EFTA02695993.pdf

Musichitand Medicine v, :epub.cor Effects of Music on Physiological and Behavioral Indices of Acute Pain and Stress in Premature Infant; Clinical Trial and Literature Review Mark Jude Tramo, Miriam Lense, Caitlin Van Ness, Jerome Kagan, Margaret Doyle Settle and Jonathan H. Cronin Music and Medicine 2011 3: 72 DOI: 10.1177/1943862111400613 The online version of this article can be found at: http://mmd.sagepub.com/content/3/2/72 Published by: OSAGE hun wwww sancin:• :)n5 corn On behalf of: I 1\I \ I International Association for Music and Medicine Additional services and information for Music and Medicine can be found at: Email Alerts: 1p://mmd.sagepub.cornrcgiralerts Subscriptions: .m://mmd.Sagepub.comrsubscriptionS Reprints: hIlpliwww.sagepub.comtpumalsReprints.nav Permissions: http://www.sagepub.comfjoumalsPermissions.nav EFTA_R1_02048096 EFTA02695993 Armies Mean and Modem* 3(2) 72-83 Effects of Music on Physiological and {LL The Author(s) 2011 Rennet' and perm anon uplink tonepuonalsPermtssions nay Behavioral Indices of Acute Pain and OOt. 10.1177:1943862111400613 InenEmmidtagepub corn Stress in Premature Infants: Clinical ®SAGE Trial and Literature Review Mark Jude Tramo, MD, PhD", Miriam Lense, MSI '2'4, Caitlin Van Ness, MSTI '2, Jerome Kagan, PhD25, Margaret Doyle Settle, RNC, MSN", and Jonathan H. Cronin, MD6 Abstract Infants in intensive care units often undergo medically necessary heel-stick procedures. Because the risks of administering analgesics and anesthetics are often thought to outweigh the benefits, there remain no proven means of ameliorating the pain and stress these infants suffer, particularly during procedures. This study examined the controlled use of recorded vocal music to attenuate physiological and behavioral responses to heel stick In 13 premature infants via an experimental design. In both instances, infants exposed to music and infants in the control group, heart rate, and respiration rate increased during the heel-stick procedure (P's — .02) and nearly all infants cried. During a 10-minute recovery following the heel stick, heart rate, and crying significantly decreased in infants exposed to music (P = .02) but not in unexposed infants. Controlled music stimulation appears to be a safe and effective way to ameliorate pain and stress in premature infants following heel sticks. Keywords NICU music, infant, heart rate, pain, heel suck, premature Introduction demonstrate behavioral habituation to auditory stimuli." Heart rate (HR) changes in response to music have been observed Ample empirical evidence indicates that music stimulates in fetuses of 28 to 38 weeks gestational age.20 Functional cognitive, emotional, and sensoritnotor processing across widely Magnetic Resonance Imaging (fMRI) studies indicate left tem- distributed brain regions.I.2 The strong physiological and emo- poral lobe activation to sound in fetuses at 33 weeks gestational tional effects ofmusic on many listeners" and the wealth ofqua- age.2' Newborns demonstrate ERP sensitivity to 10% devia- litative and quantitative findings provided by music therapy"' tions in tonal frequency.22 Approximately, 12% of the US motivate the development of standardized protocols for the use births are premature (ie, before 37 weeks)23; about half of these of music in a wide range of clinical settings. Prospective, require immediate hospital admission and many need long- randomized-controlled clinical trials are needed in order to eluci- term care. Neurological complications of premature birth date how music's effects can be harnessed to ameliorate suffering include learning disorders in as many as 2 in 5 school-age and, possibly, decrease morbidity and mortality independent of, and additive to, benefits related to therapist skills." Two recent Cochrane Database reviews have examined the use of music for I The Institute for Musk & Brain Science, Health & Medicine Program, Boston, pain relier and end-of-life care." MA, USA Neonates, especially premature infants, constitute a needy Harvard University Mind/Brain/Behavior Initiative. Cambridge. MA, USA population of patients who might benefit from the implementa- 'Departments of Neurology and of Ethnomusicology. David Geffen School of Medicine & Herb Alpert School of Music, UCLA, Westwood, CA. USA tion of standardized protocols incorporating music for analge- 4 Vanderbilt Kennedy Center. Vanderbilt University. Nashville. TN, USA sia, stress reduction, and auditory enrichment." 