Glatiramer acetate treatment persistence - but not adherence - in multiple sclerosis patients is predicted by health-related quality of life and self-efficacy: a prospective web-based patient-centred study (CAIR study).

Background: In patients with relapsing remitting multiple sclerosis (RRMS) the persistence of and adherence to disease modifying drug (DMD) treatment is inadequate. To take individualised measures there is a need to identify patients with a high risk of non-persistence or non-adherence. As patient-related factors have a major influence on persistence and adherence, we investigated whether health-related quality of life (HRQoL) and self-efficacy could predict persistence or adherence.

Methods: In a prospective web-based patient-centred study in 203 RRMS patients, starting treatment with glatiramer acatete (GA) 20 mg subcutaneously daily, we measured physical and mental HRQoL (Multiple Sclerosis Quality of Life-54 questionnaire), functional and control self-efficacy (Multiple Sclerosis Self-Efficacy Scale), the 12-month persistence rate and, in persistent patients, the percentage of missed doses. HRQoL and self-efficacy were compared between persistent and non-persistent patients, and between adherent and non-adherent patients. Logistic regression analysis was used to assess whether persistence and adherence were explained by HRQoL and self-efficacy.

Results: Persistent patients had higher baseline physical (mean 58.1 [standard deviation, SD] 16.9) and mental HRQoL (63.8 [16.8]) than non-persistent patients (49.5 [17.6]; 55.9 [20.4]) (P = 0.001; P = 0.003) with no differences between adherent and non-adherent patients (P = 0.46; P = 0.54). Likewise, in persistent patients function (752 [156]) and control self-efficacy (568 [178]) were higher than in non-persistent patients (689 [173]; 491 [192]) (P = 0.009; P = 0.004), but not in adherent vs. non-adherent patients (P = 0.26; P = 0.82). Logistic regression modelling identified physical HRQoL and control self-efficacy as factors that explained persistence. Based on predicted scores from the model, patients were classified into quartiles and the percentage of non-persistent patients per quartile was calculated: non-persistence in the highest quartile was 23.4 vs. 53.2% in the lowest quartile. Risk differentiation with respect to adherence was not possible. Based on these findings we propose a practical work-up scheme to identify patients with a high risk of non-persistence and to identify persistence-related factors.

Conclusions: Findings suggest that pre-treatment physical HRQoL and control self-efficacy may identify RRMS patients with a high risk of early discontinuation of injectable DMD treatment. Targeting of high-risk patients may enable the efficient use of persistence-promoting measures.

Trial Registration: Nederlands Trial Register code: NTR2432 .