Objectives: To determine the degree of noncompliance with antidepressant treatment in the Alberta population and to investigate the reasons for noncompliance.
Method: We used data from the Alberta Mental Health Survey, a telephone survey conducted in 2003 (n = 5323 adults), to produce population-based estimates of the frequency of noncompliance and the reported reasons for noncompliance.
Results: Reported noncompliance was 41.7% (95% confidence interval [CI], 36.9% to 46.6%) for respondents taking 1, 2, or 3 antidepressants. Noncompliance for those taking 1 antidepressant was 42.0% (95%CI, 36.9% to 47.2%), whereas noncompliance for those taking 2 or 3 antidepressants was 39.4% (95%CI, 26.7% to 53.6%). Among respondents currently taking at least one antidepressant, 64.9% (95%CI, 57.4% to 71.7%) reported that forgetfulness was the most common reason for noncompliance. Of respondents taking 1 medication, 64.1% (95%CI, 56.0% to 71.4%) reported forgetfulness as did 71.3% (95%CI, 48.3% to 86.8%) of those taking 2 or 3 medications. Both the frequency of noncompliance and the reported reasons for noncompliance were independent of sex and age.
Conclusion: Our study replicates prior reports that indicate that noncompliance is common with antidepressant treatment. Forgetting to take medication is the most important reported reason for this noncompliance.
(Can J Psychiatry 2006;51:719-722)
Information on funding and support and author affiliations appears at the end of the article.
Clinical Implications
* Noncompliance with antidepressant treatment is a major clinical issue.
* Frequently cited reasons for noncompliance, such as "feeling better" or "side effects," are relatively unimportant.
* Simple memory aids such as dated blister packs, automated voice mail reminders, and patient education might help address this important issue.
Limitations
* These findings are derived from cross-sectional data, so we cannot confirm causal direction of associations.
* To confirm generalizability, we would need Canada-wide data.
* The sample size was too small to examine specific classes of antidepressants.
Key Words: antidepressant, noncompliance, forgetfulness
Noncompliance with prescribed medication is a major issue in medicine (for example, 1-3). Because of the authoritarian connotation of the word compliance, other terms are sometimes preferred, for example, adherence. Since antidepressants must be taken for extended periods of time, in some cases for life, noncompliance may have a significant impact on the individual (for example, relapse) as well as a broader economic and social impact. A metaanalysis revealed that patients suffering from depression are 3 times more likely to exhibit noncompliance with medical recommendations than patients without depression (4). Knowledge in this field is important because it can potentially lead to the development of interventions to improve compliance.
The general consensus is that noncompliance with antidepressant prescriptions is a major problem for depression treatment but that predictors are not well understood. Noncompliance manifests itself in several ways, including failure to fill a prescription, failure to take any medication, early discontinuation of medication (dropout), and failure to regularly take prescribed dosages. A recent survey provides an opportunity to describe noncompliance with antidepressant prescriptions in a population-based sample of adults in Alberta.
Methods
The AMHS was a cross-sectional, random digit dialing-based telephone survey that collected data from 5383 adults. Eligible individuals from various households were contacted and the participation rate among them was 77%. The detailed methods of this survey and its potential vulnerability to bias are described elsewhere (5).
We cross-tabulated raw data and calculated population estimates with 95%CIs, using weighted survey commands in Stata Software, (Release 8; Stata Corp, College Station, TX, 2003). We first determined how many respondents were taking one or more antidepressant medications. Then we examined the prevalence of reported noncompliance for antidepressant use in this subpopulation. Finally, we tabulated the prevalence of self-reported reasons for noncompliance. The latter 2 stages depended on how participants answered 2 questions:
* "When you take antidepressants, are there any days when you took less than you were supposed to?"
Answer: yes or no.
* "Why did you take less antidepressant medication than you were supposed to?"
