Impact of Treatment-Related Beliefs on Medication Adherence in Immune-Mediated Inflammatory Diseases: Results of the Global ALIGN Study

Introduction Medication adherence is critical in chronic immune-mediated inflammatory diseases (IMIDs) and could be affected by patients’ treatment-related beliefs. The objective of this study was to determine beliefs about systemic medications in patients with IMIDs and to explore the association of those beliefs and other factors with adherence. Methods This was a multi-country, cross-sectional, self-administered survey study. Included were adults diagnosed with one of six IMIDs receiving conventional systemic medications and/or tumor necrosis factor inhibitors (TNFi). Patients’ necessity beliefs/concerns towards and adherence to treatments were assessed by the Beliefs about Medicines Questionnaire and four-item Morisky Medication Adherence Scale. Correlation of patients’ beliefs about treatment and other factors with adherence were evaluated by multivariable regression analyses. Results Among studied patients (N = 7197), 32.0% received TNFi monotherapy, 27.7% received TNFi–conventional combination therapy, and 40.3% received conventional medications. Across IMIDs, high adherence to systemic treatment was more prevalent in TNFi groups (61.3–80.7%) versus corresponding conventional treatment groups (28.4–64.7%). In at least four IMIDs, greater perception of the illness continuing forever (P < 0.001), of the treatment helping (P < 0.001), and more concerns about the illness (P < 0.01), but not clinical parameters, were associated with higher treatment necessity beliefs. Higher treatment necessity beliefs, older age, Caucasian race, and TNFi therapy were associated with high medication adherence in at least four IMIDs. Conclusions Treatment necessity beliefs were higher than concerns about current medication in patients with IMID. Illness perceptions had a greater impact on treatment necessity beliefs than clinical parameters. Older age, greater treatment necessity beliefs, and TNFi therapy were associated with high self-reported medication adherence in at least four IMIDs. Trial registration ACTRN12612000977875. Funding AbbVie. Electronic supplementary material The online version of this article (doi:10.1007/s12325-016-0441-3) contains supplementary material, which is available to authorized users.

According to the World Health Organization, adherence to long-term therapy is defined as the degree to which a person's use of medication, following a diet, and/or adopting lifestyle changes is consistent with agreed-upon recommendations from a healthcare provider [13]. Many Table 1). Additional patient characteristics are summarized in Table S2.

Disease Characteristics
The mean duration of disease varied across the six IMIDs (Table 1) Table S2.

Specific Beliefs Towards Current IMID Medications
BMQ-Specific Necessity subscale scores indicated a relatively high perceived need for current treatment ( Fig. 1a; Table S3a). Numerically higher mean scores were reported for patients receiving TNFi (either as monotherapy or as a combination therapy) compared with conventional therapy alone. BMQ-Specific Concerns subscale scores were lower than those observed for the BMQ-Specific Necessity subscale, and were in a similar range across the three treatment groups (Fig. 1a, b; Tables S3a, b).

Illness Perception and Depressive Symptoms
Patients understood the chronic nature of their diseases, as reflected by high mean scores across IMIDs (8.5-9.0) in response to the BIPQ question ''How long do you think your illness will continue?'' Patients gave high mean BIPQ scores (8.3-8.5) for how much they thought that their treatment could help their illness. The lowest BIPQ scores were regarding how much their illness affected their life, how much they experienced symptoms, and how much their illness affected them emotionally (Table S5).
The percentage of patients with signs and symptoms suggestive of depression (PHQ-2 total score C4) ranged from 13.7% (CD) to 20.5% (RA ; Table S6). to be highly adherent was greater for the TNFi component than for the conventional therapy component (Fig. 2). The lowest rate of highly adherent patients was observed in the AS population receiving conventional medication, either alone (28.4%) or in combination therapy (33.1%; Fig. 2).  illness concern and emotional impact of illness scales (all IMIDs) were associated with higher BMQ-Specific Concerns scores, while lower BIPQ coherence score (all IMIDs) and lower BIPQ treatment control scores (all IMIDs except AS) were associated with lower treatment concerns (Table S8). ''Complete treatment response'' and treatment type were the only disease-or treatment-related factors with a significant impact on treatment necessity beliefs or treatment concerns in four or more IMIDs.

TNFi in combination therapy (all IMIDs) and
TNFi monotherapy (AS, PS, CD, and UC) were associated with greater BMQ-Specific Necessity scores. Increased depressive symptoms were associated with higher BMQ-Specific Concerns and Necessity scores, and Caucasian race was associated with lower treatment concerns.

DISCUSSION
Various ''patient-supporting'' approaches (e.g., telephone follow-up and supportive care) have been established and tested in order to improve medication adherence. Many studies on the impact of patient support methods have been small in size and produced mixed results [35,36]. To devise successful programs and to provide targeted adherence support, it is crucial to understand key drivers for enhanced treatment adherence in patients with IMIDs.  [19,20].
In ALIGN, perceptions about the necessity of the medication rather than concern beliefs affected high medication adherence in a majority of the IMIDs. These observations are consistent with results from four [19,20,37,38] of the six [19,20,28,[37][38][39]  Across all six IMIDs, patients with higher medication necessity beliefs were those who believed more strongly that their condition would be longer lasting and was more controllable by treatment, and who had greater concerns about their illness. Direct associations of the perception of the illness to continue longer/forever (in RA) and the perception of the treatment to help (in PS and CD) with higher medication adherence, and of experiencing more symptoms with lower adherence, were seen with certain IMIDs but not all, indicating a rather indirect association of several BIPQ items with treatment adherence via their effect on treatment necessity beliefs in the majority of IMIDs. In addition, in at least four of the six IMIDs, older age and Caucasian race were associated with better medication adherence. Age or race/ethnicity as factors affecting medication adherence were noted in patients with RA or other immune diseases in previous studies [18,[39][40][41][42][43][44]. In our study, Caucasians in general harbored significantly less concerns and overuse beliefs but not higher treatment necessity beliefs compared with non-Caucasians.
One limitation of the ALIGN study was that illness perception, beliefs about medication, risk of depression, and adherence results were based on self-reported outcomes that may be influenced by self-presentational and recall biases. Patients may overestimate the extent of their adherence in an attempt to ''please the doctor,'' or if they believe that admitting to non-adherence may result in adverse judgments.
However, in our study patients' responses were confidential; thus, the risk of overestimating adherence to ''please the doctor'' was reduced.