch of cry, duration of cry, and heart rate.
Participants The study was carried out in the neonatal unit and cardiac surgical unit at the authors’ institution. The neonatal unit is a tertiary referral center with 22 medical-surgical beds that provides care for neonates and infants up to 6 months of age. The cardiac surgical unit is one of the Volume 31, Number 5
main referral centers in Australia for infants and children with congenital heart defects. A total of 20 infants undergoing 24 heel lance procedures were studied over a 10-week period between November 1999 and January 2000. Two of the infants were receiving ventilation support. The remaining 18 infants were mainly convalescing after surgery.
Methods Conduct of this study complied with the ethical requirements of the institution. Infants were recruited on the day of study if a blood test by heel lance had been ordered and if they had not received neuromuscular blockade drugs in the last 48 hours. Two of the authors, DH and CE, carried out all assessments simultaneously and independently, and were blinded to each other’s scores. Equipment used in the study was a dictaphone to record crying times and characteristics and an electronic digital clock timer to time the duration of the procedure and to ensure that pain assessments were documented at the correct time points. All infants were connected to a Hewlett Packard Transport Monitor; model M1276A (Hewlett Packard, Now Phillips, Andover, MA) for measurement of heart rate. The following behavioral and physiologic parameters were used to measure pain in this study: 1. A limited subset of the NFCS (brow bulge, eye squeeze, nasolabial furrow, and open lips) (adapted with permission from Grunau & Craig, 1987). Each parameter was scored 1 if present and 0 if absent. 2. A modified version of the IBCS (to allow feasible scoring at the bedside) (Craig et al., 1993). The modified scoring system consisted of (a) limbs— sweeping, waving arms and legs, or kicking legs; and (b) hands/toes—clenched or splayed. Both parameters were scored 1 if present or 0 if absent. 3. Crying characteristics: duration of crying and pitch of cry. Pitch of cry was scored on a 0 to 4 scale, with 4 being scored for a high-pitched cry, 3 for a mediumpitched cry, 2 for a low-pitched cry, and 1 for an altered respiratory pattern compared to baseline. Zero was scored if there was no cry. The “silent cry” facial expression, based on the silent cry description by Sparshott (1996), was included to enable scoring of crying characteristics in the nonverbal infants with endotracheal tubes in situ. The silent cry was scored on a 0 to 4 scale, ranging from no cry to altered respiratory pattern, graded as minimal, moderate, or maximum effort. 4. Heart rate. Observations were taken at baseline, immediately before the infants underwent the heel lance procedures. Baseline observations of behavioral state (Prechtl, 1974), heart rate, facial and body movement scores, and crying were documented. September/October 2002
Upon heel lance, the authors (DH, CE) independently hand-recorded measurements of the infant’s facial expressions and body movements. The heart rate was recorded at the beginning of each time point. The behavioral parameters and heart rate were then recorded again at 30 seconds, 1 minute, and each subsequent minute until the blood collection was completed, as denoted by application of the adhesive bandage. The same parameters were recorded for the 3 minutes following completion of the procedure. For the infants with endotracheal tubes in situ, the effort of silent cry, as determined by facial expressions, was also scored at these time points. During this period, the bedside nurses and parents, if present, were asked to continue with any normally instituted comfort measures. After all the scoring was completed, the crying characteristics were coded and timed from the audiotape. DH and CE independently scored the pitch of cry on a 0 to 4 scale. Duration of the first cry (until a 5-second pause) was then measured from the audiotape. The duration of total crying time during the procedure was calculated as a percentage of the duration of the heel lance procedure. Crying time following the procedure also was expressed as a percentage of the 3-minute observation period.
Data Analysis Microsoft Excel, version 2000 (Microsoft Corporation, Redmond, WA) was used to analyze the descriptive statistics. Lin’s concordance correlation coefficient (Newton, 2000) was used to examine the strength of agreement between the two raters on the three behavioral parameters (Stata 6.0 Statistical Software, Stata Corporation, College Station, TX). The Kruskal-Wallis test was used to examine analysis of variance with regard to sleep states and pain response.
