Objectives

Beta-Thalassaemia-Major is a life-long genetic haemoglobin disorder where patients require intensive regimens, including frequent blood transfusions and daily chelation therapy. Despite the introduction of oral chelators, non-adherence continues to be prevalent. Understanding psychosocial correlates of chelation adherence is important when considering potential interventions to improve adherence as they may be potentially modifiable. Previous studies of the relationships between psychosocial factors and chelation adherence have neither been theory driven nor assessed theoretically important variables such as self-efficacy and outcome expectancies. In addition, there has been little investigation of situational determinants of adherence (cognitive, behavioural or affective). This study, informed by the Health Action Process Approach, investigated within-participant correlates of oral chelation adherence on a daily (episodic) basis.

Design

The study used a within-participant cross sectional design. Thirty seven adult participants with Beta-Thalassaemia Major were recruited from clinics at two hospitals, of which 31 were able to identify an adherent and non-adherent episode within the last two months. Sampling was systematic.

Main outcome measures

A structured interview was developed to assess cognitive, behavioural and affective situational variables related to recent adherent and non-adherent episodes. The main cognitive variables (situational self-efficacy and outcome expectancies) were derived from The Health Action Process Approach (HAPA). Behavioural situation variables included substance use, day of the week, activity, location, access to medication, reminder cues and social context. Affective situational variables were mood and bodily pain.

Results

Positive outcome expectancies and higher self-efficacy together significantly predicted adherent episodes (p<0.001), however, only self-efficacy independently predicted adherent episodes (β=-.0.26, SE=0.11, p=0.018). This relationship persisted when potentially confounding behavioural factors were controlled (difficulty in accessing medication, location and whether alone) (See Table 1).

Table 1

Exploratory regression models including additional potential confounding variables

Model A: Logistic regression including potential confounder of location

βStd. Err.P>|z|95% Confidence Interval
Outcome Expectancies -0.006 0.375 0.987 -0.742 - 0.729 
Self Efficacy -0.251 0.121 0.038 -0.488 - -0.013 
Location 0.203 1.045 0.846 -1.845 - 2.252 
βStd. Err.P>|z|95% Confidence Interval
Outcome Expectancies -0.006 0.375 0.987 -0.742 - 0.729 
Self Efficacy -0.251 0.121 0.038 -0.488 - -0.013 
Location 0.203 1.045 0.846 -1.845 - 2.252 

Model B: Logistic regression including potential confounder of activity

Coefficients.Std. Err.P>|z|95%Confidence Interval
Outcome Expectancies 0.064 0.318 0.839 -0.558 - 0.687 
Self Efficacy -0.251 0.112 0.026 -0.471 - -0.031 
Activity -0.394 0.883 0.656 -2.124 - 1.337 
Coefficients.Std. Err.P>|z|95%Confidence Interval
Outcome Expectancies 0.064 0.318 0.839 -0.558 - 0.687 
Self Efficacy -0.251 0.112 0.026 -0.471 - -0.031 
Activity -0.394 0.883 0.656 -2.124 - 1.337 

Model C: Logistic regression including potential confounder of difficulty accessing medication

Coefficients.Std. Err.P>|z|95% Confidence Interval
Outcome Expectancies -0.016 0.322 0.960 -0.648 - 0.616 
Self Efficacy -0.218 0.112 0.051 -0.438 - 0.001 
Difficulty accessing medication 0.774 0.992 0.435 -1.170 - 2.718 
Coefficients.Std. Err.P>|z|95% Confidence Interval
Outcome Expectancies -0.016 0.322 0.960 -0.648 - 0.616 
Self Efficacy -0.218 0.112 0.051 -0.438 - 0.001 
Difficulty accessing medication 0.774 0.992 0.435 -1.170 - 2.718 

Affective factors were not related to adherence.

Conclusion

Findings were consistent with the HAPA and suggested the importance of situational cognitive factors, particularly self-efficacy, to chelation adherence. These findings add to the evidence of the effect of self-efficacy on medication adherence across medical conditions and extend previous work in emphasising situational variation in self-efficacy. Adherence interventions should consider situational variables (cognitive and behavioural) that are potentially modifiable.

Disclosures:

No relevant conflicts of interest to declare.