E = mc 2: Education (E), medication (m), and conditional cash (c 2) to improve uptake of antiseizure medications in a low‐resource population: Protocol for randomized trial

Abstract Objective Most people with epilepsy (PWE) could live seizure‐free if treated with one or more antiseizure medications (ASMs). The World Health Organization (WHO) estimates that 75% of PWE in low‐resource settings lack adequate antiseizure treatment. Limited education surrounding epilepsy and the out‐of‐pocket costs of ASMs in particular pose barriers to managing epilepsy in resource‐poor, low‐income settings. The aim of this study is to implement and test a novel strategy to improve outcomes across the epilepsy care cascade marked by (1) retention in epilepsy care, (2) adherence to ASMs, and (3) seizure reduction, with the measured goal of seizure freedom. Methods A randomized, double‐blinded clinical trial will be performed, centered at the Ignace Deen Hospital in Conakry, Republic of Guinea, in Western Sub‐Saharan Africa. Two hundred people with clinically diagnosed epilepsy, ages 18 years and above, will receive education on epilepsy and then be randomized to (i) free ASMs versus (ii) conditional cash, conditioned upon return to the epilepsy clinic. Participants will be followed for 360 days with study visits every 90 days following enrollment. Significance We design a randomized trial for PWE in Guinea, a low‐resource setting with a high proportion of untreated PWE and a nearly completely privatized healthcare system. The trial includes a conditional cash transfer intervention, which has yet to be tested as a targeted means to improve outcomes for people with a chronic neurological disorder. The trial aims to provide an evidence base for the treatment of epilepsy in such settings. Plain Language Summary We present a clinical trial protocol for a randomized, blinded study of 200 people with epilepsy in the low‐resource African Republic of Guinea, providing an educational intervention (E), and then randomizing in a 1:1 allocation to either free antiseizure medication (m) or conditional cash (c 2) for 360 days. Measured outcomes include (1) returning to outpatient epilepsy care, (2) adherence to antiseizure medications (ASMs), and (3) reducing the number of seizures. This study is an initial look at giving small amounts of cash for desired results (or “nudges”) for improving epilepsy outcomes in the sub‐Saharan African and brain disorder contexts.


| INTRODUCTION
4][5][6] Most recently, the World Health Assembly adopted the Intersectoral Global Action Plan on Epilepsy and Other Neurological Disorders ("Draft Intersectoral global action plan on epilepsy and other neurological disorders 2022-2031," 7 ).This includes targets to treat more people with neurological disorders with essential medicines including ASMs for PWE.
Cascades of care are being developed to assess the delivery of care for noncommunicable diseases, similar to the approach seen in HIV/AIDS. 8In a cascade, specific targets are identified including: (1) ensuring >90% of PWE are aware of the diagnosis as a brain disorder, (2) retaining >90% of PWE in epilepsy care, (3) ensuring >80% of PWE take quality, appropriate chosen and dosed ASMs, and (4) ensuring seizure freedom in 70% of people treated. 9Of these, seizure freedom (step four) is the outcome most valued by PWE and is associated with a self-reported quality of life equivalent to people without epilepsy.Given the high efficacy of ASMs, access to ASMs would lead to seizure freedom in a majority of PWE.
Access to ASMs remains a challenge to seizure prevention and epilepsy treatment in low-resource settings.Epilepsy care in the Republic of Guinea, a low-income country in Western Sub-Saharan Africa with a 2020 gross national income of 1020 USD per capita, 10 reflects a lowfunctioning health system with few epilepsy-trained clinicians, and high volumes of patients.Guinea has few health facilities and a nearly completely privatized healthcare system, necessitating costs to be paid "out of pocket."PWE often go without ASMs due to multiple factors at the level of the patient, provider, and health system.
2][13][14][15] Several negative consequences of untreated and undertreated epilepsy have been observed including unintentional death; missed opportunities for school attendance of Guinean children with epilepsy 12 ; poor cognitive performance; frequent unintentional injuries; epilepsy stigma 16 ; high rates of traditional healing injuries including cuttings, instrumentation, and animal sacrifices 11 ; and large losses of personal wealth on ineffective treatment options. 17Similar to other low-resource locations, a high number of PWE do not take ASMs and have suffered serious consequences as a result, including death from unintentional injury or status epilepticus. 13sh (c 2 ) for 360 days.Measured outcomes include (1) returning to outpatient epilepsy care, (2) adherence to antiseizure medications (ASMs), and (3) reducing the number of seizures.This study is an initial look at giving small amounts of cash for desired results (or "nudges") for improving epilepsy outcomes in the sub-Saharan African and brain disorder contexts.

