Methods

Comparing the Effectiveness of Three Treatments for Reducing Fatigue among Patients with Multiple Sclerosis—The COMBO-MS Trial - NCBI Bookshelf An official website of the United States government Here's how you know The .gov means it's official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you're on a federal government site. The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely. Log in Show account info Close Account Logged in as: username Dashboard Publications Account settings Log out Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation Bookshelf Search database Books All Databases Assembly Biocollections BioProject BioSample Books ClinVar Conserved Domains dbVar Gene Genome GEO DataSets GEO Profiles GTR Identical Protein Groups MedGen MeSH NLM Catalog Nucleotide OMIM PMC Protein Protein Clusters Protein Family Models PubChem BioAssay PubChem Compound PubChem Substance PubMed SNP SRA Structure Taxonomy ToolKit ToolKitAll ToolKitBookgh Search term Search Browse Titles Advanced Help Disclaimer NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health. Comparing the Effectiveness of Three Treatments for Reducing Fatigue among Patients with Multiple Sclerosis—The COMBO-MS Trial Tiffany J. Braley , MD, MS and Anna L. Kratz , PhD. Author Information and Affiliations Authors Tiffany J. Braley , MD, MS 1 and Anna L. Kratz , PhD 2 . Affiliations 1 Department of Neurology, Division of Multiple Sclerosis and Neuroimmunology, University of Michigan, Ann Arbor 2 Department of Physical Medicine and Rehabilitation, University of Michigan, Ann Arbor Washington (DC): Patient-Centered Outcomes Research Institute (PCORI) ; 2023 Oct . Copyright and Permissions Copyright © 2024. University of Michigan. All Rights Reserved. This book is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License which permits noncommercial use and distribution provided the original author(s) and source are credited. (See https://creativecommons.org/licenses/by-nc-nd/4.0/ Structured Abstract Background: Fatigue affects 90% of people with multiple sclerosis (MS) and is a well-recognized risk factor for many social and economic challenges. Given the limited number of available treatments for MS fatigue, optimization of existing treatments is imperative. Objective: The objectives of this pragmatic, analyst-blinded comparative effectiveness trial ( NCT03621761 ) were to compare the benefits and harms of a commonly accepted behavioral strategy (telephone-based cognitive behavior therapy [CBT]), a commonly used pharmacologic treatment (modafinil), and a combination of both therapies for fatigue in a representative sample of people with MS. Methods: Adults with a diagnosis of clinically definite MS (all subtypes) and clinically significant fatigue (Fatigue Severity Scale score ≥4) were randomly assigned (1:1:1) to receive CBT monotherapy (8 active sessions plus up to 2 maintenance sessions; n = 114), modafinil monotherapy (target dose, 100-400 mg per day as needed; n = 114), or a combination of both therapies (n = 108) for 12 weeks. The primary outcome was fatigue impact (Modified Fatigue Impact Scale [MFIS] score). Secondary outcomes were subjective fatigue intensity and fatigability measured with 7 days of ecological momentary assessment (EMA; real-time self-report in the lived environment) using numeric rating scales and actigraphy. Outcomes were assessed at baseline and 12 weeks. Effect modifiers of interest, including depression (Patient Health Questionnaire-8 score), sleep (Sleep Hygiene Index scores, Epworth Sleepiness Scale scores, sleep duration, and presence of known or suspected obstructive sleep apnea), disability level (Expanded Disability Status Scale score), and MS subtype, were also assessed at baseline. Responder analyses (proportion who experienced a 10-point or higher reduction in MFIS score) and scores on the Patient Global Impression of Change (PGIC) scale, a self-reported measure of overall change in activity, symptoms, function, and quality of life, were also assessed at 12 weeks. Overall treatment effect on fatigue outcomes (as change scores) was assessed with multiple linear regression models, with imputation for missing data in MFIS models. Potential effect modifiers of interest were assessed in separate multiple linear regression models with interaction terms. Mean PGIC scores were compared with Kruskal-Wallis tests. Results: A total of 336 participants completed the study; 114, 114, and 108 were randomly assigned to the CBT, modafinil, and combination therapy arms, respectively, and all were included in intention-to-treat analyses. Of the 336 randomly assigned participants, 256 (76.2%) were women, 286 (85.1%) were White, and 239 (71.1%) had relapsing MS. The mean (SD) baseline MFIS scores were 52.6 (13.7) (N Missing = 10 [8.77%]), 53.2 (14.4) (N Missing = 7 [6.14%]), and 52.3 (14.1) (N Missing = 8 [7.41%]) for CBT, modafinil, and combination therapy, respectively. At 12 weeks, relative to baseline, telephone-based CBT, modafinil, and combination therapy were associated with statistically significant and clinically meaningful mean (SD) MFIS score reductions of 15.2 (11.9), 16.9 (15.9), and 17.3 (16.2) points, respectively. In unadjusted models of treatment effect on change in total MFIS score at 12 weeks (following imputation of missing values), relative to combination therapy, the mean difference in change in MFIS score was 1.26 (95% CI, −2.91 to 5.42; P = .55) for CBT and −0.56 (95% CI, −4.68 to 3.56; P = .79) for modafinil. Reductions of similar magnitude were seen in EMA fatigue intensity and fatigability scores. More than two-thirds of participants in each treatment group saw at least a 10-point reduction in their overall MFIS score. In multiple linear regression models (adjusted for age, sex, and baseline levels of anxiety, pain, physical activity, and the fatigue outcome), 12-week reductions in both MFIS and EMA measures of fatigue intensity did not significantly vary by treatment intervention. Mean (SD) PGIC scores were higher in absolute terms for combination therapy (5.1 [1.6]) compared with CBT monotherapy (4.7 [1.6]; P = .07) and were significantly higher compared with modafinil monotherapy (4.5 [1.7]; P < .01). Disability level (Expanded Disability Status Scale score) significantly moderated the treatment effect on EMA measures of fatigue intensity, with modafinil showing the most robust benefits for those with lower disability levels ( P = .014). Sleep hygiene significantly moderated treatment effect on MFIS scores in complete case models (ie, no data imputation; P = .0327), with CBT showing the greatest benefit for fatigue in patients with poorer sleep hygiene, but statistical significance of this interaction diminished with imputation for missing data. Daytime sleepiness (Epworth Sleepiness Scale score) demonstrated direct effects on both MFIS and EMA fatigue intensity change scores (higher levels of sleepiness at baseline were related to greater declines in fatigue impact and intensity) but did not significantly modify treatment effect. Both CBT and modafinil were well tolerated with good study adherence and low frequency of treatment discontinuation. There were no significant differences between treatment arms in terms of adherence or treatment discontinuation; 26 patients (9%) withdrew from modafinil and 6 (2.6%) withdrew from CBT. Conclusions: Telephone-based CBT, modafinil, and combination therapy each resulted in a clinically meaningful decrease in fatigue impact and intensity in people with MS. Combination therapy was not associated with fatigue reduction above and beyond fatigue reduction achieved by either monotherapy; however, combination therapy was associated with greater perceived global benefits in functional outcomes and quality of life. Furthermore, treatment effect may vary by clinical characteristics, including sleep hygiene and disability level. Clinicians should consider phenotypic differences and goals of treatment (targeted symptom therapy vs overall functional outcomes) when selecting a treatment intervention to offer a more personalized approach to fatigue management. Limitations: The pragmatic design of this study did not allow for a placebo arm; however, the effect size of each treatment arm exceeded effect sizes seen in placebo arms of prior fatigue trials of similar duration. The decision to report all effect modifier results, regardless of statistical significance, rather than conduct multiple comparison corrections may raise concern about the risk for type I error. The number of participants with progressive MS or greater levels of disability were slightly lower than projected, which could have affected our heterogeneity of treatment effect analyses for disability. Similarly, the low proportion of marginalized or underrepresented participants with MS could have affected the generalizability of our findings. Background Multiple sclerosis (MS) is a chronic central nervous system disorder characterized by inflammation, myelin destruction, and axonal degeneration. This condition affects approximately 1 million people in the United States and is the leading cause of nontraumatic disability among young adults. 1 , 2 Fatigue is the most common chronic symptom experienced by persons with MS, affecting up to 90% of patients at some point during the disease course. 3-5 Nearly half of people with MS describe fatigue as their most disabling symptom 6 because of its impact on activities of daily living, social interactions, quality of life (QOL), standard of living, and employment. 4 , 7-12 Despite its prevalence and impact, MS fatigue remains immensely challenging to manage, in part because of decisional uncertainty regarding whether existing treatment modalities should be utilized and for which patients. The decisional uncertainly surrounding the treatment of MS fatigue has been fueled by 3 major research gaps: (1) lack of pragmatic comparative effectiveness trials to assess both nonpharmacologic and pharmacologic treatments; (2) lack of in-depth examination of potential treatment effect modifiers in fatigue intervention trials; and (3) lack of novel, reliable, patient-centered fatigue outcome measures. A critical need exists to identify targeted, patient-centered MS-fatigue management strategies that account for diversity among people with MS. Existing fatigue interventions include behavioral and pharmacologic strategies. Among behavioral treatments, cognitive behavior therapy (CBT) promotes effective self-management skills, including adaptive thought processes and behaviors (ie, ”coping skills”), goal-setting, and behavioral activation strategies for engaging in valued activities. Cognitive behavioral therapy has been shown to ameliorate MS fatigue in both placebo- and active comparator-controlled trials 11 , 13-17 through various modes of delivery, including in-person, 18 web-based, 15 and group-based 19-21 delivery. More recently, telephone-delivered CBT for MS symptom self-management has been shown to be an effective means of expanding access to this treatment. 22 Despite promising evidence that CBT is an effective monotherapy for many individuals with MS, it remains unknown whether combining CBT with other therapies could improve outcomes. Medications are frequently employed off-label to treat MS-related fatigue. 23 Modafinil is a safe, well-tolerated, and effective wake-promoting agent that is FDA approved for the treatment of excessive daytime sleepiness caused by obstructive sleep apnea, shift work disorder, and narcolepsy. Modafinil is also one of the most commonly used medications for MS-related fatigue in clinical practice. Although several studies, including those that used placebo comparators, have demonstrated the effectiveness of modafinil for MS-related fatigue, 24-26 limitations of previous nonpragmatic trials have precluded prescribing recommendations or FDA approval for MS-related fatigue. 27-30 Consequently, off-label modafinil use for MS fatigue lacks empirical support, which has contributed to uncertainty about when it should be used, and for which patients. To date, modafinil has not been compared with CBT directly or studied in combination with CBT. Additional crucial questions remain regarding which patients are most likely to respond to CBT or modafinil. Even the most encouraging trials of CBT or modafinil for fatigue have not fully accounted for clinically relevant effect modifiers, including depressive symptoms, sleep disturbances, and disability level, which are each highly prevalent in MS and closely associated with fatigue. To optimize personalized approaches to MS fatigue, these traits and their potential effects on responsiveness to fatigue interventions require further study. The present study was developed in response to a PCORI request for comparative effectiveness research that addresses decisional uncertainty in treatment selection for symptom management in MS; namely, research that “compares two or more treatment options for specific symptoms in people with MS where uncertainty exists.” 31 This report describes the rationale, aims, hypotheses, design, methods, and results of a pragmatic, randomized comparative effectiveness trial to compare the real-world benefits and harms of CBT, modafinil, and a combination of both treatments for MS-related fatigue in the context of common, clinically relevant potential effect modifiers. To enhance patient centeredness, the trial used novel, patient-centered fatigue measures in a sample of people with MS to provide actionable data for clinical practice and policy for the treatment of MS fatigue. Specific Aims Aim 1. Compare the effectiveness of 3 therapies: CBT monotherapy, modafinil monotherapy, and CBT + modafinil combination therapy on patient-reported fatigue impact (primary outcome), fatigue intensity, and fatigability among individuals with MS-related fatigue. Aim 2. Test whether depression, sleep disturbances, or MS disability level modifies comparative treatment responsiveness across treatment arms in terms of fatigue impact (ie, heterogeneity of treatment effects). Aim 3. Compare adverse events (AEs), side effects, treatment adherence, and patient dropout rates among the 3 treatment arms and patient subgroups of interest. Tested Hypotheses Aim 1, hypothesis 1. CBT + modafinil will more effectively ameliorate fatigue impact (primary outcome), fatigue intensity, and fatigability than either monotherapy in persons with significant MS-related fatigue. Aim 2, hypothesis 1. CBT + modafinil and CBT monotherapy will more effectively ameliorate fatigue impact than modafinil monotherapy for patients with higher levels of depression. Aim 2, hypothesis 2. CBT + modafinil and modafinil monotherapy will more effectively ameliorate fatigue impact than CBT monotherapy will in patients with known or suspected sleep disturbances not related to poor sleep hygiene (which would be expected to respond more favorably to CBT). Aim 2, hypothesis 3. CBT + modafinil and modafinil monotherapy will more effectively ameliorate fatigue impact than CBT monotherapy in subgroups with higher Expanded Disability Status Scale (EDSS) scores and in those with progressive MS compared with those with relapsing-remitting MS. Aim 3, hypothesis 1. CBT + modafinil combination therapy will be associated with higher treatment adherence rates than either monotherapy across the entire study sample. Participation of Patients and Other Stakeholders Aligned with PCORI methodology standards, the COMBO-MS study team was composed of both researchers and a group of stakeholders who were distinct from study participants. Stakeholders participated throughout all phases of the research operations, from study conception through dissemination of findings. 