Real-world treatment preferences among health care providers in the United States in selecting disease modifying therapies for patients with multiple sclerosis: a discrete choice experiment

Abstract

Aims

Health care providers (HCPs) treating multiple sclerosis (MS) in clinical practice have numerous disease-modifying therapies (DMTs) to consider when evaluating treatment options. This study assessed the treatment preferences of HCPs in the United States, both direct (explicit) and derived (explicit and implicit), when selecting MS DMTs based on clinical and logistical treatment attributes.

Materials and methods

A 45-minute web-enabled questionnaire was administered to HCPs who manage patients with MS to assess the importance of treatment attributes. HCPs were recruited through an online panel. This study examined treatment attributes relevant to treatment decisions in MS, with a focus on the burden to HCPs and their staff, as well as HCP attitudes toward various aspects of MS care such as diagnosis, treatment prioritization, and ease of initiating or switching DMTs. The study also employed a discrete choice experiment (DCE) to assess direct and derived treatment preferences.

Results

The study recruited 145 HCPs. Direct assessments (a score of greater than 7.0 was considered important) suggested that safety (mean importance rating = 7.8/9) and relative risk reduction in relapses (7.6/9) and disability progression (7.5/9) were most important when selecting DMTs. In contrast, derived importance from the DCE (higher points corresponding to greater importance) suggested that logistical attributes such as dose frequency (mean relative attribute importance = 17.5%), dose titration (10.3%), formulation (9.4%), and volume of calls (9.1%) were important considerations, along with efficacy (16.5%), safety (9.8%), and gastrointestinal tolerability (9.4%).

Limitations

This study may have been subject to selection bias due to the application of eligibility criteria, the convenient sampling recruitment methodology, and recruitment of HCPs with internet access.

Conclusion

In the direct assessment, clinical attributes were chosen as the most important treatment attributes by HCPs. However, in the DCE, derived treatment decisions rated logistical attributes as also being as important in treatment choice.

PLAIN LANGUAGE SUMMARY

In this study, researchers aimed to understand what multiple sclerosis (MS) neurologists, nurse practitioners, and physician assistants think is most important when choosing medicines for their patients. They surveyed 145 health care providers (HCPs) in the United States for this study. The HCPs reported that safety and reducing the risk of relapses and disability were most important when selecting medicines. Additionally, the researchers used a method called a discrete choice experiment to determine the relative importance of medication characteristics to HCPs. They found that additional factors, such as how often the medicine needs to be taken, how it is given, and how easy it is to use, were also very important. The study may not represent the opinions of all HCPs due to the number of participants and participation criteria.

Keywords:

• Multiple sclerosis
• disease modifying therapy
• discrete choice experiment
• treatment preferences
• treatment prioritization

JEL CLASSIFICATION CODES:

• I12
• I1
• I
• I11

Introduction

Multiple sclerosis (MS) is a chronic, demyelinating, inflammatory disease with a heterogenous disease course^1–9. The prevalence of MS in the United States of America (US) in 2010 was estimated to be around 309.2 per 100,000 persons, with the highest prevalence in the 55–64 year age group¹⁰. Health care providers (HCPs) make individualized treatment decisions based on the patient’s clinical profile and patient preferences. HCP preferences in the current MS treatment landscape have not been extensively evaluated and understanding HCP preferences may inform the clinical or logistical attributes important for MS treatment decision making.

Several disease modifying therapies (DMTs) have come to market in the past two decades including interferon beta, glatiramer acetate, teriflunomide, dimethyl fumarate, diroximel fumarate, natalizumab, fingolimod, alemtuzumab, mitoxantrone, cladribine, ocrelizumab, ofatumumab, ublituximab, siponimod, ozanimod, and ponesimod^11–16. The utilization of the proper agent for a given patient or patient population is becoming more complex as new mechanisms, routes of administration (ROAs), safety, and efficacy profiles are developed¹⁷. Prior studies have consistently demonstrated the importance of certain clinical attributes driving treatment preferences among HCPs and patients with MS. These attributes include reducing disease progression, minimizing side effects, and considering the frequency and method of administration^18–25. HCPs have traditionally considered these clinical and safety related factors when making treatment choices. However, there is a growing recognition of non-clinical factors that also play a significant role in decision making among HCPs²⁶. HCPs are likely to take into consideration both clinical (e.g. safety, efficacy) and logistical (e.g. patient monitoring, burden to patient, caregiver, or HCP, administrative challenges for clinical staff) treatment attributes when choosing treatments for their MS patients^27–30. This research aims to understand the relative impact that clinical and logistical attributes have on HCP decisions when making treatment choices for their MS patients.

