Psychological Interventions in Managing Postoperative Pain in Children: a Systematic Review Fiona Davidsona,*, Stephanie Snowa,B, Jill A

Comprehensive Review

Psychological interventions in managing postoperative pain in children: a systematic review Fiona Davidsona,*, Stephanie Snowa,b, Jill A. Haydenc, Jill Chorneya,b,d

Abstract Pediatric surgeries are common and painful for children. Postoperative pain is commonly managed with analgesics; however, pain is often still problematic. Despite evidence for psychological interventions for procedural pain, there is currently no evidence synthesis for psychological interventions in managing postoperative pain in children. The purpose of this review was to assess the efficacy of psychological interventions for postoperative pain in youth. Psychological interventions included Preparation/education, distraction/ imagery, and mixed. Four databases (PsycINFO, PubMed, EMBASE, and Certified Index to Nursing and Allied Health Literature) were searched to July 2015 for published articles and dissertations. We screened 1401 citations and included 20 studies of youth aged 2 to 18 years undergoing surgery. Two reviewers independently screened articles, extracted data, and assessed risk of bias. Standardized mean differences (SMDs) and 95% confidence intervals (CIs) were calculated using RevMan 5.3. Fourteen studies (1096 participants) were included in meta-analyses. Primary outcome was pain intensity (0-10 metric). Results indicated that psychological interventions as a whole were effective in reducing children’s self-reported pain in the short term (SMD 520.47, 95% CI 520.76 to 20.18). Subgroup analysis indicated that distraction/imagery interventions were effective in reducing self-reported pain in the short term (24 hours, SMD 520.63, 95% CI 521.04 to 20.23), whereas preparation/education interventions were not effective (SMD 520.27, 95% CI 520.61 to 0.08). Data on the effects of interventions on longer term pain outcomes were limited. Psychological interventions may be effective in reducing short-term postoperative pain intensity in children, as well as longer term pain and other outcomes (eg, adverse events) require further study. Keywords: Pain, Postoperative pain, Psychological interventions, Children, Adolescents

1. Introduction potentially respiratory depression,31 thereby limiting safe and In 2009, there were over 200,000 inpatient surgeries in children effective doses. Thus, whereas analgesics are an important part and adolescents in the United States.38 Surgeries are often of postoperative pain management, children continue to expe- painful and anxiety provoking and can cause significant distress in rience significant pain and there is a need for additional pain children before and after the procedure. Implementing appropri- intervention. ate pain management interventions is a key factor in improving Pain is a subjective experience that is influenced by psycho- 10 patient care, yet research shows that pediatric pain is generally logical factors. Psychological interventions, such as cognitive– not well managed in hospital44 or at home.50 Postoperative pain behavioural therapy techniques, have been shown to be effective can negatively impact children in a variety of ways, including in reducing pain intensity and distress in medical procedures 43,49 difficulties with eating and fluid intake, sleep disturbances, and other than surgery (eg, venipuncture and immunizations). For changes in behaviour.12 example, cognitive–behavioural therapy interventions including Postoperative pain is most commonly treated with analgesics, distraction, hypnosis, preparation, and behavioural relaxation but even when maximally treated with medications, pain is often strategies have been found to be effective for managing needle- still problematic.45 Analgesics, particularly opioids, can also have related pain and anxiety in children and adolescents.49 side effects, such as nausea, vomiting, constipation,21 and The mechanism by which psychological interventions are proposed to operate may differ depending on the technique Sponsorships or competing interests that may be relevant to content are disclosed used. For example, distraction redirects an individual’s attention at the end of this article. to an alternative stimulus, thus interfering with the processing of 7,8 a Department of Psychology and Neuroscience, Dalhousie University, Halifax, NS, pain. In a related process, techniques involving imagery, Canada, b Centre for Pediatric Pain Research, IWK Health Centre, Departments of hypnosis, relaxation, and suggestion have been proposed to c d Community Health and Epidemiology and, Anesthesia, Pain Management and function by allowing attention to become highly focused, which Perioperative Medicine, Dalhousie University, Halifax, NS, Canada allows for alterations in awareness of sensations including pain.42 *Corresponding author. Address: Department of Psychology and Neuroscience, The mechanism by which provision of pain-specific information Dalhousie University, 1355 Oxford St, Halifax, NS B3H 4J1, Canada. Tel.: 902-401- 3757. E-mail address: [email protected] (F. Davidson). aids in pain management is hypothesized to be rooted in self- Supplemental digital content is available for this article. Direct URL citations appear regulation theory, whereby patients use past experience and 19 in the printed text and are provided in the HTML and PDF versions of this article on available information to regulate themselves. By providing the journal’s Web site (www.painjournalonline.com). specific information about what to expect regarding pain, it is PAIN 157 (2016) 1872–1886 hypothesized that patients are able to habituate to the thoughts © 2016 International Association for the Study of Pain and images associated with the painful event, and thus have 19 http://dx.doi.org/10.1097/j.pain.0000000000000636 better response.

1872 F. Davidson et al.·157 (2016) 1872–1886 PAIN®

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited. September 2016· Volume 157· Number 9 www.painjournalonline.com 1873

Given the evidence for psychological interventions in reducing (5) mixed. Operational definitions of intervention type are pain intensity and distress for procedural pain, it is expected that provided below. Of note, there were no interventions that met these interventions will also be effective in managing post- the criteria for standalone progressive muscle relaxation, deep operative pain intensity. Whereas there are several empirical breathing, and/or biofeedback; therefore, these are not in- studies available on the topic, there are inconsistencies in the cluded in the results section. literature around the use and effectiveness of specific psychological interventions for managing children’s postoperative pain 2.1.2.1. Distraction intensity. For example, where some studies have found that Distraction interventions direct attention away from the painful interventions are helpful in reducing postoperative pain, others stimulus/experience toward a more enjoyable stimulus (eg, have found no difference between interventions and typical watching movies/videogames and listening to music).5 management. Furthermore, the interventions used, delivery characteristics, as well as pain measurement tools differ across 2.1.2.2. Imagery/hypnosis/suggestion studies, making it difficult to draw broad conclusions. There are currently no systematic reviews on the role of psychological Interventions characterized by states of highly focused attention interventions in managing postoperative pain intensity in children. during which awareness, sensations, affective response, and 42 Therefore, the primary objective of this review is to synthesize the perceptions of stimuli (eg, pain) are altered. available research on the use of psychological interventions in managing postoperative pain intensity in children. The first aim of 2.1.2.3. Progressive muscle relaxation/deep breathing/ this study was to examine whether psychological interventions biofeedback reduce pain as compared to standard care. Primary outcomes Interventions that include a motor response and target physio- were child self-reported and nurse-/parent-reported pain in- logical arousal. These interventions shift their attention away from tensity assessed across the postoperative period (short/ the painful stimulus/experience.8 moderate term and long term). Secondary outcomes included observed pain behavior, analgesic use, and adverse events. The 2.1.2.4. Preparation/education second aim of this study was to examine the efficacy of different Interventions where specific information about pain and types of psychological interventions (eg, preparation/education expectations of pain with the intent of changing cognitions/ and distraction/imagery) and delivery characteristics of these expectations and teaching coping strategies was provided.27 interventions (eg, parent/nurse delivered vs self-directed). Over- These interventions must include specific information about all, this review seeks to answer the question: Is the addition of pain and pain management, including realistic expectations psychological interventions to typical management more effective about pain, rather than general education about surgery than typical management alone in reducing pain intensity process/expectations. experienced by children who have undergone surgery? 2.1.2.5. Mixed 2. Methods Interventions that fall under cognitive–behavioural techniques We used systematic review methods, including meta-analysis, and fit across the above categories. following recommendations of the Cochrane Collaboration,16 30 and PRISMA-reporting guidelines to evaluate psychological 2.1.3. Types of comparisons interventions for postoperative pain in the pediatric population. We included studies where the comparison group, as identified by authors, was either (1) a standard care group where children 2.1. Inclusion criteria received standard pre- and post-operative care only, and no 2.1.1. Types of participants intervention, or (2) an attention control group where children received all aspects of care that the treatment group received (eg, The population of interest in this study was children and received information about treatment and met with health care adolescents undergoing surgery. We included studies of children providers), but did not receive the intervention. and adolescents aged 2 years to age 18 years. If there were adequate numbers of studies, we planned to examine the treatment effect for each of following subgroups: preschool (age 2.1.4. Types of outcomes 2-5 years), school-aged (6-12), and adolescent (13-18). Surgery All studies must have measured pain intensity as an outcome was defined as a skin-breaking procedure requiring general following surgery. Acceptable measures of pain intensity included anesthesia. We included studies investigating interventions in validated outcome measures (Table 1) as the primary out- both ambulatory surgeries and surgeries requiring hospitalization. come.40,46,51,52 Self-report, nurse or parent reports, or validated behavioural tools were included. Analgesic use was analyzed as a secondary outcome if data were available. We planned to 2.1.2. Types of interventions examine short-term (within 24 hours postsurgery), moderate- We included psychological interventions that could be generally term (24-72 hours postsurgery), and long-term (greater than defined as cognitive, behavioural, or combined cognitive– 72 hours) time points separately. That said, there were only 2 behavioural interventions. We identified 5 categories of studies in the moderate time frame, so we chose to combine interventions to capture the range of interventions that have them with the short-term data. The results of the short-term been studied in the management of postoperative pain: (1) outcomes did not change as a result of including the moderate- distraction, (2) imagery, hypnosis, and/or suggestion, (3) term studies. If more than 1 relevant measurement within each progressive muscle relaxation, deep breathing, and/or bio- follow-up period was available from a study report, only the feedback, (4) preparation/education specific to pain, and earliest measurement in each range was extracted.

