Complementary, Holistic, and Integrative Medicine: Music for Procedural Pain

Although potentially important for a patient’s well-being, many medical procedures, including routine immunizations, are painful. Insufficient pain control during procedures can produce long-term negative effects, especially in the very young. (1) Pharmacotherapy is used commonly to treat pediatric pain; however, because of concerns about adverse effects, (2,3,4) many health care professionals and patients are interested in using nonpharmacologic options as well as or in place of medications. One potential option is the use of music or music therapy during a painful procedure.The formal definition of music therapy is “the skillful use of music and musical elements by an accredited music therapist to promote, maintain, and restore mental, physical, emotional, and spiritual health.” (5) Pediatric research has investigated the effects of both music therapy, in which a music therapist administers the intervention (active music therapy), and simply playing prerecorded music for the participant (passive music therapy). This review summarizes the evidence from randomized controlled trials (RCTs) of the use of music to reduce procedural pain in pediatric patients. Because pain is closely related to anxiety and distress, these measures are included when provided. We will look first at older children and then at neonates subject to procedural pain.The use of music for alleviating pain and anxiety in children (excluding neonates younger than 1 month) was summarized in a 2008 systematic review of 19 RCTs (5 trials of active and 14 trials of passive music therapy) that included a total of 1513 participants, ages 8 months to 20 years, who were undergoing procedures, including oral surgery, venipuncture, elective surgery, intramuscular injections, magnetic resonance imaging, general anesthesia, colposcopy, bone marrow transplantation, and dressing change in burn patients. (6) Outcomes included pain alone (8 trials, n=882), anxiety alone (6 trials, n=324), or both pain and anxiety (5 trials, n=279).A meta-analysis of the RCTs that measured pain demonstrated a statistically significant benefit of music therapy in general over controls (5 trials with the combined approach [n=465]; standardized mean difference [SMD], −0.39; 95% CI, −0.66 to −0.11; P=.005). A meta-analysis of the 5 RCTS for relief of anxiety also favored music therapy (5 trials with the combined approach [n=284]; SMD, −0.39; 95% CI, −0.76 to −0.03; P=.03).A subgroup analysis comparing active and passive music therapy found that although passive music therapy had a significant effect, active therapy did not; however, there were only 2 active trials included in the analysis. As noted by the authors, the quality of the studies was generally poor, primarily because of a lack of allocation concealment, which is a method used to minimize bias in how participants are assigned to study groups and involves masking of outcome assessors as to whether active or passive music therapy was used. In addition, heterogeneity among studies was moderate to substantial. Despite these shortcomings, there was evidence that supported the use of music to reduce pain and anxiety.Effect sizes that were found were considered small to moderate. Clinical implications were positive, as demonstrated by increased patient compliance and decreased sedative use. Because of the clinical heterogeneity among studies, the authors could not specify for which procedures music therapy will be the most effective.Since the publication of this review, we have identified 4 additional RCTs of music used to ease pediatric procedural pain. These trials analyzed the effect of music on pain during various procedures, including venipuncture (2 trials), lumbar puncture (1 trial), and intramuscular injection (1 trial); all confirmed the benefits noted in earlier trials.In the first venipuncture study, 108 patients ages 4 to 13 years were assigned to either an active music group (n=54) or a control group (n=54) during venipuncture performed without anesthetic cream. (7) In the music group, a musician met the parents and the child in the waiting room, where he sang, played music, and interacted with the child. With the permission of the child and parents, he went to the procedure room and continued to sing and play music. The musician stayed with the child and parents for 20 minutes after the procedure. In the control group only parents were present.Both distress and pain were significantly lower in the music group (P<.001 and P=.05, respectively) compared with the control group. Distress scores (Observational Scale of Behavioral Distress, lower score equals less distress) were 9.1 (SD, 8.3), 13.5 (SD, 9.7), and 8.8 (SD, 7.4) for the control group before, during and after venipuncture, respectively, compared with the music group scores of 4.9 (SD, 5.9), 6.8 (SD, 7.5), and 4.5 (SD, 5.0). Pain scores were not reported. The authors report that because