The main objective of this study was to quantitatively assess the importance of clinical and logistical treatment attributes in selection of DMTs for MS by HCPs. The degree that each studied treatment attribute impacted HCP decisions when selecting treatments for their MS patients was assessed to generate comparative evidence differentiating MS therapies. Importance was assessed via two methods – direct assessment via importance rating questions and derived assessment via a discrete choice experiment (DCE, Table 1). DCEs are often used in healthcare research to quantify the value respondents place on specific characteristics or attributes of disease interventions^31,32. These treatment attributes can have different levels that represent the possible options or choices for each attribute. A DCE study design involves the development of a grid or table of attributes and their respective levels to create hypothetical product profiles, and then subsequently testing these profiles with respondents. This study also aimed to examine the treatment-related burden of MS therapies across different burden attributes from the perspective of HCPs and their staff. In addition, we sought to quantify the relative importance of each attribute and identify what truly drives treatment selection.

Table 1. Discrete choice experiment design grid.

Attribute	Level 1	Level 2	Level 3
Patient Population	Relapsing remitting MS (RRMS)	RRMS + Primary progressive MS (PPMS)
Efficacy	35% RRR in relapses 25% RRR in disability progression	50% RRR in relapses 25% RRR in disability progression	50% RRR in relapses 40% RRR in disability progression
Safety	Risk of side effects (progressive multifocal leukoencephalopathy [PML], risk of infection, lymphopenia/thrombocytopenia, liver injury)	Risk of side effects PML, risk of infection, lymphopenia/ thrombocytopenia, liver injury Black box warning: liver injury	Risk of side effects PML, risk of infection, lymphopenia/ thrombocytopenia, liver injury Black box warning: PML
GI Tolerability Profile	Relatively good GI tolerability profile	Relatively poor GI tolerability profile	–
Formulation	Patient-administered oral capsule	Patient-administered Subcutaneous Pre-filled Syringe (ONLY SHOWN WITH ‘Dose Titration’ LEVEL 1)	Healthcare provider administered Intravenous Injection (ONLY SHOWN WITH ‘Dosing Frequency’ LEVEL 1)
Dosing Frequency	Every 6 months	Once a day	Twice a day
Dose Titration	Not required	1 week	Varied from 1 to 4 weeks
Time Required for Treatment Initiation	15-minute conversation with patient to initiate treatment	1-h conversation with patient to initiate treatment	–
Patient Monitoring	Blood tests at baseline and periodically during treatment	Blood tests at baseline and periodically during treatment; cardiac, eye, ± genetic exams prior to initiation; vaccine antibodies after first dose	Blood tests at baseline and periodically during treatment; Hep B screening prior to initiation, periodic infection assessment
Volume of Patient Phone Calls to the Clinic (regarding side effects, complaints, and/or other treatment questions)	Low to none	Moderate	High
Vaccine Recommendation (administer prior to initiation or pause treatment if vaccines needed)	No	Yes	–

Note: The attributes and levels do not indicate any ranking or prioritization from a study design standpoint and were only used to create the hypothetical treatment profiles for each choice-set.

Abbreviations. RRR, Relative Risk Reduction; PML, Progressive Multifocal Leukoencephalopathy; RRMS, Relapsing Remitting Multiple Sclerosis; PPMS, Primary Progressive Multiple Sclerosis; GI, Gastrointestinal.

Methods

Study overview

A web-based, cross-sectional quantitative survey was conducted among US HCPs who treated and managed patients with MS. Importantly, this study used various treatment attributes that may be used in treatment decisions, particularly focusing on the burden to HCPs and their staff, as well as HCP attitudes toward the current state of MS care, including diagnosis, treatment prioritization, and ease of initiating or switching DMTs.