Answer:
1. You forgot.
2. You felt better.
3. The medicine was not helping.
4. You thought the problem would get better without more medication.
5. You could not afford to pay for the medication.
6. You were too embarrassed to take the medicine.
7. You wanted to solve the problem without medication.
8. The medicine caused side effects that made you stop.
9. You were afraid you would get dependent on the medication.
10. Other reason.
This list was based on the reasons most commonly reported in pilot studies. The interviewers did not read this list to the respondents. The interviewers instead categorized the replies on the basis of the participants' spontaneous responses. The "other reason" response category was used when an answer did not fit into one of the categories or when the interviewer was unsure how to classify a response. The investigators subsequently receded these responses. Participants were allowed to give more than one answer for each antidepressant. Most gave one answer for each medication, but some gave up to 5 answers, all of which are accounted for in our analysis. In cases in which a respondent gave the same reason more than once, we counted this reason only once.
Results
Respondents taking antidepressants were taking between 1 and 3 antidepressants, but most were only taking 1. Population estimates for the proportion of Albertans taking 1, 2, and 3 antidepressants were 6.3% (95%CI, 5.7% to 7.0%), 0.8% (95%CI, 0.6% to 1.1%), and 0.1% (95%CI, 0% to 0.1%), respectively. Stratification of the data by sex and age showed that neither was related to the number of antidepressants taken (data not shown).
Noncompliance was reported by 41.7% (95%CI, 36.9% to 46.6%) of the total sample of respondents taking antidepressants (Table 1). Stratification of the data indicates that neither sex nor age predict the frequency of noncompliance (Table 1). When the data were stratified by the number of antidepressants taken, 42.0% (95%CI, 36.9% to 47.2%) of individuals taking a single medication reported noncompliance. Similarly, 39.4% (95%CI, 26.7% to 53.6%) of participants taking 2 or 3 antidepressants reported noncompliance. These data do not provide statistical evidence that noncompliance was different across these groups. Further, this result does not support the hypothesis that the complexity of the antidepressant regimen is a determinant of compliance.
It is estimated that 64.9% (95%CI, 57.4% to 71.7%) of the population forgets to take its antidepressants, making forgetting the major reason for noncompliance (Table 2). The catch-all category for "other reason," reported by 26.8% (95%CI, 20.6% to 34.2%), was the second most frequent reason for noncompliance. Common examples of these reasons for not taking medication included "ran out," "could not access it," and "did not need it." The population estimates for the 8 other categories were low. Stratification of the data for the subpopulation that forgot showed that neither sex nor age predicts forgetfulness (Table 2). Stratification by the number of antidepressants taken showed that forgetting was reported by 64.1% (95%CI, 56.0% to 71.4%) of respondents taking a single antidepressant and by 71.3% (95%CI, 48.3% to 86.8%) of those taking 2 or 3 antidepressants. The estimates did not provide statistical evidence that the degree of forgetting was different between the 2 groups.
Discussion
Our analysis suggests that noncompliance with antidepressant use is an important issue (Table 1). Contrary to expectation, the data did not suggest that compliance was lower for those taking 2 or 3 medications rather than 1. Neither sex nor age predicted the degree of noncompliance. Forgetfulness was the most frequently reported reason for noncompliance and was not dependent on sex, age, or the number of antidepressants taken (Table 2).
Forgetfulness has not been identified as a major determinant for noncompliance with antidepressant treatment (6,7). Most studies do not include forgetting as a possible reason and most are based on small clinical samples. This raises a public health issue that can be addressed by simple means, for example, dated blister packs, automated voice mail reminders, and patient education. Randomized clinical trials have provided evidence that blister packaging can increase compliance with drug treatment (8,9).
Two important issues require future study: we must identify the subpopulation actually taking antidepressants for depression rather than for other indications, and to assess its clinical significance, we must quantify the extent of noncompliance.
Funding and Support
No funding was obtained for this paper.
Acknowledgement
We thank Jeanne Williams for her assistance with Stata programming.
[Sidebar]
Abbreviations used in this article
AMHS Alberta Mental Health Survey
CI confidence interval
[Sidebar]
R�sum� : L'oubli : un r�le dans la non-observance du traitement antid�presseur
Objectifs : D�terminer le degr� de non-observance du traitement antid�presseur dans la population de l'Alberta et rechercher les raisons de la non-observance.