the outcome assessors were not masked to the intervention (they could hear whether music was playing or not), this factor could have introduced some bias into the results.In the second venipuncture study, 150 patients ages 5 to 12 years were randomly assigned to a music group, local anesthetic group (lidocaine-prilocaine [EMLA] cream), or control group (n= 50 in each group). (8) All the children wore headphones for 15 minutes before the procedure and for 5 minutes after. The music group listened to classical Indian music, whereas the other 2 groups did not listen to anything. Application of a cream to the puncture site was performed 45 minutes before the procedure. The control and music group received petroleum jelly, whereas the local anesthetic group received EMLA cream. A guardian of the child, the investigator, and an observer each independently recorded the perceived pain of the child using a visual analog scale (VAS) at 0 minutes (time of puncture), 1 minute, and 5 minutes. Where possible, the child also recorded his or her own pain scores.All observers reported the median pain scores to be higher at all time points in the control group than in the music group and the local anesthetic group. Pain scores were lower in the local anesthetic pain group than in the music group; however, this difference was statistically significant at all time points only when scored by the investigator, at the 1-minute time point when scored by the parent, and at the 5-minute time point when scored by the independent observer. This study did not measure pain scores before the intervention, so it is unclear whether baseline pain levels differed among groups and what effect this potential difference may have had on the results.In a lumbar puncture study conducted in Vietnam, 40 children with leukemia, ages 7 to 12 years, who had previously experienced at least one lumbar puncture were randomly assigned to either a music group or a control group (n=20 in both). (9) The children in the music group chose the music they wanted to hear during the procedure from an available selection. Headphones were placed on the child 10 minutes before the procedure; in the music group, music was played, whereas the control group just wore headphones. Pain (self-reported on the Numeric Rating Scale), heart rate, blood pressure, oxygen saturation, and respiratory rate were recorded before, during, and after the procedure. A Spielberger state-trait anxiety inventory was completed before and after the procedure. Ten children from each group were interviewed after the procedure. No local anesthetics or other analgesics were administered during the procedure, a reflection of the standard care offered to children in Vietnam.Pain and anxiety scores were significantly lower for the music group than the control group during and after the procedure (P<.01). Pain scores (Numeric Rating Scale, lower score equals less pain) for the control group were 1.75 (SD, 1.77), 5.65 (SD, 2.5), and 3 (SD, 2.0) before, during, and after the procedure, respectively, compared with 1.2 (SD, 1.4), 2.35 (SD, 1.9), and 1.2 (SD, 1.4) in the music group. Anxiety was lower before and after the procedure in the music group compared with the control group (P<.001).Anxiety scores (Spielberger state-trait anxiety inventory short form, lower score equals less anxiety) for the control group were 13.25 (SD, 3.7) and 13.0 (SD, 4.2) before and after the procedure, respectively, compared with 8.6 (SD, 2.8) and 8.1 (SD, 2.2) in the music group. The music group also had significantly lower heart rate readings during the procedure (P<.05) and lower respiratory rates during and after the procedure (P<.01). No significant difference was noted for blood pressure or oxygen saturation. All patients preferred for the child to listen to music during the next such procedure.Another recent study looked at musical distraction for vaccination pain in 13- to 15-year-olds receiving an intramuscular polio vaccination. (10) The 118 students were randomly assigned to 1 of 3 groups: music with headphones (n=38), music without headphones (n=41), and no music (n=39). Pain was measured using a VAS. The group that listened to music without headphones reported lower VAS pain scores compared with the other 2 groups, but the differences were not statistically significant. Using a dichotomous pain scale, those who listened to music (with or without headphones) were significantly less likely to report pain compared with those without music (odds ratio, 2.8; P=.02); however, use of headphones appeared to be less effective in reducing pain than not using headphones (odds ratio, 3.6; P=.01). This result was unexpected, and the study authors speculate that using headphones may have been too distracting or isolating for the patients and may have limited the patient’s feeling in control of the immunization or in interacting with the nurse administering the immunization. The authors comment on the lack of masking in this study