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited. 1874 F. Davidson et al.·157 (2016) 1872–1886 PAIN®

Table 1 were either translated using online systems or by colleagues with Descriptions of accepted self-report pain scales for assessing an understanding of the written language. pain intensity and behavioural observational tools.29,40,51 Self-report pain scales 2.3.3. Risk of bias within studies Faces Pain Scale (FPS) Faces Pain Scale–Revised (FPS-R) Risk of bias for individual studies was assessed using the 16 Wong-Baker FACES Pain Rating Scale (FACES) Cochrane Collaboration Risk of Bias tool. Each study was Oucher Pain Scale assessed across six variables: random sequence generation, Numerical Rating Scale of Pain (NRS-11) allocation concealment, blinding of participants and personnel, Behavioural/observational tools blinding of outcome assessment, incomplete outcome data, and Face, Legs, Activity, Cry, Consolability (FLACC) selective reporting. For each variable, reviewers assessed the risk Children’s Hospital of Eastern Ontario Pain Scale (CHEOPS) of bias using the Cochrane definitions and examples (See Tables Parents’ Postoperative Pain Measure (PPPM) 8.5a and 8.5d in the Cochrane Handbook16). For example, when Tools designed for self-report have also been used in some studies to obtain parent/nurse observational assessing selection bias, an example criterion for “low risk of bias” ratings. is reference to a random number table. A criterion for “high risk of bias” is sequence generated by rule based on the date of hospital 2.1.5. Types of studies admission.16 In the cases in which risk of bias could not be assessed because insufficient information was provided, the We included published articles and dissertations that used variable was rated as “unclear.” Risk of bias was assessed experimental methods to assign subjects to treatment and independently by each reviewer (F.D. and S.S.) and then comparison groups (including random assignment, alternating discussed to reach consensus. Results are shown in Table 2. assignment, or matched designs).

2.3.4. Missing data 2.2. Search strategy Attempts to obtain missing data were made by contacting We conducted electronic searches in 4 databases: PubMed, corresponding authors by e-mail. If no response was received PsycINFO, EMBASE, and Certified Index to Nursing and Allied within 1 week of the original request, a second and final request Health Literature from inception of the database up to and was made by e-mail. In 2 studies, data were presented as including July 2015. The PubMed search strategy was created first frequency scores on a pain scale.3,13 We recreated the data sets and then adapted for use in the other databases. Our search based on the frequency of each score and calculated the SDs. strategy included keyword and controlled language (eg, MeSH This was completed by the primary reviewer (F.D.) and was terms) for the concepts postoperative pain, psychological in- checked by the second reviewer (S.S.) for accuracy. In 1 study, tervention, pediatrics, and postoperative care. Reference lists from only the mean values were presented but not the SDs.23 review articles identified in the search were reviewed for any Therefore, the SDs were imputed from 4 articles that had similar additional relevant primary articles. There were no language methods and populations.24,25,33,34 restrictions imposed on the search. A copy of the full PsycINFO search strategy is included in Appendix A (available online as supplemental digital content at. http://links.lww.com/PAIN/A287). 2.3.5. Data synthesis We used meta-analysis to pool sufficiently homogeneous study 2.3. Data Collection and analysis results. We calculated standardized mean differences (SMDs) with 95% confidence intervals (CIs) using a random-effects model 2.3.1. Study selection for each available outcome of interest comparing intervention and We combined database search results and removed duplicate comparison groups. We calculated the I-squared value and articles; initial screening of the titles and abstracts was completed interpreted heterogeneity using guidelines from the Cochrane independently by 2 reviewers (F.D. and S.S.). Any disagreements handbook, where up to 40% heterogeneity might not be were discussed with a third reviewer (J.C.) to reach consensus. important, between 30% and 60% may be moderate, 50% to Relevant articles were reviewed in full independently by each of 90% may be substantial, and 75% to 100% is considered the reviewers, with consensus; if inclusion criteria were met, they considerable heterogeneity.16 We considered studies to provide were included for data extraction. In the case of disagreement, statistically significant evidence for the use of psychological the third reviewer was called upon to determine inclusion or interventions as an adjunct to pharmacological postoperative exclusion based on the notes of the 2 original reviewers. pain management based on type I error of 0.05. To more easily synthesize and compare outcomes, all pain intensity outcome measures were rescaled to a 0 to 10 metric. One study included 2.3.2. Data extraction more than 1 relevant intervention group.24 To include the data We extracted data using a predefined data extraction tool that from both interventions in the overall meta-analysis examining included information about study population (eg, age and sex), any psychological intervention compared with standard care, type of surgery, type of intervention, any training that occurred for the data from the 2 intervention groups were pooled based on the intervention, outcome measures, and timing considerations. methods described in table 7.7a in section 7.7.3.8 of the Data from each study were first extracted by the primary reviewer Cochrane Handbook.16 Data from the separate intervention (F.D.) into individual article data extraction forms. These forms groups were not pooled for separate analyses.24 were then checked for accuracy by the second reviewer (S.S.). To address aim 1, where data were available, separate meta- Extracted data were compiled into an electronic spreadsheet and analyses were conducted to examine the effect of all types of reviewed by 2 authors (F.D. and S.S.) before being entered into psychological interventions on child-reported and nurse-/parent- RevMan 5.3.36 Articles that were written in non-English language reported pain intensity (primary outcomes) within each time

Table 2 Risk of bias within individual studies. Study Random sequence Allocation Blinding of Blinding of outcome Incomplete outcome Selective generation concealment participants assessment data reporting Calcaterra et al.3 Low Low High High Low Unclear Charette et al.4 Low High High Unclear Unclear Unclear Crandall et al.6 High High High High Low Low Fortier et al.9 Low Low High Low Unclear Low Hatem et al.13 High High High High Unclear Unclear He et al.14 Low Low High Low Unclear High Helgad´ottir et al.15 Low Unclear High High Unclear Unclear Huth et al.18 Unclear Unclear High High Low Low Huth et al.17 Low Low High High Unclear Unclear Lambert et al.23 Unclear Unclear High High Unclear Unclear LaMontagne et al.24 Low Unclear High High Unclear Unclear Li et al.25 Low Unclear Unclear High High Unclear Nilsson et al.33 Low Low High High Low Unclear Polkki et al.34 Unclear Unclear High High Low Unclear Ull´an et al.48 Unclear Unclear High Unclear Unclear Unclear