32 Our stakeholder panel consisted of patients with MS; clinical neuroimmunologists; MS center clinical staff; a payer representative from Blue Care Network; and individuals from community advocacy groups, including regional and national representatives of the National Multiple Sclerosis Society (NMSS) ( Table 1 ). Patient stakeholders were selected on the basis of their engagement with health care professionals (HCPs) regarding their care during clinic visits, prior experience with PCORI trials, or involvement with the NMSS. Health care professional stakeholders were selected on the basis of their clinical expertise and proximity to the parent study site. Our payer stakeholder was selected on the basis of its high number of covered lives and potential to shape payer policies. Patient-centered research questions were shaped directly by HCP stakeholders' clinical experiences and feedback from patient and payer stakeholders. Decisions regarding protocol development were made with direct stakeholder input, with consensus stakeholder approval before protocol finalization. Table 1 Partner Roster, Bios, and Roles (Names Withheld for Those Who Wanted to Remain Anonymous). Development of strong partnerships with patients, community organizations, physicians, and payers to ensure that our strategies were effective and consistent with the realities of people living with MS were primary goals when assembling the stakeholder panel. Additional goals included creation of a foundation for future research and implementation of our pragmatic findings into clinical care. Collaboration with clinical neuroimmunologists and clinic staff who treat people with MS ensured that our methods were consistently pragmatic, feasible, and could be implemented clinically. Partnership with a Blue Cross Blue Shield representative provided perspective on the process for payer coverage of fatigue treatments in MS. Our partnership with leadership of the NMSS allowed us to conduct research that was consistent with the priorities and lived realities of the MS community and to disseminate findings to a broader audience of adults living with MS and their health care teams. Engagement Structure The stakeholders and study team members participated in quarterly meetings, including 2 in-person meetings held at the lead study site (University of Michigan [UM]) and 11videoconference meetings between March 2018 and February 2022. Each meeting was well attended, with 15 to 24 attendees. During these meetings, stakeholders were involved in shaping decisions regarding study operations (including continuation of research activities in an adapted manner following onset of the COVID-19 pandemic), participant recruitment strategies, enrollment and retention, monitoring study progress, interpretation of findings, and dissemination plans. To facilitate communication, trust, and thorough integration of stakeholder feedback, we identified an engagement lead: study team member Kristen Pickup, MSW. Ms Pickup tracked engagement activities, including action items and outcomes, and served as a central contact for stakeholders between meetings to facilitate and streamline communication. The study stakeholders were encouraged to contact the engagement lead with any concerns, feedback, or questions ( Figure 1 ). Meeting agendas were emailed to stakeholders before meetings and always included a glossary of common research and study-specific terms that were updated throughout the course of the study. Consistent efforts were made to use clear and plain language in materials and to promote an environment that was encouraging and respectful of individual experiences and opinions. Minutes were disseminated following each meeting with requests for confirmation of minutes or edits and additions. Relevant materials (eg, draft recruitment materials and initial versions of the COVID-19 survey) were emailed by the engagement lead to stakeholders ahead of meetings to prepare attendees for discussion and enable them to contemplate pertinent issues before providing input on key decisions. Figure 1 Stakeholder Engagement Structure. Preparing Stakeholders for Engagement Before enrollment, the research team held an in-person study kickoff meeting, during which the study principal investigators, Tiffany Braley, MD, and Anna Kratz, PhD, provided an overview of the study, reviewed the goals and mission of PCORI, and discussed stakeholder goals and expectations. At this meeting, study stakeholders were informed of the significance of stakeholder engagement to the study. They discussed their experiences and motivation for engagement. Stakeholder roles were defined as being collaborative and reciprocal in nature, with the expectation that they would provide their individual expertise, experience, and input in the research process, which was discussed in a transparent manner. Stakeholder Contributions To enhance study recruitment and retention strategies, we collaborated closely with our patient stakeholders to obtain their unique perspectives of living with MS and participating in research. We solicited and incorporated their feedback regarding content and design of recruitment materials and sought approval of recruitment strategies. We also called on the expertise of our HCP stakeholders (physicians and clinic staff) to offer their perspectives, suggestions, and assistance with recruiting participants within their MS clinics. Patient stakeholders informed peers, and clinic stakeholders informed potential study candidates about the study and made study advertisements available. To ensure we covered a diverse range of recruitment sources and mechanisms, we incorporated the input from a thought leader representative from the NMSS (patient advocacy stakeholder) who had access to critical MS patient networks, both open and closed to the public. Our intention was to incorporate the diverse perspectives of individuals across the spectrum of MS care. Stakeholder engagement was critically important in terms of data collection methods. In particular, stakeholders were heavily involved in the construct of a study-specific COVID-19 impact survey ( Appendix A ), which was developed to assess COMBO-MS participant experiences during the COVID-19 pandemic. Stakeholders provided input on survey length and the addition or removal of specific items. Stakeholder involvement in the construct of the survey allowed for a more comprehensive assessment of the unique experiences of people with MS during the COVID-19 pandemic. Stakeholders also advised on our approach to the data analysis plan. Specifically, stakeholders were engaged in discussions about important variables to include in analyses, how best to capture modafinil usage given the trial's flexible dosing schedule, and how to characterize variables such as disability status. Stakeholders provided concrete feedback and ideas that were subsequently incorporated into the analysis plan, including comparisons of modafinil dose ranges in the combination therapy group vs the modafinil monotherapy group. Methods Study Overview This was a 3-group, randomized, analyst-blinded, pragmatic trial to compare the effectiveness of telephone-based CBT, modafinil, and the combination of both treatments for the management of MS-related fatigue. Eligible consenting adults with MS who endorsed chronic problematic fatigue were randomly assigned in a 1:1:1 ratio to receive telephone-based CBT (group 1: 12 weeks of one-on-one telephone-delivered sessions of CBT; 8 weekly programmatic sessions and up to 2 maintenance sessions); pharmacologic treatment with modafinil (group 2: target dose of 100-200 mg once or twice daily as tolerated for 12 weeks); or combination treatment with CBT and 12 weeks of modafinil (group 3). Self-reported measures of fatigue impact (Modified Fatigue Impact Scale [MFIS]; primary endpoint) and potential effect modifiers, including depression, sleep (measures of sleep hygiene, excessive daytime sleepiness scale, sleep duration, and presence of known or suspected obstructive sleep apnea), and MS disability level, were assessed at baseline and at 8 weeks, 12 weeks (primary end point), and 24 weeks. For additional data on physical activity level, fatigue intensity (numerical self-reported scale), and fatigability (the ratio of fatigue intensity to physical activity level), participants also underwent 1 week of continuous actigraphy monitoring and ecological momentary assessment (EMA) data collection with a wrist-worn accelerometer before treatment (baseline) and at 12 weeks. Study design and timeline are outlined in Figure 2 . For the purposes of this report, we focus only on the primary 12-week results. Figure 2 Study Design and Timeline. Study Setting The trial aimed to enroll at least 330 participants across 2 sites: the MS Center at the UM/Michigan Medicine in Ann Arbor (lead site and data coordinating center) and the MS Center at the University of Washington (UW) Medicine in Seattle, with approximately half of the participants enrolled and treated at each site. Assuming a 10% dropout rate, we anticipated that 330 enrollees (165 per site) would yield at least 100 completers in each study arm. The UM and UW multiple sclerosis specialty clinics served as primary recruitment sites. Additional participants were recruited from other neurology practices throughout central and southeastern Michigan (by invitation) and the Pacific Northwest, UM and UW electronic clinical trials registries, advertisement through community outlets (eg, posting flyers in community centers, distributing flyers, and fundraising walks), ClinicalTrials.gov, Facebook and other social media campaigns, iConquerMS, and advertisements posted by the NMSS (advocacy study stakeholder). Given the COVID-19 pandemic, the study settings consisted of either a combination of in-person and virtual activities (before March 2020) or entirely virtual activities. Screening and Baseline Assessments Initial contact with interested volunteers occurred either over the phone or in person. Individuals who remained interested in participating after a brief description of the study were asked several prescreening questions to determine appropriateness and timing of an in-person screening and baseline visit. Volunteers who met all prescreening requirements underwent a scheduled screening and baseline visit. At the baseline visit, informed consent was obtained, and all inclusion and exclusion criteria were reviewed by the coordinator and confirmed by an examining neurologist. The participant then completed a battery of web-based self-reported and interview-based baseline measures, prior to randomization. Participants were given a wrist-worn accelerometer/actigraph enhanced with a user interface that allowed for input of real-time self-reported data (ie, the PRO-Diary) and instructions for how to use it during the home monitoring period. Depending on random assignment, participants were given instructions regarding modafinil dosing or initiation of CBT. For the purposes of this pragmatic trial, in lieu of a research pharmacy, modafinil was dispensed at a designated clinical pharmacy located at either UM or UW to reflect standard clinical care. Participants For this pragmatic trial, inclusion and exclusion criteria were chosen to maximize generalizability across the MS spectrum. Inclusion and exclusion criteria are summarized in Table 2 . Table 2 Eligibility Criteria. Interventions and Comparators or Controls Study Interventions: CBT, Modafinil, and Combination Therapy Cognitive behavioral therapy Cognitive behavioral therapy for fatigue management (see Table 3 ) was chosen as a comparator given previously demonstrated benefits for people with MS. 15 , 17 For this study, CBT was conducted by telephone and delivered on a 1:1 basis. In a prior randomized trial of the telehealth CBT intervention that formed the basis for the manual used in this intervention, 50% of fatigued participants had a clinically meaningful reduction in fatigue impact that was maintained at 6 and 12 months post treatment. 22 The original telehealth CBT treatment protocol created by co-investigator Ehde et al 22 focused on multisymptom management of fatigue, pain, and/or mood. For this study, the therapist manual and participant workbook were adapted to specifically target fatigue by decreasing content related to pain and mood and increasing content related to fatigue, including the addition of more content related to sleep hygiene and behaviors, physical activity, and activity pacing. The CBT therapists followed the treatment manual, covering core content for each treatment session. Adherence to delivering core CBT content was assessed through a rigorous fidelity rating process (described in detail later). Consistent with this core content, the manualized CBT intervention was flexibly tailored to each participant to be responsive to their unique situation (eg, adaptations to skills practice to account for participant impairments, reordering of sessions based on participant priorities or needs, and more emphasis and time on content most relevant to participant) and to work toward individual fatigue-management goals (identified in session 2), much like what is done in clinical practice. Participants who received CBT (alone or in combination) were given digital audio recordings of relaxation and mindfulness exercises to facilitate practice of these skills. Table 3 Major Components of the Manualized CBT for Fatigue Management Intervention. The manualized CBT intervention consisted of 8 active or “programmatic” sessions and 2 “maintenance sessions.” As 8 weekly sessions of CBT are often not achieved within 8 consecutive calendar weeks in real-world clinical practice (owing to participant illness, scheduling errors or conflicts, travel, etc), the protocol was set up to ensure that patients could feasibly complete the 8 programmatic sessions within a 12-week interval, even in cases of multiple missed sessions. The maintenance sessions were included for those participants who successfully completed programmatic sessions at least 1 week before the end of the 12-week treatment period to maintain gains they made during the 8 active CBT sessions. All CBT therapists had special training in CBT for symptom management and telehealth delivery and were required to pass a certification procedure to ensure competence in the delivery of the manualized intervention. The team included 2 licensed clinical psychologists at each site (UM: Dr Kratz and Deirdre Conroy; UW: Kevin Alschuler and Dawn Ehde) who supervised the study; interventionists, including licensed clinical psychologists, clinical psychology predoctoral students, and postdoctoral fellows; and master's-level social workers. Therapist training included readings, didactics, and ongoing review and supervision of recorded sessions. In addition to the CBT supervision conducted at each site, the CBT team engaged in group supervision biweekly via videoconference. The fidelity protocol included therapist manuals, protocol checklists, weekly local supervision meetings, twice-monthly cross-site group supervision meetings, and an ongoing independent review (double-coded by study staff at UM and UW [ie, each selected session was coded by staff at each site]) of randomly selected digital recordings from at least 10% of all sessions. We found that the reviewed sessions were delivered with more than 96% fidelity. These procedures ensured that patients assigned to 1 of the CBT arms received CBT as prescribed in the treatment manual. Modafinil Modafinil (2-[(diphenylmethyl) sulfinyl]acetamide) is a centrally acting oral agent that is FDA approved to enhance wakefulness in patients with excessive sleepiness associated with narcolepsy, obstructive sleep apnea, or shift work disorder. Although the precise mechanism(s) through which modafinil promotes wakefulness remain unknown, the potential benefit of modafinil for fatigue in MS has been demonstrated in 5 previous studies. Three successful open-label studies that used various subjective fatigue measures demonstrated a benefit of modafinil on fatigue intensity. 24 , 40 , 41 A single-blind, placebo-controlled study also demonstrated a significant improvement in fatigue intensity with modafinil. 26 One 8-week, double-blind placebo-controlled trial that examined the effects of modafinil on fatigue intensity showed a significant improvement in fatigue with modafinil. 33 For participants who received modafinil alone or in combination, the medication was initiated following the 7-day period of baseline actigraphy monitoring and then continued until 12-week study assessments were completed. Unlike previous trials that studied fixed doses of modafinil, we employed a flexible dosing schedule that, in combination with study team guidance for dose adjustment based on perceived benefits, risks, and side effects, allowed participants to take between 50 mg daily and 200 mg twice a day as needed (see dosing escalation and adjustment instructions). This approach was chosen to allow participants to tailor the dose according to their needs and medical picture, as would typically happen in clinical practice. The study team and stakeholders also chose to include doses up to the maximally recommended daily dose (400 mg daily) based on prior clinician and patient stakeholder experiences, published data 42 , 43 that suggest a benefit for doses up to 400 mg in patients with depression (including those with fatigue), and limitations in prior trials that have contributed to decisional uncertainty regarding optimal dosing of this medication. When modafinil was dispensed to a participant, the study coordinator reviewed the modafinil dosing instructions with them, notified the participant of their study start date, and gave them a paper medication diary, which included (1) a brief overview of dosing instructions and how to adjust dosing in case of side effects or lack of effects; (2) general administration instructions; and (3) pages to record date and time, dose, and comments (eg, side effects) for each day of their treatment period. Participants were instructed to record every dose of modafinil and to return the medication diary at the conclusion of the study. Standard dosing escalation for participants with average risk. Modafinil 100-mg tablets were provided. Participants new to modafinil were instructed to start by taking 100 mg modafinil once daily upon awakening for 1 week. Depending on perceived response, participants were allowed to increase their dose stepwise to a maximum dose of 200 mg twice daily (upon awakening and around lunchtime; maximum daily dose of 400 mg). Alternatively, to lessen known side effects of the starter dose if needed, participants were allowed to reduce their dose to 50 mg once to twice daily. Participants who had previously used modafinil and were familiar with its effects were allowed to start at their previously perceived effective dose (though were still required to wash out of this medication for at least 30 days prior to enrollment). Reduced dosing escalation for participants with potentially increased risk. For safety, a separate reduced dosing schedule was outlined for participants who were both new to modafinil and had a history of liver problems; 65 years of age or older; or currently using warfarin, phenytoin, monoamine oxidase inhibitor, or cyclosporine; as well as for participants whom the treating neurologist thought could be at increased risk for sided effects. Participants who met these criteria were advised to initially take 50 mg once daily. Depending on results, participants were allowed to adjust their dose to up to 100 mg once to twice daily (maximum daily dose, 200 mg). Participants who were previously taking stable doses of modafinil without adverse effects prior to enrollment, regardless of study criteria, could continue their regular required dose of up to 400 mg per day. Dosing reductions and treatment discontinuation rules. Participants who experienced commonly encountered intolerances to modafinil (headache, gastrointestinal side effects, anxiety, and insomnia) that did not fulfill serious AE criteria were required to stop taking modafinil at the dose that was associated with the symptoms. On a case-by-case basis as determined by the treating neurologists, upon improvement of symptoms, participants were allowed to restart the medication at half of the dose that previously led to such symptoms. Participants who continued to experience symptoms after a rechallenge with half a dose of modafinil were instructed to discontinue modafinil but encouraged to remain in the study (and for participants who received combination therapy, to continue CBT). Combination therapy The CBT + modafinil combination therapy included simultaneous engagement in both therapies for 12 weeks. Discontinuation of 1 treatment due to participant preference or safety did not preclude continuation of the other treatment through the 12-week treatment interval. Randomization and Procedures to Minimize Bias Participants were randomly assigned using a web-based treatment assignment system (http://cscar-randomization.appspot.com/), which was developed and is supported by the UM Center for Consulting for Statistics, Computing, & Analytics Research. This system implements an approach to treatment assignment using minimization methodology that was developed by Pocock and Simon. 