The Economic and societal burden of multiple sclerosis on lebanese society: a cost-of-illness and quality of life study protocol

ABSTRACT

This protocol describes the estimation of the societal costs and quality-of-life (QOL) burden of multiple sclerosis (MS) in Lebanon. This cross-sectional, prevalence-based burden-of-illness study was carried out in a premier MS center in Lebanon. We enrolled Lebanese patients aged 18 years and older who had been diagnosed with MS more than 6 months. The study uses a bottom-up approach to estimate the cost-of-illness (COI) and QOL using a retrospective face-to-face interview questionnaire. This resource utilization questionnaire was adapted to the Lebanese context by clinical and health economics experts. The methodologies used to estimate the consumption of healthcare resources, informal care, and productivity losses are well-defined and aligned with the Lebanese healthcare system. Costs are presented overall and by MS severity levels. QOL is measured using the EuroQOL (EQ-5D-5 L) and Multiple Sclerosis International Quality of Life (MusiQoL) instrument. This protocol pioneers in informing the design of future COI and QOL studies in low – and middle-income countries (LMICs), as the methods used could be applied in similar LMICs. Furthermore, we provide recommendations and discuss the challenges of conducting a high-quality burden-of-illness study in LMICs and the steps taken to meet them, using the case of Lebanon.

KEYWORDS:

• Cost-of-illness
• economic burden
• Lebanon
• multiple sclerosis
• study protocol
• utilities

1. Introduction

Multiple sclerosis (MS) is a demyelinating and neurodegenerative disease of the central nervous system [1]. MS is characterized by impairment in any functional neurological system, due to loss of the myelin sheath, which results in decreased axonal transmission and eventually in axonal disruption [2,3]. This causes an array of symptoms, including visual disruption, fatigue, problems with balance and coordination, cognitive and emotional disturbances, altered sensation, abnormal speech, bladder and bowel problems, and sexual dysfunction [2,4,5]. These symptoms have significant impact on the quality of life (QOL) of people with MS (PwMS) and their families and friends, interfering with their productivity [6], and presenting a substantial cost. Worldwide, MS affects 2.8 million people and is twice as common in women as in men [7].

To measure the burden of a disease, cost-of-illness (COI) studies are commonly used to present analyses of the economic burden of a particular health condition in a group of patients over a defined period [8,9]. Furthermore, health-related quality of life (HRQOL) and utilities in the form of quality-adjusted life years (QALYs) are increasingly being used as a health outcome in order to assess the consequences of disease for patients’ mental health, physical and social functioning, and well-being [10]. These outcomes can also be used to evaluate population-based intervention programs and as input for economic evaluations [10]. Several burden of MS studies from different countries have reported a substantial cost per patient and a decrease in HRQOL for PwMS [11–16].

A recent systematic review over the economic burden of MS [17] shed light on the absence of guidelines for conducting and reporting on COI studies of MS in low – and middle-income countries (LMICs). This review also highlighted the significant methodological variations between included studies, and their suboptimal quality, in addition to the ambiguous estimation of costs, and suggested the need for well-designed and clearly reported studies [17].

In 2008 the number of PwMS in Lebanon was estimated to be between 1,200 and 1,700, with a female to male ratio of 1.8:1.0 [18]. Although the QOL of PwMS had been assessed in Lebanon [19,20], information on the costs of MS was not studied previously. To fill the gap in research on the costs of MS and HRQOL of PwMS specifically in Lebanon and in LMICs generally, a burden-of illness study on MS was carried out in collaboration with a premier Lebanese MS center. Presenting detailed information about the different steps involved in designing a burden-of-illness study on MS may raise awareness among policymakers with regard to the burden of MS, and foster discussion on the quality and methodology of these studies in LMICs. In the absence of local guidelines for conducting and reporting on COI and QOL studies in LMICs, this protocol aims to describe the estimation of a COI and QOL utilities study for MS and to discuss challenges in conducting a burden-of-illness study and steps taken to overcome the burdens in these countries, using the case of Lebanon. This article describes the study approach and methods used. The data collection was completed in August 2021, and data cleaning and analysis have started. The results of the COI and QOL of MS will be presented in two papers.