Results Twelve female infants and 8 male infants were included in the study. Four infants were included twice with a minimum of 1 week between each study. Demographic characteristics are outlined in Table 1. In regard to the infants’ behavioral state at baseline, 12 infants were in State 1 (quiet sleep), two infants in State 2 (active sleep), seven infants in State 3 (quiet awake), two infants in State 4 (awake and active) and one infant was crying. The mean time taken for the blood collection procedure (timed from lance to bandage application) was 164 ± 94 (SD) seconds. There was a wide variability in the facial scores recorded, with low scores at baseline and an increase after the heel lance. The scores remained high throughout the procedure but were substantially reduced over the 3 minutes following the procedure (see Figure 1). Scores for body movements were similar to facial scores, with a wide vari-
JOGNN 553
2
Demographic Characteristics Characteristic
Mean (SD)
Range
Gestational age (weeks) Birth weight (kilograms) Age at day of study (days) Total length of stay (days)
36 (3) 2.52 (0.78) 29 (32) 48 (66)
28–41a 0.98–3.5 2–124 7–220
a28–32
weeks: 3 infants. > 32–36 weeks: 6 infants. > 36 weeks: 11
Body Movement Scores
TABLE 1
1
0
Baseline Lance 30 sec. 1 min. 2 min. 3 min. 1 min. 2 min. 3 min. post post post
infants.
Time Points
4
Mean Facial Scores
FIGURE 2
Mean body movement scores obtained during and following the heel lance.
3
Note. The two investigators, DH ■ and CE ■, independently scored two categories of body movements at baseline, during, and for 3 minutes following a routine heel lance procedure. The scoring system was based on a modified version of the Infant Body Coding System.
2
1
0 Baseline Lance 30 sec. 1 min. 2 min. 3 min. 1 min. 2 min. 3 min. post post post Time Points
FIGURE 1
Mean facial scores during and following the heel lance. Note. The two investigators, DH ■ and CE ■, independently scored facial expressions at baseline, during, and for 3 minutes following a routine heel lance procedure. Scores were based on the limited subset of the Neonatal Facial Coding System.
ability in response to the procedure. As illustrated in Figure 2, most infants had low body movement scores at baseline but increasingly higher scores after the lance and throughout the heel squeeze phase of the procedure. There was a reduction in body movement scores over the 3-minute period after the procedure. The crying duration differed widely among the infants. The duration of the first cry varied, from 6 infants who had no initial first cry to 13 infants crying for the duration of the procedure. Three infants continued to cry for the full 3-minute period after the procedure. The mean percentage of time spent crying was 76% (SD = 40%) during the procedure and 31% (SD = 40%) for the 3 minutes after the procedure. As illustrated in Figure 3, mean crying scores (effort of silent cry for the 2 infants with endotracheal tubes) increased from baseline and remained constant throughout the heel lance procedure, then reduced to a mean score of 0.4 by 3 minutes after the procedure. All of the infants showed an increase in their heart rates from the baseline period, in response to the heel lance (see Figure 4). The infants’ maximum mean increase 554 JOGNN
in heart rate of 18% from baseline was seen at 2 minutes into the heel lance procedure.
Interrater Agreement Interrater reliability studies examine the extent to which two or more people agree in their observations of an event (Hicks, 1996). For pain measures to be reliable, high agreement of pain ratings should be established between different people observing and scoring the same situation at the same time (Stevens, Johnston, & Grunau, 1995). Lin’s concordance correlation coefficient (Newton, 2000) was calculated to assess the strength of agreement between the two raters on the three behavioral parameters, with a score of 1 showing perfect agreement and a score of zero showing no agreement. The two authors’ (DH, CE) mean pain scores for each infant scored during and for the 3 minutes following the procedure were examined for interrater agreement. The concordance correlation coefficient estimated on the four facial expressions scored during the heel lance procedure was ρc = .95, with a 95% confidence interval of 0.90 to 0.99; for the 3 minutes following the procedure it was ρc = .98, (0.97 to 0.99). The correlation of the scores of body movements was relatively low at all time points, with a concordance correlation coefficient of ρc = .66 and a 95% confidence interval of 0.42 to 0.89 during the procedure, and ρc = .69 (0.47 to 0.90) for the 3 minutes following the procedure. There was high interrater agreement on the pitch of cry, as evidenced by a high concordance correlation coefficient and narrow confidence intervals of ρc = .97 (0.95 to 0.99) during the procedure, and ρc = .98 (0.97 to 0.99) for the 3 minutes following the procedure. Volume 31, Number 5
4
40% Percentage change
35%
Cry Scores
3
2
1
30% 25% 20% 15% 10% 5% 0%
0 Baseline Lance 30 sec.
1 min.