K E Y W O R D S
behavioral economics, clinical trial protocol, conditional cash, epilepsy, global health, sub-Saharan Africa

Key points
• Untreated epilepsy remains an important public health problem in resource-limited settings such as the Republic of Guinea.
• There have yet to be clinical trials that evaluate the role of conditional cash in treating epilepsy.
• We provide a clinical trial protocol to test conditional cash versus free antiseizure medications (ASMs) in Guinea.
• Measured outcomes will include whether participants return to outpatient epilepsy care, adhere to ASMs, and have fewer seizures.
• This trial could provide evidence for the role of conditional cash for untreated people with epilepsy in resource-limited settings.
As a response to this tragic situation, the study team proposes to compare two ways to deliver epilepsy care to people with epilepsy (PWE) in a low-income setting.Although several lines of evidence suggest that both these interventions will lead to improved outcomes, [18][19][20] they have not been directly tested in low-resource populations with low literacy.Conditional cash transfer (CCT) interventions in particular have not been systematically tested in PWE or in people with any chronic neurological disorder.CCT interventions grant individuals a sum of money on the condition that they fulfill a predetermined task.In this way, CCT interventions create a "nudge," or an economic incentive, for taking a desired course of action.These interventions differ notably from unconditional cash transfer (UCT) interventions, which simply provide direct payment to individuals without any behavioral nudges.As such, CCT interventions uniquely serve as an important lever for both alleviating the financial burden of poverty and incentivizing people to act for the benefit of their own health.
A global scoping review of randomized trials with healthfocused CCT interventions found most CCT interventions benefit health outcomes and have the potential to change patient behaviors for a range of disease areas. 21However, CCT interventions are yet to be tested for neurological conditions.3][24][25] CCT interventions could therefore enable epilepsy patients to break out of a cycle of poverty that can accompany high seizure burdens and resultantly, an inability to attend school or work.

| Trial design
A randomized, double-blinded clinical trial will be performed, including multiple healthcare facility sites in rural and urban Guinea but headquartered at the Ignace Deen Hospital in the capital city, Conakry.Two hundred PWE, clinically diagnosed by a Guinean and a U.S. neurologist, ages 18 years and above, will be enrolled and randomized 1:1 to (i) education and medication, or (ii) education and CCT (Figure 1).All participants will receive epilepsy education.Participants will be interviewed and examined by the Guinean-U.S. study team of neurologists at each visit.Follow-up visits every 90 days until day 360 (approximately 1 calendar year) will assess the intervention on participants' clinical and socioeconomic outcomes.The recurring 90-day follow-up visits are considered beneficial to participants who may not access epilepsy care in any other way while also providing a sufficiently long timeline to study the sustainability of this intervention.

| Objectives
The primary objective is to implement and test a novel strategy to improve outcomes across the epilepsy care cascade marked by (1) retention in epilepsy care, (2) adherence to ASMs, and (3) seizure reduction with the measured goal of seizure freedom.
The first aim is to compare free ASMs vs. conditional cash using the epilepsy cascade of care.We hypothesize that the provision of free medication versus conditional cash will lead to (1) similar retention in epilepsy care-90% of PWE-at 360 days (Hypothesis 1a), (2) 80% of PWE regularly taking ASMs at 360 days (Hypothesis 1b), and (3) 65% of PWE achieving seizure freedom 360 days postenrollment (Hypothesis 1c).
The secondary aim is to detail facilitators and barriers to free medication versus conditional cash to improve seizure freedom among PWE.We hypothesize that barriers and facilitators to the uptake of each intervention strategy can be learned through careful, structured documentation by healthcare workers during study visits using an investigator-designed written intake form as well as freetext responses, serving as foundational information for a future larger-scale implementation study in low-resource populations.