44 The minimization approach reduces covariate imbalances by using nonuniform assignment probabilities for the different treatment groups in order to reduce the level of imbalance following each round of treatment assignment. Randomization parameters set up a priori for this project include equal sampling rates (1:1:1), a determinism of 5, with stratification for MS subtype (relapsing or progressive). Subtype was included in the minimization procedure (a means of stratifying randomization) given the relatively high proportion of patients with progressive MS at the 2 academic centers (UM and UW) and prior published data implicating disability as a possible predictor of fatigue in MS (which relates closely to MS subtype). 45-48 Only statisticians were blinded for this study. Participants and study team members (principal investigators, co-investigators, and coordinators) were not blinded to treatment assignment. Study participants directly entered outcome data into web-based or ambulatory data collection devices, which reduced the chance of study staff biases affecting data collection and results. Data Safety and Monitoring IRBs at each site oversaw site-specific study activities, and additional IRB oversight was employed for both sites at the data coordinating center (UM). All IRB approvals were obtained prior to beginning any human participant activities. The trial was registered (ClinicalTrials.gov ID NCT03621761 ) prior to enrolling the first participant. To ensure adequate protection of the rights of participants, the study was monitored to ensure that all activities were implemented in accordance with the protocol and applicable federal and local regulations, that proper informed consent procedures were followed, and that the integrity and quality of the data were maintained. A qualified trial monitor from the Michigan Institute for Clinical & Health Research (https://michr.umich.edu/rdc/2016/6/2/study-monitoring) was appointed to monitor both study sites. Monitoring visits took place on a quarterly basis. Monitor visits were initially conducted in person but were converted to virtual visits following the onset of the COVID-19 pandemic. The site initiation and site closeout monitor visits were conducted to (1) ensure that the study team members received proper protocol training and had taken all necessary regulatory steps prior to enrolling the first participant and (2) to ensure that all participant interactions and study data collection were complete and stored in accordance with the protocol, respectively. The site initiation visit was conducted prior to enrollment, and the study closeout visit was conducted after the final participant was enrolled. In accordance with the protocol and all applicable regulatory requirements, an established monitoring plan was developed to ensure integrity of the data and compliance with good clinical practice. In addition to the clinical trial monitor, a board-certified neurologist who was not involved in the study or affiliated with either institution was appointed as a medical monitor to review AEs, provide feedback on potential protocol changes, and assist with decisions regarding treatment withdrawal and dose reduction for participant safety. Data were presented to the monitor for approval on an annual basis (in aggregate) and as needed for specific cases that were associated with safety concerns. Participants who demonstrated high suicide risk (ie, suicidal intent or plan) were excluded from the trial. A suicide safety protocol was developed, however, to use with any active participant who exhibited indications of possible self-harm at any time during contact with study personnel. For participants assigned to modafinil, rules for treatment discontinuation for safety reasons were defined a priori. Rules included pregnancy; any serious AEs; onset of a severe psychiatric episode (psychosis, severe or uncontrolled depression, or increased suicidality); suspected allergic reactions; and other physical, emotional, or behavioral signs thought to be related to the study treatment. Participants who had modafinil withdrawn for safety or ineligibility were encouraged to stay in the study for observation, continued outcomes assessment (for intention-to-treat analyses), and ongoing CBT (for the combination treatment group) until completion unless they chose not to continue. Study Outcomes Self-reported measures with evidence for reliability and validity were used in this study ( Table 4 ). Table 4 Study Measures (With Time Point and Data Collection Method). Primary Outcome Measures The Modified Fatigue Impact Scale 7 , 49 (MFIS; primary outcome measure) is a patient-reported, 21-item, validated measure of fatigue impact based on patients' recent personal experience. Higher scores indicate more fatigue. The MFIS items can be categorized into physical (9 items), cognitive (10 items), and psychosocial (2 items) functioning. Items are rated on a 0 (never) to 4 (almost always) Likert scale. Scale scores are derived by summing all responses; scores range from 0 to 84 for the full-scale score, 0 to 36 for physical, 0 to 40 for cognitive, and 0 to 8 for psychosocial subscale scores. A change of 10 points or more on the total MFIS score is considered clinically significant, 78 and this was the change threshold used for responder analyses. Secondary Outcome Measures The secondary outcome measure was an EMA (real-time self-report) of fatigue intensity with 2 items that assessed current levels of mental fatigue and physical fatigue rated on a numeric rating scale (NRS) of 0 (no fatigue) to 10 (extremely severe fatigue). These items were administered 4 times per day over the 7-day home monitoring period, at baseline, and at 12 weeks (see “ The PRO-Diary ” section for details on timing of assessment). A composite fatigue measure was created by summing the mental and physical fatigue responses; scores ranged from 0 to 20, with higher scores indicating higher fatigue intensity. A measure of perceived fatigability was created by generating a daily ratio of mean daily composite EMA fatigue to daily average activity counts per minute (accelerometer-derived value); these daily ratios were then averaged for each individual to generate an overall mean ratio of self-rated fatigue to mean physical activity; higher scores indicate greater fatigability. Given that our construction of a fatigability index from self-reported fatigue intensity and actigraph-derived physical activity is novel, there are no established cut-points for clinically significant change. The Patient Global Impression of Change 50 scale is a single item rated on a 7-point scale (from 1 = ”no change or condition has gotten worse” to 7 = ”a great deal better and a considerable improvement that has made all the difference”) that assesses respondents' global impression of change in activity limitations, symptoms, emotions, and overall QOL since beginning study treatment. Number of CBT sessions, as recorded by study therapists, and modafinil dosing, as reported by participants in a paper medication diary (daily dose entries), were collected as indication of study adherence and treatment usage patterns. Effect Modifier Measures Depression was measured using the Patient Health Questionnaire 8 (PHQ-8). 52-54 The PHQ-8 assesses the frequency of 8 depressive symptoms in the previous 2 weeks. It provides clinical cut-points ranging from no depression to severe depression (range, 0-24). Baseline scores were included in effect modification analyses. Disability level at study enrollment was measured by the Self-Reported EDSS (SR-EDSS). 51 The SR-EDSS is a validated self-reported version of the clinician-administered EDSS. 79 Scores on the SR-EDSS correspond highly with the clinician-rated version. Given the self-reported nature of the SR-EDSS, rating scores generally range from 0 to 8 in 0.5-point increments on an ordinal scale, based on the results of 7 functional systems subscores. The instrument was completed by participants at baseline, and each participant SR-EDSS score was double-scored by both UM and UW study team members. Scores that were discordant by 2 or more points were arbitrated by Dr Braley. Each participant's MS subtype (relapsing-remitting, primary progressive, or secondary progressive) was recorded from the medical record at the time of study participation. For statistical models, primary and secondary progressive subtypes were collapsed into a single group for comparison with the relapsing-remitting subtype. Sleep hygiene was measured using the Sleep Hygiene Index (SHI). 65 The SHI is a 13-item self-reported instrument that assesses practices and behaviors related to sleep hygiene using 5-point Likert scale items. A total score is calculated by summing each SHI item. Higher scores are indicative of poorer sleep hygiene. Subjective sleepiness was measured with the Epworth Sleepiness Scale (ESS). 61-64 The ESS is an 8-item questionnaire that asks about the likelihood of dozing in variably sedentary situations. This commonly used measure of subjective sleepiness, or sleep propensity, is reliable, internally consistent, and validated against objective polysomnographic measures. A score of 10 or greater suggests excessive sleepiness. Baseline ESS scores for each participant were evaluated as continuous and dichotomous (≥10) measures in effect modification analyses. Known or suspected obstructive sleep apnea (OSA) was determined by review of medical records (presence of obstructive sleep apnea diagnosis) or, in the absence of a preexisting OSA diagnosis, a STOP-Bang score of 3 or greater. The STOP-Bang Questionnaire 55 is a sensitive, reliable OSA screening tool that has been validated in several non-MS outpatient samples against overnight polysomnography, the gold-standard diagnostic tool for OSA. In a recent validation study led by Dr Braley, STOP-Bang score at a threshold score of 3 provided sensitivities of 87% and 91% to detect moderate and severe OSA, respectively; and negative predictive values of 84% and 95% to identify people with MS without moderate or severe OSA, respectively. 55 , 80 , 81 Specificities to detect moderate and severe forms of OSA in people with MS were 43% and 36%, respectively. Sleep duration (in minutes) was measured by actigraphy and calculated using the average duration of nightly sleep recorded throughout the baseline EMA monitoring period. Sleep duration was chosen given the high frequency of sleep deprivation in people with MS 82 and correlations between performance fatigue and sleep duration in prior general population studies. 83 We decided a priori to examine duration as both a continuous and dichotomous variable (using a cutoff of 7 hours based on American Academy of Sleep Medicine guidelines). 84 Covariates Models include demographic variables (self-reported age and sex, with female as the reference), study site (UM as the reference), physical activity (mean average activity counts at baseline as measured with an actigraph), characteristic pain 85 (average of pain intensity items from the Brief Pain Inventory multiplied by 100), and anxiety (Generalized Anxiety Disorder 7-item scores) as covariates. Adverse Events Per protocol, AEs that were not related to the normal course or chronic symptoms of MS (such as relapses or progression on magnetic resonance imaging scans) or events that were expected to occur in the course of daily life or through use of disease-modifying therapy (injuries, infections, disease-modifying therapy reactions, or disease-modifying therapy-induced lymphopenia) were collected during each visit, scheduled phone call, or unscheduled phone call. At each point of contact, participants were queried about any symptoms experienced during the study, with open-ended questions followed by questions directed toward events known to occur with the study treatments (including headache, nausea, nervousness, rhinitis, diarrhea, anxiety, insomnia, dizziness, dyspepsia, worsening depressive symptoms, psychosis, and increased suicidality) or suspected allergic reactions (rash; hives; mouth sores; blisters; swelling of the face, eyes, lips, tongue, or throat; trouble swallowing or breathing; fever; shortness of breath; lower extremity edema; jaundice; icterus; or dark urine). Occurrence of these issues prompted a review of treatment discontinuation rules (see “ Data Safety and Monitoring ”). Sample Size Calculations and Power Statistical Power For aim 1 (primary sample size determinant), the 21-item MFIS was the primary outcome variable of interest. The MFIS has been validated in patients with MS, is frequently used in MS clinical studies as a reliable measure of fatigue, and is sensitive to changes in intervention. 22 , 26 , 27 , 86-91 The primary outcome measure was the mean within-subject difference between baseline and 12-week MFIS values (δ-MFIS) compared between groups. Based on a power analysis of 2-sample means (combined monotherapy groups vs combination therapy, group weights 2:1) with a 2-sided α of .05 and SD of 12, a sample size of 300 participants was determined to provide 92% power to detect a δ-MFIS of 5 between the monotherapy and combination treatment groups. An MFIS score reduction of 10 is traditionally accepted as a clinically meaningful change in monotherapy trials 92 ; however, expecting that combination therapy would provide at least 50% greater improvement than the clinically significant difference achieved by either monotherapy (eg, an additional 5-point reduction in the MFIS), the trial was powered to detect a difference of this magnitude. A multiple linear regression power analysis also suggested that a sample size of 300 would provide 85% power to detect an R 2 difference of 0.04 for the interaction terms of interest (used to test heterogeneity of treatment effects), assuming the R 2 explained by the adjustment variables was at least 0.3, for our aim 2 interaction models with up to 11 covariates of interest. Time Frame for the Study Participants' activities (enrollment, study intervention delivery, and data collection) occurred between November 15, 2018 (first participant consented), and November 30, 2021 (final participant completed follow-up data collection). Data Collection and Sources Data Collection Procedures and Measures As depicted in Figure 2 , primary outcome data were collected at baseline (pretreatment) and at 8 and 12 weeks after treatment initiation. Patient-reported outcomes were collected on a wrist-worn device (see “ The PRO-Diary ”) during two 7-day home monitoring periods (baseline and week 12) and via a web-based data collection platform (Research Electronic Data Capture [REDCap]; baseline, 8, 12, and 24 weeks). The PRO-Diary The PRO-Diary (CamNTech) is a wrist-worn accelerometer enhanced with a user interface that allows for input of real-time self-reported data. The PRO-Diary was used to provide optimally sensitive assays of fatigue intensity, fatigue impact, physical activity, and fatigability (calculated using fatigue intensity and activity measures). Self-reported data were entered into the PRO-Diary 4 times a day for 7 consecutive days at baseline and week 12. Ratings were initiated by the participant upon waking (the time they woke up, not necessarily when they got out of bed) and at bedtime (“lights out,” or the time they intended to go to sleep, not necessarily the time they got into bed). An alarm alerted participants to use the touchscreen to log symptom ratings at 2 times between 11:00 am and 7:00 pm . The PRO-Diary collected physical activity data as activity counts in 15-second epochs during the home monitoring period and time-stamped self-reported