Scholarly article on topic 'Prescribing benzodiazepines in general practice: a new view of an old problem'

Prescribing benzodiazepines in general practice: a new view of an old problem Academic research paper on "Clinical medicine"

Share paper
Academic journal
OECD Field of science

Academic research paper on topic "Prescribing benzodiazepines in general practice: a new view of an old problem"


benzodiazepines in general practice: a new view of an old problem

health: An IntnrdiseipUnaryJouirtial for the Social Study of Health, Illness and Medicine

Copyright © 2007 SAGE Publications (Los Angeles, London, New Delhi and Singapore) DOI: 10.1177/1363459307074693 Vol 11(2): 181-198

Anne Rogers, David Pilgrim, Susie Brennan, Ilyas Sulaiman, Gareth Watson & Carolyn Chew-Graham

University of Manchester, Universities of Liverpool and Central Lancashire, University of Manchester, University of Manchester, University of Manchester & University of Manchester, UK

abstract General practitioner (GP) prescribing has been identified as an arena that has broad social and political implications, which stretch beyond individual outcomes for patients. This article revisits aspects of the controversy about prescribing benzodiazepines (or 'minor tranquillizers') through an exploration of contemporary views of GPs. In the 1980s the prescribing of these drugs was considered to be both a clinical and social problem, which brought medical decision making under public scrutiny. The legacy of this controversy for recent GPs remains a relatively under-explored topic. This article describes a qualitative study of GPs practising in the north-west of England about their views of prescribing benzodiazepines. The accounts of the respondents highlight a number of points about: blame allocation, past and present; clinical challenges about risk management; and deserving and undeserving patients. These GP views are then discussed in the wider context of psychotropic drug use. It is concluded that, while there has been a recent consensus that the benzodiazepines have been problematic, when they are placed in a longer historical context, a different picture is apparent because other psychotropic drugs have raised similar problems.

keywords benzodiazepines; GPs' perspectives; narratives; psychotropic medication

address Anne Rogers, NPCRDC, Division of Primary Care, University of Manchester, Oxford Road, Manchester M13 9PL, UK. [e-mail: anne.]


In the past, sociologically informed studies have highlighted the impact of the consultation on social and clinical outcomes, for patients in primary care (May et al., 2004). In particular, the prescribing of medication in primary care has been identified as an area where the legitimacy and moral authority of the doctor is enacted (Britten et al., 2004). It is also a health care arena where the power and influence of patients can be enhanced (through shared decision making) or thwarted through the embedded power imbalance between GP prescription preferences and those of recipients (Stevenson et al., 2002; Britten et al., 2004).

This article has three aims. First, a brief history will be given about the social and clinical controversy which has attended a particular area of medical practice - the prescription of benzodiazepines. This controversy peaked in the mid-1980s but the drugs were prescribed at high rates beyond that period (Medawar, 1992). Second, a qualitative study of recent GP views about the use of this group of drugs is reported. This explores the way in which contemporary practitioners view an emotive and controversial area of prescribing, one in which the actions and professional norms of GPs have been implicated in creating and maintaining a form of clinical iatrogenisis. The intention here is to explore the views of current practitioners, many of whom were not in their current post or training at the start of the controversy, but who are now dealing with its consequences in their clinical work. Third, these current views and their historical roots will be discussed in relation to the general context of psychotropic drug responses to the psychosocial features of mental health problems.

The medical rationale for benzodiazepine use and the emergent controversy

The first benzodiazepine, chlordiazepoxide (Librium) was introduced into clinical practice in 1960. This and related products (such as Valium, Xanax and Ativan) came to be prescribed in large quantities. They have been used for a variety of medical purposes, including the brief tranquillization of surgical patients, the treatment of epilepsy, as an aid to withdrawal from alcohol abuse and to calm psychotic agitation. However, their main use has been for the treatment of anxiety and insomnia. In the clinical literature, the drugs are classified as 'anti-anxiety agents', 'anxiolytics' or 'minor tranquillizers'.

The notion of 'minor tranquillizers' (rather than the drugs merely being about the removal of anxiety symptoms) is warranted for two reasons. The first is pharmacological - chlordiazepoxide was synthesized from chlor-promazine (the first 'major' tranquillizer') in 1955. The second reason for calling them 'tranquillizers' is the impact of benzodiazepines on the central nervous system and on behaviour. These drugs, like alcohol, can lead to

euphoria and disinhibition as well as sedation and sleepiness. These effects raise the probability of risky behaviour - they can impair motor tasks, create dangers when driving or using machinery and release emotions that might be expressed antisocially. Thus, the generic tranquillizing effect of the drugs does more than remove anxiety; it also dampens sensibility, impairs competence and can loosen normal social and moral inhibitions.

Like other tranquillizing drugs (for example the barbiturates and alcohol) the benzodiazepines reinforce the action of a naturally occurring neu-rotransmitter (gamma-aminobutyric acid or GABA). GABA has a general calming effect on neural activity. As a consequence, once the nervous system habituates to the artificial booster effect of GABA levels, it becomes excited if the external agent is withdrawn. A consequence of withdrawal is heightened anxiety; a 'rebound' or iatrogenic effect of benzodiazepine use.