INTERVENTIONS, AND OUTCOMES
Institutional Review Board approval will be obtained through the Ignace Deen Hospital in Guinea and the U.S. study team's institution.

| Study setting
The Ignace Deen Hospital is one of two teaching hospitals in Conakry, receiving several thousand cases of epilepsy per year.Additional recruitment and enrollment are planned to occur in Kankan, Forecariah, and Dubreka regional health facilities.There are five practicing neurologists in Guinea.Table 1 displays summary statistics from a prior cohort study we conducted with 297 Guinean PWE whose characteristics are expected to resemble future trial participants.

| Eligibility criteria
PWE will be screened in person and enrolled on a "first come, first served" basis.The goal is to enroll a person with active epilepsy.The person will have therefore experienced two or more recurrent, unprovoked seizures in the past year (inclusion criterion) and not have had any other diagnosis that can better explain the symptoms, as evaluated by the study neurologists (one Guinean and one U.S.).Specifically, patients with exclusively febrile seizures, pseudo-seizures, and other movement disorders or episodic events that are not primary epilepsy are not eligible.The study will include a brain MRI such that secondary seizures from a space-occupying lesion (secondary to tumor, stroke) can be detected and referred to usual care but not enrolled.Similarly, seizures that are provoked (e.g., alcohol withdrawal) will be queried through the intake forms and corroborated by physical examination and extant medical records.
Inclusion criteria are (1) resident of Guinea, 18 years or older, presenting to one of the study sites in Guinea; (2) able to provide informed consent or have a next of kin proxy provide informed consent when appropriate; (3) experienced two or more recurrent, unprovoked seizures in the past year; (4) able to provide a clinical history or have a next of kin provide it; (5) not taking a daily ASM for epilepsy for the past 30 days or more for any reason; (6) willing to have his/her data sent anonymously to the U.S. study site; and (7) anticipated to be able to return for follow-up visits for 360 days.
Exclusion criteria are (1) any other diagnosis that can better explain the symptoms, making the diagnosis of primary, active epilepsy unlikely; (2) unable to undergo the study procedures; (3) age less than 18 years old due to the need to measure several self-reported outcomes; (4) unable to provide informed consent (or have an appropriate proxy informed consent); (5) currently pregnant or lactating; (6) secondary seizures, for example, due to a condition that requires imminent treatment (e.g., malignancy, alcohol withdrawal); (7) critically ill; (8) exclusively pseudoseizures/nonepileptic behavioral events; (9) unwilling to take an ASMs for epilepsy; (10) already treated with ASMs for other purposes (e.g., mood, pain); (11) known to be unable to return for follow-up visits.
The investigators will monitor participants' health throughout the study and re-evaluate eligibility.For example, a woman may become pregnant or a participant may have a new illness (e.g., malaria).It will be ensured that all pregnant women have access to clinical care and an ASM (such as levetiracetam or lamotrigine) that is associated with the lowest fetal risk (whether she remains in the study or is withdrawn).Changes in health status and medical risk will be recorded, and the study team will adjudicate the event for administrative withdrawal of the participant.All withdrawn participants will have access to usual care following study discontinuation.

| Informed consent
Screening will be performed by Guinean investigators with mentorship and collaboration with the U.S. team (Appendix S1) and will include a brain MRI to rule out other diagnoses (e.g., tumors, untreated significant hydrocephalus).An estimated 50% of the enrollment will occur in Conakry and 50% at rural and peri-urban sites.

| Explanation for the choice of comparators
Participants randomized to the CCT arm will still be given an ASM recommendation and advice on purchasing ASMs consistent with the best available neurological care.The decision to create two arms will enable the study team to assess the effectiveness of each intervention without using a noninterventional control group.Using a control group comparator without any intervention was considered unethical.All participants will have access to local diagnostic testing as needed.

| Intervention description
The education intervention will be given to each participant upon study enrollment in the form of a prerecorded, standardized video, made by the investigators, lasting approximately 30 min.The video is designed for people with low literacy and will answer the following basic questions: [What is a seizure?What is epilepsy?Why does epilepsy occur?Why is epilepsy a brain disorder?What works to treat epilepsy?What does not work to treat epilepsy?Why are ASMs important?]The video will be available in local languages (Sousou and Malinké), English, and French and shown on electronic tablets.A question-and-answer period, led by the study neurologists, will follow with no limitations on the amount of time.A flyer providing basic, written information on the education intervention will be given.
ASMs will be provided free of charge and prescribed for the participants in the study arm that includes the medication intervention.The ASM choice will be determined at the discretion of the neurologist in the context of the patient's history.There is no one best ASM for all patients, and most ASMs provide similar efficacy but dissimilar side effects and safety.There will be a total of seven possible ASMs to be used.
CCT will be conditioned on the return for follow-up visits to the outpatient epilepsy clinic for scheduled appointments (±7 days every 90 days).Each visit for those in the CCT groups will entail the provision of 15 USD (given in the equivalent of Guinean Francs) to participants.This amount  Note: Gray shades denote the title of the columns.
is anticipated to be a "nudge" (i.e., mildly incentivizing) as it exceeds the general cost of transportation for a return visit in the area (~5 USD on average) and ASM (which approximately costs ~4-10 USD for a monthly dosage in Guinea). 11t is important to note that this nudge will ensure participants in the CCT arm can afford the ASM, which enables study staff to assess outcomes from behavioral change rather than affordability considerations.The transfer will also be conditional on presence at visits, not on seizure outcomes.This ensures that participants are not penalized for aspects of their illness that are beyond their control.