(short/moderate term and long term) using self-report or SDs were imputed. In the cases where the age range of the observational reports of pain (Table 1). Secondary outcomes of participants extended beyond the age range defined by our analgesic use, behavioral measures, and adverse events were inclusion criteria (eg, study included participants younger than reported in only a few studies and are therefore summarized 2 years or older than 18 years), we removed these data in narratively. Two studies1,41 were not included in any meta- a sensitivity analysis. analysis because adequate statistical data were not available. Results of these studies are reported in narrative form. To address aim 2, we planned to conduct separate meta- 2.3.7. Quality of evidence analyses for each of the 5 types of interventions at each time The overall quality of evidence available was assessed using the point. That said, meta-analyses were conducted only for Grading of Recommendations, Assessment, Development, and outcomes measured at the short-/moderate-term time point Evaluation (GRADE) approach.11,37 Quality was assessed for because of the limited number of studies with long-term data. main comparisons when possible considering study design, risk Long-term data are presented qualitatively. Given the limited of bias, inconsistency, indirectness, imprecision, and publication number of studies available and the theoretical similarity between bias. This overall GRADE rating applies to the quality of evidence distraction and imagery/hypnosis/suggestion interventions, we for particular comparisons for a specific population, intervention, combined these groups for analyses. Furthermore, preparation/ and outcome of interest. Results from GRADE assessments are education interventions involved preoperative engagement, available in the online appendix (available online as supplemental whereas the distraction/imagery interventions involved or en- digital content at http://links.lww.com/PAIN/A287). couraged postoperative engagement. Preparation/education interventions were analyzed separately. There were no studies that included progressive relaxation/breathing/biofeedback as 3. Results standalone interventions; thus, no analysis was conducted. Of 3.1. Study selection note, when studies included interventions that fell into the mixed intervention category (eg, they had components of more than one The initial search resulted in 1401 citations (see Fig. 1 for PRISMA kind of intervention), they were included only in the overall analysis flowchart), of which 136 were identified as appropriate for full and not in the separate distraction/imagery and/or preparation/ article review. Of those, 116 articles were excluded. Data were education analyses. In this aim, we also planned to conduct extracted from 20 studies (1435 participants), one of which was analyses of delivery characteristics of the interventions (eg, published in Farsi, but was translated by an academic colleague duration of intervention and teaching before surgery), but were fluent in both English and Farsi. unable because of limited sample sizes and limited reporting of these data. 3.2. Study and intervention characteristics Twenty studies included in analyses represented a total of 1435 2.3.6. Sensitivity analyses children, 767 of whom received intervention and 668 of whom Sensitivity analyses were conducted to explore the impact of our received standard care or attention control. See Table 3 for review methods. One study did not present statistical data (eg, characteristics of included studies. As per inclusion criteria, all SDs) required for inclusion in meta-analysis, and therefore the studies used experimental design and had at least 1 comparison

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited. 1876 F. Davidson et al.·157 (2016) 1872–1886 PAIN®

Figure 1. PRISMA flowchart of study identification and selection.

group. One study included data from 2 intervention groups as well intensity and 14 of the 20 studies were included in quantitative as a combined intervention group,24 and the remaining 19 studies analyses, with a total of 1096 children, 587 of whom received each included 1 intervention group and 1 comparison group. intervention, and 509 who received either standard care or Fourteen studies included a standard care comparison group attention control. and 6 studies identified attention control groups as the comparison group. The types of interventions that were included were distraction/ 3.3.1. Primary outcomes imagery interventions (13 studies), preparation/educational interventions (4 studies), and mixed/more than 1 intervention 3.3.1.1. Child self-report of pain intensity (short term) (3 studies). There were no studies that included progressive Based on 11 studies with 477 participants in the intervention muscle relaxation/biofeedback interventions. Interventions took group and 394 in the control group,3,6,13–15,17,23–25,33,34 psy- place preoperatively for 6 studies, postoperatively for 9 studies, chological interventions were statistically more effective in and both pre- and post-operatively for 5 studies. See Table 4 for reducing pain intensity compared with standard care in the short additional intervention details. Surgical procedures were in the term (SMD 520.47, 95% CI 520.76 to 20.18, P , 0.01; following specialties: otolaryngology, general surgery, plastic Fig. 3). The test for heterogeneity suggests substantial hetero- surgery, dentistry, and orthopedics. Of the 20 studies, 18 geneity (I2 5 75%; P , 0.001), which impacted the GRADE rating included pain intensity scales as the primary outcome measure for this comparison. This comparison was rated as providing very (15 studies had child self-reports, 3 studies had reports by low quality evidence because of downgrading as a result of nurses, and 1 study had reports by parents). Two of the included potentially serious risk of bias, substantial heterogeneity, and studies reported on observational measures only (completed by potentially serious indirectness given that all interventions were research assistants), and 2 included both observational meas- examined together. ures (completed by research assistant/nurses) and self-reports. A summary of the risk of bias within studies is presented in 3.3.1.2. Child self-report of pain intensity (long term) Table 2, and a summary of the risk of bias across studies is presented in Figure 2. Based on 2 studies,4,24 psychological interventions (n 5 90) were not statistically more effective than standard care (n 5 39) in reducing pain intensity in the long term (SMD 520.60, 95% CI 5 3.3. Aim 1: any psychological intervention compared with 21.24 to 0.04, P . 0.05; Fig. 4). The test for heterogeneity standard care suggests moderate to substantial heterogeneity (I2 5 59%; P . Twenty studies (1435 participants) provided evidence about the 0.05), which along with potentially serious imprecision down- effect of any psychological intervention on postoperative pain graded the GRADE rating for this comparison to that of low

