Quantifying Differences in Health Care Consumption for the Management of Multiple Sclerosis Within Privately and Publicly Insured Health Care Programs

BACKGROUND: Multiple sclerosis (MS) is a chronic and debilitating disease of the central nervous system that affects more than 570,000 persons in the United States and 2.3 million worldwide. Since most individuals experience initial symptoms between the ages of 20 and 40 years, MS can have a significant effect on health care consumption, quality of life, productivity, and employment over the long-term disease course. Opportunities exist to better understand how benefit design and other nonclinical factors can affect health care delivery and associated costs. OBJECTIVE: To observe and report variances in health care consumed for the treatment of MS in patients enrolled in privately (commercial) and publicly (Medicaid) funded health insurance programs. METHODS: In a retrospective analysis using Havas Gemini’s proprietary MS Benchmarks Disease-Modeling Process and IMS LifeLink Health Plan Claims and Longitudinal Prescriptions databases, integrated medical and pharmacy claims data were analyzed to select patients with a diagnosis of MS during the 2012 calendar year. Comorbidities were determined using ICD-9-CM codes present on medical claims. Prescription drug use was evaluated by pharmacy claims and drug-specific billing codes. RESULTS: 19,984 patients with MS were identified—18,269 from commercial payers and 1,715 from Medicaid. Although total annual costs related to the care of MS for the groups reflected a relatively small difference ($31,107 commercial; $33,344 Medicaid), costs associated with specific service categories varied greatly. Pharmacy costs were considerably less in the Medicaid group; however, inpatient and emergency room costs were as much as 5 times higher. Overall use of disease-modifying treatments (DMTs) in the Medicaid group was seen in 32.5% of patients and 52.1% in the commercial patient group. Thus, lower pharmacy costs in the Medicaid group were possibly related to lesser use of DMTs among that group of patients. CONCLUSIONS: This analysis illustrates that notable variances exist in consumption of health care resources between patients enrolled in privately and publicly funded health care programs. These variances may have additional implications relating to outcomes specific to MS.

Quantifying Differences in Health Care Consumption for the Management of Multiple Sclerosis Within Privately and Publicly Insured Health Care Programs 6. Ancillary Service Markers are documented. These represent services of interest that are directly or indirectly associated with the underlying disease and are the variables reported in the analysis. The relevant procedure codes (Current Procedural Terminology, 4th edition, primarily, but revenue code when appropriate) are detailed and documented. 7. Definitions of the terms used throughout the analysis and results are listed and defined.
When put together, this MS-specific disease model provided the framework on which the entire analysis in this article was based. The model gives the data user the ability to see how, why, and what went into the analysis and, if desired, the ability to recreate the analysis, since all of the terminology, data elements, variables, and definitions are provided and documented. (For a summary of the MS-specific disease model, see the Appendix, available in online article.) Charges and use of medical services and prescription drugs were identified and captured using the Episode Treatment Groups (ETG) software and were collected specifically for ETG 315100 (Multiple Sclerosis) during the study period of January 1, 2012-December 31, 2012. Episodes were grouped by several criteria, including demographics, benefit/plan design, presence of specific medical conditions, and use of pharmacotherapy. Economic data were broken down by the service category: inpatient and outpatient (ancillary, facility, management, and surgical); emergency room (ER); and pharmacy, depicting at which point in the health care continuum the service was received. All costs presented in this analysis represent dollar amounts of charges submitted or billed by a practitioner or institution to the health plan or insurer for payment.
Patients were selected for inclusion if they were aged 18 years or older, had at least 1 ICD-9-CM code for MS (340.xx) and were continuously enrolled during the 2012 calendar year. Patients were divided into 2 cohorts: patients with Medicaid coverage and patients with commercial insurance coverage. Patients were excluded if they did not have complete eligibility during the study period (i.e., less than 12 months of coverage) or were missing age, gender, or benefit design data. Overall annual cost differences related to MS care were relatively small between the commercial and Medicaid groups ($31,107 vs. $33, 344). However, there were notable contrasts in the prevalence of a select group of medical conditions related to MS, as well as common chronic medical conditions between disability, MS may impose a significant effect on health care consumption, quality of life, productivity, and employment over the long-term disease course. 6,8,[12][13][14] The availability of health care resources and drug therapies to MS patients may have profound implications on disease progression, health care utilization and costs, productivity, and quality of life. The purpose of this study was to observe and report variances in health care consumed for the treatment of MS in patients enrolled in privately (commercial) and publicly (Medicaid) funded health insurance programs.