M�thode : Nous avons utilis� les donn�es de l'enqu�te sur la sant� mentale de l'Alberta, un sondage t�l�phonique men� en 2003 (� = 5 323 adultes), afin de produire des estimations bas�es dans la population de la fr�quence de la non-observance et des raisons d�clar�es de la non-observance.
R�sultats : La non-observance d�clar�e �tait de 41,7 % (95 % intervalle de confiance [1C], 36,9 % � 46,6 %) pour les r�pondants qui prenaient 1, 2 ou 3 antid�presseurs. La non-observance pour ceux qui prenaient un seul antid�presseur �tait de 42,0 % (95 % IC, 36,9 % � 47,2 %), tandis que la non-observance pour ceux qui prenaient 2 ou 3 antid�presseurs �tait de 39,4 % (95 % IC, 26,7 % � 53,6 %). Parmi les r�pondants prenant pr�sentement au moins un antid�presseur, 64,9 % (95 % IC, 57,4 % � 71,7 %) ont d�clar� que l'oubli �tait la raison la plus fr�quente de la non-observance. Parmi les r�pondants prenant un seul m�dicament, 64,1 % (95 % IC, 56,0 % � 71,4 %) ont �voqu� l'oubli, tout comme 71,3 % (95 % IC, 48,3 % � 86,8 %) de ceux prenant 2 ou 3 m�dicaments. Tant la fr�quence de la non-observance que les raisons d�clar�es de la non-observance �taient ind�pendantes du sexe et de l'�ge
Conclusion : Notre �tude reproduit des r�sultats ant�rieurs qui indiquent que la non-observance est fr�quente dans le traitement antid�presseur. Oublier de prendre le m�dicament est la raison la plus importante d�clar�e pour cette non-observance.
[Reference]
References
1. Blackwell B. From compliance to alliance: a quarter century of research. Neth J Med 1996;199:140-9.
2. Cramer JA, Spilker B, editors. Patient compliance in medical practice and clinical trials. New York (NY): Raven Press; 1991.
3. Osterberg L, Blaschkc T. Adherence to medication. New Engl J Med 2005;353:487-97.
4. DiMatteo MR, Lepper HS, Croghan TW. Depression is a risk factor for noncompliance with medical treatment: meta-analysis of the effects of anxiety and depression on patient adherence. Arch Intern Med 2000;160:2101-7.
5. Kassam A, Carter C, Patten, SB. Sedative hypnotic use in Alberta. Can J Psychiatry 2006;51:287-94.
6. Demyttenaere K. Risk factors and predictors of compliance in depression. Eur Neuropsychopharmacol 2003;13(Suppl 3):S69-S75.
7. Masand PS. Tolerabiliry and adherence issues in antidepressant therapy. Clin Ther 2003;25:2289-304.
8. Simmons D, Upjohn M, Gamble GD. Can medication packaging improve glycemic control and blood pressure in type 2 diabetes? Results from a randomized controlled trial. Diabetes Care 2000;23:153-6.
9. Wright JM, Htun Y, Forman P, Ballard RC. Evaluation of the use of calendar blister packaging on patient compliance with STD syndromic treatment regimens. Sex Transm Dis 1999;26:556-63.
[Author Affiliation]
Andrew G Bulloch, PhD1, Carol E Adair, MSc, PhD2, Scott B Patten, MD, PhD3
[Author Affiliation]
Manuscript received October 2005, revised, and accepted May 2006.
1 Professor, Department of Physiology and Biophysics, and Hotchkiss Brain Institute, Faculty of Medicine, University of Calgary, Calgary, Alberta.
2 Associate Professor, Departments of Community Health Sciences and Psychiatry, and Hotchkiss Brain Institute, Faculty of Medicine, University of Calgary, Calgary, Alberta.
3 Professor, Departments of Community Health Sciences and Psychiatry, and Hotchkiss Brain Institute, Faculty of Medicine, University of Calgary, Calgary, Alberta.
Address for correspondence: Dr AG Bulloch, Department of Physiology and Biophysics, University of Calgary, Faculty of Medicine, 3330 Hospital Drive NW, Calgary, AB, T2N 4N1; bulloch@ucalgary.ca
