Background

Tackling antimicrobial resistance is considered to be an international public health challenge as there are now resistant strains of nearly all bacteria, with some bacteria becoming multi-drug resistant1. A contributing factor to this increasing resistance to antibiotics is the failure by patients to take courses of antibiotics as directed i.e. missing doses or not completing the course2. This prospective study applied the Theory of Planned Behaviour (TPB)3 to the prediction and understanding of adherence to antibiotics.

Method

A total of 249 patients (75 male, 174 female, mean age = 48.6 years, SD = 16.2) presenting with a prescription for a course of antibiotics to a participating pharmacy in Leeds were recruited to the study. Medication data (name of antibiotic, length of course, number of daily doses), demographic data (age, gender), TPB variables (intention, perceived behavioural control) and past behaviour were recorded at recruitment. Participants were then telephoned on the day after they were due to complete their antibiotic regimen to record their adherence to the medication, 241 participants provided these data. Two adherence measures were used: (i) a pill count, in which participants were asked to say how many tablets were left in the packet; (ii) self-report, using open-ended questions (e.g. How many times have you taken the antibiotics at a different time? What was the reason for this?). Logistic regression analyses were administered to explore the prediction of adherence to the antibiotics.

Results

Adherence to the antibiotics was 74.6% (pill count measure) and 65.1% (self-report measure). Good adherence (pill count measure) was predicted by low daily doses and not experiencing side effects. This model correctly classified 76% of participants. Intention, perceived behavioural control and past behaviour were not significant predictors of adherence using this measure. Good adherence (self-report measure) was also predicted by low number of daily doses, not experiencing side effects and being older. This model correctly classified 71% of participants. Using this measure of adherence, past behaviour (previously taking antibiotics as prescribed) and low perceived behavioural control were significant predictors of good adherence.

Discussion

These important medication factors are consistent with the compliance literature4. Contrary to expectations, intentions did not predict adherence. This suggests that non-adherence was non-intentional (e.g. forgetting to take tablets) rather than intentional (deliberately missing doses).

References

1 World Health Organisation (1998). Emerging and Other Communicable Diseases: Antimicrobial Resistance. Resolution of the 51st World Health Assembly, May.
2 Department of Health (1998). Antibiotics: Don't wear me out. HMSO: London, UK.
3 Ajzen, I. (1985). From intentions to actions: A Theory of Planned Behavior. In Action Control: From Cognitions to Behaviour (Edited by J. Kuhl and J. Beckmann), pp. 11-39. Springer-Verlag: Germany.
4 McGavock, H., Britten, N. & Weinman, J. (1996). A review of the literature on drug adherence. Commissioned by the Royal Pharmaceutical Society of Great Britain as part of the project Partnership in Medicine Taking.