Because of this GABA effect, the benzodiazepines were soon noted to have similar disadvantages to previously preferred sedatives and hypnotics, such as paraldehyde and the barbiturates. All of these anti-anxiety agents were associated with tolerance and thus the need for increasing dose levels. The benzodiazepines, like their predecessors, have been characterized by psychological and physiological dependence, so their prescription has led to predictable iatrogenic addiction. The latter has meant that clinical time and effort and health care resources have had to be devoted to treating iatrogenic rather than primary symptoms of mental health problems.

Thus the controversy about benzodiazepines in one sense was not new; previous anti-anxiety agents had been linked to similar problems. Healy (1997) describes a cycle of legitimacy associated with drugs that are highly prescribed for symptoms of common distress (be it anxiety, depression or their frequent co-occurrence). For example, prior to the benzodiazepines, the bromides of the 1920s gave way by the 1940s to the barbiturates. Similarly, the benzodiazepines have now given way to the antidepressants.

This is not merely about one fashionable named drug (say Valium) being displaced by another (say Prozac), although this is partially the case. It also includes unique selling points of newer drugs, which claim a legitimate advantage over past agents. For example, the benzodiazepines are not very toxic even in high doses, whereas twice the prescribed dose of the barbiturates could be fatal (though it should be noted that several weeks supply of the former, if ingested with alcohol, can still kill).

Notwithstanding this unique selling point of the benzodiazepines, rapid tolerance means that they become clinically ineffective after a very short period (around 10 days), with 58 -77 per cent of recipients reporting sedation effects of the drugs (drowsiness, lethargy and memory disturbances) (Lader et al., 1992). Thirty per cent of those taking these drugs for more than a few weeks develop withdrawal symptoms, including panic attacks, insomnia, tremor, palpitations, sweating and muscle tension (Tyrer, 1987). In a small percentage (under 5 per cent) more severe problems, including epileptic seizures and paranoid reactions, might occur (Baldessarini, 1999).

While these problems were recognized as early as the mid-1960s in the USA, and a law suit was instigated by the US government against the pharmaceutical company Roche in 1967, it was not until the 1980s that the problem was recognized politically in the UK. This culminated in the banning and suspension of some drugs. Both a limited and black list of drugs licensed for prescription was then implemented and advice on prescribing was drawn up and issued (Committee on the Safety of Medicines, 1988).

During the 1980s, prescription levels began to fall, but mainly for their use as anxiolytics. Rates of their prescription as sleeping tablets (the same drug is then called a 'hypnotic') actually increased slightly. Also, this period of heightened professional concern did not lead to the medical abandonment of the drugs. In 1980 there were 23 million prescriptions of benzodiazepines issued in England. Ten years later there were still 16 million prescriptions of the drugs issued (Tyrer et al., 1997, using government-disclosed data).

The extent of the problem was illustrated by a MORI poll conducted in 1985, which suggested that 10 million UK citizens had been prescribed benzodiazepines at some time in their lives and that 3 million people were, at that time, 'chronic' users. However, Healy (1997) makes the important point that common distress has always led to psychotropic agents (including ones prior to and since the benzodiazepines) being highly prescribed. Because neurotic distress is highly prevalent, medical practitioners consulted will tend to prescribe highly in response (compared to less common ailments). This does not negate the reality of the iatrogenic scale of benzodiazepine addiction. However, it does place it in the context of other commonly prescribed agents over time and it is a proxy measure of common neurotic misery.

During the 1980s, the scale of iatrogenic addiction prompted a popular protest movement co-ordinated by the British mental health charity MIND and the popular BBC TV programme That's Life, which led to further litigation against the drug companies (Lacey, 1991). Other social responses ensued. For example, self-help groups (such as 'Tranx') which were run by 'ex and partially withdrawn addicts', emerged specifically to deal with iatrogenic addiction ( In this respect, the UK response mirrored the problem identification and reaction that had occurred previously in the USA (Cohen, 1983).

Media attention regarding the iatrogenic effect of benzodiazepines occurred in the late 1970s and early 1980s in the USA (Cohen, 1983). In the UK, Bury and Gabe (1990) pointed up the role of the media in legitimizing the social problem status of the benzodiazepines. They identified four social process elements: the claims-making activities of medical experts; legal challenges; the role of the media; and the response of the State. Together these four processes made a contribution to this class of drugs becoming a public and governmental, rather than merely a clinical, matter.

While these wider social influences, institutions and processes had a profound impact on the use and acceptability of benzodiazepines, the

problem could be readily reduced in the public eye to one of incompetent GPs 'hooking' innocent housewives on addictive and dangerous drugs. At that time (the early 1980s) evidence emerged that women consumed twice the amount of GP-prescribed psychotropic medication as men (Cooperstock, 1978; Olson and Pincus, 1994). A complex set of social processes could then be reduced to poor GP judgement and gender vulnerability. This had the effect of both focusing blame on one profession and framing women as inherently psychologically weak. At the same time, it was evident that women would selectively use medication as part of a daily management regime to cope in difficult circumstances (Gabe and Thorogood, 1986).