1'13 The human 'Department of Psychology. Harvard University. Cambridge. MA. USA cochlea is anatomically developed by 24 weeks gestational Newborn Intensive Care Unit, Massachusetts General Hospital, Boston. MA. age.14 and auditory evoked responses have been recorded in USA premature infants as early as 26 weeks gestational age. I5.16 The Corresponding Author: results ofmany recent studies of the fetus and infant are consis- Miriam Lense. Department of Psychology. Vanderbilt University. Peabody Box tent with the notion that perceptual competence develops PO. 230 Appleton Place, Nashville, TN 37203 prenatally.'" Fetuses from 27 to 35 weeks gestational age Email: mlense©postharvard.edu ecemrdre Prom • , ro uaa. try Mint pro 0, Lint N. 20 EFTA_R1_02048097 EFTA02695994 Tram° et al 73 children,24 a 2.6 relative risk for attention deficit hyperactivity with respect to music type, intensity, presentation, dose, and disorder (ADHD),25 and a significant risk of hypothalamic- dose-interval, could counter the effects on unpredictable noise pituitary dysfunction.36 Developmentally, sensitive care that in the NICU environment and promote normal auditory and incorporates noise management as well as human contact and cognitive development via exposure to the language and music other "positive" stimuli appear to improve clinical outcome of the infant's culture. and decrease costs associated with inpatient care for prema- Acutely painful stimuli typically cause increases in HR, ture infants.22-32 respiration rate, blood pressure, plasma cortisol levels, facial The acoustic environment in which many premature infants grimacing, crying and body movements, and decreases in oxy- spend their first days-to-months of life—hospital neonatal gen saturation (O2-sap:I?" These responses, which could intensive care units (NICUs) and special care units (SCUs)— reflect an internal state of stress, can be difficult to appreciate is at once impoverished and chaotic.33 Whether lying in a bas- in the most vulnerable premature infants because their imma- sinet or enclosed in a temperature-controlled isolette on ture central nervous system precludes their ability to generate mechanical-assisted ventilation, hospitalized infants arc all the components of a stress response.67•4R Even routine pro- exposed to little in the way of speech, music, and other etholo- cedures administered to hospitalized infants have been shown gically relevant sounds important for normal language and to elicit a stress response:I t" A recent study of 430 infants social development. Even worse, the sound environment is admitted to NICUs in Paris found that on average, each infant filled with unpredictable, sometimes loud acoustic stimuli received 12 painful procedures daily during their first 2 weeks (eg, alarms indicating a potentially dangerous change in a phy- in the NICU 51 In another study, 54 infants admitted to a siological measure)." The ambient sound level in an NICU can NICU over a 3-month period experienced 3283 invasive pro- reach intensities as high as 90 dB SPL, several-fold louder than cedures.52 The majority (56%) involved the "heel-stick" (aka the ambient intrauterine intensities the infant had been accus- "heel lance," "heel prick") procedure, a painful method of tomed to (50 dB SPL).35.36 Moreover, NICU sounds contain obtaining blood for serologic analyses in which the infant's high as well as low frequencies, whereas the intrauterine envi- heel is pierced with a sterile needle and squeezed repeatedly ronment only allows low-frequency sounds (less than 250 Hz) to express blood through the puncture site. The high meta- to reach the infant. Infants born before 36 weeks may be espe- bolic demands of these repeated stressors could decrease cially sensitive, and thus vulnerable, to the effects of an impo- energy stores available for growth. Moreover, adrenocortical verished, chaotic auditory environment because their auditory responses to repeated stressful stimuli might weaken the discrimination capabilities are immamrc,37 and they remain infant's immune system and increase the risk of illness."'" unable to visually identify the sources of sound and have lim- Grunau53 has hypothesized that infants who receive frequent ited exposure to faces and visual scenes in general. The infant's medical interventions without "positive" or soothing stimuli heightened auditory sensitivity requires that physicians and may develop a low pain threshold or become hypersensitive to nurses determine the type, dose, and dose interval of acoustic touch. Recent evidence shows that 3- to 18-month preterm stimuli empirically. What the optimal auditory conditions are, babies have abnormal basal cortisol levelsS6 and that 4- and how they could be provided in the NICU environment, month-old infants have abnormal cortisol responses during remain unknown. The American Academy of Pediatrics" and pain associated with immunizations.S7 Abnormal cortisol lev- the National Association ofNeonatal Nurses" have proposed a els may be one mechanism by which early pain exposure number of procedural and technical strategies to reduce ambi- could compromise brain development.55 This, in