Lack of treatment adherence and self-medication are two of the biggest problems in antibiotic misuse among patients.1–3 Adherence has been specially studied in chronic but not in acute diseases such as infectious diseases.4 A meta-analysis found that 37.8% of patients forget to take a dose of antibiotic.5 The rate of adherence in this therapy in our country is not well known. Few studies have examined the compliance with antibiotic regimens, with figures between 40 and 60%.6,7 The 42% found in the Pan European Survey of Patients was higher than values found in Italy (34%), Belgium (18%), France (16%) and Britain (9%).8

Unlike drugs that only affect individual patients, misused antibiotics add the global risk of bacterial resistance, which jeopardizes their effectiveness.9 In Spain resistance rates are particularly high and their consequences are serious: resistant bacterial infections are associated with increased morbidity, mortality, health demands, hospitalization, medical expense and impairment of the effectiveness of treatment of future patients.10,11

The WHO Strategy for Containment of Antimicrobial Resistance encourages prescribers and dispensers to educate patients on the proper use of antibiotics and the importance of completing the prescribed treatment.12

Results of intervention studies in community pharmacy about improving treatment adherence are variable. A systematic revision conducted in 2005 concluded that it is impossible to identify an overall successful adherence-improving strategy performed by pharmacists and that more well-designed and well-conducted studies need to be performed.13

In our country there are few studies that address this question from the sphere of pharmacy; they are of local scope or with design problems, they provide written information and their results are non-significant.14,15 For this reason, we considered it important to conduct a study to evaluate the benefit of an oral educational intervention, in terms of increased treatment adherence and symptom improvement or resolution, vs. “routine pharmaceutical care”.

Of the 138 patients initially included, 12 were lost to follow-up because they could not be contacted by telephone. There were no statistically significant differences, so the 126 valid cases (62 CG vs. 64 IG) were analyzed.

The groups were comparable by the sociodemographic and clinical variables (Table 1). In the CG, the mean age was 44.8 years (SD=17.7) and in the IG it was 44.5 years (SD=18.2). There was a majority of women in both groups (69.4% and 65.6% in the CG and IG, respectively). No significant differences in the degree of knowledge were found between the groups.

Adherence to antibiotic treatment was 48.4% (CI: 36.4–60.6) in the CG and 67.2% (CI: 55.0–77.4) in the IG, with a difference of 18.8% (95% CI=15.8–34.6; p=0.033). Moderate non-compliance (more than one dose intake missing) was observed in 81.2% of patients in the CG and in 38.1% in the IG, with a difference of 43.1% (95% CI=16.4–63.1%; p=0.001).

There was no significant difference in the health perception, although it was higher in the IG, with reports of being “totally cured” in 54.7% (95% IC=42.6–66.3) in the IG and in 46.8% (95% IC=34.9–59.0) in the CG (p=0.297).

Results of the bivariate analysis are presented in Table 2. Variables with a p<0.1 were selected to be included in the regression model. The logistic regression analysis confirmed a group effect (OR 2.23; 95% CI=1.01–4.93; p=0.047). The variables that positively influenced adherence were the knowledge of the antibiotic treatment before the intervention (OR 3.06; 95% CI=1.34–6.99) and coincidence between duration of treatment indicated by physician and duration of drugs container written in the prescription (OR 2.84; 95% CI=1.16–6.99). The use of beta-lactams has a negative influence (OR 0.32; 95% CI=0.14–0.72). This model used 99.2% of the study sample. The model goodness-of-fit was good as assessed by the Hosmer–Lemeshow test (χ2 4.7; p=0.786).

The study results show that implementing an educational intervention when patients come to a pharmacy to collect antibiotics that they have been prescribed, increases adherence to the treatment. However, this does not mean that the patient will feel cured at the end of the treatment course. The educational intervention in this study was effective, with an 18.8% difference in compliance between the CG and the IG.