In addition, the study included a DCE that was used to quantify the preferences of HCPs and perceived burden for currently available therapies. The DCE methodology is a well-established approach used to understand individual preferences by presenting respondents with a series of hypothetical choices based on product attributes/profiles³³. Treatment attributes were developed based on 12 DMTs approved by the US Food and Drug Administration (FDA) that represented DMTs administered orally (dimethyl fumarate, teriflunomide, diroximel fumarate, fingolimod, ozanimod, siponimod, and cladribine), via injection (ofatumumab, glatiramer acetate, and interferons) or via infusion (ocrelizumab, natalizumab)^34–45. HCPs participating in the study were presented with various hypothetical scenarios that included varying combinations of these treatment attributes. By statistically analyzing the choices made by HCPs across these scenarios, study researchers were able to quantify and assess HCP preferences of different available therapies. All study materials were reviewed and approved by a central institutional review board (IRB) in the US (Advarra).

This study was conducted in accordance with the protocol, the current version of the Declaration of Helsinki⁴⁶, and applicable national and local requirements for good pharmacovigilance practices. This study recruited HCPs in the US and included both neurologists and advanced practice clinicians (APCs, such as nurse practitioners or physician assistants) that treat MS patients across community and academic settings.

Web-based questionnaire

The web-based questionnaire was approximately 45 min in length and was developed following good research practices. Respondents for this study answered initial screening questions to help establish their eligibility to participate in the study. Eligible respondents received the complete questionnaire immediately following screening. Consent was obtained from all respondents prior to entry into the web-based questionnaire. The questionnaire included the current diagnosis and management approach to MS, an MS DMT DCE, a deep dive into logistical burden, and clinigraphics, demographics, and the impact of the COVID-19 pandemic.

Three cognitive interviews were conducted with eligible respondents, during which the moderator hosted the web-enabled questionnaire on their computer and shared their screen with the HCPs via screensharing software. These interviews focused on subjective validation of the instrument – to ensure that all instrument sections were capturing intended data, the instrument was appropriately timed, and questions were appropriately detailed and clear for HCPs to answer accurately without the need for external assistance.

The following attributes included in the direct assessment of treatment importance were measured on a 9-point scale, where 1 was “not at all important” and 9 was “extremely important”: safety; relative risk reduction (RRR) in relapses; RRR in disability progression; ROA; ability for self-administration; minimal time required for treatment initiation; gastrointestinal (GI) tolerability; dosing frequency; minimal patient monitoring required during treatment; mechanism of action (MOA); impact on vaccinations or vaccination status; resulted in minimal volume of patient phone calls to clinic; and dose titration. Attributes were aligned based on secondary research to attributes of current DMTs^34–45.

Discrete choice experiment

Lighthouse Studio 9.8.0 (a Sawtooth software) was used to generate an orthogonal design for the DCE. Such a design ensures minimal correlation between the treatment attributes when presented as hypothetical treatment profiles. Attributes from the DCE included: safety; efficacy; formulation; time required for treatment initiation; GI tolerability profile; dosing frequency; patient monitoring; patient population; vaccination recommendation; volume of patient phone calls to clinic; and dose titration. A 14-choice-set experimental design with two hypothetical treatment profiles in each choice-set was generated using an 11-attribute grid with 2–3 levels per attribute. The design had two-way frequencies between each attribute level greater than 50 and standard errors lower than 0.05. Respondents were also offered a “neither therapy” option. The responses were used to assess the strength of preference for each attribute and level^31,32. Respondents were asked to make four decisions for each choice set:

Question #1: most preferred when selecting a DMT for your eligible MS patients.

Question #2: least burdensome to the respondent’s MS patients.

Question #3: therapy in which the respondent thought their MS patients would be most adherent.

Question #4: least burdensome to the respondent, their staff, or their practice.

Study recruitment and selection of study population

HCPs were recruited for participation in the study using convenience sampling from an online panel that was representative of the US HCP population. Those who qualified based on the eligibility criteria were allowed to participate in the study. Respondents needed to complete screening and fulfill the following criteria to qualify for participation in the study:

Inclusion criteria

1. Respondent must have been board eligible (if physician) and licensed or accredited by a board (if APC) in neurology with the following positions:

   • Neurologists.

   • APCs (nurse practitioners [NP], physician assistants [PA]) in neurology.

2. In practice post-residency for 3–35 years, inclusive.

3. Had spent at least 70% of their professional time providing direct patient care.

4. Treated at least 25 unique MS patients in an average month.

5. Treated at least five unique MS patients who were currently on DMTs in an average month.

Exclusion criteria

1. Been a participant in market research, health economic and outcomes research, or general pharmaceutical research in the last 30 days.

2. A current employee of or had an affiliation with the following:

   • Pharmaceutical manufacturer or biotech.

   • Contract research organization.

