Why do patients discontinue antidepressant therapy early? A qualitative study

Abstract

Background: Current guidelines for antidepressant prescribing are that treatment should be continued following the resolution of symptoms: six months after a first episode and for at least two years for those with previous episodes. Despite this, sub-optimal treatment duration still predominates. Patients have negative and positive views on antidepressants, which change throughout their treatment journey.

Objectives: To explore views and experiences of patients recently initiated on antidepressants (within six months), and to consider the influences on early discontinuation.

Methods: A qualitative interview study was used in four general practices in the North East of Scotland. A purposive sample of primary care patients, newly initiated on antidepressants, was interviewed to explore views and experiences with antidepressant therapy. Interviews were audio-recorded and transcribed verbatim. Thematic analysis was conducted using a consensus coding frame developed by two researchers.

Results: Twenty-nine patients participated. Three main factors influencing discontinuation were identified: ownership, knowledge and support. The treatment journey was characterized by four important time points where health care intervention may be helpful.

Conclusion: Health care professionals would benefit from exploring patient knowledge and views on depression and antidepressants at an early stage in treatment. Patients would welcome active involvement in treatment decision making, the provision of information and ongoing support.

Keywords::

• general practice
• depression
• stigma
• attitudes
• antidepressants
• adherence

KEY MESSAGE:

• Patient views on antidepressants are influenced by their perceptions of ownership, knowledge and support.

• These are subject to change and influential throughout four distinct stages within the treatment journey.

• Treatment adherence could be improved by intervention to address perceptions at these time points.

INTRODUCTION

International guidelines on the management of depression recommend a maintenance period of at least six months following the resolution of symptoms, and longer for patients who have experienced previous episodes (1–6). Many individuals follow a course of antidepressants for sub-optimal periods (7,8). In some cases, patients remain on antidepressants for longer than is necessary, but more often, they discontinue early. Studies exploring the phenomenon of early discontinuation demonstrate how patient demographics (e.g. socioeconomic deprivation) partly explain this (8,9); but patient views and beliefs, including how they see themselves in the context of taking antidepressants (10) and concerns regarding dependency (11), also influence decision making. Chakraborty et al. (2009) found patients maintained incorrect beliefs about antidepressants, despite being given explanations about the nature of depression and the rationale behind treatment. Perceptions of the presence or absence of support are also important (12,13).

Patient perceptions of causes of their depression relate to their treatment preferences. Karasz et al. (2012) found that patients who presented their depression in symptomatic terms were more likely to be offered antidepressants than those who presented in situational terms, and that few patients refused them (14).

Malpass et al. (2009) identified key periods during treatment when patients’ views on antidepressants are more susceptible to change, and intervention from a practitioner could improve treatment adherence (15). Van Geffen et al. (2011) describe two main thought patterns leading to treatment discontinuation or continuation (16). At three months, continuers had generally positive attitudes to antidepressants while discontinuers’ attitudes were negative. However, their study excluded patients who had used antidepressants in the previous year, and included patients on SSRIs for indications other than depression. It is, therefore, unclear if these views are representative of patients suffering from depression. Here, we consider all patients recently initiated on antidepressants for depression, not just those being treated with antidepressants for the first time. We explore the factors, which hinder or facilitate the continuation of therapy, and the time points throughout treatment when they are influential.

METHODS

Recruitment

A purposive sample of patients recently (last six months) initiated on antidepressants (Table 1) for the treatment of depression were recruited by inviting six general practices, representing a range of urban/rural settings to take part. Practices identified patients fulfilling the inclusion criteria (over 18 years of age, undergoing antidepressant treatment for depression) who were recruited over six months during 2011. Patients were included regardless of whether they continued to collect their prescriptions and had other co-morbid conditions or not. GPs approved patients for invitation, aided by a protocol detailing inclusion and exclusion criteria. Patients receiving antidepressants for reasons other than depression, or where the GP considered participation to be inappropriate (e.g. cognitively impaired) were excluded. The aim was to recruit a sample large enough to elicit a sufficient diversity of themes, yet small enough to conduct in-depth qualitative analysis (17). Most qualitative researchers report finding little ‘new’ themes emerging after around 20 interviews have been conducted (17). Within the time constraints of the study, it was feasible to conduct up to 30 interviews, and in so doing, access sufficient diversity of experience.

Table 1. Patient characteristics.

