INTRODUCTION
The common mental health problems of depression and anxiety are major sources of disability in the community, and are mainly treated in primary care in accordance with current UK mental health policy.1 The full range of depressive and anxiety disorders are seen in primary care, from ‘problems of living’ to severe depressive illnesses requiring hospitalisation. However, recent policy for the treatment of depression in primary care has shifted away from a concern about lack of recognition of depressive symptoms and consequently under-treatment,2,3 toward a concern with over-treatment, or more specifically, over-prescription of antidepressants and medicalisation of patients with problems of living.4–6 The prescription of antidepressants has risen dramatically over the last 20 years.7 In 2002–2003 over 26 million prescriptions were issued, costing £380 million.8
There is considerable uncertainty about whether antidepressants are beneficial to those with less severe conditions, particularly as most clinical trials of anxiety and depression are based in secondary care. Not only is there uncertainty about the benefits of antidepressants, but there is also concern about the possible serious side-effects (inducing suicidal behaviour) of some, possibly all antidepressants, in a sub-group of patients.9 This has led to draft recommendations by NICE and the Drugs and Therapeutics Bulletin that for people with mild depression antidepressants should no longer be used as the first-line treatment. It is suggested that these patients often respond to interventions such as counselling and self help.5,6 To influence GPs' prescribing policies and practices it is necessary to have an understanding of how they make decisions and the factors that shape these. The limited evidence available suggests that not only do GPs find making decisions about diagnosing10 and prescribing for depression problematic,11 but decisions and treatment options are severely constrained by lack of time and inadequate access to specialist services.12,13 Therefore, as a result of such resource constraints, GPs, in line with current guidelines,2,14–15 will inevitably prescribe antidepressants for patients presenting with symptoms of anxiety and depression. The aim of this study was to explore whether or not this is an accurate interpretation of how GPs make decisions.
How this fits in
The GPs in this study did not take the decision to prescribe antidepressants lightly and used decision rules for deciding whether or not to prescribe an antidepressant based on clinical and social criteria. The prime rule tended to be ‘wait-and-see’, which was thought to discourage inevitable prescribing. GP decision-making was shaped by: organisational constraints such as access to psychological services; by the GPs' professional responsibilities; their belief that antidepressants are often an effective form of treatment even if they only act as enablers; and patient attitudes toward depression and its management.
RESULTS
What decision rules and strategies do GPs use to decide whether or not to prescribe an antidepressant?
The GPs from all five groups described their use of what could be considered as rules in the consultation process for deciding whether a prescription for an antidepressant would be appropriate. These rules were used initially to identify the underlying reason for the consultation. The rules incorporated predictable clinical questions, such as the need to eliminate other possible physical conditions that could be causing the presenting symptoms and the identification of so called ‘classic’ symptoms of depression, particularly physiological features such as sleep disruption, appetite loss and social withdrawal. Questions about duration, severity and ascertaining the important issues for the patient were also part of the routine questioning:
‘I look for my symptoms, you know, and I ask very direct questions once I've let them talk about all the general stuff, why they came in in the first place and what's been going on at home and how are they feeling, but then I actually ask, if they haven't mentioned it already, about weight loss, you know, biological symptoms and then I talk about how long this has been going on for, because if it is a week I'm not going to be worried. If it's been 3 or 4 months or a year, then obviously I'm much more worried. Has there been a trigger or no obvious trigger, because that can make a difference, for instance a bereavement, might be quite different, they may or may not respond well [to antidepressants].’ (Dr A, Group 2.)
Other ‘non-clinical’ indicators were also used. For example sex was seen as a proxy indicator of severity in that the presence of a younger male patient at a consultation often signified a serious problem as they were perceived to be reluctant consulters:
‘Thirty-eight-year-old men don't often come to see the GP, to start with. The fact that he has actually come there is a big alarm bell … He is a much more worrying person and young men kill themselves.’ (Dr C, Group 3.)