data that were stored until the watch was returned to the laboratory for data download, cleaning, and analysis. The PRO-Diary also generated accelerometer-derived sleep variables (sleep duration, wake after sleep onset, sleep latency, and sleep efficiency). REDCap This study used REDCap, 93 , 94 a secure, password-protected, and HIPAA-compliant web-based data platform hosted by the Michigan Institute for Clinical and Health Research for data capture and storage. This protected database was accessed and maintained by study personnel only. This system features both a local and remote web-based interface, secure data transfer, and an Oracle database. Data security, patient privacy, and HIPAA requirements are a premium consideration for clinical trials research using REDCap. A complete time-stamped audit of all REDCap activity (including which study personnel access data and when) was maintained, adding to the security and fidelity of the data. Study personnel entered data into the database through administrative access to add to the self-reported data provided by participants. For self-reported data collection, participants accessed individualized study URLs to securely enter data either during clinic visits or at a time of their choosing on an internet-connected device of their choosing (for home-based assessments). Participant email addresses were used to send the follow-up survey links. Participant emails were stored in REDCap. University of Washington study staff were only able to view email addresses of UW participants, but study staff at UM (the data coordinating center) had access to email addresses for all study participants, including those from UW. Data transfer (M+Box) Transfer of accelerometer and EMA data from the PRO-Diary from UW to the UM data coordinating center was achieved via M+Box, UM's implementation of the Box.com cloud storage and collaboration service. M+Box is capable of handling protected health information and other sensitive data (eg, research data). We have used M+Box extensively in previous studies to transfer PRO-Diary data to the lead site for data cleaning and analysis. After being cleaned and scored, the PRO-Diary data were later merged with the REDCap data before analysis. Analytical and Statistical Approaches Missing Data Before conducting data imputation strategies, we assessed the extent of missing data in the study; the focus was on the MFIS survey, the primary outcome variable. Regarding the missingness of MFIS raw scores at baseline, there was just 1 participant who had completely missing data for the MFIS and raw scores and 25 participants who had missing responses (data) on 1 or more items of the MFIS (of these, 19 had data missing for just 1 item of the MFIS); 311 (92.6%) participants had complete information for all 21 MFIS items at baseline; at week 12, there were 24 participants whose records for all 21 MFIS raw scores were missing and 15 participants who had missing responses on 1 or 2 of the items of the MFIS; 297 (88.4%) participants had complete information for all 21 MFIS raw scores at week 12; and none of the 336 participants had missing data for all 21 MFIS raw scores at both baseline and week 12. To deal with missing data in MFIS models, multivariate imputation by chained equations 95 was implemented using the R package mice. 96 In this procedure, imputations for each variable with missing data were generated in a manner conditional on the other variables according to a separate imputation method, which was logistic regression for binary variables and predictive mean matching 97 for continuous variables. The number of imputed data sets was set as 10. After the imputation, multiple linear regression models were fit to each of the 10 imputed data sets, and the results (β-coefficients) were then pooled using standard methodology 98 ). The data set for imputation included the following variables: (1) raw scores on items 1 through 21 for MFIS at baseline and week 12 and (2) all other baseline covariates (main effect terms only) considered in analyses for aims 1 and 2: age, sex, anxiety, pain, activity, study site, treatment group, PHQ-8 (continuous), MS subtypes (progressive MS vs relapsing-remitting MS), EDSS (continuous), ESS (continuous), SHI (continuous), sleep deprivation (yes = 1: average sleep duration <7 hours per night and no = 0: ≥7 hours per night), and known or suspected OSA (yes = 1 and no = 0). Analyses were completed using the R software ( https://cran.r-project.org/ ). For secondary analyses, models were run for complete cases only, given the relatively small amount of missingness regarding outcome data. Statistical Analysis Plan Aim 1: Compare the effectiveness of 3 therapies (CBT monotherapy, modafinil monotherapy, and CBT + modafinil combination therapy) on patient-reported fatigue impact (primary outcome), fatigue intensity, and fatigability among fatigued individuals with MS. The primary outcome measure was the 12-week change from baseline in the mean MFIS score (δ-MFIS), compared between treatment groups. Multiple linear regression models were used. This approach has been shown to be similar to linear mixed modeling in the absence of intermediate measures and is generally regarded as the preferred approach. Further, regression models that condition on baseline measures have a simpler causal interpretation (for more discussion, see O'Connell et al 99 ). The mean difference was calculated in unadjusted models and in models adjusted for the following baseline covariates: age, sex, baseline anxiety, baseline pain, baseline MFIS score, and baseline physical activity (average activity counts from accelerometer), which were included as covariates based on their established association with fatigue in MS. Additional similar secondary regression analyses with EMA measures were run to examine treatment effect on change in EMA outcome measures of fatigue intensity and fatigability at 12 weeks; fully adjusted regression models only are presented for the secondary outcomes. Aim 2: Test whether depression, sleep disturbances, or MS disability level modifies comparative treatment responsiveness across treatment arms in terms of fatigue impact (MFIS). Heterogeneity of treatment effects of several key potential effect modifiers, including depression severity (PHQ-8 score), progressive MS subtype, disability level (EDSS score), subjective sleepiness (ESS score), confirmed or suspected OSA, and sleep hygiene (sleep hygiene index), on MFIS scores (as well as EMA fatigue intensity and fatigability) were evaluated with interaction terms in multiple linear regression models. Adjustments for multiple comparisons were not made; rather, based on recommendations for best practices, 100 , 101 all results are reported whether or not models were statistically significant. Aim 3: Compare AEs, side effects, treatment adherence, and patient dropout rates among the 3 treatment groups and patient subgroups of interest. Descriptive statistics were used to analyze AEs, including side effects (incidence, type, severity, and relatedness) and treatment usage patterns in all treatment groups (modafinil vs CBT vs combination therapy) throughout the 12-week intervention interval. We also assessed whether the degree of treatment effects measured in aim 1 was related to treatment adherence (eg, dose response). Safety analyses and adherence were summarized with descriptive statistics for all participants who received at least 1 modafinil dose or CBT, classified using the current Common Terminology Criteria for Adverse Events. Changes to the Original Study Protocol Approximately 17 months into study recruitment, in March 2020, study adaptations were initiated to accommodate new public and institutional safety protocols imposed by the COVID-19 pandemic. These changes, approved in May 2020 at the UM site and July 2020 at the UW site, allowed for virtual, no-contact participation on a 100% virtual platform. No changes to the study aims, interventions, timing of study activities, or selection of study participants were necessary: The enrollment visit, which was previously conducted in-person, was changed to a 3-way HIPAA-compliant Zoom for Healthcare video conference meeting among the study participant, the study coordinator, and the study neurologist. Informed consent procedures, which were previously conducted in-person, were changed to allow for completion either via SignNow (e-signature) or by a mailed paper consent. A new inclusion criterion requiring access to internet with video-capable smartphone or computer was added to eligibility criteria. In-person vital sign assessments were omitted. Consequently, eligibility criteria were changed to exclude participants who carried a diagnosis of hypertension within the past year that was not actively managed by an HCP. Female participants of childbearing age who were unsure of pregnancy status were mailed a home testing kit to confirm negative pregnancy status prior to enrollment. Participants randomly assigned to receive modafinil (monotherapy or combination therapy) were allowed either to have their medication mailed directly from the pharmacy (Michigan) or to pick it up from a drive-through pharmacy (Washington). Participants were mailed an Rx Destroyer and instructed to dispose of leftover medication at home as opposed to in-person at the end of the 12-week treatment period. Baseline visit materials were mailed to participants ahead of the visit. Additional materials were mailed to participants after random assignment, including the PRO-Diary along with prepaid return box and, if randomly assigned to the study drug, medication instructions, medication diary, and drug disposal solution (Rx Destroyer). With stakeholder input, we designed and deployed a COVID-19 Impact Survey to be given at baseline and at 8, 12, and 24 weeks for participants enrolled after the start of the COVID-19 pandemic. In addition, several study outcomes that were planned to be used per the original research proposal were ultimately not collected or analyzed for this report. Before study onset, the study team decided not to collect the Brief Fatigue Inventory because this measure has not been validated in people with MS and because of the strengths of the MFIS and EMA fatigue intensity (which cover fatigue impact and severity, respectively). In addition, although the original research proposal included use of the Patient Health Questionnaire-9 to assess depressive symptoms, prior to study initiation, this was changed to the PHQ-8. The PHQ-8 is a validated measure also used to assess depressive symptoms sensitively and reliably in people with MS. The primary rationale for switching was to reduce participant and study team burden triggered by the suicidality item on the PHQ-9. In the investigators' experience, this item is associated with a high number of false positives, and its omission was not thought to increase risk in study participation. Results Participants A total of 336 people with MS were enrolled and randomly assigned across both study sites. As indicated in the CONSORT diagram ( Figure 3 ), 44% of participants conducted their participation in a completely virtual manner. Figure 3 CONSORT Diagram. Baseline Analyses Baseline characteristics for the sample are shown in Table 5 . Each of the baseline covariates and treatment effect modifiers included in statistical models were balanced by randomization between treatment groups (ie, no significant group differences in terms of mean age; Generalized Anxiety Disorder 7-item scores; characteristic pain intensity; physical activity [accelerometer]; PHQ-8 scores; EDSS, ESS, SHI, or sleep duration scores; or in terms of distribution of sex, study site, sleep deprivation, OSA risk, or MS subtype). Mean (SD) age was 48.8 (11.6) years. Seventy-six percent of participants were female, and 85% were White. Twenty-eight percent had a progressive MS subtype (mean [SD] EDSS, 5.1 [1.4]). Mean (SD) baseline MFIS score was 52.7 (14.1), and 7-day mean (SD) fatigue intensity NRS was 9.0 (3.0). Mean (SD) ESS and SHI scores were 9.9 (5.4) and 16.4 (6.5), respectively. Half of participants (50.9%) had known or suspected OSA. Mean (SD) PHQ-8 score was 9.0 (4.9). Ten percent of participants had previously used modafinil for fatigue management. Table 5 Baseline Characteristics. Analyses of Treatment Group Effect Change in Fatigue Impact (MFIS)—Primary Outcome A total of 336 participants (100%) completed 12-week study activities. Results of imputed and adjusted multiple linear regression models that predicted change in total MFIS score are shown in Table 6 . At 12 weeks, all 3 treatment arms were associated with statistically and clinically significant (ie, >10 points) reductions in total MFIS score from baseline, with total mean (SD) MFIS score reductions of 15.2 (11.9) for CBT, 16.9 (15.9) for modafinil, and 17.3 (16.2) for combination therapy (spaghetti plot, Figure 4 ). Reductions in MFIS score (primary outcome) did not differ by treatment group. Relative to combination therapy, the unadjusted mean difference in the change score was 1.26 (95% CI, −2.91 to 5.42; P = .55) for CBT and −0.56 (95% CI, −4.68 to 3.56; P = .79) for modafinil. Data imputation did not affect the direction or significance of associations seen in the analyses of complete cases. Reductions in MFIS subscale scores (physical, cognitive, and psychological fatigue impact; Appendix B ) resembled those for the total score. Table 6 Adjusted Models of Treatment Effect on Change in Total MFIS Score at 12 Weeks Following Imputation of Missing Values (Pooled Results Based on 10 Imputed Data Sets, With Combination Therapy as the Reference) . Figure 4 Overall Treatment Effect on Total MFIS Score at 12 Weeks. In post hoc responder analyses, more than 65% of participants in each of the 3 treatment groups saw at least a 10-point reduction in their total MFIS score. More than half of the people in each group experienced at least a 25% reduction in their MFIS score ( Appendix C ). Change in Fatigue Intensity (EMA Fatigue Intensity)—Secondary Outcome The results of adjusted multiple linear regression models (complete cases only) that predicted change in EMA fatigue composite score and physical and cognitive subscale scores at week 12 are shown in Table 7 . All 3 treatment arms were associated with statistically significant reductions in EMA fatigue intensity score. At 12 weeks, telephone-based CBT, modafinil, and combination therapy were each associated with unadjusted mean (SD) NRS score reductions of 1.41 (2.43), 1.69 (2.33), and 2.03 (2.82), respectively. Reductions in EMA fatigue intensity did not differ by treatment group. Data imputation did not affect the direction or significance of associations seen in the analyses of complete cases (data not shown). Reductions in EMA fatigue intensity subscale scores (physical and mental fatigue intensity) resembled those for the total score. Table 7 Adjusted Models of Treatment Effect on EMA Fatigue Intensity Measures (Complete Cases Only, With Combination Therapy as the Reference) . Change in Fatigability (Ratio of EMA Fatigue Intensity to Physical Activity)—Secondary Outcome The results of adjusted multiple linear regression models that predicted change in overall (composite) fatigability and subscores of physical and mental fatigability at week 12 are shown in Table 8 . All 3 treatment arms were associated with within-group statistically significant reductions in overall (composite) and physical fatigability scores. Analyses of all available cases suggested that change in fatigability did not differ across treatment groups. However, following removal of 4 outliers (identified via graphical diagnostic procedures—Normal Q-Q, Residuals vs Leverage, Scale-Location, and Residuals vs Fitted plots), CBT showed a significantly different effect on changes in fatigability composite score (β = .007, P = .04) and mental fatigability score (β = .004, P = .02) compared with combination therapy. Table 8 Adjusted Models of Treatment Effect on Fatigability (Complete Cases Only, With Combination Therapy as the Reference) . Patient Global Impression of Change Results from a Kruskal-Wallis test indicated omnibus differences between treatment groups in 12-week Patient Global Impression of Change (PGIC) ratings ( P = .019). Combination therapy was associated with the highest average PGIC score (mean [SD], 5.08 [1.56]; median, 6); pairwise Wilcoxon rank tests indicated that differences between scores for combination and modafinil monotherapy (mean (SD), 4.50 [1.74]; median, 5) were statistically significant ( P = .019), but differences between scores for combination therapy and CBT monotherapy (mean [SD], 4.70 [1.59]; median, 5) were not significant (P = .067). Differences between the monotherapies were also nonsignificant ( P = .488) ( Appendix D ). Effect Modification Analyses (Heterogeneity of Treatment Effects), by Outcome Heterogeneity of treatment effect analyses were conducted with MFIS scores as the outcome. Given the robust effects of each treatment group on EMA measures of fatigue intensity, interaction models were repeated fatigue intensity scores. Outcome: MFIS Interaction effects of depressive symptomatology on MFIS score. In adjusted interaction models imputed for missing MFIS items, with modafinil as the reference group, depression severity (as measured by PHQ-8 score, continuous) did not significantly moderate treatment effect on MFIS scores. In other words, depression did not differentially impact the response to any of the 3 treatment arms. Interaction effects of MS Disability