In the immediate aftermath of the public outcry about benzodiazepines, various guidelines were issued to doctors concerning the treatment of anxiety and the use of this class of drugs. In line with this, research about prescribing since then suggests compliance with a much more restrictive repertoire. Currently, the level of benzodiazepine prescribing is a measure of quality used by Primary Care Trusts in England.

There is also now evidence of a shift towards patient self-regulation and active self-management by users, rather than dominance and control by pre-scribers (North et al., 1995). There remains a continuing professional and patient interest in the risk of these drugs. A recent qualitative study undertaken in Queensland, Australia suggested that the role of dependence in continued benzodiazepine use is well recognized, along with the importance of lifestyle change in its cessation. However, benzodiazepine users still ask GPs to provide advice about non-pharmacological management of their problems and the potential adverse consequences of long-term use prior to being prescribed benzodiazepines (Parr et al., 2006).

A qualitative study of current GP views about benzodiazepines

While there is emerging evidence that patients with mental health problems express a preference for seeing GPs rather than specialists (Lester et al., 2005), the impact of the benzodiazepine controversy may have shaped recent prescribing patterns among GPs and affected their general approach to primary mental health care. Certainly the dominance of secondary care views about primary mental health care and the expert role assumed by psychiatrists are likely to have reinforced uncertainties that GPs may have had in relation to operating assertively and independently (Rogers and Pilgrim, 2001). In this light, a study was conducted to explore the dilemmas the legacy of the benzodiazepine controversy has created for recent practitioners. Below, personal accounts from practising GPs are summarized and analysed thematically to shed light on this legacy.

Semi-structured interviews were carried out on a sample of 22 GPs drawn from a variety of localities across a major north-western English city. They included both male (n = 15) and female GPs (n = 7), newly qualified practitioners and those who had been practising for a considerable period

of time. GPs from a variety of practices (from single-handed to large group practices) were included in the study. The respondents were drawn from a total sampling frame of 70 GPs who participated in or host undergraduate medical education in their practices. Both university and NHS Ethics Committee approval was obtained.

As this was an exploratory qualitative study, a purposive attempt was made to capture the voices of GPs of different ages working in different clinical settings and those who had been involved in prescribing benzodi-azepines during the 1980s and in the contemporary period. However, there was a bias towards younger GPs, who had been practising for a relatively short period of time. Despite targeting a group of GPs, who were older or who had been practising for a time, only a handful of the GPs (n = 5) who had been practising in the 1980s agreed to be interviewed. In one sense this can be seen as a limitation of the study. However, this reluctance to be interviewed and failure on the part of interviewers to engage this group of GPs could have arisen for a range of reasons (e.g. being busier than their more junior colleagues or more jaded and less willing to participate in research). However, it may also reflect the blame, shame and lingering responsibility felt by longer-practising GPs. That cohort of practitioners had been caught up in the adverse publicity of the time.

Interviews, which were conducted in these various settings explored practitioners' views of the management of anxiety in general practice; the role of benzodiazepines in this management; and the types of patients that might still warrant repeat prescriptions of these drugs. The interview schedule was devised after reading relevant literature about benzodiazepine prescribing and use and it was modified in the light of emerging data.

The interviews were taped and transcribed and themes agreed between the five researchers. Interviews were carried out at two time periods, which were determined by the availability of the researchers (May- June 2001 and May - July 2002).The figure of 22 subjects occurred, when content saturation occurred in the transcripts in relation to broad headings structuring the interviews.

An initial categorization of themes was generated during the period of interviewing and moderated through discussion between the researchers. The initial categories were reduced to major and sub-themes, through reading and re-reading of the transcripts. The four main themes emerging across all of the transcripts are now summarized.

The appropriateness of prescribing benzodiazepines

For the majority of respondents, a sense of responsibility for avoiding the risks associated with past benzodiazepine use framed contemporary 'parameters' of appropriate prescribing. Caution, vigilance and regulation were emphasized, along with a sense of ownership of practitioner responsibility for rectifying past errors of judgement and failings. A few of the respondents

felt that the social and clinical problem of the drugs had been overstated (see later) but the bulk of them were very concerned to minimize their use and the risks attached.

Judgements about the blame or responsibility for the past mistakes of overuse and inappropriate use centred on norms:

To blame GPs is probably too strong but to say that they were a contributory factor is to state the obvious really, when they were the people prescribing them. But I think that they were probably prescribed in good faith - that they were the correct treatments at that time and you know only subsequently did problems come to light. So I think to apportion blame is probably not right but clearly GPs were part of the picture. (GP-8)

Although the 'blame' for past problems was not always accepted as being caused by GP action alone, respondents acknowledged transmitted responsibility for past events. Most respondents had not been trained GPs during the 1980s, but had been sensitized to the benzodiazepine problem as medical students (rather than having learnt about it subsequently through reports in medical journals or post-qualification training). The GPs reported the belief that initially it was psychiatrists who often initiated prescribing. While recently psychiatrists predominantly have focused on the management of psychosis, at the time of the emergence of the benzodiazepine problem they would regularly see outpatients with problems of anxiety and depression who were not, as now, managed in primary care.