These decision rules were also influenced by GPs' ideas about types of depression. GPs described a traditional, dichotomous model of depression: that is, social (reactive) and biological (endogenous), for distinguishing between self-limiting, understandable social misery and an organic depressive illness. However, this approach was seen to be problematic. GPs recognised that making a social/biological distinction was simplistic, as depression was centred in the context of the patients' life and hence social issues usually played a part. Thus, a bio-psycho-social approach was alluded to, although no distinct model or interpretation of such a mind-body link was clearly expressed. Understandably, GPs expressed difficulty with interpreting and treating depression perceived to be associated with social conditions or circumstances:
‘… I think for me what is the big dilemma in managing these sort of conditions in general practice is how many of them are simply sad people, with sad lives, coping with difficult life events, which will just with the passage of time and a bit of understanding work through. Or are they genuine depressive illness[es] in the biological sense that are related to chemical changes in the brain, whatever. That is what I find very difficult …’ (Dr J, Group 3.)
‘What I find is that if I ask the patient, if you dig deep enough you can often find there are actually life events, which the patients immediately presents you with. And actually that failed relationship that they are just coming out of is probably what's precipitated this low mood.’ (Dr S, Group 3.)
However, in some contexts these ideas about causation were less important, particularly as many GPs thought that antidepressants could be prescribed regardless of the cause of depressive symptoms:
‘I mean when I was a student and did psychiatry there was very much “there's reactive and there's endogenous depression and that won't respond to antidepressants and that will” but since then everything has shown that if somebody's depressed it doesn't matter what's caused it, they still need antidepressants …’ (Dr J, Group 4.)
‘Wait and see’: the preferred strategy
The preferred management option was to wait and see, but under some circumstances; antidepressants would be prescribed at the first consultation. This was when the reported symptoms were regarded as severe, persistent, biological and ‘classic’ and where delay would provide no benefit to the patient. In these contexts, biological explanations of depression dominated and the influence of social circumstances was regarded as of limited importance:
‘I think she [vignette C] would probably get a tablet from me, I have to say that she probably would because I think there is quite a bit of depression there. There is quite a lot of depressive features, the weight loss and everything, that she needs so much help.’ (Dr C, Group 3.)
‘… she's down, she's losing weight, and she's not looking after herself …’ (Dr T, Group 3.)
The preferred strategy of ‘wait and see’ was influenced by the tendency of the GPs' to want to look initially for physical explanations for the presenting symptoms. This is an easier option where there is uncertainty. Focusing the discussion on physical symptoms was also a strategy for legitimising the patient's complaint and therefore opening up the possibility of a later discussion about psychological issues:
‘I mean one of the great tools that you have in a general practice is to get people back and that is one of the main ways of dealing with uncertainty.’ (Dr D, Group 5.)
‘You're going to have to do a fair amount of questioning to get a bit more of an idea of what's going on and I think see them probably two or three times before being able to say, what do you think? … I know if I went to someone I would be horrified if they gave me a prescription for something after seeing me for 10 minutes.’ (Dr H, Group 1.)
‘Well the physical is much easier so, just explore the physical symptoms …’ (Dr C, Group 2.)
If physical explanations compete with psychological, then the physical one took priority. If, however, no physical explanation is available then psychological ones were adopted. Thus, the ‘wait and see strategy’ performed a number of functions. One of these allowed the GP time to gain a greater understanding of the condition and determine whether it was self-limiting. Another was to allow patients time to reflect on their situation and see if they were willing to consider that they may have a mental health problem and that medication may help.
The benefits of antidepressants
GPs were, on the whole, positive about the beneficial effects of antidepressants, although the potential side effects were noted:
‘Generally they're worth a try. I think when they do work they're amazing, changes people.’ (Dr E, Group 2.)
‘… I say to them they either don't help at all and you get side effects you can't tolerate or you will feel so fantastic you will come here and say “I wish I'd had them years ago. I have never felt so well in my life”.’ (Dr B, Group 2.)
Even when they felt that antidepressants were more appropriate for depression with a biological basis, GPs considered them appropriate as tools for managing different types of depressive problems. Antidepressants were believed to help people tackle their problems as it enables them to ‘stop crying all the time’ or to sleep and therefore function better:
‘Well I quite often say to them that it won't take away their social situation but it will help them sleep better so they'll have more energy to start dealing with the problems that they need to deal with …’ (Dr L, Group 4.)