| Outcomes
The primary outcome of the study is adherence to neurology clinic appointments.This is defined as the number of appointments attended as scheduled and on time by the participant, as reported by the investigator.Secondary outcomes are as follows: 1. Percent of PWE who achieve seizure freedom.2. Seizure frequency reduction: The percentage decrease in number of monthly seizures from baseline to the study endpoint of 360 days.
A complete list of outcomes can be found in Table 2.

| Participant timeline
The timeline of screening, enrolment, and interventions is provided in Table 3.The interventions will be provided at follow-up visits taking place every 90 days.The baseline visit will include a neurologist's description of seizure history, including seizure semiology, epilepsy diagnosis, and etiology of epilepsy to the degree possible given resource limitations.Structured forms will be used.Several measures will be collected during follow-up visits, including detailed seizure information, quality of life surveying, medication possession, stigma scale surveying using the Stigma Scale of Epilepsy (SSE), 26 health visit history, school/work attendance, and injury surveying.

| Recruitment
Participants will be recruited via ( Participants will be blinded to the primary study hypothesis.The Guinean site collaborators providing the intervention will be blinded to the primary study hypothesis. 5| DATA COLLECTION AND MANAGEMENT 5.1 | Plans for assessment, collection of outcomes, and data management Data will be collected from the baseline visit and followup visits (Table 3).An enrollment survey is provided in Appendix S2.All investigators will receive research ethics and data quality control training to ensure adherence to proper documentation and data collection.Data quality control measures will be implemented by the U.S. and Guinean staff on a weekly basis to ensure accuracy.Missing data will be addressed throughout the study.Use of REDCap™ with forced responses to progress to the next page will be used for surveys to avoid missingness.Other data points that are missing in a systematic way will be re-evaluated for any clinical or cultural reasons.No data imputation will occur.

| Plans to promote participant retention and complete follow-up
The study investigators will provide a careful explanation of the importance of follow-up visits before a prospective participant gives informed consent.Participants will be asked to follow up in person and reminded of upcoming visits approximately a week in advance.Participants will also be given access to a study phone number and allowed to attend their visit ±7 days of the scheduled date to accommodate personal, work, and/or public transit schedules and further promote retention.We a randomized two-arm design with an equal sample size in the arm with the provision of conditional cash and the arm with the provision of free medication.
Our sample size calculation is based on Hypothesis 1a.
To investigate Hypothesis 1a, we will first test the null hypothesis that the retention rate of the group with the provision of conditional cash is less than or equal to 90% using a non-inferiority one-sample proportion test.We will then test the null hypothesis that the provision of conditional cash will lead to a retention rate no less than the retention rate of free medication (controlling for medication type) using a noninferiority two-sample proportion test.Since there are two hypotheses, in the sample calculation, we applied the Bonferroni correction to ensure a familywise error rate of 0.05. 27Given the existing studies and preliminary results, we expect that the retention rate of the group with the provision of conditional cash will have a retention rate higher than 90%.This reflects financial incentive clinical trials in other diseases where the intervention improves retention in care, particularly studied in HIV/AIDS. 28,29ssuming it is 90%, to ensure a statistical power of 80%, with a noninferiority margin of 8.5%, the sample size of the arm with the conditional cash required is 98. 30 We will then test the null hypothesis that the provision of conditional cash will lead to a retention rate no less than the retention rate of free ASMs using a noninferior two-sample proportion test.Assuming the conditional cash group has 90% of PWE who will return at 12 months, with a noninferiority margin of 12%, to ensure statistical power of 80%, the sample size of the arm with the conditional cash required is 99.
For Hypothesis 1b, we will use a one-sample proportion test to test the null hypothesis that the percentage of PWE regularly taking ASMs at 12 months in the conditional cash arm is less than or equal to 85%, and we will use a two-sample proportion test for the null hypothesis that the percentage of PWE regularly taking ASMs in the conditional cash arm is less than or equal to the percentage of PWE regularly taking ASMs in the free medication arm.For Hypothesis 1c, we will use a one-sample proportion test to test the null hypothesis that the percentage of PWE with seizure freedom at 12 months in the conditional cash arm is less than or equal to 65%, and we will use a two-sample proportion test for the null hypothesis that the percentage of seizure freedom in the conditional cash arm is less than or equal to the percentage of seizure freedom in the free medication arm.
To leverage the longitudinal observations collected during follow-up visits at 90, 180, 270, and 360 days, we will use a generalized linear mixed effects model which allows us to study the association between multiple baseline variables (e.g., including the intervention options, education level, medication status, ASM type, distance to the study clinics) while accounting for the correlation between longitudinal observations.The study will convene a Data Safety and Monitoring Board (DSMB) quarterly throughout the trial period.The DSMB will be composed of Guinean, U.S., and other global experts, including neurologists and biostatisticians, to review adverse events, trial progress, and data on the outcomes of interest.Adverse effects of the ASMs will be recorded.The DSMB may vote to unblind and withdraw a participant in the event of a serious adverse effect.Participants will be advised of any abnormal head imaging results performed by the study and referred for appropriate clinical care.