■■ Methods Study Design
This study presents the results of a retrospective claims-based analysis using Havas Gemini's proprietary MS Benchmarks Disease-Modeling Process on data from the IMS LifeLink Health Plan Claims and Longitudinal Prescriptions databases during the 2012 calendar year. At the time these analyses were conducted, the IMS LifeLink Health Plan Claims Database contained data from approximately 60 million de-identified patients in over 100 private (commercial) and public (Medicare and Medicaid) managed care plans across the United States. Additional data used in these analyses included patient characteristics such as geographical region, age, gender, and payer type. Because of the broad reach of this data, patients were similar to the national insured population in terms of age and gender.
Gemini's proprietary disease model follows a standardized format and template to make understanding the disease process and data outputs easier. The disease model includes the following: 1. The criteria for eligibility, inclusion, and exclusion criteria for patients being included in the analysis are defined and documented. 2. Other criteria such as definitions of enrollment and history of disease are documented where relevant for a given disease process. 3. Demographic markers (age, gender, product type, payer, region, and physician specialty) are documented and defined. 4. Clinical markers, the range of associated disease processes, complications, and comorbidities that are key to a particular disease are documented, along with the relevant ICD-9-CM (International Classification of Diseases, Ninth Revision, Clinical Modification) codes. 5. Pharmacotherapy markers are identified, listing all drug categories, drug classes, and generic names of those drugs that are related to the disease process of interest and/or have been requested specifically. Where IMS LifeLink is the source of the dataset, the Healthcare Common Procedure Coding System codes and General Product Identifier codes that comprise these drugs are listed and documented.
Quantifying Differences in Health Care Consumption for the Management of Multiple Sclerosis Within Privately and Publicly Insured Health Care Programs the 2 groups, with Medicaid patients reflecting higher rates of many conditions related to MS, such as gait abnormalities (17.3% compared with 10.4%), dysesthesia (21.9% compared with 17.9%), dizziness (14.6% compared with 9.2%), and urinary incontinence (16.0% compared with 5.8%). In addition, other chronic (or comorbid) conditions, including asthma, chronic obstructive pulmonary disease, diabetes, and hypertension, were higher in the Medicaid group. Almost none of the conditions observed were more prevalent in the commercial group (Tables 2 and 3).
Although total annual costs related to the care of MS for the groups reflected a relatively small difference, charges associated with specific service categories varied greatly (Table 4). Pharmacy charges were the major cost driver in both populations, although costs were considerably less in the Medicaid group compared with the commercial group (62.5% vs 78.2%). However, inpatient and ER charges were as much as 5 times higher in the Medicaid group. The higher ER and inpatient charges in the Medicaid group may be attributed to a greater percentage of the population having single and multiple ER visits than the commercial population (21.5% vs. 6.7%) and a greater percentage of single and multiple inpatient admissions than the commercial population (6.9% vs. 1.6%; Table 5). While the percentage of patients having admissions was much higher in the Medicaid group compared with the commercial group, the average number of admissions per year between groups was similar (1.3 vs. 1.4). The Medicaid patients had a lower average length of stay, compared with the commercially insured patient cohort (8.3 vs. 12.5 days). These data suggest that Medicaid patients used ER and inpatient care at a higher rate than commercial patients and that the factors driving costs were related more to the volume of services required by the Medicaid patients.
Lower pharmacy charges in the Medicaid group may be related to the overall lower use of disease-modifying treatments (DMTs)-32.5% compared with 52.1% in the commercial group (Table 6). While self-injectable and oral product use was proportionally lower in the Medicaid group, the use of infused agents (e.g., natalizumab) was even lower. Although the proportion of patients receiving DMTs was lower in the Medicaid group, these patients received a greater number of prescriptions filled for DMTs during the year. This may suggest a greater adherence to therapy or at least a more consistent refill rate in those patients who received a prescription for a DMT.

■■ Discussion Health Care Delivery
Variances in health care delivery and utilization are affected by benefit design. This study provides a source of notable differences in health care utilization between commercially and publicly insured MS patients. These differences may be especially important regarding coordination of care and DMT use. Well-coordinated care is vital for those with chronic illnesses and those with multiple comorbid conditions, as in this

Prevalence of MS-Related Comorbidities
Quantifying Differences in Health Care Consumption for the Management of Multiple Sclerosis Within Privately and Publicly Insured Health Care Programs study; adults with chronic conditions who receive coordinated and easily accessible care have better outcomes and report higher levels of satisfaction with that care. [15][16][17][18] Medicaid coverage, compared with other insurance types such as commercial insurance, has been associated with poorer access to preventive, acute, and follow-up care. 19 Variances in treatment options may also affect disease progression, relapse, and the development of comorbid conditions, which can have a favorable or unfavorable effect on cost-effectiveness. 20 Because patients seen primarily in ER settings are often sicker and are likely unknown to the physician treating them, these patients are at greater risk for complications, readmissions, medical errors, duplicate tests, and functional deterioration that may drive up costs considerably. 15,16 In a survey by Shabas and Heffener (2005) of 92 MS patients enrolled in a Medicaid long-term managed plan, 32% of the patients were never seen by an MS specialist physician, and 18% of patients who had begun a course of prescribed MS medications were noncompliant at the time of the survey. 16 In our study, inpatient and ER charges were as much as 5 times higher in the Medicaid group.

