Chronic pain in pediatric patients is a multidimensional experience, involving interplay between nociceptive processing, affect, sociocultural context, and behavioral and cognitive mechanisms.28,84 Consequently, a variety of biopsychosocial variables including depression, anxiety, low self-esteem, sleep disturbances, fatigue, and decreased physical functioning interact to affect the functioning and health of children and adolescents living with chronic pain.23,31,38,42,47,75 Because the pediatric chronic pain population often receive a variety of ineffective treatments by nonspecialized providers,84 it is probable that treatment failure is related to limited tailoring of biopsychosocial needs. For these reasons, multimodal biopsychosocial approaches are considered gold standard for chronic pain treatment, thus interdisciplinary chronic pain programs are an ideal setting for this patient population.65 Because the spectrum of disability between pediatric patients with chronic pain widely varies,78 accurate measurement of the biopsychosocial impact of pain may help clinicians with referral processes, prioritization of care, and ongoing assessment of patient response to treatment within such chronic pain programs.
Existing systematic reviews have highlighted the psychometric qualities of single-item pain intensity scales,7,43,66 observational pain measures for children and adolescents,76 and tools that measure parent response to their child's pain.33 Reviews have also shown that most adolescent chronic pain assessment tools focus solely on the psychological domain. A comprehensive understanding of multi-item tools that measure the biopsychosocial impact of pain across multiple domains of the pediatric chronic pain experience is lacking. Such a review is needed to illuminate the ways in which biopsychosocial variables are interpreted and weighed in predicting patient complexity in the pediatric population with chronic pain, which may help to inform prioritization of care into and within interdisciplinary pediatric chronic pain programs.
The Multidimensional Biobehavioral Model of Pediatric Pain73 is a framework to support the consideration of chronic pain as a biopsychosocial phenomenon. The model was specifically developed to account for the wide variability of pain perception, pain behavior, and functional status.75 It has been used to identify the factors associated with pain intensity and functional disability in a variety of pediatric chronic pain disorders.75 The model categorizes variables of the pain experience into the following domains: (1) precipitants, including pain related diagnosis or disease, injury, stress, and/or pain-producing procedures; (2) intervening variables, including biological predispositions, family environment, cognitive appraisal, coping strategies, and perceived social support; (3) pain perception and behavior; and (4) functional status, including activities of daily living, school attendance, depressive symptoms, anxious symptoms, behavioral problems, and interpersonal relations.
1.1. ObjectivesThe specific objectives of this review were to (1) identify multidimensional biopsychosocial assessment tools used in the pediatric (ie, ages 2–18 years) population with chronic pain; (2) describe the relationships between chronic pain and the biopsychosocial domains (precipitant variables, intervening variables, and functional status variables) measured in each tool, as defined by the Multidimensional Biobehavioral Model of Pain; and (3) review the reliability and validity evidence of such tools and their biopsychosocial domains in the pediatric population with chronic pain.
2. Methods 2.1. Design and reportingWe conducted a systematic review in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 checklist54 which is outlined in Supplementary File 1 and Supplementary File 2 (available at https://links.lww.com/PR9/A207).
2.2. Eligibility criteriaMultidimensional biopsychosocial assessment tools included in this review were those that (1) included variables reflecting at least 2 of the Multidimensional Biobehavioral Model of Pain domains of pain, precipitant variables, intervening variables, and functional status variables; (2) were developed specifically to measure the impact of pain rather than general functional interference not specific to pain; and (3) intended for use in patients with primary chronic pain diagnoses (as defined by the International Classification for Disease (ICD-11) classification).72 We excluded disease-specific tools, such as those for children with sickle cell disease or cancer and parent proxy measures.