Past problems were attributed to the nature of drugs, the comparative lack of an evidence base and the way in which drug companies operate.

I think part of the responsibility lies in the drug companies that produce them in the first place and I don't think, well in those days there weren't that many trials and things into these sorts of medications and their side-effects. I don't think, well I don't think doctors knew how addictive they were at the time, 'cos otherwise I think they'd have definitely thought twice about prescribing ... (GP-10)

Medical confidence to prescribe a drug with apparently no reported adverse effects meant that its widespread utilization went unabated when the drugs were kick-started in the market after 1960. The respondents considered that this early optimistic therapeutic ethos has now been being replaced with one of cautiousness as indicated by this respondent:

I think there's been an interesting change in the last 20 years in that I think you could say that there was a certain naivety and that if you, if anybody, now that I'm 50 said to me, 'Here's a great drug for anxiety which people can take long term and doesn't have any addictive effects' I simply wouldn't believe them. But the fact was that when I qualified that was what we believed, as of course it was believed for barbiturates 20 years before that. So I think, you know, you can only keep up with what appears to be acceptable medical practice at the time, bearing in mind things change. So I don't feel particularly bad about that because it wasn't, as it were, taught to me as a no-no thing to do at medical school. (GP-7)

This older GP was in a position to have lived through the shift in consciousness about the benzodiazepines and could offer some empathy with that hindsight. By contrast, most of the other respondents were younger and had been sensitized at medical school to the 'no-no' status of the drugs.

The origin of the problems of over-prescribing and turning a blind eye to the addictive features of the drugs were not always accepted as having been the sole responsibility of GPs; psychiatrists were seen as the source of the prescription habit. The respondents were keen to emphasize that a focus on GPs has been unreasonable, given that the psychiatric profession was deemed to be responsible for initiating and legitimizing the use of the drugs.

In the more recent climate of GPs defining their clinical autonomy about mental health problems, the respondents did feel a sense of continuity about professional responsibility. The younger respondents, for temporal reasons, logically had no personal responsibility for past mistakes but they still wanted to signal their professional duty as inheritors of a problem, as well as signalling their sympathy for predecessors in primary care:

I think that my perception was that actually what the GPs were doing was following what the psychiatrists did ... because in the 70s patients with anxiety disorders would be seen by a psychiatrist. They would recommend a benzo and then the GP would continue that prescription. So the GPs got blamed but often they didn't initiate ... (GP-14)

Thus the current discourse of GPs (contra the one depicted in the 1980s) seems to be one of locating the medical difficulty in psychiatry, in the past, rather than with their primary care predecessors alone.

Managing the past, managing the present

While the causes of past problems included an analysis that implicated psychiatric colleagues, there was a clear acceptance that GPs were now the sole guardians of the moral responsibility for appropriate prescribing and for rectifying the consequences of historical prescribing. A legitimate and expected part of the GP's role now is to wean people off drugs to which they had inadvertently become addicted, for whatever reason, in the past. This is now acknowledged as a substantial clinical challenge:

The addiction is so well known about that I think we all would just try and avoid using them for that reason ... you don't need to be on them for long for problems (to arise) ... they've got a bad reputation. (GP-3)

GPs viewed initiating a programme of withdrawal as a part of their role in reversing the previous consequences of prescribing and they identified specific ways and means to achieve this end. A number of respondents described how they would go about this with patients: 'I'd just sort of sit down and negotiate and say, "you know, there are problems with this medication and let's gradually wean you off it". That straight forward (GP-2). Other

GPs made a point of how they would do this without producing withdrawal symptoms: 'We would gradually try and reduce, and certainly if they're on a mixed bag of benzos then we would say, convert them over to diazepam and gradually reduce' (GP-4). Another GP pointed out the importance of moving a long-term user of multiple benzodiazepines onto just one drug and then weaning the person off it subsequently.