| Dissemination plans
Transparency on processes and protocols and dissemination of trial data will be enabled through sharing of all survey instruments, data collection materials, data analysis plans, and de-identified data.These will be made available in academic peer-reviewed journals and institutional research websites with final data available free online.

| DISCUSSION
Our study provides a concrete plan to disaggregate issues contributing to the epilepsy treatment gap and target simple interventions addressing three major presumed reasons for the treatment gap: low epilepsy awareness, limited access to ASMs, and absolute poverty.Our approach provides a new pathway and model for clinicians and researchers to consider this problem.This study comes at a time when the international community is searching for effective mechanisms of alleviating the burden of epilepsy globally.The Intersectoral Global Action Plan on Epilepsy and Other Neurological Disorders 2022-2031 emphasizes the need for concerted, interdisciplinary action to enhance epilepsy care and treatment in the near term ("Draft Intersectoral global action plan on epilepsy and other neurological disorders 2022-2031," 7 ).CCT interventions, if proven to be effective for epilepsy care, could serve as a powerful tool for these global efforts.
There are several strengths to this proposed study.The study design is both ethical and appropriate for low-resource populations and, if successful, could be programmatically scaled to other locations.We also pose the first test of CCT, a method that has been used in other conditions, such as HIV/AIDS 31 and TB, 32 and in vaccination campaigns, to address low uptake.The use of CCT is controversial and untested in PWE but could serve as a poverty-reduction intervention in addition to improving health outcomes among PWE.Though commonly tested for other health-focused outcomes and in other regions, there is no published research on CCT programs in neurological disorders to our knowledge and only a limited sample of tested CCT interventions in West Africa. 33,34here are also several limiting factors to the study design.Though participants are blinded to the hypothesis of the study, there is a possibility of participants disclosing information on their allocation to other participants, which would reduce allocation blindness.There is also a risk of differential attrition between the study arms.Several study outcomes are participant-reported, which require clear and proactive communication with participants.
In general, large medication trials for neurological disorders on the African continent remain surprisingly rare.Testing interventions that can improve and save lives is overdue in the field of neurology and, even more so, in resource-limited settings.This randomized trial, enrolling an anticipated 200 PWE in West Africa, will establish an evidence base for a leading problem in neurological care in low-resource settings.The trial is designed for appropriateness in Guinean PWE with health equity and local standards of care considered.Inherent to the study is capacity building for epilepsy treatment and management among patients, families, and clinicians who will be actively involved in the E = mc 2 protocol.The motivation of the trial design is to improve the overall situation of epilepsy care in Guinea, that is, to do no harm and leave no patient behind.As such, all trial participants are anticipated to benefit from their allocated study intervention, even if the study hypotheses are not supported.

F I G U R E 1
Study enrollment and intervention randomization flow diagram.

T A B L E 1
Epilepsy patients, Republic of Guinea (n = 297).

1 |
Data monitoring committee

INTERVENTIONS: BLINDING 4.1 | Who will be blinded?
1) clinicians' referrals, (2) referrals through in-person meetings with traditional healers, and (3) if needed, advertisements through Radio Television Guinea and local Conakry-based radio and television stations that reach most of the Guinean population.Traditional healers will be leveraged for recruitment purposes but will not serve as investigators or otherwise monitor study interventions.Most participants will be recruited by Guinean neurologists, as in preliminary studies.T A B L E 2 Primary, secondary, and exploratory outcomes of interest.
Schedule of tests and events at each study visit.
T A B L E 3