Table 3 Characteristics of included studies. Study N (n intervention; Age range Type of surgery Outcome measure Type of intervention(s) n control) (measurement scale) Baratee 40 (20 intervention; 8-12 y Not reported Numeric Pain Scale (0-10) (1) Distraction: play; and (2) et al.1 20 control) standard care Calcaterra 41 (21 intervention; 3-14 y Orchidopexy, inguinal hernia repair, Child self-report: Faces Pain Scale (1) Distraction: music therapy; and et al.2 21 control) and circumcision (0-10); Research Assistant–rated (2) standard care FLACC (0-10) Calcaterra 40 (20 intervention; 3-17 y Orchidopexy, inguinal or umbilical Child self-report: Wong-Baker (1) Distraction: animal-assisted et al.3 20 control) hernia repair, circumcision, and FACES Pain Scale (0-10) therapy; and (2) standard care varicocele treatment Charette 40 (20 intervention; 11-19 y (only 1 Spinal fusion Brief Pain Report (0-10) (1) Mixed: guided imagery 1 et al.4 20 control) participant age 19) preparation/education; and (2) standard care Crandall 60 (30 intervention; 7-13 y Tonsillectomy with or without Child self-report: Numerical Rating (1) Preparation/education: pain et al.6 30 control) adenoidectomy Scale (0-10) education booklet; and (2) standard care Fortier et al.9 74 (33 intervention; 2-7 y ENT, general surgery, urologic, Nurse report: Numerical Rating (1) Mixed intervention: online 41 control) ophthalmology, plastic surgery, and Scale (0-10) preparation/education 1 imagery/ orthopedic distraction techniques; and (2) standard care Hatem 79 (61 intervention; 1 d-16 y Heart surgery Nurse report: Faces Pain Scale (1) Distraction: music therapy; and et al.13 18 control) (0-7) (2) attention control He et al.14 95 (48 intervention; 6-14 y ENT, orthopedic, and general Child self-report: Numerical Rating (1) Preparation/education: therapeutic 47 control) surgery Scale (0-10) play; and (2) standard care Helgad´ottir 40 (19 intervention; 3-7 y Tonsillectomy with or without Child self-report: Wong-Baker (1) Distraction; and (2) attention et al.15 21 control) adenoidectomy FACES Pain Rating Scale (0-10) control Huth et al.18 51 (27 intervention; 3-6 y Open or closed cardiothoracic Child self-report: Oucher Pain Scale (1) Preparation/education; and (2) 24 control) procedure (0-100 for older children, 6 picture standard care scale for younger children); Parent report: Oucher Pain Scale (0-100) Huth et al.17 72 (36 intervention; 7-12 y Tonsillectomy or adenoidectomy Child self-report of pain: Oucher (1) Imagery; and (2) attention 37 control) with or without myringotomy Pain Scale (0-100) control Kim et al.20 70 (37 intervention; 2-7 y Tonsillectomy, herniorrhaphy, and Research Assistant report: CHEOPS (1) Distraction: video; and (2) 33 control) excision of neck mass (2-13) standard care Lambert 52 (26 intervention; 7-19 y Mixed: spinal fusion, ortho, cardiac, Child self-report: Numeric Pain (1) Guided imagery; and (2) et al.23 26 control) and general surgery Scale (0-10) attention control LaMontagne 89 (19, 22, 24 11-18 y Spinal fusion Child self-report: Visual Analogue (1) Preparation/education; (2) et al.24 intervention; Scale (0-10) coping; (3) preparation/education 19 control) 1 coping; and (4) standard care Li et al.25 203 (97 7-12 y Mixed: circumcision, hernia, eye, Child self-report: Visual Analogue (1) Preparation/education; intervention; ENT, dental, and orthopedic Scale (0-10) therapeutic play; and (2) standard 106 control) care Martin 94 (48 intervention; 4-8 y Tonsillectomy Child self-report Wong-Baker (1) Therapeutic suggestion; and (2) et al.28 46 control) FACES Pain Scale (0-100; Nurse standard care report: FLACC (0-10) Nilsson 80 (40 intervention; 7-16 y Arthroscopy, endoscopy, extraction Child self-report: Coloured (1) Distraction: music therapy; and et al.33 40 control) of pin/nail/thread, hernia/ Analogue Scale (0-10); Nurse (2) standard care hydrocele, and superficial surgery report: FLACC Polkki et al.34 60 (30 intervention; 6-12 y Appendectomy or upper/lower limb Child self-report: Visual Analogue (1) Imagery/relaxation; and (2) 30 control) surgery Scale (0-10); Nurse report: Visual standard care Analogue Scale (0-10) Suresh 54 (35 intervention; 6-18 y Orthopedic, neurosurgical Child self-report: Faces Pain (1) Distraction: music/audio et al.41 19 control) urological, plastic, and general Scale–Revised (0-10) distraction; and (2) attention control surgery Ull´an et al.48 39 (19 intervention; 4-7 y Genital, ENT, hernia, trauma, Research Assistant report: FLACC (1) Distraction: play; and (2) 20 control) ophthalmology, GI maxillofacial, (0-10) standard care plastic, and other CHEOPS, Children’s Hospital of Eastern Ontario Pain Scale; ENT, Ear, Nose, Throat; FLACC, Face, Legs, Activity, Cry, Consolability; GI, gastro-intestinal.

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited. 1878 F. Davidson et al.·157 (2016) 1872–1886 PAIN®

Table 4 Description of intervention characteristics. Study Intervention leader Pre-op/post-op/both Duration (min) No. sessions Other details Baratee NA Post-op 30-45 1 Intervention occurred 6-24 h postsurgery et al.1 Children played with magnetic materials (eg, balls and magnets) for 30-45 min Note: this information was provided based on a translated version of this article. No other details were reported in translation Calcaterra Not reported Post-op 20 1 Music was played during awakening period postsurgery et al.2 for 20 min through ambient speakers Music was 6 tracks of slow 70-80 beats/min and fast 140-150 beats/min classical music with 2 min of silence between sequences Lighting, temperature, and noise levels also controlled Calcaterra Dog and dog handler Post-op 20 1 Intervention occurred during awakening period and et al.3 consisted of dog and handler being present for 20 min Handler tended to dog and answered questions about the dog Dog was experienced in animal-assisted intervention Charette Research nurse Pre-op and post-op 30 2 One day before surgery, participants watched a 30-min et al.4 DVD with information on postoperative pain management, demonstrations of guided imagery, and relaxation A research nurse was available to answer questions Participants were instructed to take DVD home and practice exercises at least 3 times/wk Two weeks after discharge, research nurse called participants to reinforce intervention techniques DVD consisted of 3 parts: (1) information on surgery and post-op care with analgesics, (2) strategies for turning in bed and exercises to address functioning, and (3) relaxation and imagery exercises Crandall Researcher Pre-op ? 1 Pain education booklet was specific to pain following et al.6 tonsillectomy and adenoidectomy Booklet was developed based on interviews from children previously seen for same procedures Booklet provided information on different pain experiences and individual pain experiences, definitions of pain, questions that doctors and nurses may ask about pain, review of 0-10 pain intensity scale, expected location and duration of pain, information on pain medications, nonpharmacologic pain management strategies, instructions on when to take pain medication, and when to tell someone that pain medication is not working Fortier et al.9 Parent through Web Pre- and post-op Intervention could be Intervention could be Intervention title WebTIPS is an animated children’s Web platform used by parents for as accessed as many site and multimedia parent Web site long as desired times as desired Child site was tailored based on trait anxiety and surgery and included information, modeling, and coping skills Parent site was tailored based on parent anxiety coping style, pain management attitude, and includes information and coping skills training Content covers preoperative period (home), preoperative holding area, postanesthesia care unit, and home Parents had unlimited WebTIPS access 1 wk before surgery Parents could also e-mail researchers with questions Hatem Child Post-op 30 1 Participants listened to classical music (Spring from et al.13 Vivaldi’s 4 Seasons) for 30 min postoperation through headphones attached to a CD player He et al.14 Researcher (nurse/ Pre-op 60 1 Participants and parents received 60 min of time, 3-7 d physician) before surgery They were provided with a manual about medical objects and equipment and watched a video titled “Preparation for Your Operation at the Participating Hospital” They viewed photos of the operating room and a doll was used to demonstrate surgical preparation and procedures (continued on next page)

Table 4 (continued) Study Intervention leader Pre-op/post-op/both Duration (min) No. sessions Other details

Helgadottir Researcher Intervention delivered ? 1 Parents received booklet with distraction education, as et al.15 pre-op with instruction well as received a single session of verbal instruction from for engagement post-op researcher Instructions included the following: directing children’s attention away from pain may help pain and that distraction should be tailored to interest of child; timing of the distraction is important (eg, pain is worst first thing in the morning following tonsillectomy); recommended time for active distraction between 30 and 60 min; active distraction advised when pain was severe and passive distraction advised when pain was less severe Huth et al.18 Researcher Day of surgery (pre-op) 15-20 1 Intervention booklet tailored to different age groups or first day post-op Booklet focused on pain assessment and management Parents were provided with the booklet before child’s surgery and were permitted to keep the booklet while child was hospitalized Booklet was reviewed with the parent by the researcher using a standardized script on day of surgery or first day after surgery Content included meaning of pain, children’s reactions to pain, pain assessment, pain management with medication, and nonpharmacological strategies Huth et al.17 Researcher Intervention materials 30 3 Intervention was titled “To Tame the Hurting Thing” provided pre-op and (Broome, 1994) was delivered Children and parents were provided with imagery postoperatively booklets, a videotape with deep breathing ad imagery techniques, and audiotape with deep breathing, muscle relaxation, music, and imagery suggestions Materials were provided 2 to 22 d before surgery Children and parents were instructed to practice as much as they wanted before surgery and could use the tape as much as they wanted postsurgery Encouraged to practice at least 3 times per day before surgery Kim et al.20 Anesthetist/nurse/ Pre-op Not 1 Intervention occurred during mask induction and lasted parent provided—individual until child could not respond to name for each participant Intervention consisted of viewing an animated cartoon video on a smartphone Video was discontinued when child could no longer respond to name Lambert Researcher Pre-op 30 1 Intervention consisted of a 30-min session 1 wk before et al.23 surgery Researcher taught the child ways to relax and use their imagination to learn about the surgery Children were instructed to choose a favourable image that felt good. This image was incorporated into an individually tailored relaxation exercise whereby the researcher guided the child through rehearsal of the surgery, including suggestions for healing and minimizing pain Positive language was used LaMontagne Researcher Pre-op ? 1 Participants received 1 of 3 interventions et al.24 (1) Concrete–objective information intervention consisted of providing participants with surgical information (eg, length, insertion of equipment, and postsurgery expectations) and pain information (eg, normal pain expectations, length/duration of pain, when sitting up/ ambulation would occur, and activity limitations) (2) Coping instruction intervention consisted of instruction around relaxation, deep breathing, relaxation, and positive thoughts for managing postoperative pain (3) The concrete–objective information 1 coping instruction group received both the information components described above (continued on next page)