In initiating new prescriptions, clear rules operated. The clinical rationale for initiating the drugs was narrowly defined and circumscribed. Issuing a limited number of tablets over a short period of time was emphasized, as was the need to be stringent about their use at all: 'I don't use them in anxiety because they're of such a limited benefit ... I think the indication for benzodiazepines in general practice is fairly limited really' (GP-16). Here another respondent emphasizes cautious discretion in specific cases:

I'm most likely to initiate it in, for short-term problems and would generally prescribe in very small amounts then of perhaps half a dozen tablets. So I have got patients who, as it were, regularly but infrequently have that sort of amount of medication. (GP-7)

Retaining strict control of prescribing was seen as something that was to be policed robustly. Moreover, patient-led demand was suspended in relation to this type of drug, as was the more liberal 'empirical' or 'try it and see' approach, adopted by GPs in relation to non-controversial drugs. For the respondents, it was important to convey the exceptional and deserving circumstances of prescribing. Such cautiousness served to point up the image of the benzodiazepines being handed out like 'Smarties', which had been at the heart of the charge of GP irresponsibility in the past:

Maybe they've got a fear of flying and they haven't got time to go on a phobia course or whatever and they're going on holiday in a week's time, then sure I give them a prescription. (GP-2)

I think they're very useful in certain situations of acute anxiety, particularly I suppose in terminal situations when people are dying . because they're unable to control the anxiety. (GP-11)

This last comment leads to the next significant theme about justifying the circumstances and the type of patients that warrant the drugs.

Deserving and undeserving patients

Narratives about appropriate prescribing were littered with descriptions of the type of patient who could legitimately consult and be given benzodiaz-epines. These descriptions were discussed in relation to two imperatives in tension with one another: the moral obligation to ensure a programme of humane withdrawal; and the strict need to restrict access to a wider population. This tension is managed within the daily working constraints of GPs. Rather than the 'housewife hooked on tranquillizers', a converse picture

of the 'needy' patient emerged. Moreover, undeserving groups were those who do not elicit sympathy in the public eye and implicated GPs in a felt moral and legal responsibility. GPs labelled drug addicts and alcoholics as 'undeserving' patients. Substance abuse was a key clinical feature to attend to in decision making:

So if somebody was already alcoholic they they're not gonna get a hypnotic, whatever their circumstances really ... (GP-19)

Certainly anybody with a history of addiction of alcohol abuse, drug abuse of any kind those would be the main ones [to be excluded]. (GP-17)

It's the black market and the drug scene in [the city]. We have people who come in and you don't know if they're addicted or they're going to sell them. (GP-12)

We have a practice policy that we don't prescribe benzodiazepines to drug addicts. (GP-5)

Together these respondents were highlighting several disincentives about prescribing to substance abusers. First, it is wasteful of resources. Second, it may support an illicit street market in psychotropic drugs. Third, it may unhelpfully feed an addictive habit. In the latter regard, the issue is not about prescribing psychotropic drugs in principle but prescribing the 'wrong' drug (for example, methadone is prescribed to heroin users). We return to a discussion of this point about the 'wrong' drugs later. The medical framing of patients as morally worthy or unworthy is not new but an enduring and salient theme within medical sociology (Stimson and Webb, 1975; Jeffries, 1979). But in addition to this general tendency in some medical encounters, there is a particular moral onus for the modern GP to display scrupulous prescribing rectitude, given the reputation of past colleagues when they issued benzodiazepines. They are repairing this past reputation in a context of drugs with a street value. These considerations are thus more complex than simply the wider medical tendency to label patients as worthy or unworthy.

Turning to 'deserving' patients these, by exclusion, were those who otherwise were not abusing substances and who, for clinical reasons, invited sympathy. The dying patient was noted earlier but more commonly it was the chronic user who had suffered iatrogenic addiction from the past: 'It's the stereotypical situation of the little old lady on her sleeper ... the sort of the sleeper at night' (GP-3). The following deserving case shows that in some contexts the normal aversion to offering the drugs to patients with a drink problem noted earlier can be overridden by some GPs sympathetic to the personal plight of some patients:

The commonest is people with those long-term life problems, bad unhappy marriages, maybe have a partner who's a drinker, has a drinking problem themselves. It's hard to generalize but, yeah, you know, domestic violence, dysfunctional relationships, whatever. People end up on them and stay on them. (GP-2)

Notwithstanding this personal exception about drink problems, dichoto-mization into deserving and undeserving patients was common in the accounts. Restricted use meant that these drugs were reserved for special circumstances and particular deserving patients, reversing the old image of the drugs being handed out indiscriminately for people with psychosocial problems. However, for some GPs the prescription acted as a buffer against dealing with the complexities of the latter. This respondent recognized this temptation and was critical: 'Sometimes I think short-term use just puts off confronting what the real issues are either for me or even for the patient' (GP-2). Thus, patients with overwhelming life circumstances were viewed as legitimate recipients of the benefits of tranquillizers in the absence of an alternative. Despite identifying the restricted circumstances under which prescriptions should have been offered, extreme life situations tempt GPs to issue the drugs as a means of rescuing distressed patients. The psychosocial needs of patients and the demands these placed on GPs meant that they, like their patients, felt like they had limited options.