2 min.
3 min.
1 min. post
2 min. post
3 min. post
Baseline Lance 30 sec. 1 min. 2 min. 3 min. 1 min. post
Time Points
FIGURE 3
Mean cry scores obtained at baseline, during, and for the 3 minutes following the heel lance. Note. Mean cry scores obtained at baseline, during, and for the 3 minutes following the heel lance. The two investigators, DH ■ and CE ■, independently scored crying, on a 0 to 4 scale during the routine heel lance procedure. Scores were based on pitch of cry or effort of silent cry.
Discussion The purpose of this study was to examine whether facial expressions, generalized body movements, crying characteristics, and heart rate could be used as measurements of procedural pain in sick hospitalized infants and whether acceptable interrater reliability could be established using these parameters. The bedside application of the subset of four NFCS expressions (Grunau & Craig, 1987) to score procedural pain was shown to be clinically feasible for this small group of sick hospitalized infants. The change in scores (elevation from the baseline to the lance and during the procedure, followed by reduced scores afterward) suggests that the measures were able to detect change occurring as a result of a tissue-injuring event. This is consistent with the results of other studies that have used the NFCS at the bedside (Grunau et al., 1998; Ramenghi et al., 1996; Rushforth, Griffiths, Thorpe, & Levene, 1995; Rushforth & Levene, 1994). The four distinct facial expressions were easily and rapidly identified as being either present or absent at all the observed time points. With respect to the infants’ generalized body movements, based on the IBCS (Craig et al., 1993), low scores were documented at baseline, with the scores increasing on lance and throughout the procedure, followed by a reduction in scores in the 3-minute recovery period. This suggests that body movements in this context are associated with the painful event. The duration of the three measurements of crying time showed variability between infants. This is consistent with other results, as reported by Franck and Miaskowski (1997) in their review of neonatal responses to painful stimuli. September/October 2002
2 min. 3 min. post post
Time points
FIGURE 4
Mean percentage heart rate change from baseline. Note. Infants’ heart rates were recorded at baseline, upon heel lance, at 30 seconds, 1 minute, then each minute until 3 minutes following the procedure. The mean percentage heart rate changes, with standard deviations, are shown.
The infants’ physiologic response to the heel lance procedure was sustained, as indicated by an increase in heart rate that failed to fully return to baseline levels during the 3 minutes following completion of the heel lance procedure. A mean heart rate increase of 9.6% from baseline was still evident 3 minutes after the procedure. This observation is consistent with the sustained physiologic changes reported in other studies following a heel lance (Craig et al., 1993; Stevens & Johnston, 1994). The behavioral states of infants have been shown to affect their response to painful procedures (Ballantyne et al., 1999; Grunau & Craig, 1987). However, in this study no
P
ain assessment methods validated through research are important to facilitate the use of pain reduction strategies.
correlation was found between sleep states and behavioral responses or heart rate. This lack of observed effect may have been due to the small numbers of infants in the sample. As illustrated, there was variability in the three behavioral parameters measured. Not all infants responded to the heel lance by grimacing, gross body movements, or crying, yet all infants responded physiologically, with heart rate alterations. This observation further demonstrates a need to combine both behavioral and physiologic measures when assessing pain, especially in the group of premature and sick infants who may be less able to
JOGNN 555
mount a strong behavioral response (Franck & Miaskowski, 1997). Interrater agreement on facial scores demonstrated high correlation. This result supports other studies in which high interrater agreement was demonstrated at the bedside (Grunau et al., 1998; Rushforth & Levene,
T
he relatively low interrater agreement for infants’ body movements contrasts with results of the original study evaluating the Infant Body Coding System.