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited. 1880 F. Davidson et al.·157 (2016) 1872–1886 PAIN®

Table 4 (continued) Study Intervention leader Pre-op/post-op/both Duration (min) No. sessions Other details

Li et al.25 Nurse researcher Pre-op 60 1 Intervention was a usual care plus a therapeutic play session 1 week before surgery in a group setting (5 children per group) and 1 parent per child Participants received a tour of reception, induction room, operating room, and recovery room In the operating room, children were provided with a demonstration of anesthesia induction using a doll Children were permitted to examine equipment and had the opportunity to demonstrate the procedures on the doll themselves Martin Nurse Post-op 15-30 1 Intervention was an original therapeutic suggestion script et al.28 Script was positively worded and designed specifically for children undergoing tonsillectomy Script was provided over MP3 recording to keep caregivers blind to intervention Control group listened to recording of PACU noise The intervention was initiated upon completion of the surgery. Headphones were placed on the child when surgery was completed and lasted until child was ready to wake up Nilsson Researcher and nurse Post-op 45 1 Participants in the music intervention group had soft et al.33 relaxing music (MusiCure) Music was played starting from admission to the postoperative care unit for 45 min Polkki et al.34 Nurse Post-op 20 1 Intervention consisted of children listening to an imagery and relaxation CD through earphones At the beginning of the CD, children were instructed to mentally choose a favourite place and the imagery trip encouraged the child toward this place downward At the end of the trip, the child returned upward Suresh Self-administered Post-op 30 1 Either music or audiobook intervention was chosen pre-op et al.41 Audio/music intervention was delivered in postoperative period through noise cancelling headphones for 30 min Ull´an et al.48 Parent with instruction Post-op Not reported 1 Parents were provided with written information about by a specialist in social distraction through play to relieve child pain/distress and education were provided with play materials Play material was a plush toy rabbit dressed as a doctor PACU, postanesthesia care unit.

quality. Of note, one of these studies included 1 participant who range.13,23 Overall, the results of these sensitivity analyses also was 19 years of age4; however, it was not possible to remove this revealed that the inclusion of these studies did not significantly data in a sensitivity analysis given that only 2 studies were impact the overall effect. See Table 5 for all sensitivity analyses. available for meta-analysis. 3.3.1.6. Narrative summary of studies not included in meta- 3.3.1.3. Parent/nurse report of pain intensity (short term) analyses Based on 3 studies with 90 participants in the intervention group Two studies met the inclusion criteria for this review but did not and 95 participants in the control group,9,18,34 psychological present sufficient statistical data on pain intensity to be included in interventions were not more effective than standard care (SMD 5 the above-presented meta-analyses. Results from 1 study showed 20.09, 95% CI 520.38 to 0.20, P . 0.05; Fig. 5) in the short term. that mean pain change scores were significantly larger for the The test for heterogeneity was not significant (I2 5 0%; P . 0.05) distraction intervention group when pain was assessed before and and the GRADE rating for this comparison was rated as moderate either immediately or 15 minutes after the intervention. There was because of downgrading with respect to possible indirectness. no difference between groups when pain was assessed immediately after and 15 minutes after the intervention.1 The other study 3.3.1.4. Parent/nurse report of pain intensity (long term) found that pain burden (defined as minutes multiplied by self- reported pain score) was significantly reduced in the music/audio There were no studies that assessed parent or nurse report of distraction intervention group compared with control.41 pain intensity beyond 72 hours.

3.3.1.5. Sensitivity analyses 3.3.2. Secondary outcomes We performed 3 sensitivity analyses. When we excluded the study 3.3.2.1. Analgesic use where SDs were imputed,23 interpretation of results was un- changed. The second and third sensitivity analyses were performed Eight studies reported outcome data on analgesic use. It was not excluding the 2 studies that included participants beyond the age possible to synthesize these data in meta-analysis because there

Figure 2. Risk of bias across studies for 6 factors.

was inconsistency in the kind of analgesics that were used and 3.3.2.3. Observational measures the timing of the analgesic delivery. Furthermore, there was often Four of the included studies used behavioural measures of pain; 2 more than one kind of analgesics given to patients following used only behavioural measures completed by research assis- surgery. Results are presented narratively here. Only 1 study tants,20,48 and 2 used behavioural measures completed by reported that children in the intervention group received nurse/research assistant.20,28 Data were available for meta- significantly less pain medication than those in the control group analysis for 2 studies that examined distraction at the short-term (P 5 0.05).33 The remaining 7 studies reported that there was no time point using Face, Legs, Activity, Cry, Consolability (FLACC) significant difference in the amount of pain medication received and Children’s Hospital of Eastern Ontario Pain Scale.20,48 Based by children.3,9,14,17,18,23,28 on these 2 studies with 56 participants in the intervention group and 53 in the control group, distraction was not more effective 3.3.2.2. Adverse events than standard care (SMD 5 0.15, 95% CI 520.22 to 0.53, P . Of the 20 included studies, adverse events were reported in 0.05; Fig. 6). Although data were not available for meta-analysis, 3.17,23,33 One study reported that 2 children (of 73 total a third study examined music distraction and found no significant participants) became distressed during the intervention and difference between groups on research assistant–rated pain withdrew,17 another reported that the intervention interfered with behaviours (FLACC).2 An additional study examined therapeutic one participant’s (of 52 total participants) religious beliefs,23 and suggestion and reported that pain levels as rated by nurses on the the third reported that the intervention increased sound levels in FLACC were significantly lower for children who received the staff working environment.33 intervention compared with controls when assessed 30 minutes

Figure 3. Forest plot of studies examining any psychological intervention for child self-reported pain intensity (short term). CI, confidence interval.

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited. 1882 F. Davidson et al.·157 (2016) 1872–1886 PAIN®

Figure 4. Forest plot of studies examining any psychological intervention for child self-reported pain intensity (long term). CI, confidence interval.