2. Patients and methods

2.1. Study design

This cross-sectional, prevalence-based burden-of-illness study uses a bottom-up approach, aggregating data from patients through a validated questionnaire. The study entails an assessment of health resources consumed by PwMS from a societal perspective, and of their HRQOL. The study is carried out in collaboration with the Nehme and Therese Tohme Multiple Sclerosis Center at the American University of Beirut Medical Center, a leading hospital in Lebanon.

2.2. Participants and recruitment

All Lebanese patients 18 years of age and older, who had been diagnosed with Relapsing-Remitting MS or Primary Progressive MS or Secondary Progressive MS more than six months, were invited to take part in the study. This premier MS center in Lebanon provides treatment for MS patients from all the Lebanese governorates. The Center is also a leader in MS training, clinical management, and research. Patients treated at this center are diagnosed according to the 2017 McDonald criteria [21].

A purposive sampling approach was used to recruit MS patients, with the aim of enrolling enough subjects at each stage of disease progression (defined by EDSS score). In the absence of a standard method for sample size calculation in COI studies, it was estimated that around 200 respondents would be enough, based on previous similar studies [15,22–24]. Ethical approval (SBS-2019-0268) was obtained from the Institutional Review Board associated with the MS center, and all participants provided informed consent.

2.3. Data collection

All data were collected retrospectively from patients through face-to-face interviews during clinical visits from December 2020 to August 2021. Data collection was conducted by a trained interviewer using a structured questionnaire to guide the discussion. Confidentiality was guaranteed by anonymizing recruitment; each participant was assigned an ID number.

2.4. Data collection tools

In the absence of a tool for measuring COI studies in Lebanon and LMICs, we deployed a copyrighted MS Health Resource Use Questionnaire that has been developed and improved over the past two decades and used in various similar studies in high-income countries [11,25–27]. The questionnaire was obtained from the original developers [11,25–27] and adapted to the Lebanese healthcare system by the study authors (health economics and clinical neurology experts practicing in Lebanon). Authorization to use, adapt and translate the questionnaire into Arabic was granted from the copyright owner. The questionnaire was translated into Arabic following international guidelines [28]. Both forward and backward translations were interpreted by two independent translators. Then, both versions were compared, and necessary amendments were made on the Arabic version. The questionnaire was then pilot-tested on a sample of adult Lebanese MS patients (n = 10) who were not included in the current study, and feedback was incorporated into the final version of the questionnaire. An example of the Arabic questionnaire is available in the electronic supplementary material.

The questionnaire requests information on demographic characteristics, and disease data including prevalent symptoms, information on relapses, severity of disability, workforce participation, healthcare and service consumption, and informal care. In order to ensure anonymity, collection of personal information was limited.

We investigated the health state utility values of patients with MS using the five dimensions and five levels (EQ-5D-5 L) of the EuroQOL as a generic outcome measure, and used the Multiple Sclerosis International Quality of Life (MusiQoL) instrument to address particular health-related effects of MS not included in the generic measure (EQ-5D-5 L).

2.5. Assessment for severity of disease

Following the evolution of the disease, the course of MS is described as relapsing-remitting MS, secondary progressive MS, or primary progressive MS [1]. Relapsing-remitting MS, which comprises 80–85% of initial diagnoses of MS, is typified by new or recurrent neurological symptoms (relapses) and periods of stability (remissions) [29]. Progressive MS with or without relapses may be secondary progressive MS, when it evolves after relapsing-remitting disease, or primary progressive MS when it manifests without the initial relapsing phase [30]. MS progression differs from person to person, and disability is routinely measured with the Expanded Disability Status Scale (EDSS) to assess the degree of impairment in neurological functions. This is a well-recognized method applied in clinical practice, clinical trials, and epidemiological studies [31]. EDSS is used in economic studies as it has a clear association with HRQOL assessed as both utility and as costs [26]. Subjects’ self-assessed disability is reported using a descriptive scale originating in the initial EDSS [31] and the patient-assessed Patient Determined Disease Steps (PDDS) [32]. EDSS scores range from 0, which is normal neurological functioning, to 10, which is death due to MS, and increase in accordance with the disability level. In this study, the economic burden for PwMS was collected by level of disability (EDSS) to assess patients’ disability level and describe the changes of costs and utilities with the progression of the disease. EDSS scores of 0–3 indicate mild disability, 4–6.5 indicate moderate, and 7–9 indicate severe disability, as described by other authors [11,15].