DMT Accessibility and Adherence
Common MS sequellae (i.e., fatigue, physical disability, cognitive dysfunction, and pain) are associated with considerably higher costs for MS-related health care and may greatly affect quality of life, employment, social functioning, and nursing home placement. 21,22 Controlling these sequellae through DMT compliance and coordinated care may provide for better clinical and economic outcomes, including lower risk for MS-related hospitalizations and relapse and a better quality of life. [15][16][17][18][23][24][25] Most patients with MS are covered by some form of health care insurance; however, plan designs and formulary restrictions can create access barriers for some patients. 26 The compliant use of DMTs is a core component of MS treatment and has been shown to lower the risks for MS-related hospitalizations and disease relapse, slow the progression of disability, and reduce MS-related treatment costs. [23][24][25] The 2014 Drug Trend Report found that Medicare and commercially insured clients experi-enced a similar upward trend in specialty therapy prescribing (11.5% Medicaid; 12.9% commercial); however, a nonadherence rate of 23.0% among all patients with MS was identified. 27 Our study found a significant difference in overall DMT use between the Medicaid and commercially insured groups (32.5% vs. 52.1%) but noted that a greater number of prescriptions for DMTs was filled in the Medicaid patient cohort. While a greater number of prescriptions filled may appear to indicate greater medication compliance, adherence issues are multifaceted and include lack of care continuity, lack of patient education and support, and long wait times (such as those encountered by patients who primarily rely on the ER for care). 28,29 Filling a prescription does not necessarily mean that the medication was used correctly, timely, or at all. Some patients are challenged by the physical and psychological discomfort of injected or infused DMTs, experience intolerable side effects, or have insurance coverage that limits which DMTs may be prescribed, which limits the physician's ability to switch to other more tolerable DMTs. 30 The Medicaid group of patients had greater ER and inpatient use along with significantly higher MS-related comorbidities, which may indicate poor coordination of care or a lack of patient education regarding the proper administration of DMTs. Coordinated management strategies that optimize a strong provider-patient relationship, patient communication and education, and positive reinforcement medication have a positive impact on DMT use and adherence among all patients with MS regardless of insurance provider type. 24,25,29 Limitations There are 4 key limitations to our study. First, the data do not differentiate between MS-specific and comorbidity-related ER visits. Our analysis points out the much higher ER use among Medicaid-insured patients with MS compared with those who are commercially insured (21.5% vs. 6.7%), possibly indicating worse disease control and unmet needs. However, additional studies using differentiated data are needed to distinguish MS-specific from comorbidity-related ER visits.  Second, our data do not distinguish between MS-specific charges and comorbidity-related charges. MS-related symptoms and comorbidities, such as numbness and tingling, gait abnormalities, diabetes (that may result from reduced immobility leading to weight gain), urinary incontinence, and dizziness, occurred at higher rates among Medicaid-insured MS patients. These comorbidities may influence inpatient and ER use and can be presumed to add significantly to MS-related costs. Management strategies leading to a reduction in the frequency and severity of symptoms, relapse, or comorbid conditions requiring ER or inpatient care may have a substantial effect on the economic consequences of MS. 31 A third limitation rests with the design of this study, which focused solely on an analysis of data. As such, patient-reported satisfaction regarding insurance coverage was not included. A 2012 study by Fenton et al. concluded that higher ER use resulted in lower patient satisfaction. 32 Whether or not that would have been true in our study is not known. Satisfied patients are more adherent to their plan of care, which potentially results in improved disease control and less need for emergency care. 32 Finally, because we did not interview patients, true compliance and adherence rates were not determined and, thus, remain unknown.
Further studies with analyses that control for patient demographics, comorbidities, and regional variation may help to substantiate our data findings in the charges, resource use, and use of DMTs among publicly and privately insured MS patients. Moreover, additional studies controlling for MS disease severity, copayments, and coinsurance may further our understanding of patient adherence. To avoid potential overcounting of the study populations, a more conservative future study algorithm should include only patients who (a) received at least 2 diagnoses of MS or (b) received 1 diagnosis of MS and receipt of a DMT. Finally, there is also the possibility that Medicaid patients use the ER in fundamentally different ways compared with those who are commercially insured.

■■ Conclusions
Significant variances exist in the consumption of health care for the treatment of MS in patients enrolled in privately and publicly funded health insurance programs. These variances may have additional implications relating to outcomes specific to MS. Controlling the prevalence of key medical conditions and use of drug therapy to slow progression of disability and reduce relapse through coordinated, accessible health care delivery could have a measurable effect on economic and utilization measures.