postextubation; however, this was not maintained at 50 minutes heterogeneity indicated that this comparison had substantial postextubation.28 heterogeneity (I2 5 77%; P , 0.001). Overall, this comparison was rated as providing low quality evidence due to downgrading as a result of the substantial heterogeneity and serious 3.4. Aim 2: to examine the efficacy of different types of indirectness. psychological interventions 3.4.1. Preparation/education interventions 3.4.2.2. Parent/nurse report of pain intensity (short term) Parent-/nurse-reported data were available for only 1 study, thus 3.4.1.1. Child self-report of pain intensity (short term) meta-analysis was not conducted. Based on child self-reports of short-term pain from 4 studies,3,14,24,25 preparation/education interventions (n 5 199) were 4. Discussion not effective over standard care (n 5 202) (SMD 520.27, 95% CI 520.61 to 0.08, P . 0.05; Fig. 7). The test for heterogeneity Overall, the results of this review suggest that there is evidence revealed moderate to substantial heterogeneity (I2 5 61%; P 5 that psychological interventions are at least moderately effective 0.05) impacting the GRADE rating for this comparison. Overall, in managing short-term postoperative pain intensity in children, this comparison was rated as providing moderate quality based on child self-report. Interventions included distraction, evidence due to downgrading as a result of heterogeneity. imagery, preparation/education, and mixed interventions. They ranged from 1 to 3 training sessions, with no more than 30 3.4.1.2. Parent/nurse report of pain intensity (short term) minutes per session. Whereas effect sizes were relatively small 18 (eg, less than 1 point difference on a 0-10 metric), it is important to Parent-/nurse-reported data were only available for 1 study, so note that there is a growing body of literature on the minimum meta-analysis was not conducted. clinically significant difference in pediatric pain intensity. A recent study found that for children presenting in the emergency 3.4.2. Distraction/imagery interventions department with acute pain, the minimum clinically significant difference in pain intensity was 2 (95% CI 5 1-2) on the Faces Pain Scale–Revised and a change score of 1 (95% CI 5 0.5-1.25) 3.4.2.1. Child self-report of pain intensity (short term) on the Color Analog Scale.47 Another study found that for children Results showed that distraction/imagery was effective over in the emergency department presenting with pain, the minimum standard care for short-term child self-reported pain (SMD 5 clinically significant difference pain change score on a 0-100 20.63, 95% CI 521.04 to 0.23, P , 0.01; Fig. 8) based on eight scale was 10 (95% CI 5 7-12) on the Visual Analogue Scale.35 studies with 254 participants in the intervention group and 211 Despite potentially clinically significant effects, comparisons participants in the control group.6,13,15,17,23,24,33,34 The test for ranged from high to low quality, which implies that effects should

Figure 5. Forest plot of studies examining any psychological intervention for child self-reported pain intensity (long term). CI, confidence interval.

Table 5 Results of sensitivity analyses. All psychological interventions: child report Distraction interventions: child report Hatem 1 Lambert removed 20.34 (20.55 to 20.13); P 5 0.002 20.46 (20.76 to 20.17); P 5 0.002 Hatem only removed 20.33 (20.52 to 20.14); P 5 0.0007 20.43 (20.68 to 20.17); P 5 0.0009 Lambert only removed 20.49 (20.81 to 20.18); P 5 0.002 20.69 (21.16 to 20.23); P 5 0.003

be interpreted with a degree of caution, as further research is suggest that future research and practice should consider the necessary and important in determining the true effect of importance of postoperative instruction or engagement. In both psychological interventions. subgroup analyses, parent-/nurse-reported pain was available Although meta-analysis of analgesic use data was precluded for only 1 study, and in both cases results were nonsignificant. by variability in medication types and doses, there were no The interpretation of parent-/nurse-reported data is limited by the notable differences in analgesic use between psychological small number of studies, but the distinct difference between intervention and standard care conditions. This raises important parent/nurse and child self-report will be important to address in questions for considering analgesic use as an outcome in future future studies. studies. It is quite possible, especially given shifts from providing These findings draw attention to some important consider- medication as needed, to around the clock dosing recommen- ations in the area of pain management and psychological dations, that analgesic use is influenced much more by pre- interventions. Research on brief procedures (eg, immunizations scribing patterns than pain or the efficacy of other interventions. and venipunctures) has demonstrated that distraction and Furthermore, very few studies reported on adverse events preparation interventions are effective in reducing pain and associated with their intervention or control groups. One study distress,32,39 but research on psychological interventions for did find that 2 of the 36 children in the imagery arm experienced postoperative pain in children and adolescents is limited. The pain distress and were withdrawn from the study. Although this is experienced during a single procedure (eg, venipuncture) may a relatively small number, it does draw attention to the need to differ in length and intensity as compared to postoperative pain, monitor potentially adverse events from psychological interven- and as such it is possible that the delivery of psychological tions. We generally think of these interventions as being relatively interventions may need to be adapted to more specifically target benign from a safety point of view, but this still deserves further postoperative pain (eg, more preparation or practice). study. The psychological interventions that were included in analysis 4.1. Strengths and limitations of included studies were categorized into either preparation/education or distraction/ imagery intervention. Thus, the findings of this review are As a whole, the studies included in the meta-analyses had several particularly clinically useful given that many of the interventions strengths including being moderate quality, experimental, ran- had a brief training component and they were typically self, nurse, domized trials. As part of each study design, there was also good or parent-led. In addition, the interventions that did require consideration for power issues in selecting sample sizes, as well training (eg, information books or imagery CDs) have already as randomization methods to either intervention or control group. been developed and therefore could easily be distributed to Furthermore, the majority of included studies assessed inter- hospitals more widely. ventions that fell into one category (eg, distraction/imagery), The results from subgroup meta-analyses revealed that rather than into the mixed intervention category, meaning that as distraction/imagery interventions, but not preparation/education additional primary studies are conducted it is likely that specific interventions, were effective in reducing short-term child- intervention components could be examined. reported postoperative pain in the short term compared with Some overall limitations of the included studies are the standard care. Given that there were no direct comparisons of inclusion of mixed surgeries in individual studies, as well as distraction/imagery and preparation/education interventions in- different pain intensity scales used by a variety of raters, with pain cluded in this study, we cannot conclude that one intervention is assessment occurring at a variety of times. These factors made better than another. That said, based on the descriptions of the subgroup analysis difficult and in some cases impossible. Even in interventions by study authors, distraction and imagery inter- the case in which we were able to conduct subgroup analysis ventions seemed to differ from the preparation/education in that (distraction/imagery vs preparation/education), variability in inter- they were either delivered postoperatively or had instruction for ventions may have confounded these results. This is especially children to engage in the intervention postoperatively. This may the case for preparation/education interventions in which

Figure 6. Forest plot of studies examining distraction/imagery for research assistant–completed observational scales (short term). CI, confidence interval.

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited. 1884 F. Davidson et al.·157 (2016) 1872–1886 PAIN®

Figure 7. Forest plot of studies examining preparation/education for child self-reported pain intensity (short term). CI, confidence interval.

strategies varied from a brief, relatively passive educational presented data for a range of ages, and subgroup data were booklet to more interactive therapeutic play. Furthermore, there not available. was little discussion of standardizing analgesic approaches across groups, making it difficult to assess independent 4.2. Strengths and limitations of this review contributions of psychological interventions. Standardization across the field would be beneficial for synthesis of data. In This review has several limitations that are in part due to the terms of populations, children differ developmentally in their ability difficult nature of assessing the efficacy of psychological to engage in interventions, as well as their response to pain, and interventions in children. First, we planned to examine types of as such subgroup data would be needed to examine these psychological interventions separately; however, there were variables. Although the majority of studies included child self- limited trials available for inclusion, and as such we had to pool reported pain, 4 studies used an observer rating (eg, parent or interventions together to conduct meta-analyses. In addition, we nurse). Although these were analyzed separately from the child had planned on assessing multiple aspects of intervention self-report data, it is important to note that in these 4 cases, the delivery, such as preoperative training and practice opportunities child pain intensity data were collected using measures originally and whether these had an impact on outcome. These questions designed for self-report (eg, Faces Pain Scale–Revised and would have allowed us to examine the intervention efficacy in numerical rating scale), which has implications given that these greater detail; however, these details were not frequently ratings would be theoretically different than the child self-report of reported, and as such could not be included for analysis. pain intensity experience. It is notable that very few studies had Similarly, there were not enough studies to analyze subgroups of data available to examine longer term impacts of these surgeries, and many of the studied included samples undergoing interventions. In fact, only a few studies included assessments a variety of surgeries. It is possible that the effectiveness of beyond 24 hours. Even in the cases in which long-term data were psychological interventions may depend on the type of surgery available (pain after 72 hours), these data were collected while that children undergo; however, we were not able to address this children were still in hospital, thus limiting conclusions about in the current review. Synthesizing pediatric research in particular effects of psychological interventions after relatively short-term is difficult because of the developmental differences between recovery or once children go home. Finally, many studies childhood and adolescence. To capture enough studies to

Figure 8. Forest plot of studies examining distraction/imagery for child self-reported pain intensity (short term). CI, confidence interval.