   • Market research or advertising firm.

   • FDA or government agency.

The study fielded for 5 weeks in February–March 2022 until the target sample size was reached. Respondents were enrolled in the study for up to 45 min or the time that it took to complete the web-based questionnaire. No additional follow-up was needed.

Data analysis

Statistical analyses were conducted in Lighthouse Studio 9.8.0 (a Sawtooth software) and Q Research Software 5.6. Descriptive statistics (means, medians, maximum or minimum ranges, and frequencies) were used to describe the recruited sample for the initial topic areas covered by the study instrument. A hierarchical Bayesian random-effects-only model was used on the data captured in the DCE to compute mean relative attribute importance scores for each treatment attribute and mean relative preference weights for each attribute level. Data were reported as mean (standard deviation; SD) relative attribute importance and mean (SD) relative preference weights.

Means for the product attribute ratings obtained via direct assessments and mean relative importance scores obtained from the DCE were plotted on a chart to help reveal potential differences between direct (explicit) and derived (explicit and implicit) importance of MS DMT treatment attributes.

Results

Study population – Demographics

Data was collected on 151 HCPs. A total of six respondents were removed during and after fieldwork due to poor data quality (e.g. completion of the survey in a time well below the mean length of interview, ratings with absolutely no differentiation despite different products and attributes being evaluated). A final sample of 145 HCPs participated in this study, including 100 neurologists and 45 APCs (n = 30 NPs and n = 15 PAs) (Table 2). The attributes and levels do not indicate any ranking or prioritization from a study design standpoint and were only used to create the hypothetical treatment profiles for each choice-set. Mean years in practice was 11.6 years for APCs and 18.6 years for neurologists and mean years prescribing DMTs was 7.7 years for APCs and 17.2 years for neurologists. Gender differences were observed in the study population, with 74% of neurologists identifying as male while 26% were female and 24% of APCs identifying as males while 76% were female. Data on type of practice (community hospital, private practice, academic hospital), staff assisting in reimbursement, and average patient load was also collected and is summarized in Table 2.

Table 2. Demographics of respondents.

	Advanced Practice Clinicians (N = 45)	Neurologists (N = 100)
Respondent-specific information
Average years in practice	11.6	18.6
Average years prescribing DMTs	7.7	17.2
Gender (%)
Male	24	74
Female	76	20
Type of practice
Community Hospital	38	7
Private Practice	44	60
Academic Hospital	18	32
Staff assisting in reimbursement	71	73
Average respondent patient load
All patients with neurological conditions	234	238
MS patients	78	86
MS patients on DMTs	60	75

Abbreviations. DMTs, Disease Modifying Therapies; MS, Multiple Sclerosis.

Direct assessment of treatment preferences

For direct assessments, a score of greater than 7.0 was considered important for DMT choices by HCPs. Direct assessments indicated that safety (mean importance rating = 7.8/9) and efficacy (defined as the combined “RRR in relapse” [7.6/9] and “RRR in disability progression” [7.5/9]) were the top attributes impacting overall treatment selection (Figure 1 and Table 3). Compared to safety and efficacy, HCPs considered attributes such as minimal patient monitoring required during treatment (6.5/9), results in minimal volume of patient phone calls to the clinic (5.9/9), minimal time required for initiation (6.5/9), and dose titration (5.8/9) as less important (Table 3).

Figure 1. Direct assessment of product attribute importance by HCPs rating on a 9-point scale.

Abbreviations. APC, Advanced Practice Clinicians; RRR, Relative Risk Reduction; ROA, Route of Administration; MOA, Mechanism of Action.

Table 3. Mean product attribute importance that HCPs considered for all MS patients.