A meta-analysis and a Cochrane review found interventions carried out by health professionals (physicians, nurses, pharmacists) to be effective on treatment adherence,21,22 but pharmaceutical interventions found mixed results13,23–25 and it is difficult to identify an overall successful strategy improving adherence – performed by pharmacists. In our country, there have been positive results following the intervention of hospital pharmacists in patients with other pathologies, albeit with a small sample size26. In a community pharmacy, Machuca et al. achieved a 14% increase in the antibiotics adherence.14 Andres et al. also improved adherence, although the differences were not statistically significant.15 The intervention in these studies was done with written information, and their results were not strong. The results of our study provide valuable information about the effectiveness of oral interventions on the treatment adherence in the community pharmacies, which are close and accessible to the patient, and should be used to actively work with other health professionals in the appropriate use of antibiotics.

The adherence in the IG was higher than average antibiotic therapy adherence reported in Spain by Gil et al., although it is very similar to that found by other authors.6,19 This difference may be related to the measuring method, as Gil et al. used the technique of pill counts, while the other two used the telephone interview. Factors that were not evaluated and that can influence adherence are the duration of the treatments and the level of education of the study population.

Our results are lower than those found in a meta-analysis conducted internationally,5 and the difference may be related to the high rate of resistance in our country.8 It is difficult to maintain the effectiveness of adherence interventions over time, and this can explain the better results in acute diseases.

No significant differences were found between groups with regard to patient-perceived health. This may be due to the fact that the infectious process is self-limiting in most cases and resolves without intervention in a short time. Different studies show that there is a relationship between health as perceived by patients and their adherence to treatment.19,27 It therefore seems important to analyze this variable. However, we must bear in mind that we measured the perceived health through a single direct question to the patient, so the results need to be interpreted with caution.

Patients frequently report discontinuing antibiotic therapy when they begin to feel better or when adverse events occur.28 This might be the main cause of the existence of antibiotics at home29 and of self-medication, in few cases responsible for the emergence of resistance.8,29

Logistic regression showed an association between adherence and the coincidence between the duration of treatment orally indicated by the physician and the duration of treatment written on the prescription. The mismatch between the duration of antibiotic treatment and the format in which drugs are manufactured sometimes makes more than one container necessary, and therefore more than one prescription to end the treatment. This could be a barrier for patients to successfully complete their treatment. In August 2012, the Spanish Agency for Medicines and Health Products issued a Resolution on matching therapeutic drug formats (J01 and J02 group), urging manufacturers to adapt to proper formats for these treatments.30

Logistic regression also showed that medication knowledge is correlated with medication adherence. This finding is according to a recent publication.31 Therefore improvement in the patient's knowledge would bear a positive effect on adherence, and patient education in the proposed dispensing protocol is considered to be an implicit strategy for improving compliance. Pharmacists should be encouraged to play a proactive role in large-scale health education programmes.

We evaluated patients’ baseline knowledge with a validated questionnaire – this is a strong point of our study, as it is the first validated questionnaire to specifically assess knowledge of patient medication.20

The study has the limitations inherent to using different indirect tests, which tend to overestimate adherence, although the use of two different methods can reduce the bias. A contamination effect could have occurred considering that both groups have antibiotics dispensed from the same site, and the intervention effect could have been increased. The study was conducted in a local setting, so the results cannot be extrapolated to wider settings. The patient allocation was not randomized, and this can be a design of weakness. For future studies it would be advisable to use a randomized clinical trial design, in order to eliminate this possible bias.

In conclusion, pharmaceutical intervention increased antibiotic adherence in this community pharmacy. The results suggest that knowledge is an important factor associated to adherence, and that an intervention aimed at increasing it may improve adherence. Thereby, this report also reinforces the pharmacists’ role in improving the health care system and they should be strengthened as educators in the rational use of medicines. It is necessary to conduct multicenter studies to support this thesis, to verify that the implementation of these services in the pharmacy is, not only effective, but also feasible and compatible with the rest of the activities.