‘I see antidepressants … as an enabler which is something very important — and enables people to take control themselves.’ (Dr C, Group 5.)
Some GPs said that they felt comfortable prescribing an antidepressant even without a diagnosis of depression as certain antidepressants were licensed to treat a range of mental health conditions:
‘… I think the Prozacs of this world are so good because now you've got the license to treat anxiety and panic symptoms. You don't actually have to make a diagnosis, the antidepressants are going to cover everything from mild depression to severe anxiety.’ (Dr J, Group 4.)
Furthermore, diagnostic labels of depression were sometimes written in patients' notes to fulfil administrative requirements, but were seldom used in the consultation with the patient.
Alternatives to antidepressant medication
The acknowledgement by GPs that antidepressants were beneficial did not imply that they would always be satisfied with prescribing antidepressants; and hence, their preference for a ‘wait and see’ strategy that allowed for consideration of other treatment options.
Individual interventions, such as counselling, and social interventions, such as referral to a health visitor and benefits advice, were considered as an alternative or sometimes in combination with antidepressants:
‘It's not just giving an antidepressant, there are other things that you're doing at the same time.’ (Dr J, Group 4.)
‘Often patients who say “I don't want antidepressants” end up with antidepressants and counselling.’ (Dr E, Group 5.)
‘… there are lots of other avenues that you could explore before you start giving antidepressants, like the health visitors, social support, get her to see a social worker and check that she has all the benefits coming to her.’ (Dr C, Group 1.)
There was also evidence of social stereotyping where perceived benefits and appropriateness of management options, such as counselling, were influenced by the socioeconomic status or the level of education of the patient:
‘You laughed when we mentioned counselling?’ (Facilitator.)
‘Because he's (vignette B) a 38-year-old factory worker who isn't going to want to sit around and talk about himself.’ (Dr A, Group 2.)
‘I'm not sure that people with less education or from those sorts of backgrounds would be less likely to benefit from counselling. I think they're probably more likely to benefit from counselling. I think people in professional jobs who come from wealthy backgrounds have more likely developed better coping skills than others.’ (Dr D, Group 1.)
The sex of the patient also had implications for condition management:
‘As you say quite a lot of these sort of people, like older men, probably haven't talked very much to anyone else about their problems and they find the whole experience coming along and just talking to somebody very helpful. And they then feel they can carry on and readjust things. And they don't want to carry on seeing somebody or taking antidepressants.’ (Dr D, Group 5.)
Giving men, in particular, the opportunity to talk at the first consultation about their problems was often considered to be therapeutic in its own right. However, a contrasting view was that men would prefer medication as opposed to talking therapy:
‘… but I mean men of that age with kids feeling irritable and tired are very unreceptive to the idea of just having a chat about things. They want action and the reason they've come today is probably because they've been nagged to come and now they're here they want something to go away with …’ (Dr A, Group 2.)
What factors shape GPs' decision making?
GPs' individualistic use of decision rules result in different prescribing thresholds. However, prescribing decisions were also shaped by a range of other factors, which were perceived to be out of their control. There was a perception of limited resources for psychological services in primary care and limited scope for referring to secondary care services. Hence, the GPs felt there was often no alternative regarding treatment but to prescribe an antidepressant:
‘… I can't get any other sort of support for them and then you enter into the “this is not going to make better but it might help you cope with it scenario”. I think that is a big area of prescribing. I am not particular proud that I do it. For life events … I am wrong to treat them.’ (Dr S, Group 3.)
‘Our psychiatrists perform [and] provide a psychosis-only service. The community psychiatry nurses won't take on anybody who is not psychotic. There is not the resource to do anything but prescribe. If they have a depressive illness they have to have a prescription because they are going to wait 4 months to see the practice counsellor, you know.’ (Dr J, Group 3.)
‘… in those cases I'm much more inclined to give them antidepressants. Partly because other than drugs there's nothing else. There's no counselling for people who don't speak English as their first language.’ (Dr C, Group 1.)
Time pressures from having to restrict patients to a 10-minute appointment were thought to encourage a prescription rather than other treatment options:
‘Well I would say that the group here seems to be reluctant to prescribe on the first consultation but for whatever reasons we end up doing it … but we're not rushing into it so I think we're trying to be as selective as we can but for whatever pressures, either time pressures or patient pressure, or a bit of both, perhaps do end up prescribing.’ (Dr H, Group 1.)