	Specialty		Overall sample
	APCs (A)	Neurologists (B)
Specialty ‘n’	45	100	145
Product Attribute Importance| Mean [SD] (1 to 9 scale where 1 is ‘not at all important’ and 9 is ‘extremely important’)
Safety	7.9 [1.5]	7.8 [1.6]	7.8 [1.6]
RRR in Relapses	7.5 [1.4]	7.7 [1.5]	7.6 [1.5]
RRR in Disability Progression	7.4 [1.6]	7.5 [1.7]	7.5 [1.7]
ROA	6.8 [1.5]	6.5 [1.7]	6.6 [1.7]
Ability for Self-Administration	6.8 [1.6]	6.4 [1.8]	6.5 [1.7]
Minimal Time Required for Treatment Initiation	6.6 [1.6]	6.4 [1.7]	6.5 [1.7]
GI Tolerability	6.8 [1.8]	6.2 [1.6]	6.4 [1.7]
Dosing Frequency	6.4 [1.5]	6.3 [1.8]	6.4 [1.7]
Minimal Patient Monitoring Required during Treatment	6.5 [1.6]	6.4 [1.8]	6.4 [1.7]
MOA	6.5 [1.7]	6.2 [1.8]	6.3 [1.8]
Product Does Not Impact Vaccinations or Vaccination Status	6.4 [1.7]	6.2 [1.8]	6.3 [1.8]
Results in Minimal Volume of Patient Phone Calls to the Clinic	5.9 [2.1]	5.9 [1.9]	5.9 [2.0]
Dose Titration	5.8 [1.9]	5.8 [1.8]	5.8 [1.8]

Abbreviations. APC, Advanced Practice Clinicians; RRR, Relative Risk Reduction; ROA, Route of Administration; GI, Gastrointestinal; MOA, Mechanism of Action.

Derived treatment preferences via discrete choice experiment (DCE)

For the DCE, scores higher than 9.09 were considered important for DMT choices by HCPs. As there were a total of 11 attributes and 100 points for the DCE score, attributes would all have a score of 9.09 if they were considered equally important by HCPs. Therefore, attributes with a score higher than 9.09 can be considered important for therapy preference. Derived importance from the DCE of overall preferred DMT for their eligible MS patients (Question #1) demonstrated that logistical attributes such as dosing frequency (mean [SD] relative attribute importance = 17.5% [6.7]), dose titration (10.3% [5.2]), volume of patient calls to the clinic (9.1% [4.7]), and formulation (9.4% [4.8]) were also important when selecting DMTs, along with efficacy (16.5% [8.9]), safety (9.8% [4.5]), and GI tolerability profile (9.4% [6.1]) (Figure 2).

Figure 2. Derived assessment of DCE Question #1 Output: Mean [SD] overall therapy preference of relative attribute importance.

Note: Question #1: most preferred when selecting a DMT for your eligible MS patients.

Abbreviation. GI, Gastrointestinal.

When asked to indicate their preference for a therapy that is least burdensome to patients (Question #2), HCPs indicated that attributes such as dosing frequency (27.4% [8.5]) and formulation (16.0% [6.2]) were important when choosing a DMT (Figure 3 and Supplementary Table S1). Similarly, HCPs, when asked for preference for a therapy to which patients would adhere the most (Question #3), indicated that attributes such as dosing frequency (22.0% [10.0]), formulation (18.2% [3.6]), and dose titration (12.3% [6.8]) were important when choosing a DMT (Figure 4 and Supplementary Table S2). Derived analysis from the DCE asking HCPs for preference for a therapy least burdensome to their staff and themselves (Question #4) indicated that attributes such as volume of patient phone calls to the clinic (19.5% [11.8]) and dosing frequency (15.1% [8.0]) were important (Figure 5 and Supplementary Table S3).

Figure 3. Derived assessment of DCE Question #2 Output: Mean [SD] relative attribute importance for least burdensome therapy to patients.

Note: Question #2: least burdensome to the respondent’s MS patients.

Abbreviation. GI, Gastrointestinal.

Figure 4. Derived assessment of DCE Question #3 Output: Mean [SD] relative attribute importance for therapy with the best adherence.

Note: Question #3: therapy in which the respondent thought their MS patients would be most adherent.

Abbreviation. GI, Gastrointestinal.

Figure 5. Derived assessment of DCE Question #4 Output: Mean [SD] relative attribute importance for least burdensome therapy to HCP and staff.

Note: Question #4: least burdensome to the respondent, their staff, or their practice.

Abbreviation. GI, Gastrointestinal.

Mean relative preference weights for Question #1 suggested that HCPs preferred treatments that required less frequent dosing (every 6 months), allowed flexibility of dose titration (varies from 1 to 4 weeks), resulted in a low volume of patient phone calls to the clinic, were a patient-administered oral capsule, and had better efficacy, safety, and tolerability profiles (Table 4 and Figure 6). HCPs preferred comparable treatments with regards to the most important attributes driving decisions on Questions #2 (Supplementary Table S1), #3 (Supplementary Table S2), and #4 (Supplementary Table S3).

Figure 6. Derived assessment of DCE Output Question #1: Mean relative preference weights.

Note: Question #1: most preferred when selecting a DMT for your eligible MS patients.

Abbreviations. Mos. = Months; RRR = Relative Risk Reduction; BB = Black Box; PML = Progressive Multifocal Leukoencephalopathy; SC = subcutaneous; IV = intravenous; min = minute; hr = hour; Hep B = Hepatitis B; RRMS = Relapsing Remitting Multiple Sclerosis; PPMS = Primary Progressive Multiple Sclerosis; GI = Gastrointestinal.

Table 4. HCPs overall preference for DMTs for their Eligible MS patients – Derived analysis from DCE Question #1.

Attribute	Mean relative attribute importance [SD]			Levels	Mean relative preference weights [SD]
	APCs	Neurologists	Overall		APCs	Neurologists	Overall
Patient population	4.1 [3.1]	3.6 [2.6]	3.8 [2.7]	RRMS	9.6 [26.5]	−2.3 [24.4]	1.4 [25.6]
				RRMS + PPMS	−9.6 [26.5]	2.3 [24.4]	−1.4 [25.6]
Efficacy	16.9 [8.8]	16.3 [9]	16.5 [8.9]	35% RRR in relapses 25% RRR in disability progression	−69.8 [74.6]	−73.8 [74.5]	−72.5 [74.3]
				50% RRR in relapses 25% RRR in disability progression	−6.3 [42.2]	9.9 [38.8]	4.9 [40.5]
				50% RRR in relapses 40% RRR in disability progression	76.0 [71.7]	63.9 [74.3]	67.7 [73.5]
Safety	9.3 [4.4]	9.9 [4.6]	9.8 [4.5]	Risk of side effects (PML), risk of infection, lymphopenia/thrombocytopenia, liver injury	17.4 [51.2]	30.7 [52.4]	26.5 [52.2]
				Risk of side effects (PML), risk of infection, lymphopenia/thrombocytopenia, liver injury. Black box warning: liver injury	−9.6 [40.6]	−14 [40.8]	−12.6 [40.6]
				Risk of side effects (PML), risk of infection, lymphopenia/thrombocytopenia, liver injury. Black box warning: PML	−7.8 [49.4]	−16.7 [48.2]	−13.9 [48.6]
GI tolerability profile	8.9 [6.3]	9.6 [5.9]	9.4 [6.1]	Relatively good GI tolerability profile	43.7 [41.5]	49.7 [37.4]	47.8 [38.7]
				Relatively poor GI tolerability profile	−43.7 [41.5]	−49.7 [37.4]	−47.8 [38.7]
Formulation	8 [3.8]	10 [5.1]	9.4 [4.8]	Patient-administered oral capsule	−1.3 [44.4]	29.5 [52]	19.9 [51.6]
				Patient-administered subcutaneous pre-filled syringe	15.3 [37.8]	7.6 [40.5]	10 [39.7]
				Healthcare provider administered intravenous injection	−14 [38.1]	−37.2 [40.4]	−30 [41]
Dosing frequency	17.2 [6.8]	17.7 [6.7]	17.5 [6.7]	Every 6 months	93.2 [63.2]	107.9 [51.3]	103.3 [55.4]
				Once a day	−24.7 [47.5]	−32.4 [37.9]	−30 [41.1]
				Twice a day	−68.5 [49.8]	−75.5 [40.8]	−73.3 [43.7]
Dose titration	11.9 [5.2]	9.6 [5]	10.3 [5.2]	Not required	−59.1 [46.2]	−28.2 [52.9]	−37.8 [52.7]
				1 week	24.7 [53.1]	9.8 [50]	14.4 [51.3]
				Varies from 1–4 weeks	34.4 [34.5]	18.5 [37.2]	23.4 [37]
Time required for treatment initiation	4.1 [3.2]	4.1 [2.8]	4.1 [2.9]	15-minute conversation with patient to initiate treatment	5.2 [28.2]	2 [27.3]	3 [27.5]
				1-h conversation with patient to initiate treatment	−5.2 [28.2]	−2 [27.3]	−3 [27.5]
Patient monitoring	6.3 [3.7]	6.5 [4.2]	6.4 [4.1]	Blood tests at baseline and periodically during treatment	14.9 [32.0]	7.5 [41.7]	9.8 [39]
				Blood tests at baseline and periodically during treatment; cardiac, eye, ± genetic exams prior to initiation; vaccine antibodies after first dose	9.5 [28.3]	5.3 [30.8]	6.6 [30]
				Blood tests at baseline and periodically during treatment Hep B screening prior to initiation Periodic infection assessment	−24.5 [29.5]	−12.7 [32.4]	−16.4 [31.9]
Volume of patient phone calls to the clinic	9.9 [4.7]	8.8 [4.7]	9.1 [4.7]	Low to none	37 [38.5]	30.1 [35]	32.3 [36.2]
				Moderate	3.1 [36.9]	2.2 [40.1]	2.5 [39]
				High	−40.1 [46.6]	−32.3 [41.1]	−34.8 [42.9]
Vaccine recommendation	3.5 [2.9]	3.8 [2.6]	3.7 [2.7]	No	5.5 [24.5]	8.1 [24]	7.3 [24.1]
				Yes	−5.5 [24.5]	−8.1 [24]	−7.3 [24.1]

Note: The mean relative preference weight is a measure of preference for an attribute level, relative to any other attribute level, in the context of each attribute level tested in the DCE design. While a mean relative preference weight for an attribute level does not carry any inherent meaning/interpretation, when compared with any other mean relative preference weight for another attribute level, a high value indicates a higher preference for that attribute level in selecting a treatment for MS, and a lower value indicates a lower preference. Question #1: most preferred when selecting a DMT for your eligible MS patients.

Abbreviations: SD, Standard Deviation; APC, Advanced Practice Clinicians; RRMS, Relapsing Remitting Multiple Sclerosis; PPMS, Primary Progressive Multiple Sclerosis; RRR, Relative Risk Reduction; PML, Progressive Multifocal Leukoencephalopathy; GI, Gastrointestinal; Hep B, Hepatitis B.

Comparing direct and derived assessment of treatment preferences

An analysis comparing direct versus derived attribute preference highlighted dosing frequency, formulation, GI tolerability profile, volume of patient phone calls to the clinic, and dose titration as relatively less important via direct (Quadrant 3) analysis; in contrast, these factors were shown to be important via derived (Quadrant 4) analysis (Figure 7) and are therefore implicit drivers of treatment preference.

Figure 7. Direct versus derived rated importance by DCE relative importance scores.

Note: Direct assessments are plotted on the y-axis and derived assessments are plotted on the x-axis; the mean stated preference threshold of >7.0 = high score on a 9-point scale is based on accepted convention.

Abbreviation. GI, Gastrointestinal.

Discussion

When HCPs were asked which attributes were important when choosing DMTs via direct assessment, clinical attributes such as efficacy (including RRR in relapse and RRR in disability progression) and safety were rated the highest. However, the DCE results demonstrated that, while efficacy and safety were still important, real world logistical attributes such as dosing frequency, dose titration, volume of patient phone calls to the clinic, and formulation were also important when choosing a DMT for MS. This DCE helps shed light on these logistical attributes as implicit drivers of DMT preference. A previous DCE study examining patient treatment preferences reported that patients considered dosing regimen (including frequency and ROA) as equally important attributes to safety and efficacy⁴⁷. This DCE helps validate findings from prior research from the HCP perspective to provide a holistic view of treatment preferences of two key stakeholders in the literature.

Perceptual decisions are biased by the cost to act⁴⁸. Humans make decisions based on the amount of effort required for an outcome. Decisions on which DMTs are prescribed are partly based on the level of burden for patients, caregivers, HCPs, and practice staff. Treatment burden, or the workload and effect on patient functioning and wellbeing for patients with chronic diseases, can cause a decrease in adherence to treatments^49,50. Additionally, studies have demonstrated that 16% to 24% of an HCP’s time is spent on administrative duties related to patient care, and that these affected their career satisfaction as well as their ability to deliver high-quality care^51,52. It is no surprise then that logistical attributes linked to level of burden such as dosing frequency, minimal time required for treatment initiation (i.e. dose titration), ability for self-administration (i.e. route of administration), and minimal patient monitoring required during treatment (i.e. lower volume of patient phone calls to the clinic) featured largely in importance for HCPs when considering which DMTs to prescribe to patients. While it is crucial to emphasize that all DMTs necessitate monitoring to validate treatment decisions and make necessary adjustments as required, HCPs generally prefer DMTs that require fewer periodic tests. Furthermore, results presented here suggest that HCPs preferred treatments that require less frequent dosing, such as ocrelizumab or natalizumab with dosing of every 6 months or every 4 weeks. Some DMTs had high mean relative preference weights in multiple attribute categories (e.g. diroximel fumarate); mean relative preference weights associated with diroximel fumarate were preferred for nine of the 11 attributes tested. Other DMTs such as ozanimod and dimethyl fumarate also had high relative preference weights in categories such as ROA, efficacy, safety, and tolerability^44,45.