2.6. Cost estimation

The societal perspective was adopted; this incorporates all costs irrespective of who incurs them [9], including out-of-pocket expenses and patient co-payments. We follow a macro-costing approach, whereby cost estimation contains three main steps: identification, measurement, and valuation [33].

Step 1: Identification of Costs

The COI study measures healthcare consumption considered as direct medical costs (e.g. hospitalization, consultations, costs of medication, medical tests), as well as direct non-medical costs (e.g. home and automobile modifications, professional home care, informal care provided by family and friends, patients’ travel expenses to reach healthcare facilities, and home – and community-based services), and indirect costs related to reduced productivity due to MS. Table 1 shows the cost items to be considered while calculating the COI of MS from a societal perspective. These cost items were included in most COI studies [11,34–36] and systematic reviews [15,17,37].

Table 1. Cost items considered when calculating the costs of MS from a societal perspective

Direct medical costs	Direct non-medical costs	Indirect costs
Inpatient care (overnight hospital stay)	Investments and equipment (home or car modifications, walking aids)	Productivity loss
Outpatient care (less than 24 hours)	Formal care (professional care, nurse, maid, etc.)	Short-term absence
Healthcare consultations	Informal care (home help from family or friends)	Long-term absence
Neurologist	Transportation
Internist
Urologist
Ophthalmologist
Psychiatrist
Psychologist or Counselor
General practitioner
Dietitian
Speech therapist
Physical therapist
Drugs/medication
Disease-modifying treatments
Relapse treatments
Other prescription and non-prescription drugs
Tests and investigations
MRI (brain or spine)
CT Scan (computed tomography)
Spinal tap (lumbar puncture, cerebrospinal fluid (CSF) analysis)
Evoked potential (e.g. visual)
Ultrasound Heart (echocardiography, ultrasound cardiogram, UCG)
Ultrasound Bladder
Electrocardiogram (ECG)
Blood tests

Step 2: Measurement of Costs

To minimize recall bias, data on the type and quantity of resource used were collected retrospectively following timeframes associated with each type of resource. The time periods for the healthcare resource consumption questions are 3 months, to ensure the best possible recall. Only the question related to investment in devices, equipment, and aids is collected over 12 months.

Loss of production is estimated based on patients’ information regarding their employment situation. The questionnaire included questions on employment, self-employment, unemployment, part-time working hours, and short-term and long-term absence due to MS.

For unemployed patients, the reason they were not working was requested to determine if this was due to MS or other causes. Information about early retirement is also collected from those not working due to MS.

Step 3: Valuation of Costs

Data on healthcare consumption were collected from patients via a face-to-face structured interview. However, price lists were obtained from national databases kept by the Lebanese Ministry of Public Health, the Lebanese National Social Security Fund, military and civil defense sources, private insurance, market prices, the Lebanese National Drugs Database, the Lebanese Ministry of Labor, and other official sources open to the public.

The quantity of use of each service is multiplied by its respective unit cost to obtain the cost for each health resource. To estimate the COI, it was vital to understand the specifications of the Lebanese healthcare system. As shown above, Lebanese healthcare coverage is based on several types of insurance and third-party payers. Each type of insurance has a different percentage of coverage and tariff for the same health resource. In the absence of standardization of health resource costs, unit costs are not publicly available for all healthcare resources. Accordingly, we employed a variety of methods to obtain cost data from different sources, including patient recall, national databases, and key informant interviews. Patients are asked about their insurance and third-party payer such as the National Social Security Fund, the Ministry of Public Health, the military and civil defense, private insurance, etc. Then, key informant interviews are conducted with these third-party payers to collect information on the percentages of coverage and the tariffs of health resources.