‘I think as GPs, if somebody comes to see you, you feel obliged to do something or seem to be doing something. And I think with the increasingly large numbers of patients turning up at the surgery, you feel under an obligation to do something and that something fairly quickly in the turn of an appointment, and [a] very easy thing to do is write another prescription to give them a pill and say this is going to make you better.’ (Dr J, Group 3.)
There was also evidence of a ‘prescribing culture’ in general practice, and one GP expressed concern that he could be regarded as incompetent if he did not prescribe antidepressants for depressive symptoms:
‘I feel that by mentioning them [antidepressants] I've done my bit for, you know, the prevailing culture in the profession which means to be you must prescribe antidepressants, but if you don't you're hailed incompetent …’ (Dr D, Group 2.)
When cost is an issue for the patient, GPs said that they would be more willing to explore alternative options before prescribing antidepressants. Prescribing a full 6 months supply of antidepressants at the initial consultation was an uncomfortable option for GPs, although prescribing multiple, smaller prescriptions were considered too expensive for some patients particularly as the patient was more likely to throw the prescription away if they could not afford it. GPs also felt uncomfortable prescribing if they believed the depression may be self-limiting and could improve without medication, as this would represent a waste of resources:
‘I don't want to prescribe when it's inappropriate. It's very worrying isn't it? Are we just giving out a load of antidepressants for people whose illnesses are going to get better, partly because we're seeing them and partly because depression does get better anyway and it's a huge waste of NHS resources that we're participating in?’ (Dr E, Group 1.)
Cost of prescribing was reported as important by some GPs as it influenced the type of antidepressant prescribed. These GPs said that they try the cheaper tricyclic medication first before prescribing a selective serotonin reuptake inhibitor (SSRI). Others preferred to prescribe the SSRIs owing to their arguably better side-effect profiles and cost did not shape their prescribing decisions nor did the influence of drug representatives.
The individualisation of the management of care was favoured, as different forms of treatment were perceived to be more effective for different people:
‘Some people do well with talking therapy and some people feel better just for coming and talking about it in the surgery and sometimes antidepressants might be necessary.’ (Dr T, Group 5.)
It was suggested that much depended on the attitude of the patient towards various management options — including their willingness to accept the limitations of antidepressants. Antidepressants were thought to be less likely to be effective for people who might regard them as a ‘cure’ for their problems, since such people were more likely to be dissatisfied with the outcomes. Moreover, where GPs believed life circumstances to be the origin of a psychological problem, they were more likely to encourage patients to explore other possible sources of help:
‘It's the people, they don't work too, are the people who have that attitude “fix this, make it better”, you know, often, because they are not prepared to engage perhaps if there has been an underlying problem and those are the ones that frequently are dissatisfied, discontented and stop doing it. And I think it becomes upon us when we prescribe it that we should actually emphasis[e] that this is not going to cure everything. It is not going to cure a bad marriage and an alcoholic situation.’ (Dr S, Group 3.)
‘She's (vignette C) got the most opportunities for social intervention … You can actually make some simple alterations with the existing, with health visitors, social services …’ (Dr B, Group 1.)
GPs acknowledged the possible stigma surrounding a diagnosis of depression and treatment with antidepressants. The negative consequences of a prescription for antidepressants regarding insurance applications were noted in particular:
‘For a lot of people who haven't been on antidepressants it does still carry quite a stigma.’ (Dr G, Group 1.)
‘It [antidepressants] will go on their insurance forms and it will be a label which will be attached to them forever more on every form they fill in and it is a big stigma and a big problem.’ (Dr E, Group 2.)