Previous DCE studies conducted among MS HCPs have examined only clinical attributes such as safety, efficacy, or administration of therapy^{18,20–22,24,53}. This study will be additive to the literature on MS DMT preference as it looks at the treatment selection holistically from both clinical and logistical perspectives. Scientific publications highlighting the role of logistical process attributes in driving HCP treatment preference, in concert with clinical efficacy outcomes and safety profiles, will help drive medical education and highlight the myriad factors HCPs need to consider in their real-world clinical practice. This is critical in the crowded MS DMT landscape, as HCPs need to weigh multiple options while determining optimal treatments for their patients.

This study suggests that the thoughtful selection of an appropriate therapy from a range of good therapy options weighs both clinical and logistical attributes. An oral product with a strong efficacy and good safety profile that also has a favorable dose titration, a good GI tolerability profile, and an oral formulation can reduce burden to both HCPs and patients, and both the patient and HCP aspects of therapy burden could be addressed.

Limitations of research methods

As the eligibility criteria required physicians to treat at least 25 unique MS patients in an average month and at least five unique MS patients who were receiving DMTs in an average month, the study may have been subject to selection bias. As HCPs were required to provide their retrospective use of MS products in a portion of the study instrument, respondents could have been subject to recall bias; however, given the reliance on the DCE for the primary objectives of the study, the impact of such a recall bias on the data captured should have been at a minimum. Recruitment was conducted with a focus on ensuring that the respondent sample was broadly representative of the US HCP population; however, selection bias could have played a role in recruitment. While the DCE study design and analysis were quite complex, they were still limited to a finite number of treatment attributes and may not have been fully reflective of all possible factors that HCPs consider in treatment selection. Additionally, interpretation of the results of this study was constrained by the levels of each attribute that were represented in the survey. Although every effort was made to ensure that these attribute levels reflect the ranges present for clinically available treatments, these values may not reflect the real-world process involved in clinical decision-making. The impact of the COVID-19 pandemic between 2019 and 2022 may have altered treatment choices and approaches to MS management, with effects that may only be known in the next few years. This study was limited to individuals with internet access, and therefore the results can only be generalized to this population.

Conclusions

This study demonstrates that the decision process used by HCPs in the selection of an appropriate DMT for MS patients involves consideration of a wide range of clinical and logistical attributes. In addition to clinical attributes like efficacy and safety, HCPs aim to minimize potential real-world logistical burdens on their patients and practice (e.g. dosing frequency, formulation, GI tolerability, dose titration, and volume of patient calls to the clinic). Based on the attribute levels preferred by MS DMT treaters sampled in this study, diroximel fumarate may be a compelling treatment option for HCPs.

Transparency

Declaration of funding

This study was funded by Biogen, Cambridge, MA, USA.

Declaration of financial/other relationships

DB, consultant, and speaker bureaus for Biogen, EMD-Serono, Horizon, Janssen, Bristol Myers-Squibb, Sanofi Genzyme and Genentech; MA, consultant for Biogen and TG Therapeutics; MB, consultant for Biogen and Novartis; AG, consultant for Biogen, Novartis, Sanofi Genzyme, EMD Serono and Jansen. AA, BO, and NH are employees of Trinity Life Sciences. NH and AA hold equity in Trinity Life Sciences. JBL and SLS are employees of and hold stock or stock options in Biogen. CG and FB were employees of and held stock options in Biogen at the time of the study.

Author contributions

All authors contributed to the study concept and design. Study material preparation, data collection, analyses, and manuscript development were conducted by BO, AA, and NH. FB, CG, JBL, and SLS provided overall strategic guidance, reviewed materials, and critically revised the manuscript. DB, MA, MB, and AG provided expert guidance on designing the DCE grid and reviewed the work. All authors read and approved the final published version.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgements

Iona Bartek, of Seshet Scribes LLC, provided medical writing services funded by Biogen. Data analysis support was provided by Josiah Edelblut, Harrison Malec, and Sarah Schneider from Trinity Life Sciences.