QOL06
Assessing Presence of MS-Related Symptoms As a Proxy for Disease Severity in Multiple Sclerosis Using Administrative Claims Data

Tuesday, October 26, 2021
Exhibit Hall (Rosen Shingle Creek)
Barry Hendin, MD , Phoenix Neurological Associates, Phoenix, AZ, Neurology, University of Arizona, Tucson, AZ
Richard A. Brook, MS, MBA , Better Health Worldwide, Inc., Newfoundland, NJ
Ian A. Beren, BS , WorkPartners, LLC, Loveland, CO
Nathan L. Kleinman, PhD , WorkPartners, LLC, Cheyenne, WY
Amy L. Phillips, PharmD , EMD Serono, Inc., Rockland, MA, USA, an affiliate of Merck KGaA, Darmstadt, Germany, Rockland, MA
Carroline Lobo, PhD , EMD Serono, Inc., Rockland, MA, USA, an affiliate of Merck KGaA, Darmstadt, MA, Germany



Background: Healthcare claims data lack clinically relevant variables such as disease severity.

Objectives: Assess multiple sclerosis (MS)-related symptoms as a proxy for disease severity using administrative claims data and examine comorbidities and direct medical costs across severity categories.

Methods: Published literature informed a mapping exercise with a clinical expert to identify MS-related symptoms using diagnostic/medication/healthcare procedural codes. Patients were categorized as mild/moderate/severe based on a combination of post-index symptoms: bladder/bowel (urinary/stool incontinence, visit to urologist, etc), psychiatric (anxiety/depression), cognitive (dementia/cognitive impairment), and physical function (spasticity, wheelchair/cane/walker use, etc). The proxy measure was applied to US Workpartners’ claims data (1/1/2010–12/31/2019). Eligibility criteria were: ≥3 MS-related (ICD-9-CM/ICD-10-CM: 340.xx/G35) inpatient, outpatient, or MS disease-modifying therapy (DMT) claims within a one-year period (latest claim=index date); continuous enrollment 6‑months pre-/1-year post-index; and age 18–64. Demographic and clinical characteristics (not included in the proxy assessment) and direct medical and pharmacy costs were compared across severity categories.

Results: The 1041 eligible patients were classified as mild (n=358 [34.4%]), moderate (n=491 [47.2%]), and severe (n=192 [18.4%]). Mean (standard error [SE]) age distribution was: mild 47.9(0.5); moderate 49.7(0.4); severe 50.7(0.7). Mean Charlson Comorbidity Index scores were: mild 0.40(0.05); moderate 0.68(0.06); severe 1.12(0.13; all p<0.001). The proportion (mild/moderate/severe) with gastrointestinal disease (14.2%/26.1%/43.2%), arthritis (5.9%/9.8%/15.1%), and hypertension/hyperlipidemia/diabetes (27.7%/38.5%/49.5%) increased with severity (all p<0.05). A greater proportion with moderate/severe vs mild disease had substance abuse disorder (4.3%/5.2% vs 1.7%; both p<0.05), chronic pain syndrome (4.5%/5.2% vs 1.7%; both p<0.05), and chronic lung disease (12.4%/13.5% vs 4.5%; both p=0.0001). Mean (SE) direct medical costs increased with severity (mild $9923[$1421], moderate $21,326[$2248], severe $36,876[$5058]; all p<0.005). Patients had higher non-DMT pharmacy costs with moderate ($34,912[$1779]; p<0.0001) and severe ($31,856[$2793]; p=0.0021) vs mild ($29,130[$2027]) disease.

Conclusions: Preliminary findings showed that comorbidities and direct medical/non-DMT costs increased with severity. Further validation of this proxy measure is needed.