complete meta-analyses, we opted to have a broad age range. measures relying on observations of behavior (eg, FLACC, We hoped to explore age group differences if enough studies Children’s Hospital of Eastern Ontario Pain Scale, and Parents’ were available; however, given the small number of trials, it was Postoperative Pain Measure) are most appropriate for this not possible to conduct these subgroup analyses. Furthermore, purpose. Future studies should also address treatment fidelity. the focus of this review was on postoperative pain; however, it is This may be more of a reporting issue than a methodological important to note that many of the included articles also assessed issue, but most reviewed studies did not report on treatment pre- and post-operative anxiety. Broadly, given that psycholog- adherence and fidelity. Lastly, given the wide age range included ical interventions are used in the treatment of anxiety, a possible in many of the included studies, future research should consider hypothesis is that the psychological interventions may have been limiting their samples to a developmental period or including effective in reducing some aspects of pain, such as anxiety enough information on their sample to allow for subgroup around pain, which was not captured by the outcome pain rating analyses. scales. Finally, there was evidence to suggest that distraction/ imagery interventions, as rated by children effectively decreased pain, but not education/preparation interventions. These findings 4.4. Summary and clinical implications may suggest that interventions that are implemented in the In sum, our review found that that there is some evidence to moment may be more beneficial than only providing information in suggest that psychological interventions, particularly those using advance about pain. techniques that are used postoperatively (eg, distraction and Despite these limitations, the current review has several imagery), are effective in reducing postoperative pain in children. strengths. This is the first meta-analysis addressing the question Additional empirical studies are needed to (1) further evaluate the of psychological interventions for the management of post- effect of psychological interventions on postoperative pain and operative pain in the pediatric population. Furthermore, the improve our confidence in the effect, as well as examine the methods for this review were modeled after a Cochrane review specific types of interventions, or specific components of the assessing psychological interventions for needle pain in children, interventions that are most effective. In terms of clinical and stringent inclusion criteria were used to find primary studies implications, most of the interventions included in this review on this topic. Specifically, this review limited included studies to were administered in a relatively short period, required little those with validated pain intensity measures and experimental training, and have already developed dissemination materials (eg, control groups. This review also presents an evaluation of the CDs and booklets), thus facilitating the dissemination of these quality of existing evidence, thereby providing readers with strategies. Psychological interventions, especially those imple- a context on which to judge the utility of our conclusions. Lastly, mented in the postoperative period (eg, distraction and imagery), the thorough overview of the current state of the evidence allows may be effective in decreasing children’s pain intensity and could for the identification of research directions that could significantly be incorporated into routine clinical care. advance the field (see section 4.3.).

Conflict of interest statement 4.3. Unanswered questions and future research The authors have no conflicts of interest to declare. Whereas this area of research is important, research to date is J. Chorney is supported by the Canadian Foundation for limited and thus recommendations cannot be made in terms of Innovation (Grant # 25741) and a CIHR New Investigator award what kind of psychological interventions should be included in the (#201009MSH-233268-HSA-CHAA-192333). standard postoperative care for children. Although there are some empirical studies, many studies were excluded because of quality as well as factors that did not meet specific criteria for Acknowledgements population, intervention, or outcome. Therefore, future studies The authors acknowledge Robin Parker for her support and are needed to address the effectiveness of individual types of expertise in the development and modification of the search psychological interventions. Of particular note, there were no strategies used in this review. studies that met criteria for this review that examined progressive muscle relaxation or deep breathing strategies alone. This is surprising given the common inclusion of these strategies in Appendix 1. Supplement Digital Content 49 cognitive behavioral packages for the treatment of pain. Of Supplemental Digital Content associated with this article can be particular importance, future studies should consider intervention found online at http://links.lww.com/PAIN/A287. delivery characteristics, particularly whether interventions are delivered before or after surgery. Our results suggest that active Article history: implementation of these strategies postoperatively seems to be Received 2 December 2015 important, but future studies should consider whether including Received in revised form 26 May 2016 preoperative training in these strategies would improve their Accepted 26 May 2016 efficacy. Available online 14 June 2016 Although this review included only validated pain scales, there was variability in the pain scales used. Future research should consider the use of standardized pain rating scales to allow for References easier comparisons; given evidence for validity and reliability in [1] Baratee F, Dabirian A, Yoldashkhan M, Zaree F, Rasouli M. Effect of a relatively wide age range. For younger child (ie, 4-12 years) self- therapeutic play on postoperative pain of hospitalized school age children reports of pain intensity, the Faces Pain Scale–Revised may be in pediatric surgical ward [in Farsi]. J Nurs Midwif 2011;21:57–61. 40 [2] Calcaterra V, Ostuni S, Bonomelli I, Mencherini S, Brunero M, Zambaiti E, most applicable for this purpose, whereas the numerical rating Mannarino S, Larizza D, Albertini R, Tinelli C, Pelizzo G. Music benefits on scale-11 may be most appropriate for children aged 8 years and postoperative distress and pain in pediatric day care surgery. Pediatr Rep older. For reports by others (eg, parent or nurse reports), 2014;6:5534–38.

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited. 1886 F. Davidson et al.·157 (2016) 1872–1886 PAIN®