For PwMS help from family and friends is considered to be informal care. In the absence of a universal method for valuing the hours of informal care [38], the proxy good method and opportunity cost are the most commonly used approaches [39]. In this study, the valuation of informal care is based on the proxy good method [40]. The cost per hour of informal care is calculated based on the Lebanese minimum wage for the year of costing.

Productivity losses are estimated using the human capital approach [41], where individual productivity is valued at the market price [42]. The average national gross wage in Lebanon is considered to be the cost of employment. Costs are calculated only for patients of employment age. To obtain the unit cost per hour, the average salary is divided by the number of monthly working hours for the private sector (n = 208 hours) [43].

The costs of drugs are derived from the latest version of the Lebanese National Drug Index, obtained from the Ministry of Public Health in Lebanon.

All costs are collected and calculated in Lebanese pounds (LBP), and adjusted to US$ for the year of costing values, using the World Bank purchasing power parity (PPP) [44]. When calculating costs, all data are annualized with the assumption that healthcare resource consumption for PwMS is equal in any given quarter.

2.7. Assessment of health state utility values

The EQ-5D-5 L is a standardized and validated HRQOL instrument [45,46] which is comprised of two parts: the EQ-5D-5 L descriptive system and a Visual Analog Scale (EQ-VAS) score [47]. The EQ-5D-5 L describes health status based on five questions related to mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each dimension has five levels of problems: none, slight, moderate, severe, and extreme. This instrument yields a single generic measure of HRQOL known as the ‘health utility score,’ which is used in clinical and economic evaluations [48], where 0 represents death and 1 is perfect health. The EQ-5D-5 L is available in more than 130 languages, and several countries have created their own national index value set (www.euroqol.org). We use the EQ-5D-5 L validated among Arabic-speaking MS Patients [49,50]. The original value set of the EQ-5D-5 L was established in the United Kingdom (UK) and has been widely deployed [51]. In the absence of a Lebanese index value, we calculate the health utility score by using the UK EQ-5D-5 L value set. The second part of the EuroQOL is the EQ-VAS, which is calibrated from 0 (the worst health) to 100 (the best health). However, the VAS provides additional information about the subjects’ QOL without affecting the EQ-5D-5 L index.

The MusiQoL instrument is a multidimensional HRQOL instrument created by an international steering committee comprised of neurologists, patients, and health economists. It reflects the point of view of PwMS on the disease’s impact on their daily life [52]. The questionnaire is available in 14 languages, as a disease-specific QOL scale that can be applied internationally [53]. The MusiQoL scale has 31 items divided into 9 dimensions: activities of daily living (eight), psychological well-being (four), symptoms (three), relationships with friends (four), family relationships (three), satisfaction with healthcare system (three), sentimental and sexual life (two), coping (two), and rejection (two). This instrument provides a global index score, which is calculated as the mean of the individual dimension scores. The total MusiQoL score ranges between 0 and 100, with lower scores indicating a worse QOL. The MusiQoL questionnaire was validated among Arabic-speaking MS patients in Lebanon and the Middle East and North Africa (MENA) [54].

Two additional HRQOL questions on the level of cognitive difficulties and the impact of MS on work productivity were added to the questionnaire [38]. Both questions were collected with a VAS ranging from 0 (no problem) to 10 (severe problems), to avoid adding length to the questionnaire.

2.8. Analysis

Missing data on the consumption of health resources (volume) are replaced by the mean value reported by patients. For the base-case analysis, the total annualized cost for all patients is divided by the number of included patients to estimate the mean annual cost. The same approach is used to calculate the mean annual cost for mild, moderate, and severe EDSS levels. Direct medical, direct non-medical, and indirect costs are also presented, and key drivers identified. Similarly, the mean QOL estimates are presented overall and by EDSS levels. Therefore, a sub-analysis is conducted to study the evolution of MS costs associated with the progression of the disease. The mean annual cost per patient is thus compared among the three EDSS categories, mild (0–3), moderate (4–6.5), and severe (7–9).

Several sensitivity analyses are conducted to test the robustness of our cost results, in which we utilize the minimum and maximum costs of some healthcare resources reported by third-party payers and patients. Analyses are performed using IBM SPSS Statistics version 21. Depending on the normality of the data, patient characteristics are described as relative frequencies for categorical data, and as mean (standard deviation) for continuous data.