They're unemployed ... they're people who have the sort of common ghastly life circumstances that just keep on coming, you know, it not just previous life events and one day they'll get over it, they just seem to keep on coming. Because of their situation in life they are still linked to very dysfunctional families composed of very dysfunctional people. (GP-9)

... my view is that in the 80s doctors were feeling just as they do now, that many of the emotional problems patients come with, they're relatively poorly equipped to deal with and if somebody comes along and says, 'Here's a good drug' and the patients like it, then people are tempted to use it because it treats our own pain as well as our patients' pain, 'cos we want to help people and make people feel better. So if we give people something and make them feel better, then everybody seems to be happier. (GP-5)

Occasionally I'll prescribe benzodiazepines to drug users you know. I don't think I should be absolutist about it. But I also know that I won't get any support from the drug services if I do that! (GP-20)

Thus, whether GPs adopt a firm or liberal prescribing policy, what most agree on is that the very need for psychotropic medication for common mental health problems cannot be isolated as a clinical need in the skin-encapsulated patient in the consulting room. Instead, there is recognition that life circumstances, past and present, have a pervasive impact on the affective and cognitive presentations of distressed patients.

Implications for general practice hereon

Some of the respondents offered a view about the implications of the ben-zodiazepine problem for future medical practice. One or two (exceptionally) were not as concerned as their colleagues, because they were tolerant of iatrogenic drug dependence; they saw it as part of the territory of medical practice in difficult circumstances:

I believe that the controversy has been blown up slightly out of proportion by pharmacological and medical purists who for some moral reason disapprove of drug dependency. There have been some overdoses - some people have died from taking these drugs, but by and large the drugs are harmless. (GP-21)

I think in some ways it has gone too far the other way in that GPs these days can be too afraid to use [a benzodiazepine]. I mean if the worst thing you can say about these drugs is that they have a potential for dependency, I don't think that is a good enough reason to withdraw. (GP-16)

While these were 'out-liers' in the accounts, they may also give some indication about why appeals to practitioners over the past 20 years to eliminate the prolonged use of benzodiazepines have not led to the prescribing habit withering away. The fact, noted earlier, that levels of prescribing are a quality indicator used by English Primary Care Trusts in 2005 suggests that many GPs are considered to be still deviating from expectations of best practice. They, like their patients, are resistant to change about benzodiazepine use.

Some respondents emphasized that the benzodiazepines were by no means unique as a clinical challenge. For example, a number of currently favoured drugs were felt to be similarly vulnerable in relation to iatrogenic dependence. Codeine-based analgesics, the anti-obesity drugs and Zopiclone (the 'z' drugs, a newer form of night sedation) were specifically noted and a general point was made by this respondent: 'Every drug has that potential. There isn't a drug that hasn't got potential. Every drug that I use, I always think of the possibility of dependency. Sometimes it is a case of which is the worse of two evils' (GP-14). This point was reinforced by another respondent with particular reference to the SSRIs. These are the newer 'antidepressant' drugs, which are used with patients who previously would have been given a benzodiazepine or an older 'antidepressant'. The latter were effective but toxic. A parallel is evident then between these two generations of 'antidepressants' and the problems with the previous cohort of anxiolytic prescribing. When the new 'antidepressants' were launched, they were considered to be non-addictive. However, this practitioner notes that: 'The SSRIs are very effective but there is a big problem with Seroxat. I have found for two or three years it is very difficult to get people off - and Prozac to some extent as well' (GP-12). This point about placing the benzodiazepines in the context of other psychotropic drug use leads to the next section of the article.

Out with the old and in with the new?

GP authority was undermined by legal and social campaigns of patients and pressure groups about the addictive effects of this class of drugs. The ready acceptance of blame two decades later by GPs in this study (including even younger doctors who could logically distance themselves from past) was evident. This is a departure from the victim blaming of patients, when they are held responsible for untoward behaviour, though residual aspects of victim blaming were evident in the construction of a division between worthy and

unworthy patients. However, as we have noted earlier, the classification of patients into good and bad marked a departure from previous sociological analyses which have emphasized the notion of good or bad patients aligned predominantly in relation to access to the sick role being seen as 'worthy' or 'unworthy' (Stimson and Webb, 1975; Jeffries, 1979).

One possible professional function of focusing on discredited psycho-tropic agents is that currently preferred drugs are compared favourably. It is commonplace for the current generation of GPs to justify their present prescribing norms and use the past inadequacies as a reference point of legitimacy. This pattern or cycle of prescribing legitimacy is not unique to the benzodiazepines, as we noted in the first section of this article. For example, benzodiazepines compared favourably with the more toxic barbiturates.

More recently, another generation of psychotropic drugs can demonstrate the same point. The SSRIs alluded to earlier were hailed as more effective and less toxic and dependency forming than the older 'antidepressants'. However, evidence is now coming to light that the drugs are dependency forming. They have also been linked to an increased risk of suicidal and homicidal action (Healy, 1997).

A second example can be given in relation to 'anti-psychotic' medication. The older 'major tranquillizers' or neuroleptic drugs were superseded by the newer 'atypical anti-psychotics' during the 1990s. These, more expensive, drugs were immediately deemed to be more effective in symptom control and were championed for their reduced adverse effects. Moreover, the dangers of the older 'anti-psychotics', particularly of movement disorders, acute dysphoria (dramatic drop in mood) and akathisia (inner restlessness) were now acknowledged with concern, whereas previously they had been minimized or ignored by clinicians (Brown and Funk, 1986). However, with increasing clinical use, problems have come to the fore about the peculiar dangers of the 'atypicals' (iatrogenic blood disorders and cardio-toxicity).