turn, could ent noise. One study of 30 premature infants showed that wear- contribute to learning, attentional, and behavioral problems ing earmuffs significantly increased quiet sleep timeo later in childhood." Another study of 24 very low birthweight neonates found that Although infants undergo many painful procedures and may wearing silicone earplugs significantly increased weight perceive pain more acutely than do adults, pain management gain:" However, there remains the possibility that quiet is sub- for this population is less than optimal 51.594' There is wide optimal because stimulation with music or other natural sounds variation in the use of pharmacological analgesics in N1CUs,62 would promote development while avoiding the potential dele- which increase fluid retention and bilirubin levels and routinely terious consequences of decreasing auditory and multimodal raise concerns about CNS depressant effects, including respira- stimulation (for reviews see Philbin42 and Aucott et al43). Sev- tory depression. Standardized protocols for nonpharmacologi- eral researchers have explored the potential benefits ofauditory cal analgesia are lacking, and NICU personnel may not be stimulation with music in the NICU environment."'" In their adequately trained in pain assessment, management, and pre- recent review, Hartling et a15 found that the researchers have vention.63 A better understanding of nonpharmacological treat- used a variety of musical types (eg, vocal vs instrumental, folk ments is needed to advance the development of protocols to vs classical), presentation methods (recorded vs live), and alleviate pain and stress without the risks of potential medica- acoustic environments (eg, music alone vs with intrauterine tion side efects.33 As a noninvasive, analgesic, and anxiolytic sounds) in the NICU.5 Methodological consideration, as the intervention, controlled auditory stimulation with music may authors point out, preclude a straight forward interpretation provide a treatment with a high benefit:risk ratio. The present ofhow the type ofmusic and its presentation affect physiologi- study tests the hypothesis that music attenuates physiological cal and behavioral responses. Auditory stimulation, controlled and behavioral responses to heel stick. inewooded 7.16-1,1t. by Mae. 'WU, MM, n 20" EFTA_R1_02048098 EFTA02695995 74 Music and Medicine 3(2) Table I. Age, Sex. Weight, and Apgar Scores of Participants Infant Sex Age (days) Birth weight (g) GA at birth (weeks: days) al Apgar 5 m Apgar IC Male 28 1200 30:6 2C Male 28 1260 30:6 7 8 3C Male 4 1790 340 8 9 5C Male 4 2305 34:3 8 9 6C Female 1780 34:0 8 9 7C Female 1I 1900 31:5 7 7 IT Male 16 1960 32:4 5 9 2T Male 35 1260 30:6 7 8 3T Female 4 2195 34:0 8 9 41 Female 4 2160 34:0 7 8 5T Female 4 2175 34:3 4 9 61 Male 2600 34:0 9 9 7T Male I0 1800 32:4 8 8 Abbreviations: C. control: T. treatment Methods was 7.5 days for the control group and 4 days for the treatment group. All infants had 5-minute Apgar scores of 7 or greater. The study protocol was approved by the Institutional Review There were no significant differences between groups for any Boards of the hospital and university where the study was of these background variables. The parents of all infants were conducted. English-speaking. Participants Special Care Unit Environment and Routine Care All participants were premature infants in the hospital SCU. All All infants were admitted to the SCU under the care of an admission logs and medical charts were reviewed soon after patients were admitted to the unit. We identified infants who attending perinatologist, pediatrician, or nurse practitioner. met the following selection criteria for inclusion: (I) gesta- Infants were housed in closed or open isolettes. The standard tional age less than 36 weeks and birth weight no more than SCU protocol called for serologic testing every Sunday night 2600 g, (2) not on a ventilator or receiving oxygen inhalation or Monday morning. Heel sticks typically occurred between therapy for respiratory illness, and (3) no neurological disease. the hours of 9 and 11 PM or between 4 and 6 am, just before Individual cases were reviewed with nurses caring for each night time or morning feeding, respectively. Standard infant. Parents/guardians received a recruitment letter detailing approaches to developmentally sensitive care were implemen- the study; they were given one week to decide about participa- ted throughout the infants' SCU stays (eg, swaddling, covering tion and were offered a video of their infant on DVD as a incubators to limit bright light exposure, limiting loud reward for participation at the conclusion of the study. Par- conversation). ents/guardians and nurses were instructed not to play music We measured the ambient sound level in the SCU on multi- to the infant from