3. Discussion

Although the burden of MS in high-income countries has been assessed extensively, information on the economic burden in LMICs remains largely unstudied [55,56]. While a substantial methodological variation is evident between studies of the economic burden of MS in LMICs [17], this article is to our knowledge the first study protocol to offer detailed information about the different steps involved in designing, conducting, and reporting on a burden-of-illness study in a LMIC. While this study protocol resonates in Lebanon, the approach and methods used could be applied in similar LMICs if adapted to the country’s own healthcare system and specifications. Finally, it may inform the design of other future COI and QOL studies in LMICs, as it could be tailored to the investigated illness and context.

This study faces two sets of challenges. The first set is common among LMICs, while the second comes from the country-specific context and related obstacles. The following is a discussion of common challenges and the steps taken to overcome them. The main strength of this protocol is that we used several validated methods. Then, we explore the Lebanese-specific context and discuss pertaining local challenges.

First, in the absence of a local health economic guideline for conducting and reporting on COI studies, this study protocol clearly describes adopted methodologies, as well as the reporting of costs by severity levels. A recent systematic review 57 underlined the scarcity of country-specific economic evaluation guidelines in LMICs, and the need to improve the methodological framework of existing ones. In this study, we adopted the societal perspective by incorporating all cost burdens imposed on society irrespective of who incurs them. This societal perspective is preferred by economists [42,58,59], in particular because it minimizes the potential biases and underestimation of the total cost burden which may be present in narrower views [60]. The COVID-19 pandemic highlighted the value of adopting a societal perspective by capturing the broader impacts on sectors outside health in economic studies [61]. Costs included in this study are measured retrospectively using a bottom-up approach. COI studies can be ‘top-down’ or ‘bottom-up.’ Both approaches have limitations, however, the choice relies on the study question. The bottom-up approach estimates costs based on information from individuals who have the disease; this approach may include questions on informal care, transportation, and productivity losses not often found in registries [62]. The bottom-up approach is considered suitable for chronic diseases [63]. The authors fully recognize the limitations of retrospective data collection, due to potential recall bias. Thus, the questionnaire uses a recall period of only 3 months for most questions, following which costs are annualized. This choice is aligned with existing research recommendations for estimating costs due to illness [64]. However, annualizing costs, especially in an unstable economic situation, might create pronounced bias, since resource consumption is transformed to one year assuming no quarterly variations in use of resources. Furthermore, the interview-based questionnaire results in far fewer missing responses. In addition, the range of input options was deliberately limited for many questions to prevent invalid answers as much as possible. Nevertheless, incomplete questionnaires are excluded from the analysis.

Second, in the absence of a tool for collecting data on the utilization of local resources, the questionnaire deployed in this study was translated into Arabic and adapted to the Lebanese healthcare system by clinical and health economics experts. This questionnaire could be used to perform other COI studies in Lebanon and Arabic-speaking countries, if specific changes based on the illness to be studied were made, and if the questionnaire were adapted to the country’s own healthcare system.

Third, recruitment methods may influence the sample; i.e. enrolling participants during clinical visits tends to a sample of patients who are in treatment for mild MS. Typically, highly disabled patients seldom visit the MS center due to their disability, and this is particularly the situation during the COVID-19 pandemic. Thus, we enrolled sufficient participants for each EDSS score, while ensuring that at least 10% of the sample consists of patients with a severe EDSS level. Our results cannot be extrapolated without weighting for the actual MS severity distribution in Lebanon.

Fourth, information on consumption of resources and associated costs was not publicly available. However, identification, measurement, and valuation of costs, as well as sensitivity analyses are in alignment with the Lebanese healthcare system and specifications. In the absence of standardization and publicly available data on the costs of health resources, we employ a variety of methods to obtain cost data from various sources. To enhance comparability, we use the method recommended by Cochrane (https://handbook-5-1.cochrane.org/) in presenting costs’ results; this method uses the PPP to convert the cost estimates of a target currency to a fixed US dollar price year.