Our inclusion criteria for studies describing these tools included the following: (1) The population exclusively involved pediatric patients (ages 2–18) with primary idiopathic chronic pain diagnoses/locations. (2) The outcomes of the study focused on the relationship between pain intensity and the multidimensional biopsychosocial items captured by the tool under investigation. (3) We excluded systematic reviews, meta-analyses, case studies, abstracts, and qualitative studies that did not include a psychometric outcome, such as a Cronbach alpha. We also excluded studies where the participant group included 50% or more of youth with secondary pain diagnoses related to an organic disease process such as cancer, sickle cell disease, juvenile rheumatoid arthritis, neurofibromatosis, or postsurgical pain.
2.3. Search strategy and information sourcesThe search strategy included 2 phases. Phase 1 searching focused on tool identification and was conducted in 2 measurement databases: PsychTEST and Health and Psychosocial Instruments (HAPI). Phase 2 searching focused on study identification, which involved a measure-forward search through 2 citation databases: Scopus and Web of Science. After eligible tools were selected in phase 1, a citation search of their development article was then conducted during phase 2. This search strategy was led by a librarian employed at The Children's Hospital of Eastern Ontario and was PRESS reviewed by a librarian employed at The Ottawa Hospital Research Institute. Both phases of this search were first conducted in February 2020 and repeated in February 2022. Details of the search strategy can be found in Supplementary File 3, available at https://links.lww.com/PR9/A207.
2.4. Study selectionA detailed instruction manual was developed based on the study eligibility criteria to guide the screening and retrieval process for phase 1 and phase 2 searches. One reviewer screened all results from the phase 1 search and a second reviewer confirmed inclusion or exclusion decisions made. All tools identified in the reference articles were listed in a Microsoft Excel document, which organized tools based on eligibility and included data on tool properties (ie, tool name, development reference, biopsychosocial domains/variables, and how they mapped to the Multidimensional Biobehavioral Model of Pediatric Pain).
After the citation analysis of the development articles for each included tool conducted in phase 2, all articles were uploaded to Covidence software to their respective project (ie, each tool represented its own project in Covidence). Two reviewers independently screened titles and abstracts. Full-text citations that met eligibility were then independently reviewed by 2 reviewers who further searched the text for additional eligible tools that may have been missed in the phase 1 search. Reviewers met on a biweekly basis to discuss discrepancies in eligibility assessments. All discrepancies were considered minor and were resolved. Details justifying elimination for excluded tools are listed in Supplementary File 4 and for excluded citations in Supplementary File 5 (available at https://links.lww.com/PR9/A207).
2.5. Data collectionA detailed data extraction instruction manual and data collection form was developed based on the study outcomes and contextual data. The data extraction process and form were piloted by 2 reviewers on 5 studies to ensure reliability of the data extraction instructions. Minor revisions were made. Two reviewers independently extracted the data based on explicit reporting of the following:
(1) Study characteristics, which included authors' names, year of publication, study purpose, population demographics, and methods (ie, study design). (2) Reliability evidence (as defined by the Standards for Educational and Psychological Testing).1 All reliability evidence per the standards was possible for extraction. (3) Validity evidence (as defined by the Standards for Educational and Psychological Testing).1 All validity evidence per the standards was possible for extraction. (4) Clinical utility, as defined by Smart, 2006,63 which included data on tool appropriateness, accessibility, practicality, and acceptability.Results from data extraction were compiled into summary tables, which were iteratively refined to best prepare for data synthesis and narrative description.
2.6. Methodological qualityMethodological quality and bias of all studies were assessed by 2 reviewers independently based on their study design and was guided respectively by the Quality Assessment and Validity Tool for Cross-sectional Studies,19 the Quality Assessment and Validity Tool for Before and After/Cohort Design Studies,65 and the Revised Cochrane Risk of Bias Tool for Randomized Trials.35 Studies were concluded to be weak, moderate, or strong based on the quality assessment tools.