the time of enrollment to the experimental ple occasions. A Quest Technologies Impulse Sound Level procedure. Meter Model 2700 was held at the head of an empty, open isol- Written consent was obtained from the parents of 14 infants. ette. During daytime hours, the ambient sound level was Participants were pseudorandomized into the treatment group approximately 62 dBA; at night, 56 dBA. (N = 7 [4 males]) and control group (N =7 [5 males]) irrespec- tive of sex and ethnicity. In total, 6 participants were from 3 fraternal twin pairs and 2 participants were from fraternal Auditory Stimulation triplets; in these cases, one sibling was assigned to the treat- We listened to several commercially available CDs of lullabies ment group and the other to the control group. Data for one sung by females in English. We selected one (SRT Music male infant in the Control Group were excluded from the anal- Groups') in which the lullabies were performed with simple ysis because of a protocol violation: a parent played recorded accompaniment at moderate tempo. music to him several hours a day. Each infant was tested indi- For each patient, the total music stimulation time was vidually in her/his isolette. 10 minutes. We avoided starting the music shortly before and The age, sex, birthweight, gestational age, and Apgar during the heel stick because we did not want the patient to asso- scores of each participant are listed in Table I. The median ciate music with the painful stimulus. We used a 10-minute gestational age for the control group was 33 weeks, one day; window of observation because this provided a sufficient time for the Treatment Group, the median gestational age was window for normalization or near-nonnalization of pain- 34 weeks. The median postdelivery age at the time of study induced changes in outcome variables. 000M040•0 Own nnvi tatrp.b by IAA Tivamo as MO* 2S. EFTA_R1 _02 048099 EFTA02695996 Tramo et al 75 Recordings of 3 complete songs and part of a 4th were pre- sample points and 2 seconds sample points for each dependent sented in the 10-minute music stimulation window: (I) "Row variable. Row Row Your Boat" (duration = 3 minutes, 9 seconds), (2) "Baa Baa Black Sheep" (3:00), (3) "Are You Sleeping" Behavioral Responses (2:23). and (4) the first 1:28 of "Rock a Bye Baby" (original song length of 3:06). Each song began with a short instrumental Behavioral responses were recorded before, during, and after introduction (range = 6-26 seconds) followed by a sung mel- the heel-stick procedure with a Samsung SCD-23 digital video ody whose pitches ranged from E3 to C59 on the equal- camera. At the start of the procedure, the camera was mounted tempered scale (fundamental frequencies = 164.8-554.4)Iz, on a tripod at the foot of the isolette; after the nurse completed A4 = 440 Hz) and a moderate tempo of 84 to 88 beats per min- the heel-stick procedure, the camera was moved closer to the ute in 4/4 or 3/4 meter. Each recording contained only 3 to infant alongside of the isolette. Digital videos were converted 4 instrumental voices with timbres varying among electric to QuickTime movie files for offline analysis. We initially piano, electric organ, glockenspiel, and synthetic sounds. We aimed to code whether or not each of the following behaviors chose traditional Western lullabies sung in English by a female occurred before, during, and after the heel-stick procedure: because we thought they would have the highest probability of (I) eye-opening, (2) head movements, and (3) crying. How- achieving a beneficial effect for our Western, English-speaking ever, we were unable to reliably code eye-opening and head population, and because a female voice (ie, mother's voice) is movements. Behavioral data could not be collected during the one most frequently heard prenatally in the womb. More- blood collection for 3 babies because the nurse blocked the over, lullabies include both music and speech sounds, have camera's view; in general, reswaddling of infants following cross-cultural significance in parent-infant communication, heel stick compromised the observations of changes in beha- and have been shown to improve longer-term endpoints, such vior. Behavioral data from one infant was lost due to equipment as weight gain in hospitalized infants.65 malfunction. Each CD track was converted to a monophonic mp3 file and uploaded onto an Apple iPod. Stimuli were presented at Experimental Procedure an intensity of approximately 70 dBA using one JBL Duet speaker placed in the sagittal midline at the foot of the infant's Figure I depicts the timeline of the heel-stick procedure, audi- isolette, approximately 50 cm from her/his head, outside the tory stimulation (for the treatment group), and data collection. field of the heel-stick procedure. The advantages of using a sin- The heel-stick