Antidepressant drugs included
SSRIs	Fluoxetine, Fluvoxamine, Citalopram,Escitalopram, Paroxetine, Sertraline
Tricyclic-like	Mianserin, Trazodone, Lofepramine
SNRIs and other	Duloxetine, Mirtazapine, Reboxetine, Venlafaxine
Interviewee characteristics
Female n (%)	20 (69)
Median age (IQR)	49 (40–56.5)
Co-morbidities (% yes)	24 (83)
First ever episode (% yes)	12 (42)
Risk factors for recurrent depression (% yes)^a	17 (59)
Previous treatment with ADs (% yes)	17 (59)
Median no. GP appointments (IQR)	6 (5.00–8.00)
Median no. GP appointments for depression (IQR)^b	5 (3.00–6.00)

^aThese include three or more previous episodes, pre-existing dysthymia, severe episodes, seasonal patterns, family history, poor response to continuation therapy, co-morbid anxiety, substance misuse.

^bThese are appointments where the records clearly state antidepressant medication was reviewed and/or mental well-being was discussed.

Ethical approval

Ethical approval was obtained from the North of Scotland Research Ethics Service (NoSRES) Ref 11/SO801/6.

Patients gave written consent for access to their medical records to enable extraction of data on medication use, consultations in the last six months and treatment history. Anonymous audit data on non-participants was collected by another researcher (AC) working on behalf of the Scottish Primary Care Research Network to assess the representativeness of interviewees in terms of age and sex.

Interviews

Face-to-face qualitative interviews were conducted in general practices, the University research centre or participants’ homes depending on individual preference. Interviews were audio-recorded and lasted approximately thirty minutes with the shortest lasting eight minutes and the longest being one hour and thirteen minutes. A semi-structured interview schedule (Supplementary Appendix to be found online at http://www.informahealthcare.com/doi/abs/10.3109/13814788.2013.838670) was developed based on a review of the relevant literature exploring patient attitudes to antidepressants. This was used to explore circumstances surrounding initiation of therapy; treatment history; views and beliefs about depression and antidepressants; experiences during GP consultations; and reasons for continuation/discontinuation with therapy.

Analysis

Qualitative interview data were transcribed verbatim and uploaded onto NVIVO. Data was analysed according to the ‘Framework’ method; a systematic process of sifting, charting and sorting material according to key issues and themes (18). Data was analysed inductively to identify themes in the main stages of data reduction, data display, conclusion drawing and verification. An initial coding theme was developed and revised as more transcripts were coded and more themes emerged. A sub-set of transcripts was analysed independently by another researcher (IC), and some existing ‘nodes’ (coding themes) were revised and new ones added.

RESULTS

Recruitment

Patient disposition is shown in Figure 1. Participants did not differ significantly from non-respondents (n = 174) in terms of age, sex and history of treatment for depression.

Figure 1. Recruitment flow diagram.

Participant characteristics are presented in Table 1. Based on of self-reported treatment status, interviewees were classified as ‘treatment continuers’ (n = 20) or ‘discontinuers’ (n = 6) as defined by Van Geffen or as ‘contemplators’ (n = 3), who were considering discontinuation (16). The treatment journey was characterized by four independent stages; three corresponding to those identified by Malpass (2009) ‘when patients experience a return to function,’ ‘adverse side effects,’ ‘latency period’ (i.e. before antidepressant effects emerge and medication is perceived as ineffective) plus an additional stage, ‘initiation of therapy’ (15).

The experiences and views of these groups (to antidepressants in general and reasons for discontinuation/continuation) are discussed in relation to three central themes emerging from the data, which are influential throughout each treatment stage. These are ‘perceived ownership,’ ‘knowledge’ and ‘support.’ ‘Ownership’ refers to the level of responsibility patients had for the management of their depression; ‘knowledge’ relates to their understanding of depression and antidepressants; and ‘support’ to their perceived level of support from their GP, family and friends.

Initiation of treatment. In general, continuers took ownership of the decision to seek help:

... now that I've been to the doctor and the doctor said yeah you have a problem and everything ... I can sort of go right, there is a bit of problems here I can do something about that [3005—continuer].

Although for some, while they recognized the symptoms, they needed encouragement from others to go:

I’ ve got a good friend who is a CPN [Community Psychiatric Nurse], she was supporting me and she said I really think you should be going you know and speak to the doctor... I decided that yes I would go [2001—continuer].

Discontinuers and contemplators tended to place ownership of this decision onto others:

He [her husband] was awful keen for me to do everything, aye for it not to be his fault [4007—discontinuer].