on MFIS score: – EDSS score. In adjusted interaction models imputed for missing MFIS items, with modafinil as the reference group, EDSS score (continuous) did not significantly moderate treatment effect on MFIS scores. – MS subtype. In adjusted interaction models imputed for missing MFIS items, with modafinil as the reference group, MS subtype (progressive vs relapsing MS) did not significantly moderate treatment effect on MFIS scores. Interaction effects of sleep disturbances on MFIS score: – Sleepiness. In adjusted models imputed for missing MFIS items, with modafinil as the reference, ESS score (both continuous and dichotomized scores with a cutoff of ≥10) had a significant direct effect on change in total MFIS scores at 12 weeks (β = −.3899 and P = .0087 for continuous ESS effect on total MFIS change score; β = −4.2257 and P = .0061 for dichotomized ESS effect on total MFIS change score; see Table 9 ). When interaction terms were added to the models, however, ESS did not significantly moderate effects of specific treatments on MFIS score. – Sleep hygiene. In adjusted interaction models of complete cases, with CBT as the reference, sleep hygiene significantly moderated treatment effect on MFIS scores in complete case models (β = .6996, P = .0327), with CBT showing the greatest benefit for fatigue in patients with poor sleep hygiene, which varied statistically from the benefit of modafinil monotherapy. Statistical significance of this interaction, however, diminished with imputation for missing data (β = .3861, P = .2009). See Table 10 and Figure 5 . – Known or suspected OSA. In adjusted interaction models imputed for missing MFIS items, with modafinil as the reference group, the presence of known or suspected OSA as a dichotomous measure did not significantly moderate treatment effect on MFIS scores. – Sleep duration. In adjusted interaction models imputed for missing MFIS items, actigraphy-based sleep duration as a dichotomized measure (cutoff <7 hours per night on average) did not significantly moderate treatment effect on MFIS scores. Table 9 Imputed Model With Modafinil as the Reference Group, Showing Direct Effects of ESS Score (Continuous) Change in MFIS Scores at 12 Weeks . Table 10 Heterogeneity of Treatment Effect Analyses for Interactions Between SHI Score and Treatment Group on MFIS Scores (CBT as Reference Group) . Figure 5 Line Plot (With Shaded 95% CIs) Showing Predicted Reduction in MFIS Total Score From Baseline to Week 12 Based on SHI Score (Higher Scores Indicate Poorer Sleep Hygiene; Score Is Mean Centered), by Treatment Group. Outcome: EMA Measures of Fatigue Intensity Interaction effects of depressive symptomatology on EMA fatigue intensity score. In adjusted interaction models of complete cases, using modafinil as the reference group, interaction effects of depression (continuous PHQ-8 score) as a treatment effect modifier on EMA fatigue intensity approached statistical significance ( P = .0597), with modafinil showing the greatest reduction in fatigue intensity for patients with more severe depression ( Table 11 , Figure 6 ). Interaction effects of disability on EMA fatigue intensity score: – EDSS score. In adjusted models of complete cases, using modafinil as the reference, disability level (EDSS score as a continuous measure) significantly moderated treatment effect on EMA measures of fatigue intensity, with modafinil showing the most robust benefits for those with lower disability levels ( P = .0140). In participants with higher levels of disability, all 3 treatments were associated with similar reductions in fatigue ( Table 12 , Figure 7 ). – MS subtype. In adjusted models of complete cases, using modafinil as the reference, MS subtype (progressive vs relapsing MS) did not significantly moderate treatment effect on EMA fatigue intensity scores. Interaction effects of sleep disturbances, on MFIS score: – Sleepiness. In adjusted models of complete cases, with modafinil as the reference, ESS score (both continuous and dichotomized scores with a cutoff of ≥10) had a significant direct effect on change in EMA fatigue intensity scores at 12 weeks (β = −.0678 and P = .0199 for continuous ESS effect on total MFIS change score; β = −.8878 and P = .0034 for dichotomized ESS effect on total MFIS change score; see Table 13 ). The ESS did not significantly moderate effects of specific treatments on EMA fatigue intensity scores with addition of interaction terms to the models. – Sleep hygiene. In adjusted interaction models of complete cases, with CBT as the reference group, sleep hygiene as measured by the SHI did not significantly moderate treatment effect on EMA fatigue intensity scores. – Known or suspected OSA. In adjusted interaction models of complete cases, with modafinil as the reference group, the presence of known or suspected OSA as a dichotomous measure did not significantly moderate treatment effect on EMA fatigue intensity scores. – Sleep duration. In adjusted interaction models of complete cases, actigraphy-based sleep duration as a dichotomized measure (cutoff of <7 hours per night or <6 hours per night on average) did not significantly moderate treatment effect on MFIS scores. Table 11 Heterogeneity of Treatment Analyses for Interactions Between PHQ-8 Score and Treatment Group, on EMA Fatigue Intensity Score (Modafinil Group as Reference Group, Complete Cases Only) . Figure 6 Line Plot (With Shaded 95% CIs) Showing Predicted Reduction in EMA Fatigue Intensity Composite Based on PHQ-8 Score by Treatment Group. Table 12 Heterogeneity of Treatment Analyses for Interactions Between EDSS Score and Treatment Group on EMA Fatigue Intensity Score (Modafinil Group as Reference Group, Complete Cases Only) . Figure 7 Line Plot (With Shaded 95% CIs) Showing Predicted Reduction in EMA Fatigue Intensity Composite Score Based on EDSS Score, by Treatment Group. Table 13 Complete Case Model With Modafinil as the Reference Group, Showing Direct Effects of ESS Score (Continuous) Change on EMA Fatigue Intensity at 12 Weeks . Analyses of Adherence and AEs Study Withdrawals A total of 8 participants withdrew from the study: 2 (1.7%), 4 (3.5%), and 2 (1.8%) from the CBT, modafinil, and combination groups, respectively (see CONSORT diagram). The primary reason for withdrawal was lack of time and interest. Treatment Withdrawals A total of 26 people (7.7%) discontinued at least 1 of the assigned interventions (see CONSORT diagram). The most common reason for modafinil discontinuation (20 participants) was intolerable side effects, occurring in 9% of those prescribed modafinil. Six participants who were prescribed CBT (2.6%) discontinued it, citing time conflicts as the most common reason for discontinuation. Adherence to CBT Most participants assigned to receive CBT in either the CBT monotherapy (97 [85.1%]) or the combination therapy arm (97 [89.8%]) completed all 8 of the active (nonmaintenance) CBT sessions. The mean number of sessions completed was 7.25 (SD, 2.01; median, 8) for CBT monotherapy and 7.57 (SD, 1.43; median, 8) for combination therapy (a nonsignificant difference; t [220] = 1.37; P = .17). Most participants completed either 1 (64 [29.1%]) or 2 (96 [43.6%]) maintenance sessions; 60 participants randomly assigned to receive CBT (27.3%) completed no CBT maintenance sessions. Modafinil Usage (Combination and Monotherapy Groups) Of the 222 individuals randomly assigned to either CBT monotherapy or combination therapy, 139 (62.6%) returned the paper medication diary. On average, participants took the study drug on 78 out of the 84 days of prescribed therapy. Per medication diary data, total daily modafinil doses ranged from 25 mg to 400 mg for all modafinil recipients. The most common daily doses used were 100 mg daily (49% of recorded doses) and 200 mg daily (30% of recorded doses) across all diary entries. Modafinil was more commonly used only once per day (69%) than twice daily (22% of daily doses). No doses or missing data accounted for 9% of diary entries. In terms of average daily dose per participant, 81% of participants used average doses that ranged between 100 mg and 200 mg per day. Twelve percent of participants used an average dose of more than 200 mg per day. These estimates were similar across the combination and monotherapy groups ( Table 14 ). Table 14 Modafinil Usage Patterns Across All Modafinil Recipients and by Treatment Group. Adverse Events Overall, all 3 treatment interventions were well tolerated with no related serious AEs. A total of 0, 27, and 32 AEs were determined as probably or definitely related to treatment in the CBT, modafinil, and combination groups, respectively. The most common AEs (summarized in Table 15 ) that occurred in at least 5% of participants in at least 1 treatment group (regardless of relatedness) were insomnia and anxiety (modafinil and combination only) and headache. Table 15 Adverse Events for 336 Randomly Assigned Participants. Discussion Summary of Results In people with MS with chronic, problematic fatigue, telephone-based CBT, modafinil, and combination therapy were each associated with clinically significant reductions in fatigue impact, as measured by the MFIS; fatigue intensity, as measured by EMA; and subjective fatigability (the ratio of EMA fatigue intensity to accelerometer-measured physical activity); however, no differences in fatigue impact reduction were detected between groups. Contrary to expectations, combination therapy was not associated with a benefit in fatigue impact above and beyond what was achieved with either CBT or modafinil monotherapy. Although MFIS- and EMA-based absolute score reductions did not vary statistically by treatment group across this sample, heterogeneity-of-treatment-effects analyses suggested differential treatment effects based on several clinical characteristics. Disability level (EDSS score) significantly moderated treatment effect on EMA measures of fatigue intensity, with modafinil showing the most robust benefits for those with lower levels of disability. Effects of depression (PHQ-8 score) as a treatment effect modifier on EMA fatigue intensity approached statistical significance, with modafinil showing the greater reduction in fatigue intensity for participants with more severe depression, but treatment effects were similar across groups in those with low depression scores. Sleep hygiene significantly moderated treatment effect on MFIS scores in complete case models, with CBT showing the greatest benefit for fatigue in patients with poor sleep hygiene, but the statistical significance of this interaction diminished with imputation for missing data. Participants with more severe subjective sleepiness (ESS score) experienced greater reductions in MFIS and EMA fatigue intensity scores that did not vary by treatment group. Global QOL as measured by PGIC scores was highest for combination therapy. All treatment modalities were well tolerated with a low number of AEs. Results in Context Comparative effectiveness data from this pragmatic trial show that telephone-based CBT, modafinil, and combination therapy—delivered in a patient-centered manner—offer similar clinically significant benefits for MS-related fatigue in people with MS, with high patient acceptance and good tolerability. Although no therapy was superior to another across the entire sample, the effectiveness of each treatment on fatigue symptoms could vary by clinical characteristics, including depression, sleep hygiene, and disability level. More broadly, combination therapy, while not superior to CBT or modafinil monotherapy for reduction in fatigue impact, was more frequently associated with global benefits in self-reported functional outcomes and QOL. Potential to Affect Health Care Decision-Making These data suggest that modafinil and CBT, as either monotherapy or combination therapy, should be considered as potential treatment options for people with MS with chronic problematic fatigue. These findings justify measures to facilitate coverage of and access to these treatments. Patient characteristics, including personal goals of care, comorbid symptoms and medical comorbidities, and QOL, should inform treatment selection. Fatigue is a complex problem that varies by personal experience and clinical characteristics, highlighting the need for patient-centered approaches that take these factors into account when considering both pharmacologic and nonpharmacologic treatment approaches. In addition, a critical need exists to explore the effects of real-world delivery of these treatments in a manner that includes customized dosing and the option for combination therapy. Consistent with this idea, experts in the field have called for the use of combined therapies that include both medication and behavioral (eg, CBT) interventions to treat MS-related fatigue. 102-104 Studies of other chronic health conditions, such as major depressive disorder 105 , 106 and fibromyalgia, 107 suggest that combination therapies of both pharmacologic and behavioral approaches can have synergistic effects and offer the best chance of long-term success for symptom management. 108 Ours is the first trial to test the effectiveness of such combination treatments for MS fatigue. The assessment of combination therapies for fatigue is particularly crucial when considering the overall milieu of day-to-day challenges and chronic symptoms experienced by people with MS, who may benefit from multimodal treatments. As demonstrated by our group and others, symptoms associated with MS frequently cluster and are temporally associated. 109-112 Perhaps for this reason, the effect of combination treatment on isolated fatigue, while similar for each intervention as monotherapy, was largest on overall perceived treatment effect per PGIC scores. These findings support the consideration of mood, activity limitations, and overall QOL when choosing treatment for fatigue. Relevant to implementation, patient-centered dosing ranges and schedules are important factors that have been insufficiently incorporated into fatigue intervention research. To this end, the flexible dosing and treatment schedules included in our trial deserve comment. In addition to clinically significant treatment effects, the modafinil-containing treatment arms were each associated with low treatment discontinuation and good tolerability. The dosing range and schedules in this study (which were informed by real-world clinical use and shaped by patient stakeholder input) were designed to reflect dose adjustments in clinical practice in order to maximize benefits and minimize risks and harms. These data highlight the importance of personalized dose titrations and close communication between patients and HCPs when prescribing modafinil. It is noteworthy that approximately 20% of participants in our study used doses outside the 100-mg to 200-mg dosing range. Although modafinil 100 mg or 200 mg formulations are frequently prescribed in MS clinic settings, in the investigators' experience, many patients are not given the option of dose titrations lower than 100 mg to enhance tolerability or beyond 200 mg for those with higher dosing needs. Additional guidelines for HCPs and patients regarding the importance of patient-centered dose adjustment could enhance the benefit/safety ratio. Similarly, timing of dosing is a critically important factor that can influence the effectiveness and tolerability of many medications. Dosing guidelines that allow patients reasonable flexibility to time modafinil administration around their schedule and peak fatigue levels could optimize benefits. For example, a person who typically awakens refreshed around 9:00 am but begins to experience fatigue at around 2:00 pm may benefit most from timing their dose around noon on days when fatigue is anticipated, as opposed to “every morning” to allow modafinil levels to peak when this medication is most needed. Conversely, given the long half-life of modafinil, patients should also be cautioned to avoid late afternoon doses of modafinil, which can interfere with sleep onset or worsen sleep quality. Our findings of exceptionally high adherence to the CBT intervention provide lessons about the importance of delivering behavioral health care in an accessible format. Despite the potential benefits of behavioral health care for fatigue, a high percentage of people with MS with significant fatigue have limited access to behavioral health care that promotes symptom self-management. 113-115 Self-management interventions for MS are inaccessible to many individuals living in rural regions or with other barriers to frequent in-person appointments because of fatigue, driving restrictions, or limited access to transportation. 