procedure was performed by Registered Nurses gle speaker playing a monophonic recording (less space, fewer caring for the SCU patients. Stimuli were presented and data wires) outweighed the advantages of using 2 speakers playing a collected by I of the 2 investigators (ML or CV). who were stereo recording, in our opinion, because we did not hypothe- trained in experimental psychology, acoustic calibration, and size that this difference in the spatial mix of the music would music. influence the results. The iPod was placed on a small dock at First, with the infant at rest, undisturbed in her/his isolette, the bedside. we recorded baseline HR, RR, and O2-sat data and started the video. Second, during the prepuncturc handling period, a nurse prepared the infant's heel with a warm pad followed by an alco- Physiological Responses hol swab. There were differences among nurses with respect to For each infant, HR, respiratory rate (RR), and O2-sat were preparation routine and pre and post heel-stick swaddling. In continuously monitored before, during, and after the heel- all. 2 infants in the control group and 2 in the treatment group stick procedure using a GE Medical Clinical Information were swaddled at baseline and remained so throughout the Center Pro system. During 1 to 2 minutes before the heel- heel-stick procedure and recovery period. One infant in the stick procedure, throughout the procedure. and during the first control group and 2 in the treatment group were not swaddled 4 minutes postprocedure, a trained observer (ML or CV) at baseline and remained unswaddled throughout the heel-stick recorded at least 4 measurements of each of the 3 dependent procedure and recovery period. One infant in the Control variables per minute. At the beginning of the study, data were Group was swaddled at baseline, unswaddled during handling recorded every IS seconds online in real time by reading the and blood collection, and remained unswaddled during the output of the HR, RR, and Orsat monitors. About halfway recovery period. One infant in the control group was not through the study, we were able to analyze the output of the swaddled at baseline, remained unswaddled during handling monitors offline, which recorded data every 2 seconds, after and blood collection, and then swaddled during postpuncture data collection was finished. Finally, during the last half of the handling and the recovery period. A total of 3 infants in the post heel stick epoch, data were sampled at 3 points: 5, 7, and treatment group were swaddled at baseline, unswaddled during 10 minutes postprocedure. For the purpose of population data handling and blood collection, and reswaddled during post- analyses, data collected preprocedure, during the procedure, puncture handling and the recovery period. Four infants in the and ≤4 minutes postprocedure were calculated using a bin control group and 4 in the treatment group were given a pacifier width of 15 seconds. We checked that there were no significant during prepuncturc handling; 2 of the pacifiers given to the differences between the means calculated from 15 seconds control group and all 4 given to the treatment were sweetened 000M00000 Mom vro :cut. •.• by M0t limo 4:a~ 05. 2011 E FTA_Ri _02048100 EFTA02695997 76 Music and Medicine 3(2) BASELINE HANDLING BLOOD COLLECTION L HANDLING RECOVERY infant at rest Iscel.mek prcparation: erase punctures skin of heel lied lance instruments removed infant undisturbed and in isolate: clan unswaddliag, IrJ warmer, with blade, squeezes heel to from isolate. tem.:addling at nal in InOICHC data collection alcohol swabbing express blood min I -I2 min 2.15 Min 20s-2min In min Intedian 655) (median62s) (median 270s) (median 43s) Wormer Ike lied is AI sic begins Music ends in place puncture bandaged eatment (Ttratmein ;tour) (Troup) F sure I. Experimental procedure. with sucrose. Blood collection began with the nurse puncturing There was no significant change in O2-sat from baseline the skin of the prepped heel using a sterile, spring-loaded blade; to blood collection (WSRT, Z = -0.14, P = .89). The nadir squeezing of the heel to collect blood into a tube followed. Fol- 02-sat fell below 90% for only 3 infants. There was no lowing blood collection and subsequent handling, including significant change in 02-sat CV (WSRT, Z= —1.07. P = .29). bandaging of the puncture site, the infants in the treatment Figure 3A-C illustrates HR, RR, and 02-sat population group were stimulated with music for 10 minutes. data, respectively, collected during blood collection and the 10-minute recovery period (Figure 3). In the treatment group, Statistical Analysis there was a significant decrease in HR across the 2 epochs (WSRT Z = —2.37, P = .02). On average, HR decreased After collecting data from 13 infants over 8 months, we examined 17%. In 