Overall, continuers tended to take ownership of the actual decision to initiate treatment:

I didn't ask to go on them but I had that in my head you know if she offered me to go on these, on something then I would, so she did offer them [2001—continuer].

Discontinuers and contemplators, however, were inclined to attribute ownership of this decision to their GP and were less accepting:

I felt quite embarrassed about it because I didn't realise they were antidepressants she just said it was something to pick me up ... like why do I need antidepressants? None of my family have ever needed antidepressants [1010—discontinuer].

Knowledge about the causes of depression and mechanism of antidepressants early on appeared to relate to how patients subsequently managed their depression. Discontinuers tended to attribute their depression to external causes:

I suspect my depression was more reactive [3008—discontinuer].

Alternatively, continuers said they had sought help to deal with their symptoms.

Patients with recurring depression felt no explanation was given because of their previous experience. Others could not remember details of the information given.

Reflecting on their initial appointment, continuers tended to feel their GP was supportive:

The doctor has acknowledged that you have got a problem and you need help, you are not just crazy or something [3005—continuer].

In contrast, those who later discontinued recalled being disappointed in the limited support they had received:

I just thought, well it's a waste of time because they are not listening to me, they are not understanding what I'm saying [3003—discontinuer].

Some patients (regardless of treatment status) felt well supported by family and friends, but were selective in terms of who they discussed their treatment with:

I was quite open with my husband but I haven't told my parents because my mum has depression [4001—continuer].

Some spoke of lack of understanding within their family and workplace:

Some of the employers, ... one of them I spoke to was okay but the other one was probably thinking ‘get over it’ [4008—continuer].

Return to function. A similar pattern was evident when patients began to feel better. Continuers spoke of making a conscious decision to continue with their medication. Discontinuers stopped because they felt better:

I think it's just sort of got me over the worst of it, I think, that's how I was starting to think perhaps I can come off of them [4004— discontinuer].

Others considered stopping to see how they felt:

As long as I am feeling good, I know I shouldn't question it but you start to wonder how will I feel if I come off them? [3007— continuer].

Beliefs about the aetiology of their depression had implications for discontinuation at this stage:

I just felt perhaps now is the time to wean me off a bit, I have come to terms with a lot [4006—continuer].

It was clear that intervention from their GP at this point could help prevent early discontinuation:

Although things were explained quite clearly at the beginning I think at this point [now I feel better] I would have benefitted from a bit more support or just knowing a bit more about what I am doing [2004—continuer].

Experience of side effects. ‘Ownership’ was an important factor at this stage. Those who felt in control of the management of their depression seemed more willing to tolerate the side effects:

My sister said I had put on weight, she said you probably should stop your antidepressants. I said no [1003].

Most participants were given information about side effects. In some cases this prompted them to discontinue:

They just said that your moods will change, and you will have an upset tummy and your side effects. I thought, no, I'm not going through that [3002—discontinuer].

Others discontinued their medication in the past because of fears of developing a dependency:

I just felt that I shouldn't really be, well, always depending on tablets [4010—continuer].

Stage four: Latency period. During this stage (characterised by perceived lack of efficacy of antidepressants), continuers seemed prepared to wait and give the medication time to work:

Yeah, I was really, really surprised, although I did have my doubts, but no, I will persevere, and I'm glad I did actually [1008—continuer].

Discontinuers understanding of this tended to be limited, suggesting they had not been given sufficient explanation of this from their GP:

No I just, I actually felt more depressed, I just wanted to sleep, I just wanted to go to my bed and it [antidepressant] was encouraging me to do that ... there was a spell that it wasn't working [3003—discontinuer].

The amount of support patients felt they had from their GP varied. Continuers said they had regular reviews:

... if there's just a little niggle, you can talk about it, a little peace of mind, and then as I said they monitor you, they don't give you automatic prescriptions ... [3001—continuer].

Although, there were others who felt relatively unmonitored:

I've kind of been left to it, she said to make sure that I came back, I don't know what would have happened if I had just stopped them, and not come back [1008—continuer].

Some continuers took responsibility for the management of their condition and made changes to their medication without seeking the advice of their GP:

I will know myself, if it's not working then to increase it again [4010—continuer].

DISCUSSION

Main findings

‘Ownership,’ ‘knowledge’ and ‘support’ had important differential impacts throughout the four stages of the treatment journey. Patients who experienced joint decision-making seemed to be more accepting of decisions made and generally had positive attitudes towards antidepressant treatment. In contrast, those who felt treatment decision-making had been done on their behalf had more negative attitudes and seemed more prone to discontinuation. Different levels of understanding about depression and antidepressants were evident. Misunderstandings related to how quickly antidepressants would be effective, side effects and when/how to discontinue. Worries about becoming ‘addicted’ were also common. There were also important differences in perceived levels of support and suggestions of poor communication between GPs and patients.

