Abstract
Purpose:
Chronic pain experienced by children and adolescents represents a significant burden in terms of health, quality of life, and economic costs to U.S. families. In 2015, the Boston Medical Center (BMC) Interdisciplinary Pain Clinic initiated an Integrative Medicine (IM) team model to address chronic pain in children. Team members included a pediatrician, child psychologist, physical therapist, acupuncturist, and massage therapist. Children were referred to the pain clinic from primary care and specialty services within BMC, the largest safety-net hospital in the northeastern United States. For this observational assessment, consent and assent were obtained from parents and pediatric patients. Individualized treatment plans were recommended by the IM team.
Methods:
Self-reported survey and electronic medical record data were collected about socioeconomic demographics, pain, use of medical and IM services, and quality of life. The authors compared health and quality of life indicators and costs of care for each participant from the year before entering the project with these same indicators for the subsequent year.
Results:
Eighty-three participants were enrolled. Participants ranged in age from 4 to 22 years (mean 14.7 years). Eighty percent of the group were females. Forty-two percent of the sample were white, 30% were Hispanic/Latinx, and 28% were African American. Primary types of pain were abdominal (52%), headache (23%), musculoskeletal (18%), and other (7%). Quality of life indicators improved (p = 0.049) and pain interference decreased (Wilcoxon p = 0.040). Major economic drivers of cost were emergency department (ED) visits, inpatient hospitalizations, and consultations with medical specialists. For the 46 participants who completed the project, the following total cost savings were noted: $27,819 (surgeries), $17,638 (ED visits), $25,033 (hospitalizations), and $42,843 (specialist consults). No adverse events were reported.
Conclusion:
The authors' experience demonstrated that the use of IM approaches in an interdisciplinary team approach is safe, feasible, and acceptable to families. Considerable cost savings were observed in the area of surgical procedures, hospitalizations, and consultations with specialists.
Introduction
Pediatric chronic pain is an emerging global health problem with estimates of 20%–35% of children and adolescents being affected. 1 –3 Chronic pain frequently interferes with everyday functioning of children in significant ways. 4 As the authors' group has shown previously, children experiencing chronic pain have increased stress, including challenges with school attendance, social interactions, bullying, and parental stress. 5 Chronic pain in children is often characterized by high rates of functional impairment and increased health care utilization. 6 If pain is not meaningfully addressed, children may also develop lifelong health-related problems.
Chronic pain among children costs the economy more than the costs related to pediatric heart disease, cancer, and diabetes combined. 7 These costs include medical services that are defined as health care resource utilization associated with any condition incurred from inpatient admissions, emergency department (ED) visits, diagnostic tests, outpatient procedures (e.g., endoscopies and colonoscopies), and laboratory tests. Furthermore, parents of children who use these services are often required to take time off work and are more likely to experience decreased wages. 8 –11
Owing to the costly nature of pediatric chronic pain, there is growing evidence about adopting comprehensive interdisciplinary treatment approaches that include pharmacologic therapies, psychological therapies, lifestyle changes, and nonpharmacologic approaches such as Integrative Medicine (IM) therapies. 12,13 Besides economic benefit, many pediatric pain clinics that utilize these approaches are also effective in addressing pain-related disabilities, school attendance, pain levels, and psychological distress. 14 –16 However, many hospitals and health systems that care for children hesitate to offer the full range of these services, especially to patients receiving Medicaid whose services are reimbursed at low rates. 17,18
This issue is even more pressing among children whose families have low socioeconomic status (SES) and are covered by public health insurance or have no insurance. 19 Today much of the treatment of pediatric chronic pain takes place in primary care settings where reassurance and medications are the mainstays of treatment. Many nonpharmacologic treatments such as acupuncture and mind–body techniques may not be easily accessible or covered by health insurance. 20 Families with low SES often lack access to local or financial resources that can be used to treat their children's chronic pain due to barriers with transportation to appointments, out-of-pocket expenses, and delays with referral processes. 21
The aims of this observational assessment are threefold: (1) to describe the patient populations' demographics and primary pain types, (2) to determine changes in pain interference and quality of life associated with the use of IM services, and (3) to assess the potential changes in expenditures related to hospital costs, ED use, and visits to medical specialists within a cohort of pediatric patients with chronic pain at Boston Medical Center's (BMC) Interdisciplinary Pediatric Pain Clinic.