6 of the 7 infants (86%) in the treatment group, HR our data using nonparametric statistics (Wilcoxon Signed-Rank decreased by 10 bpm or more. In the control group, there was Test) to compare the physiological-dependent variables before, no significant change in HR (WSRT Z = —1.15, P = .25), aver- during, and following the heel-stick procedure. Persistence or age HR decreased only 6%, and only 3 of the 6 infants (50%) cessation of crying from the heel-slick to the recovery period showed an HR decrease of 10 bpm or more. No significant was compared between the treatment and control groups using change in HR variability during recovery versus blood collection Pearson x 2 test. Given that we had tested our working hypothesis was found for either the treatment group (WRST Z = -0.68, at this juncture. we ceased enrollment of additional infants. P = .50) or the control group (WSRT Z= —1.07,P =.29). There was no significant change in RR or RR CV from blood collection Results to recovery in either the treatment group (respectively, WSRT 2 = -0.34. P =.74; 2 = -0.08, P = .93) or the control group Physiological Results (Z = —1.36, P = .17; = —0.94. P = .35). Figure 2A-C illustrates HR, RR. and O2-sat population data, There was no significant change in 02-sat or 02-sat CV respectively, collected before and during heel stick and blood from blood collection to recovery in either the treatment group collection. There was a significant increase in HR from base- (respectively, WSRT Z = -0.17, P =.86; Z = -0.81, P = .42) line to blood collection (Wilcoxon Signed-Rank Test [WSRT], or the control group (Z = —0.21, P = .83; Z = —0.37, P = .71). Z = —2.36, P = .02); on average. HR increased 19%. All To further depict the time course of HR, RR, and 02-sat data infants showed an HR increase of at least 5 beats per minute across the various epochs of data collection, we present the (bpm); in 8 (62%), HR went above normal limits (>160 bpm). results from a single premature infant in the treatment group There was no significant increase in HR coefficient of variation (Figure 4). These are representative ofdata collected: (I)before (CV) across the 2 epochs (WSRT. Z = —0.53, P = .60). and during heel stick and blood collection for the entire study There was a significant increase in RR from baseline to population of 13 infants and (2) after blood collection for the blood collection (WSRT, Z = —2.24, P = .02]; on average, population of 7 infants in the treatment group, who received RR increased 39%. All infants showed an increase of at least controlled auditory stimulation with vocal music during the 5 inspirations per minute (ipm); in 10 (77%), RR went above 10-minute recovery period. This patient was a 2.6-kg male twin 40 ipm. There was also a significant increase in RR CV (WSRT born at 34 weeks gestation with 1-min and 5-min Apgar scores Z = —2.37, P =.02). of 9. He was admitted to the SCU after delivery with a OsonlosOnd Own rrnd 'tr byInn Tio on Won a5.2011 EFTA_R1_02048101 EFTA02695998 Teraina et al 77 A. 180 170 160 150 140 130 120 Pre-handkng Handling Blood Collection Post.sock handling Epoch 8. 60 50 40 30. 20. 10 0 Pra.handling Handling Blood Collection Post-stick handling Epoch C. 100 98 96 94 Pre-handling Handling Blood Coleman Posl-slick handing Epoch Figure 2. Populadon data collected before and during heel-stick procedure. Error bars represent ± I standard error from the mean. diagnosis of prematurity. The heel-stick procedure was per- began to decline after blood collection when the heel was formed during postnatal day one to check bilirubin levels. Ten bandaged. The heel remained unswaddled throughout the minutes before skin puncture, his foot was unswaddled and postpuncture-handling period and recovery period. During prepped with a warming pad. In the minutes before skin punc- the 10 minutes of auditory stimulation with vocal music, HR ture, the patient was lying quietly with his eyes closed and no continued to decline until approximately 100 seconds post- head movements; the HR ranged from 118 to 132 bpm (mean = handling, when it reached a plateau of 125 bpm, near the base- 121 bpm; CV = 17%). During the 33 seconds ofhandling prior line mean of 121 bpm. He remained quiet with his eyes closed. to skin puncture. HR rose to 141 bpm; he remained quiet and and he made a total of only 5 brief head movements. still with eyes closed. Immediately upon skin puncture, he Changes in RR paralleled those of HR (Figure 413). At base- began to cry: by 10 seconds postpuncture, Hit was 153 bpm, line, mean RR was 24.5 ipm with a CV of 4.5%. During pre- by 30 seconds it was above the normal limit of 160 bpm puncture handling, RR rose slightly, but immediately after (161 bpm), and by one minute it was 178 bpm (47% above the skin puncture, when the infant began to cry, RR rose precipi- baseline mean). During blood collection, mean HR was tously and became highly variable. During blood collection, 175 bpm, and the CV rose to approximately 5 times what it was RR peaked at 83 ipm approximately 100 seconds postpuncture, at baseline; the peak HR was 192 bpm (59% above baseline) at and mean RR rose to 46.3 ipm, 89% above baseline, with a 31% 150 seconds postpuncture. The infant cried for more than increase in CV. Within 100 seconds after initiation of auditory 3 minutes, until he was given a pacifier; his eyes remained stimulation, RR decreased, though its mean and variance closed, and no head movements were discernable. Heart rate remained elevated relative to baseline. 