Strengths and limitations

This study provides insight regarding concepts, which relate to treatment discontinuation noted by others (15,16), and pinpoints where in the treatment journey they may be influential. It also highlights the similarity between the reasons patients give for the decision to initiate treatment in the first place and the reasons they give for early discontinuation. Maxwell et al. (2005) report similar findings where women felt they had a ‘moral obligation’ to seek help to manage their symptoms due to the adverse effect it was having on their family life, but also subsequently felt pressured to discontinue due to the negative social beliefs about antidepressants (13).

It may be of concern that those unwilling to participate in the study may have been the most likely to subsequently discontinue early since participants were predominantly continuers. However, the views and experiences of continuers in relation to their previous episodes of treatment when they discontinued early are also considered. A further potential limitation of the study is that the treatment journey and patient decision making was not considered in relation to symptom severity. It had been planned to explore the relationship between depression severity and treatment continuation using PHQ 9 (Patient Health Questionnaire) (19) scores recorded in general practice records. However, few were recorded. Ideally, this should have been assessed by the research team at the same time as interviews were conducted. There may also have been a relationship between continuation/discontinuation and having co-morbid conditions. Although this information was collected, this was not explored.

Relation to existing literature

The theme of ‘ownership’ has been discussed previously in terms of developing a concordant relationship between GPs and patients (15,20); and lack of shared decision making between patient and GP has been associated with discontinuation (16). Schofield et al. (2011) describes how patients become experts on their condition as they go through multiple treatment episodes, taking increased ownership of the management of their condition based on their previous experience (21). Karasz et al. (2012) discuss how patients work with their health care providers but drive treatment initiation based on how they present their depression (14). Those who present in social terms are often not offered antidepressants or are offered in more ‘vague’ terms suggesting anticipation of refusals. This suggests that both patients and health care professionals have the shared belief that depression rooted in social causes will not respond to treatment with antidepressants.

The importance of knowledge or ‘mental health literacy’ has previously been noted as a predictor of medication adherence (22). Patients with strongly held convictions about the causes of their depression can be reluctant to engage with interventions that do not match their conceptualizations (23). Some patients who attributed their depression purely to external influences, expected to discontinue medication once the stressors were removed. This further highlights the need for more education about the pharmacological treatment of depression. Negative attitudes towards antidepressants evident including fears of dependency have been observed previously and are known to influence discontinuation (12,16). Beliefs that medicines cause harm, may be addictive, or cause side effects, correlate with poor adherence (23). Health care professionals need to challenge these beliefs through discussion.

Perceived levels of support from family, friends and GPs were important at all stages of treatment. The importance of partner involvement in treatment decisions and patient preference for ongoing GP support throughout treatment has also been demonstrated (25,26).

Implications for clinical practice and future research

Health care professionals could benefit from examining patient knowledge and understanding about depression and antidepressants at an early stage. Patients require accurate information throughout their treatment even if they have been treated with antidepressants previously. This is of course true for almost any consultation but due to the nature of depression, where patients may find it difficult to concentrate during consultations, there may be a particularly high rate of early discontinuation with subsequent relapse. Specific gaps in understanding, which were evident, suggest there should be a focus on how long antidepressants may take to be effective, what changes patients may expect, effective management of side effects and the importance of continuation for six months after symptoms resolve. It may also be useful to provide written information patients can refer back to. Patient views about the causes of their depression may also relate to their treatment preferences and the likelihood they will adhere to treatment, so it is important that this is explored at an early stage. Further research should evaluate the development and provision of ongoing information for patients being treated for depression. The use of technology, such as text messaging may be a valuable support in this endeavour (27).

Conclusion

Three new concepts ownership, knowledge and support have been identified and, their changeable nature and the time points when they are likely to be influential discussed. These results could be used to inform further development of the guidelines for the management of depression in primary care, which may lead to improvements in adherence and the prevention of relapse.

ACKNOWLEDGEMENTS

The authors should like to thank all the practice staff and patients who have contributed to this study and the financial support of NHS Research Scotland (NRS), through the Scottish Primary Care Research Network.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

FUNDING

This study was funded by NHS Quality Improvement Scotland.

ETHICAL APPROVAL

Ethical approval was obtained from NRES, REC Reference 11/SO801/6.