DISCUSSION
Summary of main findings
Our findings lend support to evidence from research in the UK and Europe in a number of ways.10,12,13,20 First, it confirms previous research that whatever treatment and management options GPs preferred to follow they were subject to constraints, such as the limited time available to talk with the patient and the lack of availability of psychological services. There was, however, little evidence that the emotional burden of managing the problems of depressed patients had a major influence on their treatment practices.13 Second, the GPs in this study value antidepressant medication and support its use for major depression, but as other research has shown, they also expressed discomfort with prescribing for what could be self-limiting conditions of emotional distress brought on by social circumstances.10 GPs acknowledged the difficulty in distinguishing between social misery and despair, and depression, although a prescription would be likely if symptoms persisted, irrespective of the nature of the problem. However, the evidence pointed to additional factors that shape GPs' prescribing behaviour, such as perceived patient attitudes toward depression and antidepressant medication. The decision to prescribe was also influenced by considerations of cost to providers and patients, the GPs' obligation to treat and a professional culture, which is perceived to encourage prescribing.
Strengths and limitations of the study
The novel feature of this study was its specific focus on the neglected issue of the reasoning behind decisions to prescribe antidepressants. However, these themes raised by the GPs appear to reflect the general prescribing principles outlined by Taylor back in 1978,22 which question whether a given drug is ‘necessary, effective, safe and economic’ to warrant prescription and identified general rules, although the GPs gave individual accounts of the consultation process. These rules provide the basis for GPs' decisions to: prescribe immediately, ‘wait and see’ if the symptoms resolve and to use other management options apart from, or in combination with, antidepressants. Furthermore, this study has shown that these decision rules are based on both clinical (for example, classic signs such as weight loss and sleep disturbance) and social (for example, males were generally being perceived as more seriously depressed) criteria.
The methodology (focus groups) used to explore the decision-making was also relatively novel but is associated with some limitations.21 The first is whether individual face-to-face interviews may have enabled GPs to talk more specifically and freely about prescribing in the context of their own practice. We acknowledge that group norms could strongly influence opinions expressed and inhibit the expression of divergent opinions. However, as all the participants were recruited from different practices the potential for constraints created by ‘group thinking’ was limited and some variation in individual accounts was evident. The second possible limitation is self-selection, as it is likely that the group participants were more motivated and interested in mental health issues than their colleagues who did not volunteer to take part. Although more than half of the GPs did not have any formal psychiatric training and only one had a specialist qualification, the participants would all have access to peer-reviewed journals and were primarily women, who have more interest in mental health issues and are more likely to use counselling techniques than their male counterparts.23 The GPs participating in this study were also different to the population of GPs working in and around the Bristol area in that they were predominantly from training practices. Thus, the evidence presented here may exaggerate ‘watchful waiting’ as it is more in line with educational policy on depression management. Self-selection does limit the transferability of results and perhaps a more cautious conclusion is required that there is a group of GPs who are reluctant to prescribe antidepressants, rather than GPs, as a group, are reluctant to prescribe.
Implications for clinical practice
The evidence from this study provides little support for the view that GPs take the easy option of prescribing antidepressants in the face of uncertainty. Evidence suggested that the GPs were cautious in their prescribing behaviour. This was reflected in the favoured strategy of ‘wait and see’ which gave both the GP and the patient time to recognise and negotiate agreement on a possible cause of the problem and an appropriate, acceptable means of resolution in an environment of supportive care from the GP.
The apparent discrepancy between the views expressed by the GPs and the volume of antidepressants being prescribed could perhaps be explained by sample biases or that GPs are prescribing at the same rate but to a larger number of patients now identified as having depression. Alternatively, despite the GPs' reluctance to readily prescribe antidepressants, the propensity to do so may be encouraged by: GPs' beliefs that anti-depressants are beneficial (a belief primarily informed by experiential knowledge); by a lack of time available to talk with the patient (the problems of communication between doctor and patients with depression have been well documented);24 and by a lack of access to alternative treatment options including counselling and other psychological services such as cognitive behavioural therapy.
The evidence suggests GPs would support the initiative behind recent draft guidelines regarding watchful waiting.5,6 This guidance, however, needs to be clear about what constitutes mild depression and address the question of prescribing to patients who are experiencing social adversity. There is growing concern about the appropriateness of antidepressant prescribing in such cases. However, GPs need to be convinced that alternatives to antidepressants are at least as effective for patients with so called ‘mild depression’.6 Furthermore, unless resources such as psychological therapy are made readily accessible and are equitably distributed there is little alternative to antidepressant prescribing.