1994). Unlike the scoring of distinct facial expressions, the scoring of the infants’ body movements was more difficult, and there was relatively low agreement between the two scorers. The scorers had varying interpretations regarding degrees of body movements. This was in contrast to the original study that tested the IBCS, in which interobserver reliability was 0.83, as assessed by the proportion of agreement on actions obtained by the two coders relative to the total number of actions they coded (Craig et al., 1993). The IBCS study, however, was conducted using video recordings of the event, with coders having the opportunity to closely examine preselected segments of the videotapes. This may well have contributed to the higher interrater agreement reported in their study. In the current study, high interrater agreement was obtained between scorers on the pitch of cry. However, as noted in the literature review, the acoustic properties of cry in hospitalized infants may be altered as a result of disease processes. Thus, an infant’s cry may not be a reliable or valid measure of degree of response to pain. Based on the results of this pilot study, modifications have been made to the pain assessment method. Pain assessment parameters to be used in a subsequent trial will include the same four NFCS expressions, in addition to measurement of the duration of first cry and the total crying time during and following the heel lance. Because there was poor correlation between body movement scores during and following the procedure, this behavioral parameter will be omitted. As previously discussed, pitch of cry may have poor validity as a measurement of pain in the sick hospitalized infant and therefore will also be omitted from the subsequent score. However, crying times will be retained, although this parameter cannot be measured in intubated infants. Measurement of the infants’ heart rate will be retained. Oxygen saturation, which has been measured in several studies of infants’ responses to pain (Craig et al., 1993; Johnston et al., 556 JOGNN
1995; Stevens & Johnston, 1994; Stevens, Johnston, & Horton, 1993), also will be included in the subsequent study.
Limitations This study included a relatively small number of infants. They had a wide range of postgestational ages and hospitalization experiences that may have added to the normal variability of their responses to pain. The age of the infant has been shown to affect both physiologic and behavioral responses to pain (Craig et al., 1993; Johnston et al., 1995), and hospitalization experiences may alter behavioral responses to pain in infants who have been exposed to prolonged periods of neonatal intensive care (Anand, Coskun, Thrivikraman, Nemeroff, & Plotsky, 1999).
Clinical Implications Pain assessment is an important step toward the recognition of pain and the use of routinely planned pain reduction strategies. The bedside utility of the modified NFCS was established in a diverse patient population and, when combined with other behavioral parameters plus the measurement of heart rate changes, was an effective measure of infants’ responses to the noxious stimulus of a heel lance. A combination of behavioral and physiologic parameters can be utilized at the bedside to capture hospitalized infants’ divergent responses to pain resulting from an acute noxious stimulus. Pain assessment methods studied and validated through research are important to enable and encourage measurement of effectiveness of pain reduction strategies and should be a routine component of nursing assessment.
Acknowledgments This study was funded, in part, by the Nursing Research Management Group at the Royal Children’s Hospital, Melbourne. REFERENCES Anand, K. J., Coskun, V., Thrivikraman, K. V., Nemeroff, C. B., & Plotsky, P. M. (1999). Long-term behavioral effects of repetitive pain in neonatal rat pups. Physiology & Behavior, 66(4), 627-637. Anand, K. J., & McGrath, P. (Eds.). (1993). Pain in neonates (Vol. 5). Amsterdam: Elsevier. Anand, K. J., Sippell, W. G., & Aynsley-Green, A. (1987). Randomised trial of fentanyl anaesthesia in preterm babies undergoing surgery: Effects on the stress response. Lancet, 1(8524), 62-66. Ballantyne, M., Stevens, B., McAllister, M., Dionne, K., & Jack, A. (1999). Validation of the Premature Infant Pain Profile in the clinical setting. Clinical Journal of Pain, 15(4), 297303.