00iniaidal Prom inn :cat, by WV% Vino <a IWO A. 2011 EFTA_R1_02048102 EFTA02695999 78 Music and Medicine 3(2) recovery period, whereas 2 of the 4 infants in the Control &180 ■ Control Group continued to cry during the recovery period. There was ■ Treatment 170 a trend for a significant x2-test comparing cessation vs continuation of crying in the Treatment vs Control group ire (x2 = 2.667, P = .10) though the small sample size renders the significance test questionable. We were unable to reliably code iso eye opening and head movements owing to logistical problems (ie, view blocked by nurse, variability among nurses in swad- 1140 dling and pacifier use [as described previously)). 130 Discussion 120 Blood Collection Recovery The bed-stick procedure is routinely used to obtain blood for serologic analyses in small babies who lack peripheral venous B. 70 ■contra access. The results from our study population of 13 premature • Treatment infants demonstrate that the procedure precipitates sudden increases in HR, RR, and crying that peak within seconds and are sustained for several minutes after delivery of the acutely painful stimulus. We tested our working hypothesis that controlled audi- tory stimulation with vocal music attenuates physiological and behavioral signs of stress evoked by heel stick. We chose traditional Western lullabies as our auditory stimulus because we thought they might have the highest prob- ability of achieving a beneficial effect. Lullabies include both music and human vocal sounds, including words, and are Blood Coledlon OCOVe ethologically and ethnologically relevant owing to their rich cross-cultural history in parent-infant communication. In the previous clinical studies, lullabies have been shown to acceler- ate weight gain in hospitalized infants." The results show a significant decrease in mean HR over the 10-minute postprocedure period in the treatment group (17% mean HR decrease) but not the control group (6% mean HR decrease). There were no effects of vocal music on RR or O2-sat. Qualitative differences between the treatment and con- trol groups were also observed for crying. We found no signif- icant difference between the treatment and the control groups with respect to procedure length, so it is unlikely that it contrib- uted to the observed effects on HR and behavior. Blood Collection Recovery Several limitations of our study may hamper its applicability to the general population of premature infants undergoing heel Figure 3. Population data collected during blood collection and stick. Our population size was small. We studied the effects of recovery. music on a single heel stick; our study does not address the potential benefits of repeated music stimulation for the patients Oxygen saturation data for this infant are shown in in NICU undergoing frequent heel sticks. There was variability Figure 4C. Immediately after skin puncture and initiation of across the clinicians with respect to swaddling, pacifier use, crying, O2-sat declined and its variance increased. With audi- and sucrose use. Our results may have been influenced by the tory stimulation, O2-sat immediately began to rise and its var- fact that a higher proportion of infants in the treatment group iance decreased. Oxygen saturation reached a plateau of (57% vs 33% in the control group) received sucrose before and approximately 98.5% within 100 seconds. during the heel-stick procedure. However, the response to the heel stick was equally robust in both groups, and HR increased as much in the treatment group as in the control group during Behavioral Results the heel-stick procedure. Our behavioral analyses were limited Behavioral data could be reliably recorded in 9 infants: 4 in the to crying due to logistical difficulties with video recording in Treatment Group, 5 in the Control Group. During the heel-stick the SCU. Future studies would benefit from using multiple procedure, 8 out of the 9 cried. Crying ceased in all 4 infants cameras from different angles or a mobile camera to improve who received auditory stimulation with vocal music during the the behavioral data collection. With only one sm