Development of a preference-based measure for Multiple Sclerosis: the Preference-Based Multiple Sclerosis Index (PBMSI)

Assessing health-related quality of life (HRQL) has moved to the forefront of clinical research andis considered a crucial endpoint of clinical interventions. One approach to assessing HRQL isthrough the use of health profiles. Health profiles are analyzed by sub-scale, where each sub-scalerepresents a domain of health. These measures do not provide information on the relativeimportance attached to each domain. As a result, the domains cannot be combined into an overallscore, and a trade-off cannot be made between domains when evaluating the effectiveness ofinterventions. Another approach to measuring HRQL is through the use of preference-basedmeasures. Not only do these measures provide descriptive information on the various dimensionsof health, but also provide a value for each. They have the advantage of leading to a single numberthat balances gains in one domain against losses in another. When linked to life-expectancy, theyprovide measures of quality adjusted life years (QALY) and are used to make decisions about thecost-effectiveness of interventions. The best known preference-based measures are the HealthUtilities Index (HUI), the EuroQol-5D (EQ-5D) and the Short Form-6D (SF-6D). However, thechallenge of using such generic preference-based measures in people with Multiple Sclerosis (MS)is that they may not capture all domains of health relevant to the disease and the domain weightingis based on the values from the naive general population.Therefore, the overall objective of this PhD thesis is to take important steps towards developing aPreference-Based Multiple Sclerosis Index (PBMSI) for use as a global outcome in clinical andcost-effectiveness studies for MS.To do this, a systematic review of HRQL outcomes in MS interventions was carried out andidentified that an imporant source of heterogeneity in the literature arises from the many differentmeasures used and domains evaluated (Manuscript 1). As preference-based measures reduce someof the heterogeneity by yielding one value across mutliple domains of health, the content ofgeneric preference-based measures was assessed in light of the domains identified as beingimportant to people with MS (Manuscript 2), and a review of their psychometric properties wascarried out (Manucript 3). Results revealed that these generic measures were missing severaldomains that were affected by MS, such as walking, fatigue and cognition, identifying ameasurement gap. Making use of a rich data source (that I had previously collected as part of myMSc), optimally performing items targeting the important MS domains were identified and tested10for their discriminatory capacity with respect to known groups with differing disability(Manuscript 4). This study yielded a set of 5 bilingual items (English and French) ready for testingfor comprehension and wording using cognitive interviewing with a sample of 22 people with MS(Manuscript 5). An item met criteria for acceptability after 3 to 4 rounds of interviews.The final step in this thesis was to elict preferences for different health states generated throughcombinations of items, using two different standard methods of preference elicitation which areknown to have conceptual and practical differences (Standard Gamble and Rating Scale).Manuscript 6 presents the results of this preliminary investigation in a sample of 61 patients withMS. The results indicate that the Standard Gamble is difficult for patients to understand andproduces higher values than the Rating Scale. The scoring algorithm developed based on each ofthe methods yielded vastly different results. Although the Standard Gamble is a classical techniqueof measuring preferences using decision making, it was not practical in this patient population. Onthe other hand, the Rating Scale is more suitable for the population but the values are not choicebased potentially limiting their use for economic evaluation of interventions.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.