and speculate that this factor may have introduced some bias (ie, the nurses were not masked).We describe the results of 4 RCTs that explored the effects of music on perceived pain in neonates during a medical procedure. Two of the studies investigated pain during circumcision, whereas the other 2 investigated pain during heel prick.In the first study, 58 neonates were randomly assigned to 1 of 3 groups. (11) Group 1 was played a classical music tape (n=25), group 2 was played a tape of intrauterine sounds (n=15), and group 3 was a usual care group (n=18). None of the neonates received anesthetic. Heart rate, heart rhythm, blood pressure, transcutaneous oxygen (tcPo2) level, alertness (Brazelton Neonatal Assessment Scale), and facial expressions (Maximally Discriminative Facial Movement Coding System) were measured to determine pain levels in the infants. Means and SDs were calculated for all the variables and compared with reference ranges.Neither of the interventions resulted in significant differences in the measurements, and the authors concluded that music and intrauterine sounds do not seem to be effective but that more testing should be performed. The sample size in this study may simply have been too small to detect a treatment effect, even if one had been present.In a second study by the same authors, 121 infants age 2 to 9 days were randomized to 5 intervention groups (n=20 in each group) and 1 control group, which received usual care (no analgesia was provided) (n=21). (12) For the intervention groups, usual care was supplemented with a classical music tape, tape of intrauterine sounds, pacifier, classical music and pacifier combination, or intrauterine sounds and pacifier combination. Infants received intervention as soon as they were put on the restraining board, and the intervention was removed as soon as they were taken off the restraining board. Heart rate, heart rhythm, blood pressure, and tcPo2 were measured during each circumcision step.Few significant differences were noted among groups. The most significant difference was that tcPo2 levels were higher during some circumcision steps in the intrauterine sounds and pacifier combination group. This effect was attributed to the shorter crying time in this group compared with the other groups. The authors concluded that music and intrauterine sounds do not seem to lessen the amount of pain felt during circumcision.In the first RCT, 27 neonates undergoing heel prick were randomized to 3 intervention groups and 1 control group. (13) The interventions (nonnutritive sucking [NNS], music, or a combination of the 2) were given for 5 minutes immediately after the heel prick. Baseline measures were taken 1 minute before the procedure, and readings were taken every minute during the intervention and for an additional 8 minutes after the interventions were removed. A standard pacifier was used in the NNS group, whereas a tape of intrauterine sounds and soothing music was played for the music group; both were given to the combination group. Heart rate, tcPo2, and pain behavior (Neonatal Infant Pain Scale) were measured.Results indicated that tcPo2 levels were increased the most and that pain behavior was decreased the most in the combination group. Although statistical significance was not achieved at all time points, the authors concluded that music alone had the largest overall effect on heart rate, whereas combined music and NNS had the largest overall effect on tcPo2 and pain level. This study was designed to detect a moderate effect size and may have been unable to detect smaller changes. An additional limitation is that masking was not part of this study.In the second RCT, 60 preterm neonates were randomized to 3 groups (n=20 each). (14) Group 1 was given only a pacifier, group 2 was given a pacifier that activated music when the infant sucked on it, and group 3 received no intervention. Each intervention was started 3 minutes before and continued 3 minutes after the heel stick. Heart rate, respiratory rate, blood oxygen saturation, behavior state, and stress level was recorded before, during, and after the procedure.Although most of the physiologic data were inconclusive, significant differences were observed in behavior state and stress levels (Assessment of Preterm Infant’s Behavior using Continuous Response Digital Interfaces) for both the pacifier and combination groups. The combination group, in comparison with the control group, had the greatest differences. For example, during the procedure, behavior state scores (mean and SD) for the pacifier and combination groups were significantly lower (better) at 128.5 (SD, 29.0) and 127.4 (SD, 32.9), respectively, compared with the control group score of 176.2 (SD, 40.1) (P<.05). Similarly, stress scores during the procedure were lower, indicating less stress, at 59.4 (SD, 47.8) and 50.0 (SD, 46.9), respectively, compared with 125.1 (SD, 50.6) (P<.05).

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.