Methods
Setting
This study was conducted at BMC, a not-for-profit academic medical center, which is the largest safety-net hospital in New England. Approximately 70% of the patients come from medically underserved populations. The data for this study were collected through self-reported surveys and chart review from the patients within BMC's Interdisciplinary Pediatric Pain Clinic.
Participant recruitment
Potential participants were children and adolescents with ages ranging from 4 to 22 years who were receiving care at BMC's Interdisciplinary Pediatric Pain Clinic for chronic pain. Non-English-speaking patients were also included and translation was provided by medical interpreters. For children under the age of 18, parents or guardians provided informed consent for their child to join the study, although these children provided assent for themselves. Informed consent was also obtained from participants over the age of 18 years by trained research assistants (RAs) who were responsible for all consent and assent procedures. Participants could be referred to the IM clinic by clinicians from various specialties within BMC (e.g., gastroenterology, primary care, hematology, rheumatology, and neurology) or from other clinics in the Boston area.
Intervention
BMC's Interdisciplinary Pediatric Pain Clinic provides medical and psychological services, patient education, and support for managing chronic, recurrent, and complex pain. The clinic is staffed by a pediatric hematologist and oncologist, a licensed psychologist, a nurse educator, a licensed clinical social worker, a physical therapist, a licensed acupuncturist, a certified clinical aroma therapist, a registered dietician, a licensed massage therapist, and a karate instructor. All practitioners had training and/or experience working with children. The Clinical Team (pediatric doctor, pediatric psychologist, nurse educator, and licensed social worker) performed an initial assessment.
The clinical team would then formulate an integrative treatment plan with medical and behavioral interventions as well as other recommendations, which could include referrals to physical therapy, acupuncture, massage therapy, manipulation, Reiki therapy, aromatherapy, dietary supplements, nutrition consults, karate classes, and mind–body therapies (meditation, biofeedback, body scans). After the initial evaluation, the participant would meet with the pediatric psychologist for additional evaluations and family assessment. Appointments for therapies at BMC would be arranged and scheduled.
Since this was not a clinical trial, types of interventions were not randomly assigned or compared. The Clinical Team made recommendations based on previous experience of the IM Pediatric Clinic.
Data collection
Upon arrival to the Interdisciplinary Pediatric Pain Clinic, parents and children were approached by study RAs for informed consent/assent. After the patient was enrolled in the study, they or their parents were asked to fill out an intake questionnaire. Participants would follow up with the RA at ∼3 months in person or by phone.
Outcome measures
Self-reported outcomes
After informed consent was obtained, questionnaires collected information, including age, sex, race, ethnicity, primary type of insurance, and types of pain.
Quality of life was assessed using the Pediatric Quality of Life Inventory (PEDSQL) Child Form, which consists of four scales (Physical Functioning, Emotional Functioning, Social Functioning, and School Functioning). Items are reverse scored and linearly transformed to a 0–100 scale. Higher scores indicate better quality of life. Per each scale, the mean is computed as the sum of the items over the number of items answered. The estimate for the PEDSQL represents the mean value as computed by determining the sum of the items and dividing this by the number of items on the scale.
Pain-related factors were assessed using the
Chart review
The authors extracted the following information for each participant in the previous 1-year period: number of ED visits, hospitalizations, surgical procedures, and consultations with medical specialists. RAs retrieved these data through electronic medical record (EMR) review. This was completed using BMC's EMR system, EPIC Systems Information Technology.
Economic outcomes
Medical economic data included the following costs: surgical procedures, visits to the ED, inpatient hospitalizations, and consultations with medical specialists. Although other costs, such as medications, medical devices, and nutritional dietary supplements, were incurred by families, data on these expenses were not available. Costs of IM interventions were not factored into the economic analyses because these were provided free of charge as they were funded by a teaching clinic and philanthropy.
Specific dollar amounts related to direct costs of the major drivers of medical care (surgical procedures, ED visits, inpatient stays, and consultations with specialists) were not directly available in the authors' database. Using data collected on these interventions, however, the authors used cost estimates from third-party payers for the Boston MA area during the 2016–2017 period. 23 –28
Data were collected on medical procedures throughout the course of participants' enrollment in the study and for up to 1 year after enrollment. Baseline cost data consisted of costs incurred in the 1-year period before study enrollment. Total costs for major medical procedures were determined according to primary pain category (i.e., abdominal, headache, and musculoskeletal, or other). Average costs at baseline and at study completion were determined and reported according to pain category. Estimates of costs also included ranges of low cost, average cost, and high cost that were provided from insurance source references.
Statistical analysis
Descriptive statistics were used to analyze survey data and chart review information. Means, standard deviations, frequencies, and percentages were calculated for participants' demographics and descriptive information. Two-sample t test and Wilcoxon rank-sum test were used to compare PEDSQL and Pain Interference from baseline to follow-up using a statistical significance level of 0.05.
All quantitative analyses were performed using SAS 9.3 software and p-values were compared with a significance value (α) of 0.05.
Economic analysis
Cost analysis was conducted from the perspective of the health care system. Information on medical interventions, procedures, and consultations was obtained directly from participants' EMR. These data were based on self-report of participating families. Although economic data on medical interventions, procedures, and consultations were not directly available from the EMR, cost estimates were obtained through use of Massachusetts-based insurances sources as described in Economic Outcomes section.
The major contributors to medical costs comprised the areas of interest for economic comparison. Sources of cost data provided averages and ranges for each type of procedure and intervention.
Costs of procedures were summed according to participants' principal pain category as well as according to type of procedure. Sums at baseline were compared with sums at completion of each participant's time in the study. The differences were calculated by subtracting Total Costs at Follow-Up from Total Costs at Baseline. The authors also calculated average costs per participant according to pain classification.
This study was approved by the Institutional Review Board of the Boston University School of Medicine.
Results
One hundred twenty-one children were referred to the pediatric pain clinic; 89 parents and children were screened for the study by an RA. Five families were not interested in joining the study, and one child under the age of 18 years was unable to be assented; thus, 83 children were enrolled. Forty-six participants completed both pre- and postsurvey (data for cost analysis); however, 32 and 33 participants, respectively, were able to complete the PEDSQL Child form and the PROMIS Pediatric Pain Interference Scale.
Baseline demographics are stratified by pain type as presented in Table 1. These pain types comprised abdominal (52%), headache (23%), and musculoskeletal (18%), or other (7%). Participants' mean age was 14.7 years, and 80% of participants identified as female. Forty-three percent of participants identified as white, 26% as African American, and 31% as other. Thirty percent of participants identified as Hispanic, and 73% of all the participants had Medicaid as their primary insurance.
Demographics at Baseline by Pain Type (Survey Data) (n = 83)
Other category includes sickle cell disease (n = 2), chest pain (n = 2), epidermolysis bullosa (n = 1), diabetic peripheral neuropathy (n = 1).
At baseline, upon enrolling in the authors' project, medical costs for the entire cohort of 83 participants were estimated at $312,389 (range: $249,912–$374,866). Estimate of total costs related to surgical procedures (endoscopies, colonoscopies, and appendectomies) was $72,046 (range: $57,637–$86,455). Total costs associated with visits to the ED were $106,812 (range: $85,450–$128,174). Inpatient hospitalization costs were estimated at $86,776 (range: $69,421–$104,131). Costs related to consultations with medical specialists were estimated at $46,755 (range: $37,404–$56,106).
Table 2 outlines cost estimates at baseline (with associated high and low estimates from the insurance sources) according to type of pain reported; these represent costs incurred in the previous 1-year period related to type of medical intervention. The largest contributor to these baseline costs was visits to the ED, which totaled $106,812 (range: $85,450–$128.174). Although specific costs of the integrative services were unavailable due to the multiple revenue streams, including philanthropy that funded the program, ∼1245 U of service (visits for physical therapy, acupuncture, and massage therapy) were provided.
Cost Estimates for Medical Interventions at Baseline (n = 83)
Cost data are based on utilization of the interventions listed in the previous year before entering the Pediatric Integrative Medicine project. Costs ascertained through insurance estimates as described in Economic Outcomes section.
Endoscopy, 1bColonoscopy, 2 Appendectomy, 3 Cost of ED visits, 4 Cost of Hospitalizations, 5 Cost of Consults with Specialists: All these estimates are based on cost figures from citations 23–28.
Ranges refer to low and high values of estimates for each procedure.
Values are bolded to promote ease of reading. These bolded values indicate average costs.
Changes in PEDSQL scores and Pain Interference scores are represented in Table 3. There was a significant increase in Total PEDSQL (p = 0.0493) from baseline to follow-up. There was also a significant reduction in Total Pain Interference (p = 0.0405) from baseline to follow-up.
Comparison of Overall Quality of Life and Pain Interference Scores for Study Participants Who Completed Follow-Up Questionnaires
Not all participants completed follow-up questionnaire despite repeated attempts to contact them.
For the 46 participants who completed enrollment and were followed for up to 1 year postproject, costs were significantly reduced. The 21 participating patients with abdominal pain incurred baseline costs of $66,530 (range: $50,743–$76,934). After participation in the project, their corresponding medical costs were $21,346 (range: $15,275–$27,416). This represents a 66% reduction in costs of major medical interventions for abdominal pain participants. Twelve participating patients experiencing musculoskeletal pain incurred medical costs of $60,369 (range: $50,532–$78,688), whereas the costs of their postproject medical interventions totaled $16,420 (range: $9,523–$23,336), representing a 73% cost reduction.
The 10 participating patients with headache as their primary pain had initial medical expenses of $25,586 (range: $13,919–$33,276). After participation in the project, major medical expenditure cost was reduced to $8,008 (range: $4,576–$11,440), a reduction of 69%.
Figure 1 summarizes these findings.

Comparison of baseline and follow-up costs of medical interventions (surgical procedures, visits to ED, inpatient hospitalizations, and consultations with medical specialists) baseline costs represent costs of procedures in the previous year before participants joined the project. Follow-up costs include costs of procedures related to the time of starting the project up to 1 year following their time in the project (n = 46). Values represent average costs of procedures as described in Economic Outcomes section. ED, emergency department.
Table 4 shows the comparison between baseline and follow-up costs categorized according to type of intervention for the 46 patients who completed the program. Using the EMR the authors were able to obtain complete information on utilization (surgical procedures, ED visits, hospitalizations, and consultations with medical specialists) for 46 participants. If EMR did not contain information for the full 1-year period, the authors did not include this participant's involvement in the economic analysis.
Comparison of Baseline Versus 1 Year Postcompletion Costs
Values are bolded to promote ease of reading. These bolded values indicate average costs.
ABD refers to abdominal pain.
MS refers to musculoskeletal pain.
HA refers to headache.
ED, use of emergency department; Hosp, inpatient hospitalization; Spec, consultations with medical specialists; Surg, surgical procedures.
Table 5 summarizes the average cost comparisons for the 46 participants for whom complete data on surgical procedures, visits to the ED, hospitalizations, and specialist consults were available in EMR for the 1-year period before enrollment in this study as well as 1 year after involvement in the study.
Comparison of Baseline Average Cost/Patient Versus Average Cost/Patient for 1 Year Post-Treatment (n = 46)
Includes 46 participants for whom data were available for both the 1-year period preceding their involvement in the project and the 1-year period after their participation in the project. Costs represent the sum of cost estimates for surgeries, ED visits, hospitalizations, and consultations with specialists.
BMC, Boston Medical Center.
No adverse events related to the project were reported throughout or after the study period.
Discussion
This study assessed baseline demographics, changes in pain and quality of life, and economic effects related to treatment in an Interdisciplinary Pediatric Pain Clinic providing services to a low-SES, racially diverse group of children experiencing chronic pain. After treatment in the Interdisciplinary Pediatric Pain Clinic the authors observed decreased need for costly health care interventions. Results are promising because they suggest that more effective management of pain and its comorbidities can lead to fewer ED visits and inpatient hospitalizations.
Pain is considered the most prevalent symptom among pediatric patients presenting to the ED in the United States, particularly in the context of chronic pain complaints. Costs associated with inpatient services are considerably greater than costs associated with outpatient visits. If pain is being well managed through treatment delivered as outpatient visits, it is more likely that patients would not present to the ED nor be admitted for lengthy inpatient stays. Effective pain management may also improve patients' ability to engage in academic and social activities, thus potentially improving their overall daily functioning and quality of life. The authors observed this in this study.