Finally, several challenges to measuring HRQOL were identified; these include selecting instruments, the absence of a national index value set, and considering factors which impact the results. Existing instruments measure HRQOL, including generic and disease-specific measures [65], both of which we use. We selected the EuroQOL as it is one of the most commonly used generic questionnaires for measuring HRQOL, and it is comparable across diseases. In the absence of a local index value, it was necessary to define the source of the EQ-5D-5 L value set to calculate the health utility. In complement with the EQ-5D-5 L, we opted for the MusiQoL as our MS-specific questionnaire, because it includes two dimensions (symptoms and psychological well-being) not represented by the Multiple Sclerosis Quality of Life-54 (MSQOL-54) nor by the MS Quality of Life Inventory (MSQLI) [53]. The associations with HRQOL and socioeconomic instabilities [66], as well as with spirituality [67], have been studied previously. Therefore, the impact of socioeconomic instabilities in Lebanon on QOL utilities, as well as the influence of religion and spirituality on societal perceptions of illness, are considered in the interpretation of the results.

The country-specific context and associated factors vary widely across countries and are likely to affect the costs and HRQOL. These factors include healthcare-specific structures [68], assessment of consumption of healthcare resources, informal care, productivity losses [69,70], reimbursement policies [71], and other cultural and socioeconomic aspects [72]. This study was conducted amid a triple disaster in Lebanon, consisting of the COVID-19 pandemic, the drastic economic and financial crisis, and the explosion of Beirut’s port. Lebanon has been trapped in full-scale emergency crises since 2019, and the pandemic has worsened the situation to previously unseen levels [73]. The country confronts crippling debt and fiscal crises, in addition to the deterioration of its currency [73]. The Lebanese Pound is pegged to dollar rates and has lost more than 80% of its total value in the exchange markets, which saw average salaries plummet 84% by the end of 2020. The financial crisis has resulted in a dollar shortage, which restricted the supply of important medical supplies and reduced the consumption of health resources [74]. The massive explosion at the Beirut port has imposed major consequences on the wellbeing of people and on the healthcare system, and added an extra burden on an already suffering economy. These factors are likely to impact the utilization of healthcare and also QOL, both of which are considered in the interpretation of the study results. The financial crisis and the pandemic have forced lower consumption of some healthcare resources, especially during lockdown periods when access to emergency services was limited. These crises, including the wide range of COVID-19 consequences, added to the burden of MS patients, making it difficult for them to distinguish between initial MS factors adversely impacting their QOL and the harsh Lebanese living conditions.

Since economic crises are recurrent phenomena in LMICs [75], resilience and transparency in presenting the research process should generate valid results. The detailed and transparent reporting of cost units and health utilities are markers of the reliability of the burden-of-illness estimate, and the quality of reporting is crucial to facilitating comparison of studies’ methodologies and outcomes.

4. Conclusion

In the absence of local guidelines for conducting and reporting on COI and QOL studies in Lebanon particularly and LMICs generally, this study protocol provides evidence on methodologies adopted that can be used as a guideline for future studies. Moreover, we expect our findings to foster discussion on the quality and methodology of burden-of-illness studies in Lebanon and other LMICs. The results of this study on the economic burden of MS will draw the public’s attention to the precise economic impact that MS imposes on the Lebanese society. The findings of both the COI and QOL studies could be used to conduct an economic evaluation of MS interventions in Lebanon. Developing a cost analysis manual and a local index for QOL in Lebanon can enable an accurate and systematic way to conduct future economic burden-of-illness studies in the country.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Author contributions

All authors were involved in the concept and design. J Dahham drafted the manuscript. The Multiple Sclerosis Health Resource Use Questionnaire was adapted to the Lebanese healthcare system by all the study authors (J Dahham, R Rizk, M Hiligsmann, C Daccache, S Evers, I Kremer health economics experts) and (S Khoury and H Darwish clinical neurology experts practicing in Lebanon). All authors reviewed and edited the manuscript and approved the final version of the manuscript. All authors agree for the final version to be published and to be accountable for all aspects of the work.

Reviewer disclosures

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

Supplemental data

Supplemental data for this article can be accessed here.

Funding

This paper was not funded.