Comparisons with these other drugs demonstrate that rather than being an exception, the benzodiazepines are the rule because psychotropic agents are inherently crude or 'dirty' drugs (Fisher and Greenberg, 1997). Any drug that has a powerful impact on the brain is likely to have extensive physiological, experiential and behavioural effects, because they tend to have a 'blunderbuss' rather than 'sniper' impact.

Some of this blunderbuss effect will have adverse consequences for patients in the form of 'side-effects' (a term misleadingly implying a marginal or minimal outcome), which may be more dominant in their personal significance than any therapeutic benefit. In the case of psychotropic agents, they invariably cause some degree of temporary functional impairment and they even may cause permanent structural changes to the major organs of the body (Breggin, 1993).

Moreover, all of the classes of psychotropic drug invented and marketed since the purported 'pharmacological revolution' of the 1950s have tended to be portrayed as magic bullets. Terms such as 'anti-depressant', 'anti-anxiety'

or 'anti-psychotic' imply targeted action with a scientific rationale, whereas most psychotropic agents have been discovered opportunistically, often driven by market forces exploited by the pharmaceutical industry and crudely prescribed. It is common for the classes of drugs just cited to be given to patients with a diagnosis other than that indicated. For example, 'antipsychotics' are given to anxious patients, anxiolytics are used for acute psychotic agitation and 'anti-depressants' are given to anxious patients.

According to some commentators (Moncrieff and Crawford, 2001; Moncrieff et al., 2005) the psychiatric profession has preferred reified descriptions of drug action, such as 'anti-depressant' because they justify its jurisdiction over mental disorder and legitimize dubious diagnostic categories. However, it is less clear, at this stage, whether GPs will follow this professional strategy, as they take more and more responsibility for mental health problems in the community. The pragmatic need to respond to the range of psychosocial features of distress and madness with biomedical treatments connects all types of psychotropic drugs. A bio-medical response to distress and madness will inevitably and paradoxically be both inadequate and yet justifiable within a societal norm of psychosocial problems being presented for amelioration or resolution to medical experts. In these difficult circumstances, prescribers will operate their own version of situated rationality. Because of their clinical autonomy, GPs may both share and constitute clinical norms on the one hand and differ from one another at times on the other (as was evident in the data presented).

Given that all psychotropic agents (prescribed or recreational, legal or illicit) expose their recipients to biological, psychological and social risks, value judgements are made and applied, by those with the power to do so, to divide 'right' from 'wrong' drugs. An obvious aspect of this is the one mentioned earlier about past-prescribed (bad) and present-prescribed (good). Another is prescribed (good) and non-prescribed (bad). Another is legal (good) and illegal (bad). The social-cognitive reallocation of a psychotropic agent from one category to another disrupts existing norms and so creates a risk to norm maintenance. Examples here are the controversies about the medicinal use of cannabis and the sale of prescribed drugs (like methadone and the benzodiazepines) on the streets.

Thus the end of this article brings us up against a wider sociological consideration, which is beyond its scope to explore fully - the role of GPs as risk managers on behalf of a moral order, which is now a 'risk society' (Beck, 1992) and preoccupied by the 'politics of anxiety' (Turner, 1991). The arena of contention about drug use and the semantics it contains reveals a good/bad dichotomy reflecting this context of risk management (Swartz et al., 1998). For example, do people 'use', 'abuse' or 'misuse' 'substances'? At what point is alcohol a 'bad' rather than a 'good' habit? Why is the use of methadone by opiate addicts good but the use of heroin bad? These questions suggest that the moral, logical or empirical isolation of particular drugs (like benzodiazepines) is not warranted. Moreover, attempts at such

isolation may obscure our understanding of the precarious state of norm maintenance around the human intake of substances, past or present, prescribed or not, legal or illegal, that affect thoughts, feelings and actions.


This article started by exploring what appeared to be a straightforward assumption - that one group of psychotropic drugs, the benzodiazepines, requires special considerations about clinical risk management in primary care, because they are an unusual example of an iatrogenic legacy from primary care colleagues. The study of GP views began to problematize this assumption in a number of ways. This was a qualitative study intended to illuminate ways of thinking in a particular group of GPs - as it turned out a self-selected group who were not always directly responsible for prescribing benzodiazepines in the 1980s. However, within this context, most seemed to take the clinical risks seriously and own the responsibility for current good practice, as well as for the rectification of inherited clinical iatrogenesis.

In this respect, there appears to have been a cultural shift in the way in which GPs have viewed the risks associated with benzodiazepine use. Giddens (1991) and Beck (1992) have both asserted that our recent 'risk society' is associated with a greater consciousness of risk than in the past. Although since the 1990s GPs have acknowledged the problem of over-prescribing as a multifactorial one with complex and psychological roots (Hamilton et al., 1990), GP narratives from this study problematized their professional role and risks of a technological problem more than their predecessors. Based on past behaviour the latter were (until the crisis over prescribing was brought to a head) more inclined to dismiss or underrate the risk that arose from what was seen as an innovative and welcome development in science and technology.