116 With these realities in mind, even before the onset of the COVID-19 pandemic (which introduced an urgent need for virtual delivery of health care), CBT was delivered via telephone to increase access and convenience to participants in this study. Significant effects of CBT on fatigue impact, fatigue intensity, and fatigability in this study build on the previous research finding that telephone-delivered CBT for MS symptom management can have robust and durable (maintained at 12 months post-treatment) positive effects on fatigue, pain, and depressed mood. 22 The CBT intervention was delivered flexibly and tailored to each participant but also maintained high fidelity to the treatment manual. Delivery mimicked routine clinical practice while simultaneously ensuring consistency with the planned treatment across study sites, therapists, and participants. The main difference between the delivery of CBT in this trial and how it might be administered in routine clinical practice was the monitoring for fidelity to the treatment manual; in routine clinical care, it is up the clinician to self-monitor adherence to a specific treatment regimen. This flexible application of a manualized CBT intervention should aid in dissemination and implementation of CBT for fatigue in MS. 117-119 As with other manualized CBT interventions, the one tested in this study provides a treatment framework based on individual participant goals and describes basic strategies (relaxation, working with thoughts, etc) that are core components of a standard CBT intervention. A short-term goal of this research team is to effectively disseminate the treatment manual, ideally with training on how to flexibly administer the core component. Even without access to the actual study treatment manual, however, a CBT-trained clinician could deliver the intervention in a way that is equivalent to the treatment in the study simply by administering the components described in Table 3 of this report. Lessons Learned Patient and community stakeholder participation in this trial was critical. Our ability to recruit and retain a heterogenous group of patients, especially during the onset of the COVID-19 pandemic, was in large part the result of active and consistent stakeholder engagement. Given the personalized nature and variability of fatigue, engagement of patient stakeholders who shared their experience with this symptom enriched the research beyond anticipated benefits. Furthermore, patient and HCP stakeholders allowed for the assessment of fatigue in a more patient-centered manner. Indeed, inclusion of EMA fatigue measures offered a more real-time assessment of fatigue intensity that was endorsed by stakeholders for its patient-centeredness. Our previous research has shown that fatigue is highly variable within-person, both day-to-day and within-day 110 ; this variability can make it difficult for participants and patients to report their fatigue on recall measures that ask the person to estimate typical or average levels of fatigue over a period of weeks or months. In addition to presenting a cognitive response burden, they are subject to recall bias or memory decay, 120 which are important given the cognitive problems associated with MS. With symptoms like fatigue that are highly variable, repeated measures over consecutive days provide the most reliable assay for clinical trials such as this. 121 Ecological momentary assessment fatigue measures coupled with activity counts captured from the PRO-Diary also allowed for calculation of subjective fatigability. Fatigability, a whole-person construct that reflects how fatigued a person becomes as a result of activity, has also emerged as a crucial outcome measure that, to date, has been underused in clinical research. 122-125 Assessment of fatigability was deemed critical to thoroughly evaluating the treatment effects in our study. If a treatment resulted in improved fatigue along with a decline in activity, examining fatigue severity in isolation could give a false impression of positive treatment effects. Similarly, if a treatment results in no changes to fatigue intensity but does result in increased activity, the net effect would be positive. Assessment of fatigability was also supported by study stakeholders, who appreciated that this construct reflects functioning in the context of fatigue. To our knowledge, this is the first study to include subjective fatigability as an outcome measure in interventional fatigue research, adding important information about how each intervention contributes to fatigue in the context of activity and daily life. This trial, which overlapped with the onset of the COVID-19 pandemic, also shed new light on the benefits and feasibility of conducting pragmatic research trials remotely. Although not all interventional studies are amenable to a fully virtual platform, conduct of this trial during the pandemic provided the impetus to identify creative and, in fact, more patient-centered solutions for the conduct of interventional studies in a manner that mimicked virtual health care to the greatest extent possible. Generalizability We believe that our findings offer some of the most generalizable data to date to demonstrate the effectiveness of modafinil and CBT for MS fatigue. The sample closely represented the communities served at both study sites. We believe this was due to several factors. In contrast to previous fatigue intervention trials, 26-28 , 33 , 40 , 126 our eligibility criteria did not include a disability level cutoff or upper age limit, and we minimized exclusion of specific medical and psychiatric conditions as much as possible. The virtual platform of the trial as well as stakeholder suggestions for inclusive recruitment also allowed for a wider catchment area for recruitment, thus allowing participants living in rural, suburban, and urban communities to enroll. Each of these measures ensured a more clinically representative sample. Subgroup Analyses and Heterogeneity of Treatment Effects Our data suggest that phenotypic differences should also be considered when selecting a treatment intervention in order to offer a more personalized approach to fatigue management. Disability level and depression intensity assessed in the current study could offer insight into if and when modafinil monotherapy might offer the most optimal benefit for fatigue. Interactions between disability level and treatment group on real-time EMA fatigue measures—which, compared with legacy instruments, offer a more patient-centered means to detect change in fatigue—suggest that modafinil may offer greater benefits than CBT monotherapy or combination therapy for patients with lower levels of disability, whereas all 3 treatments offer similar overall benefit for patients with greater disability. Although reasons for this interaction require further study, components of CBT, which include energy management strategies (ie, activity pacing), may have greater importance for patients with greater disability who need daily periods of lower-impact activity or rest for recovery. The stimulating effects of modafinil, while beneficial for those who may not be as reliant on pacing, could alter perceptions of stamina, inclining the user to push through sustained or rigorous activities and defer necessary rest periods. Similarly, trends toward an interaction between depression and treatment effect on EMA fatigue measures invite speculation that modafinil (relative to CBT monotherapy) might offer an important activating effect that may be more conducive to daily life in patients with greater depressed mood, which has been demonstrated in prior studies of people with depression without MS. In terms of other potential effect modifiers, our heterogeneity-of-treatment-effects analyses of sleep hygiene also deserve comment. Results in this study with interaction models and complete case analysis allow speculation that CBT (relative to modafinil) could offer greater reductions in MFIS score for patients who exhibited poorer sleep hygiene ( Figure 5 ). Although these effects diminished with imputation of missing data, the initial findings support a priori hypotheses and clinical observations that delivery of long-acting, wakefulness-promoting agents, such as modafinil, could further worsen sleep quality in patients with behavior-based sleep problems, thereby leading to diminished benefits from behavior-based treatment approaches that include sleep hygiene education. 127 Our group has shown that sleep disturbances, including insomnia, which can be exacerbated by modafinil use, have the potential to cause fatigue in addition to sleepiness in people with MS. 82 , 128 Decisions regarding modafinil use should be informed by an assessment for comorbid sleep disturbances. Despite relatively modest correlations between ESS score and fatigue measures used in this sample, sleepiness (in a direct manner) also predicted change in MFIS score at 12 weeks. Although further research is needed to understand this relationship, daytime consequences of sleep disturbances, including both sleepiness and fatigue, 118 , 119 could respond to these treatments. That said, as excessive daytime sleepiness often signals an underlying sleep disorder, HCPs should not consider the treatments studied as a substitute for appropriate referrals to a sleep specialist if an untreated sleep disorder is suspected. Our study design and primary aims did not allow for the measurement of OSA severity, effects of or compliance with positive airway pressure treatment, or composite measures of sleep in relation to treatment effect. Further analyses that explore the presence and severity of specific sleep disorders as well as the presence of sleep treatments could offer additional insight. Study Limitations We acknowledge several study limitations. Despite data trends to suggest differential effects of treatment depending on disability level, lower than anticipated recruitment of people with progressive MS subtypes could have diminished our power to see more robust interactions between disability and treatment effect. In addition, power was relatively low to detect the influence of depression or sleep hygiene as effect modifiers on treatment outcomes. Although our sample was reflective of people with MS served in our community, our trial also had a relatively low number of marginalized and underrepresented participants; our sample largely consisted of individuals who identified their race as White. Interventions were delivered through 2 large academic medical centers; therefore, findings may not be generalizable to interventions delivered outside of similar systems. Delivery of online assessments and transition to virtual visits could have excluded lower groups with lower socioeconomic status who may not have reliable computer access. Similar to our study sample, our stakeholder panel primarily consisted of White individuals. Future work should prioritize creation of stakeholder panels that maximize diversity, which could enhance recruitment of participants who are Black, Indigenous, and people of color and from underrepresented and marginalized groups. Because of the pragmatic nature of this trial, use of a placebo study group, which would deviate from clinical practice, was not included. We cannot rule out the possibility that placebo effect could have contributed to our findings. That said, the magnitude of treatment effect across treatment groups (MFIS score reduction of 15-17 points) exceeded what has been observed in placebo groups of prior modafinil randomized clinical trials and CBT trials that included standard care arms. 22 , 27 , 126 , 129 , 130 For example, in another recently published study that also used the MFIS as a primary outcome, 126 neither modafinil nor 2 other pharmacological therapies were superior to placebo for MS fatigue. That said, the magnitude of MFIS reduction with modafinil use in our trial exceeded estimates from the modafinil and placebo arms of this prior study. Although reasons for this difference are speculative, differences in dosing regimens and medication adherence rates between studies could have maximized modafinil's impact in our study beyond prior observed placebo effects. Furthermore, our effect modification analyses revealed differential treatment effects across treatment groups, which would support trait-specific benefits of modafinil that exceed an across-sample placebo effect. Finally, the effects of modafinil in our trial did not statistically differ from treatment effect sizes seen with CBT monotherapy, which has been demonstrated as an effective treatment for MS fatigue in multiple prior studies. We assessed multiple different fatigue constructs, including fatigue impact, fatigue intensity, and fatigability. Taken together, these measures are considered a strength of the study. However, our measure of fatigability reflected the ratio of perceived fatigue intensity relative to physical activity levels captured over the same time frame 122 , 131 , 132 and does not necessarily represent a person's level of performance fatigability or the level of decline in performance or energy output in the context of activity or work. 123 Future research should assess physical and mental fatigability with standardized testing paradigms to assay levels of performance fatigability. We decided to use a complete case analysis for secondary outcomes rather than imputed models, which could introduce bias in situations when the missing data mechanism is not missing completely at random. Future Research Our findings suggest several avenues for future research. Treatment effects on ratings of perceived global change (PGIC scores), which suggest an increased benefit of combination therapy, were different from effects on fatigue-specific outcomes. This raises the possibility that examining patient-centered outcomes, including social functioning, perceived cognitive functioning, leisure activities, and self-efficacy for managing symptoms, may reveal differential treatment effects on these important secondary outcomes. Further, we can apply the same heterogeneity-of-treatment-effects analyses to these secondary outcomes and the PGIC data to examine whether treatments lead to better outcomes for some patients based on depression status, sleep dysfunction, and/or disability level. These data are available for future examination. Ability to detect durability of treatment effects beyond the 12-week interval was not the main focus of this study; however, observational fatigue data, including mean difference in MFIS between 0 and 24 weeks, and assessment of clinical fatigue treatments that participants subsequently pursue in the clinical setting following the 12-week interventional period will be assessed between treatment groups and reported separately. Significant site effects suggest that treatment effects were stronger at one study site than the other ( Table 6 ). Future work will focus on uncovering potential differences in patient characteristics, study processes and procedures, and implementation of study treatments at the study sites. 133 , 134 Innovations in trial design allow for more personalized and adaptive delivery of treatments when combinations of treatments are being examined. Two such examples are the Multiphase Optimization Strategy and the Sequential Multiple Assignment Randomized Trial designs. 135 Both of these designs use randomized experimentation to produce more robust and valid inferences. Future research in larger samples of people with MS and fatigue should use these trial designs to provide more advanced information on how to match treatments to patients based on individual characteristics and serial responsiveness to treatments. 135-137 Conclusions Telephone-based CBT, modafinil, and combination therapy each significantly ameliorated fatigue impact, fatigue intensity, and perceived fatigability in people with MS, but combination therapy was more frequently associated with global benefits in self-reported functional outcomes and QOL, as reflected by the PGIC score (a secondary outcome in our study). Treatment effect may vary by clinical characteristics, including sleep hygiene and disability level. Phenotypic differences and goals of treatment (targeted symptom therapy vs overall functional outcomes) should be considered when selecting a treatment intervention in order to offer a more personalized approach to fatigue management. References 1. Stern M, Sorkin L, Milton K, Sperber K. Aging with multiple sclerosis. Phys Med Rehabil Clin N Am. 2010;21(2):403-417. doi:10.1016/j.pmr.2009.12.008 [ PubMed : 20494285 ] [ CrossRef ] 2. Wallin MT, Culpepper WJ, Campbell JD, et al. The prevalence of MS in the United States: a population-based estimate using health claims data. Neurology. 2019;92(10):e1029-e1040. doi:10.1212/WNL.0000000000007035 [ PMC free article : PMC6442006 ] [ PubMed : 30770430 ] [ CrossRef ] 3. Krupp L. Fatigue is intrinsic to multiple sclerosis (MS) and is the most commonly reported symptom of the disease. Mult Scler. 2006;12(4):367-368. doi:10.1191/135248506ms1373ed [ PubMed : 16900749 ] [ CrossRef ] 4. Krupp LB, Alvarez LA, LaRocca NG, Scheinberg LC. Fatigue in multiple sclerosis. Arch Neurol. 1988;45(4):435-437. doi:10.1001/archneur.1988.00520280085020 [ PubMed : 3355400 ] [ CrossRef ] 5. Minden SL, Frankel D, Hadden L, Perloffp J, Srinath KP, Hoaglin DC. The Sonya Slifka Longitudinal Multiple Sclerosis Study: methods and sample characteristics. Mult Scler. 2006;12(1):24-38. doi:10.1191/135248506ms1262oa [ PubMed : 16459717 ] [ CrossRef ] 6. Bakshi R. Fatigue associated with multiple sclerosis: diagnosis, impact and management. Mult Scler. 2003;9(3):219-227. doi:10.1191/1352458503ms904oa [ PubMed : 12814166 ] [ CrossRef ] 7. Fisk JD, Pontefract A, Ritvo PG, Archibald CJ, Murray TJ. The impact of fatigue on patients with multiple sclerosis. Can J Neurol Sci. 1994;21(1):9-14. [ PubMed : 8180914 ] 8. Finlayson M, Winkler-Impey M, Nicolle C, Edwards J. Self-care, productivity and leisure limitations of people with multiple sclerosis in Manitoba. Occup Ther. 1998;65(5):299-308. doi:10.1177/000841749806500508 [ CrossRef ] 9. Merkelbach S, Sittinger H, Koenig J. Is there a differential impact of fatigue and physical disability on quality of life in multiple sclerosis? J Nerv Ment Dis. 2002;190(6):388-393. doi:10.1097/00005053-200206000-00007 [ PubMed : 12080209 ] [ CrossRef ] 10. Yorkston KM, Bamer A, Johnson K, Amtmann D. Satisfaction with participation in multiple sclerosis and spinal cord injury. Disabil Rehabil. 