Volume 31, Number 5
Bucher, H. U., Moser, T., von Siebenthal, K., Keel, M., Wolf, M., & Duc, G. (1995). Sucrose reduces pain reaction to heel lancing in preterm infants: A placebo-controlled, randomized and masked study. Pediatric Research, 38(3), 332335. Carbajal, R., Chauvet, X., Couderc, S., & Olivier-Martin, M. (1999). Randomised trial of analgesic effects of sucrose, glucose, and pacifiers in term neonates. British Medical Journal, 319(7222), 1393-1397. Chang, H., Anderson, G. C., & Wood, C. E. (1995). Feasible and valid saliva collection for cortisol in transitional newborn infants. Nursing Research, 44(2), 117-119. Craig, K. D., Whitfield, M. F., Grunau, R. V., Linton, J., & Hadjistavropoulos, H. D. (1993). Pain in the preterm neonate: Behavioural and physiological indices. Pain, 52(3), 287299. Franck, L. S., & Miaskowski, C. (1997). Measurement of neonatal responses to painful stimuli: A research review. Journal of Pain & Symptom Management, 14(6), 343-378. Fuller, B. F. (1999). Acoustic discrimination of three types of infant cries. Nursing Research, 40(3), 156-160. Grunau, R. E., & Craig, K. D. (1987). Pain expression in neonates: Facial action and cry. Pain, 28, 395-410. Grunau, R. E., Oberlander, T., Holsti, L., & Whitfield, M. F. (1998). Bedside application of the Neonatal Facial Coding System in pain assessment of premature neonates. Pain, 76(3), 277-286. Hadjistavropoulos, H. D., Craig, K. D., Grunau, R. E., & Johnston, C. (1994). Judging pain in newborns: Facial and cry determinants. Journal of Pediatric Psychology, 19(4), 485-491. Hicks, C. (1996). Undertaking midwifery research. A basic guide to design and analysis. New York: Churchill Livingstone. Johnston, C. C., Sherrard, A., Stevens, B., Franck, L., Stremler, R., & Jack, A. (1999). Do cry features reflect pain intensity in preterm neonates? A preliminary study. Biology of the Neonate, 76(2), 120-124. Johnston, C. C., Stevens, B. J., Franck, L. S., Jack, A., Stremler, R., & Platt, R. (1999). Factors explaining lack of response to heel stick in preterm newborns. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 28, 587-594. Johnston, C. C., Stevens, B. J., Yang, F., & Horton, L. (1995). Differential response to pain by very premature neonates. Pain, 61(3), 471-479. Newton, H. J. (Ed.). (2000). Stata technical bulletin (Vol. 43). Stata Press, College Station, TX: Stata Corporation. Prechtl, H. F. R. (1974). The behavioural states of the newborn infant (a review). Brain Research, 76, 185-212. Ramenghi, L. A., Wood, C. M., Griffith, G. C., & Levene, M. I. (1996). Reduction of pain response in premature infants using intraoral sucrose. Archives of Disease in Childhood, 74, F126-F128.
September/October 2002
Rushforth, J. A., Griffiths, G., Thorpe, H., & Levene, M. (1995). Can topical lignocaine reduce behavioural response to heel prick? Archives of Disease in Childhood Fetal & Neonatal Edition, 72(1), F49-F51. Rushforth, J. A., & Levene, M. I. (1994). Behavioural response to pain in healthy neonates. Archives of Disease in Childhood Fetal & Neonatal Edition, 70(3), F174-F176. Sparshott, M. M. (1996). The development of a clinical distress scale for ventilated newborn infants: Identification of pain and distress based on validated behavioural scores. Journal of Neonatal Nursing, 2(2), 5-11. Stevens, B., Johnston, C., Petryshen, P., & Taddio, A. (1996). Premature Infant Pain Profile: Development and initial validation. Clinical Journal of Pain, 12(1), 13-22. Stevens, B. J., & Franck, L. (1995). Special needs of preterm infants in the management of pain and discomfort. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 24, 856-862. Stevens, B. J., & Johnston, C. C. (1994). Physiological responses of premature infants to a painful stimulus. Nursing Research, 43(4), 226-231. Stevens, B. J., Johnston, C. C., & Grunau, R. V. E. (1995). Issues of assessment of pain and discomfort in neonates. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 24, 849-855. Stevens, B. J., Johnston, C. C., & Horton, L. (1993). Multidimensional pain assessment in premature neonates: A pilot study. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 22, 531-541.
Denise Harrison has a joint role as a research assistant and clinical nurse educator at the Royal Children’s Hospital, Melbourne, Victoria, Australia. Cheryl Evans is a clinical nurse specialist in the Neonatal Unit, Royal Children’s Hospital, Melbourne, Victoria, Australia. Linda Johnston is an associate professor of nursing with The Victorian Center for Nursing Practice Research, The University of Melbourne, and the Royal Children’s Hospital, Melbourne, Victoria, Australia. Peter Loughnan is a consultant pediatrician, in the Neonatal Unit, Royal Children’s Hospital, Melbourne, Victoria, Australia. Address for correspondence: Denise Harrison, RN, RM, MN, 7th floor South East Building, Royal Children’s Hospital, Melbourne, 3052 Victoria, Australia. E-mail: [email protected]. unimelb.edu.au.
JOGNN 557