2.7. SynthesisThe results from this review were synthesized descriptively. A meta-analysis was not appropriate because of the heterogeneity of studies describing the psychometric qualities of included multidimensional tools. Our synthesis involved a description of the relationships between pain and the biopsychosocial domains and subsequent variables that were measured across multidimensional tools. Our analysis focused on the validity evidence of each tool as it related to other variables. Significant bivariate and multivariate relationships were highlighted and tabulated across each pain domain and subsequent variable defined by the Multidimensional Biobehavioral Model of Pediatric Pain.73 Reported measures of tool reliability were tabulated in a descriptive form. Cronbach alphas were considered adequate if over and/or above 0.70.25 A sensitivity analysis was not appropriate because only 1 weak study was included in this review, and results were unlikely to change based on its removal.
3. Results 3.1. Tool and study selectionIn the phase 1 search, 614 reference articles were screened for eligible tools. From this, 14 of 93 identified tools met eligibility criteria. Five of those tools were excluded because we could not locate the tool despite attempts to contact original authors. Details justifying exclusion for all tools reviewed can be found in Supplementary File 4 (available at https://links.lww.com/PR9/A207). In the phase 2 search, 1029 titles and abstracts, and 973 full-text articles were screened across 9 tools. This led to further exclusion of 3 tools that did not yield outcome data. Therefore, 6 tools were included in our synthesis, which generated a total of 64 eligible studies. Details of search results can be found in Figure 1. Details justifying exclusion of studies can be found in Supplementary File 5 (available at https://links.lww.com/PR9/A207).
Figure 1.:Search results. HAPI, Health and Psychosocial Instruments; PROMIS, Patient-Reported Outcomes Measurement Information System.
3.2. Included toolsA summary of tool characteristics, including number of variables and domains as well as respective mapping to the Multidimensional Biobehavioral Model of Pediatric Pain73 is shown in Table 1. Variables reflecting the precipitant pain experience domain were not included in any of the 6 included tools. Results did not generate significant outcomes regarding clinical utility of tools, and thus information pertaining to clinical utility is described narratively below.
Table 1 - Tool characteristics. Tool Development article No. of items No. of domains Scoring method Biopsychosocial domains Precipitant variables Intervening variables Functional variables BATH Adolescent Pain Questionnaire (BAPQ) Eccleston et al.16 61 7 5-point Likert scale 0 44 items across 5 domains 17 items across 3 domains PROMIS Pediatric Pain Interference Scale (PPPI) Varni et al.74 8 N/A 4-point Likert scale 0 3 items 5 items Child Activity Limitations Questionnaire (CALQ) Hainsworth et al.32 21 N/A 5-point Likert scale 0 4 items 18 items Pain Interference Index (PII) Wicksell et al.80 6 N/A 7-point Likert scale 0 2 items 4 items Pain Experience Questionnaire (PEQ) Hermann et al.34 15 N/A 7-point Likert scale 0 3 items 12 items Pain-Related Problem List for Adolescents (PRBL-A) Weel et al.82 18 4 3-point Likert scale 0 5 items across 1 domain 13 items across 3 domainsAmong the 64 included studies, 46 were cross-sectional studies, 9 were cohort studies, 5 were randomized controlled trials, 3 were nonrandomized before and after studies, and 1 was a qualitative study. Most studies were conducted in the setting of tertiary-level outpatient pediatric chronic pain programs (n = 39) and included intensive inpatient pediatric chronic pain programs (n = 5), other speciality clinics (n = 8), research clinics (n = 1), schools (n = 2), and a data registry (n = 1). Nine studies did not specify their setting. Six studies were found to be secondary analyses of other included studies, and therefore, sample characteristics were not duplicated in our synthesis. Across the 64 studies included, a total of 19,429 participants aged between 6 and 19 years. Most participants across studies that reported sex and ethnicity were female (median 73%) and Caucasian (median 85%). Most participants in all studies had primary chronic pain diagnoses unrelated to an underlying condition, whereas 11 studies included less than 50% of participants with secondary chronic pain diagnoses and 7 studies included a description of comorbid mental health diagnoses. Study characteristics are reported in Table 2.
Table 2 - Study characteristics. Tool Study characteristics Sample characteristics Article Study type Setting Sample size Age (y)
留言 (0)