2012;34(9):747-753. doi:10.3109/09638288.2011.619615 [ PubMed : 22004285 ] [ CrossRef ] 11. Beckerman H, Blikman LJ, Heine M, et al. The effectiveness of aerobic training, cognitive behavioural therapy, and energy conservation management in treating MS-related fatigue: the design of the TREFAMS-ACE programme. Trials. 2013;14:250.3751829. doi:10.1186/1745-6215-14-250 [ PMC free article : PMC3751829 ] [ PubMed : 23938046 ] [ CrossRef ] 12. Janardhan V, Bakshi R. Quality of life in patients with multiple sclerosis: the impact of fatigue and depression. J Neurol Sci. 2002;205(1):51-58. doi:10.1016/s0022-510x(02)00312-x [ PubMed : 12409184 ] [ CrossRef ] 13. Asano M, Berg E, Johnson K, Turpin M, Finlayson ML. A scoping review of rehabilitation interventions that reduce fatigue among adults with multiple sclerosis. Disabil Rehabil. 2015;37(9):729-738. doi:10.3109/09638288.2014.944996 [ PubMed : 25066069 ] [ CrossRef ] 14. Asano M, Finlayson ML. Meta-analysis of three different types of fatigue management interventions for people with multiple sclerosis: exercise, education, and medication. Mult Scler Int. 2014;2014:798285. doi:10.1155/2014/798285 [ PMC free article : PMC4052049 ] [ PubMed : 24963407 ] [ CrossRef ] 15. Pottgen J, Moss-Morris R, Wendebourg JM, et al. Randomised controlled trial of a self-guided online fatigue intervention in multiple sclerosis. J Neurol Neurosurg Psychiatry. 2018;89(9):970-976. doi:10.1136/jnnp-2017-317463 [ PubMed : 29549193 ] [ CrossRef ] 16. De Groot V, Beckerman H, Group TTS. Rehabilitation to treat MS-related fatigue: the TREFAMS research programme. Mult Scler J. 2016;22:40-41. 17. van den Akker LE, Beckerman H, Collette EH, et al. Cognitive behavioral therapy positively affects fatigue in patients with multiple sclerosis: results of a randomized controlled trial. Mult Scler. 2017;23(11):1542-1553. doi:10.1177/1352458517709361 [ PubMed : 28528567 ] [ CrossRef ] 18. van den Akker LE, Beckerman H, Collette EH, Eijssen IC, Dekker J, de Groot V. Effectiveness of cognitive behavioral therapy for the treatment of fatigue in patients with multiple sclerosis: a systematic review and meta-analysis. J Psychosom Res. 2016;90:33-42. doi:10.1016/j.jpsychores.2016.09.002 [ PubMed : 27772557 ] [ CrossRef ] 19. Thomas PW, Thomas S, Kersten P, et al. One year follow-up of a pragmatic multi-centre randomised controlled trial of a group-based fatigue management programme (FACETS) for people with multiple sclerosis. BMC Neurol. 2014;14:109. doi:10.1186/1471-2377-14-109 [ PMC free article : PMC4046846 ] [ PubMed : 24886398 ] [ CrossRef ] 20. Thomas S, Kersten P, Thomas PW, et al. Exploring strategies used following a group-based fatigue management programme for people with multiple sclerosis (FACETS) via the Fatigue Management Strategies Questionnaire (FMSQ). BMJ Open. 2015;5(5):e008274. doi:10.1136/bmjopen-2015-008274 [ PMC free article : PMC4620165 ] [ PubMed : 26486976 ] [ CrossRef ] 21. Thomas S, Thomas PW, Kersten P, et al. A pragmatic parallel arm multi-centre randomised controlled trial to assess the effectiveness and cost-effectiveness of a group-based fatigue management programme (FACETS) for people with multiple sclerosis. J Neurol Neurosurg Psychiatry. 2013;84(10):1092-1099. doi:10.1136/jnnp-2012-303816 [ PMC free article : PMC3786656 ] [ PubMed : 23695501 ] [ CrossRef ] 22. Ehde DM, Elzea JL, Verrall AM, Gibbons LE, Smith A, Amtmann D. Efficacy of a telephone-delivered self-management intervention for persons with multiple sclerosis: a randomized controlled trial with a one-year follow-up. Arch Phys Med Rehabil. 2015 Nov;96(11):1945-1958.e2. doi:10.1016/j.apmr.2015.07.015 [ PubMed : 26254948 ] [ CrossRef ] 23. Khan F, Amatya B, Galea M. Management of fatigue in persons with multiple sclerosis. Front Neurol. 2014;5:177. doi:10.3389/fneur.2014.00177 [ PMC free article : PMC4163985 ] [ PubMed : 25309504 ] [ CrossRef ] 24. Brioschi A, Gramigna S, Werth E, et al. Effect of modafinil on subjective fatigue in multiple sclerosis and stroke patients. Eur Neurol. 2009;62(4):243-249. doi:10.1159/000232927 [ PubMed : 19672078 ] [ CrossRef ] 25. Niepel G, Bibani RH, Vilisaar J, et al. Association of a deficit of arousal with fatigue in multiple sclerosis: effect of modafinil. Neuropharmacology. 2013;64:380-388. doi:10.1016/j.neuropharm.2012.06.036 [ PubMed : 22766394 ] [ CrossRef ] 26. Rammohan KW, Rosenberg JH, Lynn DJ, Blumenfeld AM, Pollak CP, Nagaraja HN. Efficacy and safety of modafinil (Provigil) for the treatment of fatigue in multiple sclerosis: a two centre phase 2 study. J Neurol Neurosurg Psychiatry. 2002;72(2):179-183. doi:10.1136/jnnp.72.2.179 [ PMC free article : PMC1737733 ] [ PubMed : 11796766 ] [ CrossRef ] 27. Stankoff B, Waubant E, Confavreux C, et al. Modafinil for fatigue in MS: a randomized placebo-controlled double-blind study. Neurology. 2005;64(7):1139-1143. doi:10.1212/01.WNL.0000158272.27070.6A [ PubMed : 15824337 ] [ CrossRef ] 28. Rammohan KW, Lynn DJ. Modafinil for fatigue in MS: a randomized placebo-controlled double-blind study. Neurology. 2005;65(12):1995-1997; author reply 1995-1997. [ PubMed : 16382509 ] 29. Kraft GH, Bowen J. Modafinil for fatigue in MS: a randomized placebo-controlled double-blind study. Neurology. 2005;65(12):1995-1997; author reply 1995-1997. [ PubMed : 16380634 ] 30. Sheng P, Hou L, Wang X, et al. Efficacy of modafinil on fatigue and excessive daytime sleepiness associated with neurological disorders: a systematic review and meta-analysis. PLoS One. 2013;8(12):e81802. doi:10.1371/journal.pone.0081802 [ PMC free article : PMC3849275 ] [ PubMed : 24312590 ] [ CrossRef ] 31. Patient Centered Outcomes Research Institute (PCORI). PCORI Funding Announcement: Treatment of Multiple Sclerosis. October 12, 2015. Accessed July 17, 2019. https://www ​.pcori.org ​/sites/default/files ​/PCORI-PFA-2015-Cycle-3-Multiple-Sclerosis.pdf 32. Selby JV, Forsythe L, Sox HC. Stakeholder-driven comparative effectiveness research: an update from PCORI. JAMA. 2015;314(21):2235-2236. doi:10.1001/jama.2015.15139 [ PubMed : 26595059 ] [ CrossRef ] 33. Lange R, Volkmer M, Heesen C, Liepert J. Modafinil effects in multiple sclerosis patients with fatigue. J Neurol. 2009;256(4):645-650. doi:10.1007/s00415-009-0152-7 [ PubMed : 19367356 ] [ CrossRef ] 34. Ledinek AH, Sajko MC, Rot U. Evaluating the effects of amantadin, modafinil and acetyl-L-carnitine on fatigue in multiple sclerosis--result of a pilot randomized, blind study. Clin Neurol Neurosurg. 2013;115 Suppl 1:S86-S89. doi:10.1016/j.clineuro.2013.09.029. [ PubMed : 24321164 ] [ CrossRef ] 35. Durelli L, Verdun E, Barbero P, et al. Every-other-day interferon beta-1b versus once-weekly interferon beta-1a for multiple sclerosis: results of a 2-year prospective randomised multicentre study (INCOMIN). Lancet. 2002;359(9316):1453-1460. doi:10.1016/s0140-6736(02)08430-1 [ PubMed : 11988242 ] [ CrossRef ] 36. Kappos L, Radue EW, O'Connor P, et al. A placebo-controlled trial of oral fingolimod in relapsing multiple sclerosis. N Engl J Med. 2010;362(5):387-401. doi:10.1056/NEJMoa0909494 [ PubMed : 20089952 ] [ CrossRef ] 37. Hauser SL, Waubant E, Arnold DL, et al. B-cell depletion with rituximab in relapsing-remitting multiple sclerosis. N Engl J Med. 2008;358(7):676-688. doi:10.1056/NEJMoa0706383 [ PubMed : 18272891 ] [ CrossRef ] 38. Confavreux C, O'Connor P, Comi G, et al. Oral teriflunomide for patients with relapsing multiple sclerosis (TOWER): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet Neurol. 2014;13(3):247-256. doi:10.1016/S1474-4422(13)70308-9 [ PubMed : 24461574 ] [ CrossRef ] 39. Hauser SL, Bar-Or A, Comi G, et al. Ocrelizumab versus interferon beta-1a in relapsing multiple sclerosis. N Engl J Med. 2017;376(3):221-234. doi:10.1056/NEJMoa1601277 [ PubMed : 28002679 ] [ CrossRef ] 40. Nagels G, D'hooghe MB, Vleugels L, Kos D, Despontin M, De Deyn PP. P300 and treatment effect of modafinil on fatigue in multiple sclerosis. J Clin Neurosci. 2007;14(1):33-40. doi:10.1016/j.jocn.2005.10.008 [ PubMed : 17138067 ] [ CrossRef ] 41. Zifko UA, Rupp M, Schwarz S, Zipko HT, Maida EM. Modafinil in treatment of fatigue in multiple sclerosis. Results of an open-label study. J Neurol. 2002;249(8):983-987. doi:10.1007/s00415-002-0765-6 [ PubMed : 12195441 ] [ CrossRef ] 42. Abolfazli R, Hosseini M, Ghanizadeh A, et al. Double-blind randomized parallel-group clinical trial of efficacy of the combination fluoxetine plus modafinil versus fluoxetine plus placebo in the treatment of major depression. Depress Anxiety. 2011;28(4):297-302. doi:10.1002/da.20801 [ PubMed : 21456039 ] [ CrossRef ] 43. Fava M, Thase ME, DeBattista C. A multicenter, placebo-controlled study of modafinil augmentation in partial responders to selective serotonin reuptake inhibitors with persistent fatigue and sleepiness. J Clin Psychiatry. 2005;66(1):85-93. doi:10.4088/jcp.v66n0112 [ PubMed : 15669893 ] [ CrossRef ] 44. Pocock SJ, Simon R. Sequential treatment assignment with balancing for prognostic factors in controlled clinical trial. Biometrics. 1975;31(1):103-115. [ PubMed : 1100130 ] 45. Rooney S, Wood L, Moffat F, Paul L. Prevalence of fatigue and its association with clinical features in progressive and non-progressive forms of Multiple Sclerosis. Mult Scler Relat Disord. 2019;28:276-282. doi:10.1016/j.msard.2019.01.011. [ PubMed : 30639830 ] [ CrossRef ] 46. Zhang Y, Taylor BV, Simpson S, Jr., Blizzard L, van der Mei I. Patient-reported outcomes are worse for progressive-onset multiple sclerosis than relapse-onset multiple sclerosis, particularly early in the disease process. Eur J Neurol. 2019;26(1):155-161. doi:10.1111/ene.13786 [ PubMed : 30133059 ] [ CrossRef ] 47. Ghajarzadeh M, Jalilian R, Eskandari G, Sahraian MA, Azimi A, Mohammadifar M. Fatigue in multiple sclerosis: relationship with disease duration, physical disability, disease pattern, age and sex. Acta Neurol Belg. 2013;113(4):411-414. doi:10.1007/s13760-013-0198-2 [ PubMed : 23616230 ] [ CrossRef ] 48. Mills RJ, Young CA. The relationship between fatigue and other clinical features of multiple sclerosis. Mult Scler. 2011;17(5):604-612. doi:10.1177/1352458510392262 [ PubMed : 21135018 ] [ CrossRef ] 49. Fisk JD, Ritvo PG, Ross L, Haase DA, Marrie TJ, Schlech WF. Measuring the functional impact of fatigue: initial validation of the fatigue impact scale. Clin Infect Dis. 1994;18 Suppl 1:S79-S83. doi:10.1093/clinids/18.supplement_1.s79 [ PubMed : 8148458 ] [ CrossRef ] 50. Hurst H, Bolton J. Assessing the clinical significance of change scores recorded on subjective outcome measures. J Manipulative Physiol Ther. 2004;27(1):26-35. doi:10.1016/j.jmpt.2003.11.003 [ PubMed : 14739871 ] [ CrossRef ] 51. Bowen J, Gibbons L, Gianas A, Kraft GH. Self-administered Expanded Disability Status Scale with functional system scores correlates well with a physician-administered test. Mult Scler. 2001;7(3):201-206. doi:10.1177/135245850100700311 [ PubMed : 11475445 ] [ CrossRef ] 52. Kroenke K, Strine TW, Spitzer RL, Williams JB, Berry JT, Mokdad AH. The PHQ-8 as a measure of current depression in the general population. J Affect Disord. 2009;114(1-3):163-173. doi:10.1016/j.jad.2008.06.026 [ PubMed : 18752852 ] [ CrossRef ] 53. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606-613. doi:10.1046/j.1525-1497.2001.016009606.x [ PMC free article : PMC1495268 ] [ PubMed : 11556941 ] [ CrossRef ] 54. Dhingra SS, Kroenke K, Zack MM, Strine TW, Balluz LS. PHQ-8 Days: a measurement option for DSM-5 Major Depressive Disorder (MDD) severity. Popul Health Metr. 2011;9:11. doi:10.1186/1478-7954-9-11 [ PMC free article : PMC3108293 ] [ PubMed : 21527015 ] [ CrossRef ] 55. Chung F, Yegneswaran B, Liao P, et al. STOP questionnaire: a tool to screen patients for obstructive sleep apnea. Anesthesiology. 2008;108:812-821. doi:10.1097/ALN.0b013e31816d83e4 [ PubMed : 18431116 ] [ CrossRef ] 56. Yang M, Morin CM, Schaefer K, Wallenstein GV. Interpreting score differences in the Insomnia Severity Index: using health-related outcomes to define the minimally important difference. Curr Med Res Opinion. 2009;25(10):2487-2494. doi:10.1185/03007990903167415 [ PubMed : 19689221 ] [ CrossRef ] 57. Thorndike FP, Ritterband LM, Saylor DK, Magee JC, Gonder-Frederick LA, Morin CM. Validation of the insomnia severity index as a web-based measure. Behav Sleep Med. 2011;9(4):216-223. doi:10.1080/15402002.2011.606766 [ PubMed : 22003975 ] [ CrossRef ] 58. Morin CM, Belleville G, Belanger L, Ivers H. The Insomnia Severity Index: psychometric indicators to detect insomnia cases and evaluate treatment response. Sleep. 2011;34(5):601-608. doi:10.1093/sleep/34.5.601 [ PMC free article : PMC3079939 ] [ PubMed : 21532953 ] [ CrossRef ] 59. Bastien CH, Vallieres A, Morin CM. Validation of the Insomnia Severity Index as an outcome measure for insomnia research. Sleep Med. 2001;2(4):297-307. doi:10.1016/s1389-9457(00)00065-4 [ PubMed : 11438246 ] [ CrossRef ] 60. Benes H, Kohnen R. Validation of an algorithm for the diagnosis of Restless Legs Syndrome: The Restless Legs Syndrome-Diagnostic Index (RLS-DI). Sleep Med. 2009;10(5):515-523. doi:10.1016/j.sleep.2008.06.006 [ PubMed : 18823819 ] [ CrossRef ] 61. Johns MW. Sleepiness in different situations measured by the Epworth Sleepiness Scale. Sleep. 1994;17(8):703-710. doi:10.1093/sleep/17.8.703 [ PubMed : 7701181 ] [ CrossRef ] 62. Johns MW. Daytime sleepiness, snoring, and obstructive sleep apnea. The Epworth Sleepiness Scale. Chest. 1993;103(1):30-36. doi:10.1378/chest.103.1.30 [ PubMed : 8417909 ] [ CrossRef ] 63. Johns MW. Reliability and factor analysis of the Epworth Sleepiness Scale. Sleep. 1992;15(4):376-381. doi:10.1093/sleep/15.4.376 [ PubMed : 1519015 ] [ CrossRef ] 64. Johns MW. A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep. 1991;14(6):540-545. doi:10.1093/sleep/14.6.540 [ PubMed : 1798888 ] [ CrossRef ] 65. Mastin DF, Bryson J, Corwyn R. Assessment of sleep hygiene using the Sleep Hygiene Index. J Behav Med. 2006;29(3):223-227. doi:10.1007/s10865-006-9047-6 [ PubMed : 16557353 ] [ CrossRef ] 66. Osborne TL, Raichle KA, Jensen MP, Ehde DM, Kraft G. The reliability and validity of pain interference measures in persons with multiple sclerosis. J Pain Symptom Manage. 2006;32(3):217-229. doi:10.1016/j.jpainsymman.2006.03.008 [ PubMed : 16939846 ] [ CrossRef ] 67. Osborne TL, Jensen MP, Ehde DM, Hanley MA, Kraft G. Psychosocial factors associated with pain intensity, pain-related interference, and psychological functioning in persons with multiple sclerosis and pain. Pain. 2007;127(1-2):52-62. doi:10.1016/j.pain.2006.07.017 [ PubMed : 16950570 ] [ CrossRef ] 68. Cleeland CS, Ryan KM. Pain assessment: global use of the Brief Pain Inventory. Ann Acad Med Singap. 1994;23(2):129-138. [ PubMed : 8080219 ] 69. Spitzer RL, Kroenke K, Williams JB, Lowe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med. 2006;166(10):1092-1097. doi:10.1001/archinte.166.10.1092 [ PubMed : 16717171 ] [ CrossRef ] 70. Shulman LM, Velozo C, Romero S, Gruber-Baldini AL. Comparative study of PROMIS self-efficacy for managing chronic conditions across chronic neurologic disorders. Qual Life Res. 2019;28(7):1893-1901. doi:10.1007/s11136-019-02164-2 [ PMC free article : PMC7071345 ] [ PubMed : 30915674 ] [ CrossRef ] 71. Gruber-Baldini AL, Velozo C, Romero S, Shulman LM. Validation of the PROMIS® measures of self-efficacy for managing chronic conditions. Qual Life Res. 2017;26(7):1915-1924. doi:10.1007/s11136-017-1527-3 [ PMC free article : PMC5479750 ] [ PubMed : 28239781 ] [ CrossRef ] 72. Lee MJ, Romero S, Velozo CA, Gruber-Baldini AL, Shulman LM. Multidimensionality of the PROMIS self-efficacy measure for managing chronic conditions. Qual Life Res. 2019;28(6):1595-1603. doi:10.1007/s11136-019-02116-w [ PMC free article : PMC6525059 ] [ PubMed : 30806873 ] [ CrossRef ] 73. Amtmann D, Bamer AM, Kim J, Chung H, Salem R. People with multiple sclerosis report significantly worse symptoms and health related quality of life than the US general population as measured by PROMIS and NeuroQoL outcome measures. Disabil Health J. 2018;11(1):99-107. doi:10.1016/j.dhjo.2017.04.008 [ PubMed : 28442320 ] [ CrossRef ] 74. Becker H, Stuifbergen A, Lee H, Kullberg V. Reliability and validity of PROMIS Cognitive Abilities and Cognitive Concerns scales among people with multiple sclerosis. Int J MS Care. 2014;16(1):1-8. doi:10.7224/1537-2073.2012-047 [ PMC free article : PMC3967698 ] [ PubMed : 24688349 ] [ CrossRef ] 75. Cella D, Lai JS, Nowinski CJ, et al. Neuro-QOL: brief measures of health-related quality of life for clinical research in neurology. Neurology. 2012;78(23):1860-1867. doi:10.1212/WNL.0b013e318258f744 [ PMC free article : PMC3369516 ] [ PubMed : 22573626 ] [ CrossRef ] 76. Gershon RC, Lai JS, Bode R, et al. Neuro-QOL: quality of life item banks for adults with neurological disorders: item development and calibrations based upon clinical and general population testing. Qual Life Res. 2012;21(3):475-486. doi:10.1007/s11136-011-9958-8 [ PMC free article : PMC3889669 ] [ PubMed : 21874314 ] [ CrossRef ] 77. Godin G, Jobin J, Bouillon J. Assessment of Leisure-time exercise behavior by self-report: a concurrent validity study. Can J Public Health. 1986;77(5):359-362. [ PubMed : 3791117 ] 78. Kos D, Duportail M, D'Hooghe M, Nagels G, Kerckhofs E. Multidisciplinary fatigue management programme in multiple sclerosis: a randomized clinical trial. Mult Scler. 2007;13(8):996-1003. doi:10.1177/1352458507078392 [ PubMed : 17623738 ] [ CrossRef ] 79. Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology. 1983;33(11):1444-1452. doi:10.1212/wnl.33.11.1444 [ PubMed : 6685237 ] [ CrossRef ] 80. Chung F, Subramanyam R, Liao P, Sasaki E, Shapiro C, Sun Y. High STOP-Bang score indicates a high probability of obstructive sleep apnoea. Br J Anaesth. 2012;108(5):768-775. doi:10.1093/bja/aes022 [ PMC free article : PMC3325050 ] [ PubMed : 22401881 ] [ CrossRef ] 81. Singh M, Gavidia R, Dunietz GL, et al. Validation of an obstructive sleep apnea symptom inventory in persons with multiple sclerosis. Mult Scler. 2022;28(2):280-288. doi:10.1177/13524585211013014 [ PMC free article : PMC8627523 ] [ PubMed : 34048308 ] [ CrossRef ] 82. Braley TJ, Segal BM, Chervin RD. Obstructive sleep apnea and fatigue in patients with multiple sclerosis. J Clin Sleep Med. 2014;10(2):155-162. doi:10.5664/jcsm.3442 [ PMC free article : PMC3899317 ] [ PubMed : 24532998 ] [ CrossRef ] 83. Pichard LE, Simonelli G, Schwartz L, Balkin TJ, Hursh S. Precision medicine for sleep loss and fatigue management. Sleep Med Clin. 2019;14(3):399-406. doi:10.1016/j.jsmc.2019.05.006 [ PubMed : 31375208 ] [ CrossRef ] 84. Watson NF, Badr MS, Belenky G, et al. Recommended amount of sleep for a healthy adult: a joint consensus statement of the American Academy of Sleep Medicine and Sleep Research Society. Sleep. 2015;38(6):843-844. doi:10.5665/sleep.4716 [ PMC free article : PMC4434546 ] [ PubMed : 26039963 ] [ CrossRef ] 85. Von Korff M, Ormel J, Keefe FJ, Dworkin SF. Grading the severity of chronic pain. Pain. 1992;50(2):133-149. doi:10.1016/0304-3959(92)90154-4 [ PubMed : 1408309 ] [ CrossRef ] 86. Kos D, Kerckhofs E, Carrea I, Verza R, Ramos M, Jansa J. Evaluation of the Modified Fatigue Impact Scale in four different European countries. Mult Scler. 2005;11(1):76-80. doi:10.1191/1352458505ms1117oa [ PubMed : 15732270 ] [ CrossRef ] 87. Kos D, Kerckhofs E, Nagels G, et al. Assessing fatigue in multiple sclerosis: Dutch modified fatigue impact scale. Acta Neurol Belg. 2003;103(4):185-191. [ PubMed : 15008502 ] 88. Larson RD. Psychometric properties of the modified fatigue impact scale. Int J MS Care. 2013;15(1):15-20. doi:10.7224/1537-2073.2012-019 [ PMC free article : PMC3883028 ] [ PubMed : 24453758 ] [ CrossRef ] 89. Rietberg MB, Van Wegen EE, Kwakkel G. Measuring fatigue in patients with multiple sclerosis: reproducibility, responsiveness and concurrent validity of three Dutch self-report questionnaires. Disabil Rehabil. 2010;32(22):1870-1876. doi:10.3109/09638281003734458 [ PubMed : 20345240 ] [ CrossRef ] 90. Amtmann D, Bamer AM, Noonan V, Lang N, Kim J, Cook KF. Comparison of the psychometric properties of two fatigue scales in multiple sclerosis. Rehabil Psychol. 2012;57(2):159-166. doi:10.1037/a0027890 [ PMC free article : PMC3422656 ] [ PubMed : 22686554 ] [ CrossRef ] 91. Tellez N, Rio J, Tintore M, Nos C, Galan I, Montalban X. Does the Modified Fatigue Impact Scale offer a more comprehensive assessment of fatigue in MS? Mult Scler. 2005;11(2):198-202. doi:10.1191/1352458505ms1148oa [ PubMed : 15794395 ] [ CrossRef ] 92. Khanna D, Pope JE, Khanna PP, et al. The minimally important difference for the fatigue visual analog scale in patients with rheumatoid arthritis followed in an academic clinical practice. J Rheumatol. 2008;35(12):2339-2343. doi:10.3899/jrheum.080375 [ PMC free article : PMC3155760 ] [ PubMed : 19004044 ] [ CrossRef ] 93. Wright A. REDCap: A tool for the electronic capture of research data. J Electron Resour Med Libr. 2016;13:197-201. doi:10.1080/15424065.2016.1259026 [ CrossRef ] 94. Harvey LA. REDCap: web-based software for all types of data storage and collection. Spinal Cord. 2018;56(7):625. doi:10.1038/s41393-018-0169-9 [ PubMed : 29977003 ] [ CrossRef ] 95. Azur MJ, Stuart EA, Frangakis C, Leaf PJ. Multiple imputation by chained equations: what is it and how does it work? Int J Methods Psychiatr Res. 2011;20(1):40-49. doi:10.1002/mpr.329. [ PMC free article : PMC3074241 ] [ PubMed : 21499542 ] [ CrossRef ] 96. van Buuren S, Groothuis-Oudshoorn K. mice: Multivariate imputation by chained equations in R. Journal of Statistical Software. 2011;45(3):1-67. doi:10.18637/jss.v045.i03 [ CrossRef ] 97. Little R. Missing-data adjustments in large surveys. J Bus Econ Stat. 1988;6(3):287-296. doi:10.2307/1391878 [ CrossRef ] 98. Rubin DB. Multiple Imputation for Nonresponse in Surveys. John Wiley & Sons, Inc; 2004. 99. O'Connell NS, Dai L, Jiang Y, et al. Methods for analysis of pre-post data in clinical research: a comparison of five common methods. J Biom Biostat. 2017;8(1):1-8. doi:10.4172/2155-6180.1000334 [ PMC free article : PMC6290914 ] [ PubMed : 30555734 ] [ CrossRef ] 100. Cox DR. A remark on multiple comparison methods. Technometrics. 1965;7(2):223-224. 101. Rothman KJ. No adjustments are needed for multiple comparisons. Epidemiology. 1990;1(1):43-46. [ PubMed : 2081237 ] 102. Comi G, Leocani L. Assessment, pathophysiology and treatment of fatigue in multiple sclerosis. Expert review of neurotherapeutics. 2002;2(6):867-876. doi:10.1586/14737175.2.6.867 [ PubMed : 19810920 ] [ CrossRef ] 103. Comi G, Leocani L, Rossi P, Colombo B. Physiopathology and treatment of fatigue in multiple sclerosis. J Neurol. 2001;248(3):174-179. doi:10.1007/s004150170222 [ PubMed : 11355149 ] [ CrossRef ] 104. Krupp LB. Fatigue in multiple sclerosis: definition, pathophysiology and treatment. CNS Drugs. 2003;17(4):225-234. doi:10.2165/00023210-200317040-00002 [ PubMed : 12665396 ] [ CrossRef ] 105. Parikh SV, Segal ZV, Grigoriadis S, et al. Canadian Network for Mood and Anxiety Treatments (CANMAT) Clinical guidelines for the management of major depressive disorder in adults. II. Psychotherapy alone or in combination with antidepressant medication. J Affect Disorders. 2009;117:S15-S25. doi:10.1016/j.jad.2009.06.042 [ PubMed : 19682749 ] [ CrossRef ] 106. Segal Z, Vincent P, Levitt A. Efficacy of combined, sequential and crossover psychotherapy and pharmacotherapy in improving outcomes in depression. J Psychiatry Neurosci. 2002;27(4):281-290. [ PMC free article : PMC161662 ] [ PubMed : 12174737 ] 107. Ang DC, Jensen MP, Steiner JL, Hilligoss J, Gracely RH, Saha C. Combining cognitive-behavioral therapy and milnacipran for fibromyalgia: a feasibility randomized-controlled trial. Clinical J Pain. 2013;29(9):747-754. doi:10.1097/AJP.0b013e31827a784e [ PMC free article : PMC4185390 ] [ PubMed : 23446065 ] [ CrossRef ] 108. Vitiello MV, Rybarczyk B, Von Korff M, Stepanski EJ. Cognitive behavioral therapy for insomnia improves sleep and decreases pain in older adults with co-morbid insomnia and osteoarthritis. J Clin Sleep Med. 2009;5(4):355-362. [ PMC free article : PMC2725255 ] [ PubMed : 19968014 ] 109. Kratz AL, Murphy SL, Braley TJ. Pain, fatigue, and cognitive symptoms are temporally associated within-but not across-days in multiple sclerosis. Arch Phys Med Rehabil. 2017;98(11):2151-2159. doi:10.1016/j.apmr.2017.07.003 [ PMC free article : PMC5660935 ] [ PubMed : 28729169 ] [ CrossRef ] 110. Kratz AL, Murphy SL, Braley TJ. Ecological momentary assessment of pain, fatigue, depressive, and cognitive symptoms reveals significant daily variability in multiple sclerosis. Arch Phys Med Rehabil. 2017;98(11):2142-2150. doi:10.1016/j.apmr.2017.07.002 [ PMC free article : PMC5660933 ] [ PubMed : 28729168 ] [ CrossRef ] 111. Motl RW, Weikert M, Suh Y, Dlugonski D. Symptom cluster and physical activity in relapsing-remitting multiple sclerosis. Res Nurs Health. 2010;33(5):398-412. doi:10.1002/nur.20396 [ PubMed : 20725947 ] [ CrossRef ] 112. Motl RW, Suh Y, Weikert M. Symptom cluster and quality of life in multiple sclerosis. J Pain Symptom Manage. 2010;39(6):1025-1032. doi:10.1016/j.jpainsymman.2009.11.312 [ PubMed : 20434872 ] [ CrossRef ] 113. Turner AP, Chapko MK, Yanez D, et al. Access to multiple sclerosis specialty care. PM R. 2013;5(12):1044-1050. doi:10.1016/j.pmrj.2013.07.009 [ PubMed : 23920332 ] [ CrossRef ] 114. Buchanan RJ, Stuifbergen A, Chakravorty BJ, Wang S, Zhu L, Kim M. Urban/rural differences in access and barriers to health care for people with multiple sclerosis. J Health Hum Serv Adm. 2006;29(3):360-375. [ PubMed : 17571473 ] 115. Elrod CS, DeJong G. Determinants of utilization of physical rehabilitation services for persons with chronic and disabling conditions: an exploratory study. Arch Phys Med Rehabil. 2008;89(1):114-120. doi:10.1016/j.apmr.2007.08.122 [ PubMed : 18164340 ] [ CrossRef ] 116. National Multiple Sclerosis Society. Health Policy Fact Sheet: Living with Multiple Sclerosis in Rural America. National Multiple Sclerosis Society; 2012. 117. Wilson GT. Manual-based treatments: the clinical application of research findings. Behav Res Ther. 1996;34(4):295-314. doi:10.1016/0005-7967(95)00084-4 [ PubMed : 8871362 ] [ CrossRef ] 118. Kendall PC, Frank HE. Implementing evidence-based treatment protocols: flexibility within fidelity. Clin Psychol (New York). 2018;25(4):e12271. doi:10.1111/cpsp.12271 [ PMC free article : PMC6329472 ] [ PubMed : 30643355 ] [ CrossRef ] 119. Kendall PC, Gosch E, Furr JM, Sood E. Flexibility within fidelity. J Am Acad Child Adolesc Psychiatry. 2008;47(9):987-993. doi:10.1097/CHI.0b013e31817eed2f [ PubMed : 18714195 ] [ CrossRef ] 120. Shiffman S, Stone AA, Hufford MR. Ecological momentary assessment. Ann Rev Clin Psychol. 2008;4:1-32. doi:10.1146/annurev.clinpsy.3.022806.091415 [ PubMed : 18509902 ] [ CrossRef ] 121. Heapy A, Dziura J, Buta E, Goulet J, Kulas JF, Kerns RD. Using multiple daily pain ratings to improve reliability and assay sensitivity: how many is enough? J Pain. 2014;15(12):1360-1365. doi:10.1016/j.jpain.2014.09.012 [ PubMed : 25283469 ] [ CrossRef ] 122. Eldadah BA. Fatigue and fatigability in older adults. PM R. 2010;2(5):406-413. doi:10.1016/j.pmrj.2010.03.022 [ PubMed : 20656622 ] [ CrossRef ] 123. Kluger BM, Krupp LB, Enoka RM. Fatigue and fatigability in neurologic illnesses: proposal for a unified taxonomy. Neurology. 2013;80(4):409-416. doi:10.1212/WNL.0b013e31827f07be [ PMC free article : PMC3589241 ] [ PubMed : 23339207 ] [ CrossRef ] 124. Stokes MJ, Cooper RG, Edwards RH. Normal muscle strength and fatigability in patients with effort syndromes. BMJ. 1988;297(6655):1014-1017. doi:10.1136/bmj.297.6655.1014 [ PMC free article : PMC1834796 ] [ PubMed : 3142592 ] [ CrossRef ] 125. Enoka RM. Mechanisms of muscle fatigue: central factors and task dependency. J Electromyogr Kinesiol. 1995;5(3):141-149. doi:10.1016/1050-6411(95)00010-w [ PubMed : 20719645 ] [ CrossRef ] 126. Nourbakhsh B, Revirajan N, Morris B, et al. Safety and efficacy of amantadine, modafinil, and methylphenidate for fatigue in multiple sclerosis: a randomised, placebo-controlled, crossover, double-blind trial. Lancet Neurol. 2021;20(1):38-48. doi:10.1016/S1474-4422(20)30354-9 [ PMC free article : PMC7772747 ] [ PubMed : 33242419 ] [ CrossRef ] 127. Kumar R. Approved and investigational uses of modafinil : an evidence-based review. Drugs. 2008;68(13):1803-1839. doi:10.2165/00003495-200868130-00003 [ PubMed : 18729534 ] [ CrossRef ] 128. Braley TJ, Chervin RD, Segal BM. Fatigue, tiredness, lack of energy, and sleepiness in multiple sclerosis patients referred for clinical polysomnography. Mult Scler Int. 2012;2012:673936. doi:10.1155/2012/673936 [ PMC free article : PMC3539354 ] [ PubMed : 23316361 ] [ CrossRef ] 129. Ford-Johnson L, DeLuca J, Zhang J, Elovic E, Lengenfelder J, Chiaravalloti ND. Cognitive effects of modafinil in patients with multiple sclerosis: a clinical trial. Rehabil Psychol. 2016;61(1):82-91. doi:10.1037/a0039919 [ PubMed : 26654280 ] [ CrossRef ] 130. Möller F, Poettgen J, Broemel F, Neuhaus A, Daumer M, Heesen C. HAGIL (Hamburg Vigil Study): a randomized placebo-controlled double-blind study with modafinil for treatment of fatigue in patients with multiple sclerosis. Mult Scler. 2011;17(8):1002-1009. doi:10.1177/1352458511402410 [ PubMed : 21561959 ] [ CrossRef ] 131. Schnelle JF, Buchowski MS, Ikizler TA, Durkin DW, Beuscher L, Simmons SF. Evaluation of two fatigability severity measures in elderly adults. J Am Geriatr Soc. 2012;60(8):1527-1533. doi:10.1111/j.1532-5415.2012.04062.x [ PMC free article : PMC3419324 ] [ PubMed : 22860899 ] [ CrossRef ] 132. Kratz AL, Murphy SL, Braley TJ, et al. Development of a person-centered conceptual model of perceived fatigability. Qual Life Res. 2019;28(5):1337-1347. doi:10.1007/s11136-018-2093-z [ PMC free article : PMC7395299 ] [ PubMed : 30604341 ] [ CrossRef ] 133. Lane-Fall MB, Curran GM, Beidas RS. Scoping implementation science for the beginner: locating yourself on the “subway line” of translational research. BMC Med Res Methodol. 2019;19(1):133. doi:10.1186/s12874-019-0783-z. [ PMC free article : PMC6599376 ] [ PubMed : 31253099 ] [ CrossRef ] 134. Taylor SP, Kowalkowski MA. Using Implementation Science-Guided Pilot Studies to Assess and Improve the Informativeness of Clinical Trials. J Gen Intern Med. 2021;36(2):533-536. doi:10.1007/s11606-020-06220-3 [ PMC free article : PMC7878642 ] [ PubMed : 32918201 ] [ CrossRef ] 135. Collins LM, Murphy SA, Strecher V. The multiphase optimization strategy (MOST) and the sequential multiple assignment randomized trial (SMART): new methods for more potent eHealth interventions. Am J Prev Med. 2007;32(5 Suppl):S112-118. doi:10.1016/j.amepre.2007.01.022 [ PMC free article : PMC2062525 ] [ PubMed : 17466815 ] [ CrossRef ] 136. Almirall D, Lizotte DJ, Murphy SA. SMART Design issues and the consideration of opposing outcomes: discussion of “Evaluation of Viable Dynamic Treatment Regimes in a Sequentially Randomized Trial of Advanced Prostate Cancer” by Wang, Rotnitzky, Lin, Millikan, and Thall. J Am Stat Assoc. 2012;107(498):509-512. doi:10.1080/01621459.2012.665615 [ PMC free article : PMC3607391 ] [ PubMed : 23543940 ] [ CrossRef ] 137. Lei H, Nahum-Shani I, Lynch K, Oslin D, Murphy SA. A “SMART” design for building individualized treatment sequences. Ann Rev Clin Psychol. 2012;8:21-48. doi:10.1146/annurev-clinpsy-032511-143152 [ PMC free article : PMC3887122 ] [ PubMed : 22224838 ] [ CrossRef ] Related Publications •. Kratz AL, Alschuler KN, Ehde DM, et al. A randomized pragmatic trial of telephone-delivered cognitive behavioral-therapy, modafinil, and combination therapy of both for fatigue in multiple sclerosis: The design of the “COMBO-MS” trial. Contemp Clin Trials. 2019;84:105821. doi:10.1016/j.cct.2019.105821 [ PubMed : 31400515 ] [ CrossRef ] Acknowledgments The study team recognizes and is grateful for the support of the National MS Society, the committed panel of study stakeholders, PCORI, and all of the study participants who made this work possible. Research reported in this report was funded through a Patient-Centered Outcomes Research Institute® (PCORI®) Award (MS-1610-36980). Further information available at: https://www.pcori.org/research-results/2017/comparing-effectiveness-three-treatments-reducing-fatigue-among-patients-multiple-sclerosis-combo-ms-trial#project_study_registation_information Appendices Appendix A. COVID-19 Survey (PDF, 135K) Appendix B. Treatment Effect on MFIS Subscale Scores, by Treatment Arm (PDF, 396K) Appendix C. Frequency of Treatment Responders, by Treatment Arm (PDF, 62K) Appendix D. Patient Global Impression of Change Scores, by Treatment Arm (PDF, 81K) Original Project Title: A randomized controlled trial of telephone-delivered cognitive behavioral-therapy, modafinil, and combination therapy of both interventions for fatigue in multiple sclerosis PCORI ID: MS-1610-36980 ClinicalTrials.gov ID: NCT03621761 Suggested citation: Braley TJ, Kratz AL. (2024). Comparing the Effectiveness of Three Treatments for Reducing Fatigue among Patients with Multiple Sclerosis—The COMBO-MS Trial . Patient-Centered Outcomes Research Institute (PCORI). https://doi.org/10.25302/10.2023.MS.161036980 Disclaimer The [views, statements, opinions] presented in this report are solely the responsibility of the author(s) and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute® (PCORI®), its Board of Governors or Methodology Committee. Copyright © 2024. University of Michigan. All Rights Reserved. This book is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License which permits noncommercial use and distribution provided the original author(s) and source are credited. (See https://creativecommons.org/licenses/by-nc-nd/4.0/ Bookshelf ID: NBK619813 PMID: 41428831 DOI: 10.25302/10.2023.MS.161036980 Share Views PubReader Print View Cite this Page Braley TJ, Kratz AL. Comparing the Effectiveness of Three Treatments for Reducing Fatigue among Patients with Multiple Sclerosis—The COMBO-MS Trial [Internet]. Washington (DC): Patient-Centered Outcomes Research Institute (PCORI); 2023 Oct. doi: 10.25302/10.2023.MS.161036980 PDF version of this title (2.1M) In this Page Background Participation of Patients and Other Stakeholders Methods Results Discussion Conclusions References Related Publications Acknowledgments Appendices Other titles in this collection PCORI Final Research Reports Related information NLM Catalog Related NLM Catalog Entries PMC PubMed Central citations PubMed Links to PubMed Recent Activity Clear Turn Off Turn On Comparing the Effectiveness of Three Treatments for Reducing Fatigue among Patie... Comparing the Effectiveness of Three Treatments for Reducing Fatigue among Patients with Multiple Sclerosis—The COMBO-MS Trial Your browsing activity is empty. Activity recording is turned off. Turn recording back on See more... Follow NCBI Twitter Facebook LinkedIn GitHub NCBI Insights Blog Connect with NLM Twitter Facebook Youtube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov