The Economic and Humanistic Burden of Pediatric-Onset Multiple Sclerosis

Background: Multiple sclerosis (MS) is a chronic inflammatory autoimmune disease of the central nervous system. Pediatric-onset MS (POMS), defined as onset of MS before 18 years of age, is estimated to account for 2% to 5% of the MS population worldwide. Objectives: To conduct a literature review focused on the healthcare resource utilization and cost as well as quality-of-life (QOL) outcomes among patients with POMS. Methods: We conducted a systematic literature review of English-language studies published after September 2010 in MEDLINE and Embase to describe the global economic healthcare resource utilization and costs and humanistic (QOL) burden in patients with POMS. Results: We found 11 studies that reported on healthcare resource utilization, cost, or insurance coverage and 36 studies that reported on QOL outcomes in patients with POMS. Patients with POMS had higher rates of primary care visits (1.41 [1.29-1.54]), hospital visits (10.74 [8.95-12.90]), and admissions (rate ratio, 4.27 [2.92-6.25];OR, 15.2 [12.0-19.1]) compared with healthy controls. Mean per-patient costs in the United States were $5907 across all settings per year of follow-up between 2002 and 2012; mean costs per hospital stay were $38 543 (in 2015 USD) between 2004 and 2013. Three studies reported psychosocial scores between 71.59 and 79.7, and 8 studies reported physical health scores between 74.62 to 82.75 using the Pediatric Quality of Life Measurement Model (PedsQLTM). Twelve studies used the PedsQL™ Multidimensional Fatigue Scale. Mean scores on the self-reported general fatigue scale ranged from 63.15 to 78.5. Quality-of-life scores were lower than those of healthy controls. Discussion: Our review presents a uniquely broad and recent overview of the global economic and humanistic burden of patients with POMS. Additional research on healthcare resource utilization and cost would provide a more robust understanding of the economic burden in this population. Conclusions: Healthcare resource utilization and costs are high in this population, and patients report reduced QOL and significant fatigue compared with healthy children and adolescents.


BACKGROUND
Multiple sclerosis (MS) is a chronic inflammatory autoimmune disease of the central nervous system with onset of symptoms usually occurring in adults between 20 to 40 years old. 1,2 Pediatric-onset MS (POMS), defined as onset of MS before 18 years of age, is estimated to account for 2% to 5% of the MS population worldwide. 1,3 Incidence and prevalence vary by geography and patient age. A recent systematic literature review and meta-analysis reported that incidence ranges from 0.05 to 2.85 per 100 000 children across 12 countries (in North America, Eu-rope, Middle East, and Asia); pooled global incidence was reported to be 0.87 per 100 000 children. 4 Prevalence ranged from 0.69 to 26.92 (pooled: 8.11) per 100 000 children. 4 Incidence and prevalence are higher in the United Arab Emirates, where incidence was reported to be 2.30 per 100 000 children between 10 to 14 years old and 7.20 per 100 000 adolescents between 15 to 19 years old, and prevalence was reported to be 30.7 per 100 000 children. 5 Children have higher relapse rates and a greater burden of lesions detected by magnetic resonance imaging than adults, 6,7 but disease progression tends to be slower. Due in part to the younger age at disease JOURNAL OF HEALTH ECONOMICS AND OUTCOMES RESEARCH onset, children reach disability milestones and secondary progression earlier than adults. Studies have also shown patients with POMS report reduced health-related quality of life (QOL) compared with healthy controls. 8 Similar to adults, children with POMS may also experience significant fatigue (some estimates report 20%-75% of patients), 9 resulting in further psychological challenges and reduced QOL. 10 Recent reviews have focused on characterizing how the disease course differs from adult-onset MS 1,11,12 and guiding physicians on disease management. 1 Reviews on healthcare costs of adult patients with MS have also been conducted in Spain 13 and the United States, 14 but no global reviews have explored the economic burden (eg, healthcare resource utilization and costs) in patients with POMS. Studies on QOL outcomes can provide a clear description of how MS impacts emotional, social, and physical functioning of pediatric patients. Reviews of QOL studies in POMS have focused on fatigue, 10 comparing neuropsychological correlates of adults vs children with MS, 8 and cognitive function. 9,15 This literature review focuses on the healthcare resource utilization and cost as well as QOL outcomes among patients with POMS. Reporting on these outcomes together can provide a comprehensive overview of the societal and individual costs of the disease, including the decrement in QOL.

METHODS
As part of a larger study, we conducted a systematic literature review to identify published information on the epidemiology, treatment patterns, and clinical, humanistic, and economic disease burden of 10-to 17-year-old patients with POMS. We searched MEDLINE (via PubMed) and Embase on September 27, 2020, to identify studies written in English, published in the last 10 years (September 27, 2010-September 27, 2020), that presented data on human patients with POMS between 10 to 17 years old from any country, and included at least 1 outcome of interest (epidemiology, treatment patterns, morbidity, mortality, comorbidities, QOL, healthcare resource utilization, healthcare costs, and health insurance coverage). We present studies that reported data on healthcare resource utilization, healthcare costs, health insurance coverage, and QOL to present all societal and individual costs (economic and humanistic) of POMS.
MeSH and text words associated with POMS (eg, "pediatric onset multiple sclerosis"[tw], "POMS"[tiab], "pediatric multiple sclerosis" [ Table S1. Researchers experienced in literature reviews screened articles in 2 phases: an initial title and abstract screen followed by a full-text screen. Articles that stratified data by the pediatric population and were not conference abstracts or pharmacokinetics/pharmacodynamics, case studies, or in vitro studies were included. References of these included articles and relevant literature reviews on POMS were mined (ie, reviewed and screened) to ensure no key articles were missed. We conducted the review using DistillerSR (Evidence Partners, Ottawa, Canada), a systematic review program. From all studies that met the inclusion criteria, we abstracted the study design, patient and control group population (including age ranges and MS type if specified), data source, country, dates of data collection, and relevant data on each outcome of interest. Supplemental Figure S1 depicts the number of articles screened and included.

RESULTS
Our full search resulted in 4599 unique articles, of which 202 were included in the review. Eleven studies reported on economic burden, including healthcare resource utilization and cost and/or health insurance coverage (Supplemental Table S2) and 36 studies reported on QOL outcomes (Supplemental Table S3), which are described in this manuscript. Some studies reported on more than 1 of these outcomes. Key findings across all studies are highlighted in Table 1.

Economic Burden
Eleven studies presented data on healthcare resource utilization, healthcare cost, and/or health insurance coverage (Supplemental Table S2). Studies most often collected data from patient populations in the United States (n = 4), followed by Canada (n = 2), Denmark (n = 2), Brazil (n = 1), and Switzerland (n = 1); 1 study collected data from multiple countries (United States, Italy, Russia, Argentina, France, Canada, Tunisia, and Venezuela). Seven studies included data only on patients at least 18 years old (diagnosed with MS as children), and 2 studies included patients who were slightly older (≤19 years) or slightly younger (≤17 years). The remaining 2 studies included patients of any age but stratified data by patients 18 years old or younger. Five studies included control groups, most commonly age-and sex-matched healthy controls. Outcomes from each study are presented in Tables 2 and 3.
Six studies [16][17][18][19][20][21] (from multiple countries) reported on inpatient hospitalizations with varied lengths of follow-up. For example, 1 study reported 0.5 hospital admissions per year among 92 patients in Denmark, 16 while 2 studies that collected data in the United States reported 1.2 20 and 6.0 17 mean hospital admissions per patient over the course of a 10-year study period (using the Pediatric Health Information System database, which includes data from many tertiary-care pediatric hospitals across the United States, and within a hospital system in the state of Utah, respectively). Other studies reported on the percentage of patients (out of varying sample sizes) hospitalized in their study cohorts (eg, 16.5% of patients [n = 659] were hospitalized during a 1-year baseline period 19 20 Three studies reported on healthcare costs; none compared the costs of patients with POMS with healthy controls. 17,20,23 Maia Diniz et al 23 reported that in Brazil, between 2000 and 2015, the mean annual cost per patient was $12 295.33 in 2017 US dollars (USD), or $14 638.05 in 2022 USD. 24 Wright et al, 20 using data from 1 statewide hospital system in the United States, reported a mean per-patient cost of $26 523 across all settings, $22 052 for outpatient visits, $3591 for inpatient stays, and $880 for emergency visits (not including provider professional fees) over a mean follow-up period of 4.49 years in 2014 USD, or $32 781, $27 255, $4438, and $1087 in 2022 USD, respectively. 24 Using the Pediatric Health Information System database, Lavery et al 17 reported a mean cost of $38 543 per hospital stay in 2015 USD, or $47 679 in 2022 USD. 24 Two studies using data from the United States identified that 58% 25 and 67% 26 of their patients had private insurance, respectively. Ross et al 25 additionally reported that 26% of patients had Medicaid and 7% had no health insurance coverage.

Humanistic Burden
Thirty-six studies reported on QOL outcomes in patients with POMS (Supplemental Table S3). Studies collected data from Canada (n = 9), the United States (n = 6), or both (n = 3), as well as Italy (n = 7), Germany (n = 4), and the United Kingdom (n = 2). There was 1 study from each of Brazil, France, Netherlands, Norway, and Serbia. Thirteen studies included patients 18 years old and under, 1 study included patients who were slightly older (<19 years old), and 1 study included patients who were slightly younger (<16 years old). The remaining studies either included children and adolescents (eg, 4-18 years old, n = 13) or only adolescents (eg, 12-18 years old, n = 7). One study did not specify the age range of its pediatric population. Twenty-three studies included control groups, most commonly age-and sex-matched healthy controls.
Studies reported data on QOL, including fatigue-related outcomes, using 10 different tools: The Pediatric Quality of Life Measurement Abbreviations: FSS, Fatigue Severity Scale; GLTEQ, Godin Leisure Time Exercise Questionnaire, NR, not reported; OR, odds ratio; RR, rate ratio. The left column reports outcomes when ≥2 studies presented that outcome (allowing us to present a range). The right column presents outcomes compared with healthy controls. Definition of healthy controls varied by study (eg, healthy matched controls, age-and sex-matched healthy controls). a Other than patient insurance type, no 2 studies reported on the same healthcare costs and therefore no ranges can be presented. No studies compared costs with healthy controls. b Overall, higher mean scores (out of 100) indicate better health-related QOL, although no standard categories exist. 33 Outside the MS literature, Beverung et al 35 classified a score of 81 to 100 as "better quality of life" and below 60 as "impaired quality of life." c No standard categories exist for the FSS. Higher scores represent greater fatigue severity. d No 2 studies reported on the same other humanistic outcomes and therefore no ranges can be presented.   35 Only 2 studies reported total scores. One study from Italy reported a total PedsQL TM 4.0 summary score of 80.3 at baseline and 80.7 at 1-year follow-up. 27 A study from the Netherlands reported that 36% of patients 4 to 17 years old (n = 22) had abnormal PedsQL TM 4.0 total functioning scores (defined as <1 SD below the mean of healthy agematched children). 31 All 8 studies 18,27-33 reported on the physical health component of the PedsQL TM 4.0, which ranged from 74.62 to 82.75 at baseline; 3 studies 18,27,28 reported on the psychosocial summary score, which ranged from 71.59 to 79.7 at baseline. One study from France reported that 20% of their patients (n = 26) had scores under 75 on the physical component (considered to be poor QOL). 30 Another study from the Netherlands reported that 45% and 46% of patients (n = 22) had abnormal physical and psychosocial summary scores, respectively. 31  64] from the same Italian center). Additionally, compared with age-matched healthy controls, a larger proportion of patients with POMS reported severe fatigue (19.8% vs 2.9%) using patient self-reported data from Germany and Austria. 33 Seven studies reported on fatigue using the FSS. Mean scores can be calculated 2 ways, and studies in our review reported both. The first method involves averaging numerical responses across all 9 items (ranging from 1 to 7) with higher scores representing greater fatigue severity. 37 While our review did not include any studies that compared the FSS with healthy controls, the mean score for healthy individuals is 2.3, and a score of 4 or more is considered indicative of problematic fatigue. 38 One study from Italy included in our review reported a mean FSS on this scale of 2.5 to 3.3. 39 The second method is based on the sum of each of the 9 items' individual scores (yielding results ranging from 9 to 63). Although no standard categories exist for this scoring scale, an older study on fatigue in patients with MS (not included) reported the median for healthy controls on the FSS to be 21.5. 40 Three studies in our review (2 from Italy and 1 from the United States) reported the FSS on this scale, with means ranging from 27.1 to 30.06. [41][42][43] Four studies reported on fatigue using the percentage of patients self-reporting fatigue (range, 23%-61.1%) 44-47 with 1 study from Canada finding a larger proportion of patients with POMS reporting chronic fatigue (43.8% vs 0.0%) compared with age-matched healthy controls. 47 One study reported on fatigue using the Modified Fatigue Impact Scale, presenting a mean of 32.52 among patients with POMS compared with 23.10 among ageand sex-matched healthy controls. 48 Other QOL outcomes: Two studies included in our review, both from the United Kingdom, used a qualitative design. 44   were $5907 across all settings per year of follow-up between 2002 and 2012, 20 and mean costs per hospital stay were $38 543 between 2004 and 2013. 17 Given that incidence, prevalence, and costs of POMS vary by geography (eg, lower per-patient costs were reported in Brazil 23 compared with the United States 17,20 ), it is difficult to generalize the global economic burden. We did not identify any published reviews on healthcare resource utilization or costs in patients with POMS. The US total cost of $5907 ($7524 in 2022 USD) 24 per patient per year of follow-up is a low estimate because this calculation excluded provider professional fees. Furthermore, these costs did not account for insurance coverage. However, this cost is still higher than overall costs in the US pediatric population, which has been reported to be less than $3000 ($3755 in 2022 USD) 24 per child across all settings and age groups. 55 Reviews based on adult patients with MS reported higher costs in the United States than our pediatric findings (total mean allcause healthcare costs for adults ranged from $8528 to $54 244 per patient per year in 2011 14 ) and lower costs in Spain (mean annual cost per patient of €30 050, 13 18,27,28 and physical health scores between 75 to 83. 18,[27][28][29][30][31][32][33] Studies of children with cancer reported similarly low mean psychosocial scores (73) and even lower mean physical health scores (72). 57 Compared with healthy controls reported in our review, both patients with POMS and those with cancer have lower scores than healthy controls (mean psychosocial scores between 79 to 89 18 and physical health scores between 82 to 84 57 ). Patients also reported problematic fatigue (FSS scores of patients in studies included in our review ranged from approximately 27 to 30, compared with a median score of 21 for healthy controls in a 2003 study of MS patients 40 ). In adults with MS, symptoms of fatigue have been shown to impact disease course and functional outcomes. 58 In children and adolescents already experiencing the typical challenges of growing up, reduced QOL, including impaired mental, physical, emotional, and academic functioning, may have a greater impact on life trajectories than in adults with MS. 8 Our review of QOL studies found similar articles to Carroll et al, 10 a recent systematic review on fatigue and its association with clinical, psychological, and social factors in children and adolescents with POMS. Six of the 12 studies included in the prior review were identified in ours. Those which were not included were either published prior to our search timeline (4 studies) or focused on cognitive tests or behavioral health outcomes (2 studies) that we did not consider to be aligned with our QOL outcomes. As a result, their study concluded similar findings to ours, namely that fatigue is a concerning and common symptom in children and adolescents with POMS. This review has limitations. While our goal was to report on patients 10 to 17 years old (other studies included wider age ranges), we included all studies that presented data on the pediatric population. Our results are not specific to adolescents and include data from some very young patients with POMS. Second, while studies reported on different types of MS (eg, relapsing remitting vs primary progressive), we did not examine nor report on these differences. Lastly, we did not assess the quality of the studies included in our review or whether any author bias was present.
Our search identified gaps in the current literature and yielded ideas for future research. We found fewer studies in our review that reported on healthcare resource utilization and costs than on QOL outcomes. POMS is costly, and additional studies, such as studies utilizing healthcare claims, should further estimate the costs associated with POMS. Further, studies that consistently report costs over the same time frame (eg, per year) would help compare findings across publications. Only 2 studies reported on patient insurance coverage. Studies exploring patient insurance coverage and potential lack of coverage would provide a more complete picture of the economic burden these young patients and their families face. Furthermore, given the reduced QOL in these patients, clinical trials should continue to measure QOL and fatigue, documenting how therapies may help address these outcomes. Lastly, while several studies reported on PedsQL TM and FSS measures, the QOL studies reported on a wide range of instruments. More consistent reporting of a limited number of instruments would have made comparing results more informative.  44 UK Five themes emerged from interviews with patients with MS 6-18 y: (1) emotional reactions to fatigue and its impact, (2) the lived experience of fatigue and impact on daily activities, (3) uncontrollability and unpredictability of fatigue (uncontrollability, uncertainty, and lack of knowledge), (4) fi nding a balance (concern about well-being, future), (5) social support and disclosure

CONCLUSION
Our review presents a uniquely broad and recent overview of the global economic and humanistic burden of patients with POMS. While no articles presented both costs and QOL, our review demonstrates that healthcare resource utilization and costs are high in this population, and patients report reduced QOL and significant fatigue compared with healthy children and adolescents. This decrease in QOL and increase in economic burden in children and adolescents may have further reaching consequences than those in older patients. Children this age are already experiencing challenges typical of this difficult development stage, burden is more likely to affect the entire family rather than the patient alone, and consequences may impact the child's potential for future achievements. Funding: This work was funded by Sanofi.