The authors' evaluation indicates important savings related to costly medical interventions. From a public health standpoint, before assessing changes in costs, the authors want to be assured that the approach and interventions studied are safe, provide clinically favorable outcomes, and are acceptable to families and pediatric patients. The authors' project fulfilled these public health requirements and also demonstrated sizeable reduction in costs, while also favorably affecting quality of life and decreased pain interference.
For the group of 46 participants who completed follow-up, costs of surgical interventions, visits to the ED, hospitalizations, and consultations with specialists decreased without jeopardizing their health and well-being. The additional benefits conferred by using an integrative approach may include improved quality of life as well as decreased interference of pain in daily life.
Although the authors' analysis was conducted from the health care system perspective and did not include costs incurred by families or other opportunity costs, other investigators have offered estimates related to lost work. Evans and colleagues estimate that the cost of lost parental missed work due to youth chronic pain decreased by $11,040 per participating family when comparing the 12 months before to the 12 months after admission to a pediatric rehabilitation program. 13
Studies indicate that parents in low-SES families show elevated anxiety as they experience more stressors and environmental challenges. 29 Some parents suffer from their own chronic pain, which may be a risk factor for children's chronic pain and lower quality of life. 30 A recent review found that children whose parents have chronic pain are more likely to report greater pain complaints as well. 31,32
Although some may argue that chronic pain in children may resolve itself within the time framework of this study, other investigators have found that chronic pediatric pain is not likely to be resolved on its own. 33 Thus, the authors do not expect a “regression to the mean” type effect in this sample.
The authors' research also identified areas in which to devote future research efforts. Visits to the ED, hospitalization, and cost of consultations with medical specialists were identified as the main contributors to high medical costs in this pediatric cohort; these results are concordant with the findings of other researchers. 34 In particular, focusing on pain-related visits to the ED could have important favorable economic consequences. Enhanced communication between Family Medicine providers, surgeons, and other medical specialists may promote early identification of pediatric patients who may benefit from integrative approaches.
Limitations
Information on health care utilization may be underestimated since these utilization data were gathered only from the EMR and self-report at BMC. Data from other Boston-area hospitals were not collected, and it was unknown if participants visited or were admitted to other hospitals. Another limitation was that this study was unable to collect post-follow-up data on the full group of 83 participants despite multiple attempts to reach them by phone or written communication. Furthermore, the authors' evaluation only captured cost estimates for up to a single year after IM pain clinic service. Because of having funding resources for providing IM care in this project, the authors did not collect data related to utilization of IM services.
Costs of IM services provided were unavailable due to multiple factors. Because of the comprehensive nature of the services and involvement of multiple departments within the hospital (viz. Pediatrics, Family Medicine, and Pediatric Psychology) as well as allied training programs (Boston University School of Medicine and New England School of Acupuncture) the authors were unable to gather relevant cost information. Furthermore, services were provided through a number of grants.
Because this represents a significant limitation of this study, the authors acknowledge the study's inability to estimate costs of IM services. In lieu of reporting this information, the authors provided the number of units of service that were provided to participants. Cost information from commercial insurers was not available, a limitation faced not only by this study but also by U.S. researchers, in general. 35
It is not clear whether the long-term benefits of IM pain clinic participation continued over time. It would be advantageous for future studies to measure these effects over longer timeframes. Future studies with larger sample sizes are needed to best analyze health care cost data. Cost parameters not observed in this study included medication costs, costs of visits to community clinics or other off-site centers not affiliated with this project, and costs to families associated with transportation and missed work. These costs are not insignificant and are worthy of attention and inclusion in future studies.
The authors report costs according to 2017 estimates to preserve original levels of third-party reimbursement. To adjust for inflation from 2017 to 2021, estimates can be modified by multiplying by 10.84% and adding this figure to the authors' cost estimates. 36,37
Conclusion
Using an integrative interdisciplinary approach combined with integrative therapies, psychological counseling, and pharmacologic methods can safely and effectively address pediatric patients' pain and related health status. Findings from the authors' Interdisciplinary Pediatric Pain Clinic support that over the course of 1 year, participation in such services may significantly improve quality of life and lessen pain interference. Reduced costs related to high levels of medical care utilization suggest that an Interdisciplinary Pediatric Pain Clinic may be an effective intervention for racially diverse children with chronic pain who come from economically challenged families.
Footnotes
Author Disclosure Statement
No competing financial interests exist.
Funding Information
Funding for the project was provided by the Marino Foundation.