[3] Calcaterra V, Veggiotti P, Palestrini C, De Giorgis V, Raschetti R, Ljungman G, Palermo T, Rappaport BA, Rhodes T, Schechter N, Scott Tumminelli M, Mencherini S, Papotti F, Klersy C, Albertini R, Ostuni S, J, Sethna N, Svensson OK, Stinson J, von Baeyer CL, Walker L, Weisman Pelizzo G. Post-operative benefits of animal-assisted therapy in pediatric S, White RE, Zajicek A, Zeltzer L. Core outcome domains and measure for surgery: a randomized study. PLoS One 2015;10:e0125813. pediatric acute and chronic/recurrent pain clinical trials: PedIMMPACT [4] Charette S, Fiola JL, Charest M, Villeneuve E, Theroux ´ J, Joncas J, Parent recommendations. J Pain 2008;9:771–83. S, Le May S. Guided imagery for adolescent post-spinal fusion pain [30] Moher D, Liberati A, Tetzlaff J, Altman DG; The PRISMA Group. Preferred management: a pilot study. Pain Manag Nurs 2015;16:211–20. reporting Items for systematic reviews and meta-analyses: the PRISMA [5] Cohen LL, Cousins LA, Martin SR. Procedural pain distraction. In: Statement. BMJ 2009;339:b2535. McGrath PJ, Stevens BJ, Walker SM, Zempsky WT, editors. Oxford [31] Niesters M, Overdyk F, Smith T, Aarts L, Dahan A. Opioid-induced textbook of paediatric pain. 1st ed. Oxford: Oxford University Press, 2014: respiratory depression in paediatrics: a review of case reports. Br J 553–59. Anaesth 2013;110:175–82. [6] Crandall M, Lammers C, Senders C, Braun JV, Savedra M. Children’s [32] Nilsson S, Enskar K, Hallqvist C, Kokinsky E. Active and passive distraction pre-operative tonsillectomy pain education: clinical outcomes. Int J in children undergoing wound dressings. J Pediatr Nurs 2013;28:158–66. Pediatr Otorhinolaryngol 2008;72:1523–33. [33] Nilsson S, Kokinsky E, Nilsson U, Sidenvall B, Enskar K. School-aged [7] Dahlquist LM, McKenna KD, Jones KK, Dillinger L, Weiss KE, Sonntag children’s experiences of postoperative music medicine on pain, distress, Ackerman C. Active and passive distraction using a head-mounted and anxiety. Pediatr Anesth 2009;19:1184–90. display helmet: effects on cold pressor pain in children. Health Psychol [34] Polkki ¨ T, Pietila ¨ AM, Vehvilainen-Julkunen ¨ K, Laukkala H, Kiviluoma K. 2007;26:794–801. Imagery-induced relaxation in children’s postoperative pain relief: [8] DeMore M, Cohen LL. Distraction for pediatric immunization pain: a randomized pilot study. J Pediatr Nurs 2008;23:217–23. a critical review. J Clin Psychol Med Settings 1995;12:281–91. [35] Powell CV, Kelly AM, Williams A. Determining the minimum clinically [9] Fortier MA, Bunzli E, Walthall J, Olshansky E, Saadat H, Santistevan R, significant difference in visual analog pain score for children. Ann Emerg Mayes L, Kain ZN. Web-based tailored intervention for preparation of Med 2001;37:28–31. parents and children for outpatient surgery (WebTIPS): formative evaluation [36] Review manager (RevMan) [computer program]. Version 5.3. Copenhagen: and randomized controlled trial. Anesth Analg 2015;120:915–22. The Nordic Cochrane Centre, The Cochrane Collaboration, 2014. [10] Gatchel RJ, Peng YB, Peters ML, Fuchs PN, Turk DC. The biopsychosocial [37] Schunemann ¨ H, Brozek ˙ J, Guyatt G, Oxman A; The GRADE Working approach to chronic pain: scientific advances and future directions. Psychol Group. GRADE handbook for grading quality of evidence and strength of Bull 2007;133:581–624. recommendations, 2013. [11] GRADEpro GDT: GRADEpro guideline development tool [software]. [38] Sømme S, Bronsert M, Morrato E, Ziegler M. Frequency and variety of McMaster University, Hamilton, ON, Canada. 2015 (developed by evidence inpatient pediatric surgical procedures in the United States. Pediatrics Prime, Inc). 2013;132:e1466–72. [12] Hamers JPH, Abu-Saad HH. Children’s pain at home following (adeno) [39] Spafford PA, von Baeyer CL, Hicks CL. Expected and reported pain in tonsillectomy. Eur J Pain 2002;6:213–19. children undergoing ear piercing: a randomized trial of preparation by [13] Hatem TP, Lira PIC, Mattos SS. The therapeutic effects of music in parents. Behav Res Ther 2002;40:253–66. children following cardiac surgery. J Pediatr (Rio J) 2006;82:186–92. [40] Stinson JN, Kavanaugh T, Yamada J, Gill N, Stevens B. Systematic [14] He HG, Zhu L, Chan WCS, Liam JLW, Li HCW, Ko SS, Klainin-Yobas P, review of the psychometric properties, interpretability and feasibility of Wang W. Therapeutic play intervention on children’s perioperative self-report pain intensity measure for use in clinical trials in children and anxiety, negative emotional manifestation and postoperative pain: adolescents. PAIN 2006;125:143–57. a randomized controlled trial. J Adv Nurs 2015;71:1032–43. [41] Suresh BS, De Oliveira GS, Suresh S. The effect of audio therapy to treat [15] Helgadottir ´ HL, Wilson ME. A randomized controlled trial of the postoperative pain in children undergoing major surgery: a randomized effectiveness of educating parents about distraction to decrease controlled trial. Pediatr Surg Int 2015;31:197–201. postoperative pain in children at home after tonsillectomy. Pain Manag [42] Syrjala KL, Abrams JR. Hypnosis and imagery in the treatment of pain. In: Turk Nurs 2014;15:632–40. DC, Gatchel RJ, editors. Psychological approaches to pain management, [16] Higgins JPT, Green S. Cochrane handbook for systematic reviews of practitioner’s handbook. New York: The Guilford Press, 2002:187–209. interventions version 5.1.0. The Cochrane Collaboration 2011. Available [43] Taddio A, Appleton M, Bortolussi R, Chambers C, Dubey V, Halperin C, at: http://www.cochrane-handbook.org. Accessed January 2013. Hanrahan A, Ipp M, Lockett D, MacDonald N, Midmer D, Mousmanis P, [17] Huth MM, Broome ME, Good M. Imagery reduces children’s post- Palda V, Pielak K, Pillai Riddell R, Rieder M, Scott J, Shah V. Reducing the operative pain. PAIN 2004;110:439–48. pain of childhood vaccination: an evidence-based clinical practice [18] Huth MM, Broome ME, Mussatto KA, Morgan SW. A study of the guideline. CMAJ 2010;182:1168–1198. effectiveness of a pain management education booklet for parents of [44] Taylor EM, Boyer K, Campbell FA. Pain in hospitalized children: children having cardiac surgery. Pain Manag Nurs 2003;4:31–9. a prospective cross-sectional survey of pain prevalence, intensity, [19] Jaaniste T, Hayes B, von Baeyer CL. Providing children with information assessment and management in a Canadian pediatric teaching about forthcoming medical procedures: a review and synthesis. Clin hospital. Pain Res Manag 2008;13:25–32. Psychol Sci Pract 2007;14:124–43. [45] Taylor A, Stanbury L. A review of postoperative pain management and the [20] Kim H, Jung SM, Yu H, Park S. Video distraction and parental presence challenges. Curr Anaesth Crit Care 2009;20:188–94. for the management of preoperative anxiety and postoperative [46] Tomlinson D, von Baeyer CL, Stinson JN, Sung L. A systematic review of behavioural disturbance in children: a randomized controlled trial. faces scales for the self-report of pain intensity in children. Pediatrics Anesth Analg 2015;121:778–84. 2010;128:e1168–98. [21] Kokki H. Current management of pediatric postoperative pain. Expert Rev [47] Tsze DS, Hirschfeld G, von Baeyer CL, Bullock B, Dayan PS. Clinically Neurother 2004;4:295–306. significant differences in acute pain measured on self-report pain scales in [22] Deleted in proof. children. Acad Emerg Med 2015;22:415–22. [23] Lambert SA. The effects of hypnosis/guided imagery on the [48] Ullan ´ AM, Belver MH, Fernandez ´ E, Lorente F, Badia M, Fernandez ´ B. postoperative course of children. J Dev Behav Pediatr 1996;17:307–10. Theeffectofaprogramtopromoteplaytoreducechildren’spost- [24] LaMontagne LL, Hepworth JT, Cohen F, Salisbury MH. Cognitive- surgical pain: with plush toys, it hurt less. Pain Manag Nurs 2014;15: behavioral intervention effects on adolescents’ anxiety and pain following 273–82. spinal fusion surgery. Nurs Res 2003;52:183–90. [49] Uman LS, Birnie KA, Noel M, Parker JA, Chambers CT, McGrath PJ, [25] Li HCW, Lopez V, Lee TLI. Effects of preoperative therapeutic play on Kisely SR. Psychological interventions for needle-related procedural pain outcomes of school-age children undergoing day surgery. Res Nurs and distress in children and adolescents. Cochrane Database Syst Rev Health 2007;20:320–32. 2013;10:CD005179. [26] Deleted in proof. [50] Unsworth V, Franck LS, Choonara I. Parental assessment and [27] MacLaren JE, Cohen LL. Interventions for paediatric procedure-related management of children’s postoperative pain: a randomized clinical pain in primary care. Paediatr Child Health 2007;12:111–16. Available at: trial. J Child Health Care 2007;11:186–94. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2528901/. Accessed [51] von Baeyer CL, Spagrud LJ. Systematic review of observational February 2014. (behavioral) measures of pain for children and adolescents aged 3-18. [28] Martin S, Smith AB, Newcomb P, Miller J. Effects of therapeutic PAIN 2007;127:140–50. suggestion under anesthesia on outcomes in children post [52] von Baeyer CL, Spagrud LJ, McCormick JC, Choo E, Neville K, tonsillectomy. J Perianesthes Nurs 2014;29:94–106. Connelly MA. Three new datasets supporting use of the numerical [29] McGrath PJ, Walco GA, Turk DC, Dworkin RH, Brown MT, Davidson K, rating scale (NRS-11) for children’s self-reports of pain intensity. PAIN Eccleston C, Finley GA, Goldschneider K, Haverkos L, Hertz SH, 2009;143:223–27.

Copyright Ó 2016 by the International Association for the Study of Pain. Unauthorized reproduction of this article is prohibited.