Notwithstanding this shift in perspective about the risks of benzodi-azepines, their prescription has not become a total taboo in primary care. In common with the findings of the study by Parr et al. (2006) undertaken in an Australian context, British GPs were knowledgeable and skilled in describing steps required for a successful cessation of benzodiazepine use. None the less, not only did the majority of the respondents plead for the 'deserving' patient, a few even minimized the problem of iatrogenic addiction, feeling that 'medical purists' had overdrawn the clinical and social problems associated with the drugs. This tolerance of addiction in turn raises some broader questions about whether benzodiazepines are that unusual. It may also account for why health service managers still struggle to bring the problem of benzodiazepine use under acceptable local political control.

Comparisons with other drugs (prescribed or self-administered) suggest that the benzodiazepines are not the exception but the rule in a number of ways. They, like other psychotropic drugs, are rather crude and dependency forming and are compared unfavourably with past alternatives, which are

now discredited. Benzodiazepines, like other drugs, are part of a set of impermanent dichotomizations. Forty-five years ago benzodiazepines were generally 'good' but now they are generally 'bad' (with some exceptions pointed up earlier). Such divisions could be pointed up about all of the other classes of prescribed psychotropic agents.

Moreover, the scandal surrounding the benzodiazepines and highlighted by the mid-1980s in the UK was much delayed, when the evidence about them is viewed with hindsight. Given that the first lawsuit about them appeared in 1967, it could be argued that undergraduate and postgraduate medical education was in a position to alter radically the trajectory of prescription levels, as early as 1970. But this did not occur.

Again the benzodiazepines are not unique in this regard. The same point about over-use and under-criticism can be found in the cases of 'anti-depressants' and 'anti-psychotics'. In their prescribed form, they are all utilized in a bio-medical arena in which the individual manifestations of the psychosocial causes and consequences of distress and madness are risk-managed. The enormity and sometimes futility of this task ensure that inadequate and imperfect medical measures are used in response to mental health problems. They also ensure that corrective criticism of such inadequacies and imperfections is slow to develop and difficult to implement.


Baldessarini, R.J. (1999). Psychopharmacology. In A.M. Nicholi (Ed.), The Harvard

guide to psychiatry, pp. 76 - 98. Harvard: Belknap. Beck, U. (1992). The risk society: Towards a new modernity. London: Sage. Breggin, P. (1993). Toxic psychiatry. London: HarperCollins. Britten, N., Stevenson, F., Gafaranga, J., Barry, C. and Bradley, C. (2004). The expression of aversion to medicines in general practice consultations. Social Science & Medicine,59(7), 1495 - 503. Brown, P. and Funk, S.C. (1986). Tardive dyskinesia: Barriers to the professional recognition of iatrogenic disease. Journal of Health and Social Behavior, 27(2), 116 - 32.

Bury, M. and Gabe, J. (1990). Hooked? Media responses to tranquillizer

dependence. In P. Abbott and G. Payne (Eds.), New directions in the sociology of health, pp. 87 - 103. London: Falmer Press. Cohen, S. (1983). Current attitudes about the benzodiazepines: Trial by media.

Journal of Psychoactive Drugs, 15(1-2), 109 - 13. Committee on the Safety of Medicines. (1988). Benzodiazepines, dependence and

withdrawal symptoms. Current Problems, 21. Cooperstock, R. (1978). Sex differences in psychotropic drug use. Social Science

and Medicine, 1(12), 179 - 86. Fisher, S. and Greenberg, R.P., Eds. (1997). From placebo to panacea: Putting

psychiatric drugs to the test. New York: Wiley. Gabe, J. and Thorogood, N. (1986). Prescribed drug use and the management of everyday life: The experience of white working class women. Sociological Review, 34(4), 737 -72.

Giddens, A. (1991). Modernity and self identity: Self and society in the late modern age. Cambridge: Polity.

Hamilton, I.J., Reay, L.M. and Sullivan, F.M. (1990). A survey of general practitioners' attitudes to benzodiazepine overprescribing. Health Bulletin, 48(6), 299 - 303.

Healy, D. (1997). The anti-depressant era. London: Harvard University Press.

Jeffries, R. (1979). 'Normal rubbish': Deviant patients in casualty departments. Sociology of Health and Illness, 1(1), 90 - 109.

Lacey, R. (1991). The MIND complete guide to psychiatric drugs. London: MIND.

Lader, M., Beaumont, G. and Bond, A. (1992). Guidelines for the management of patients with generalized anxiety. Psychiatric Bulletin, 16,560 - 5.

Lester, H.E., Tritter, J.Q. and Sorohan, H. (2005). Patients' and health professionals' views on primary care for people with serious mental illness: Focus group study. British Medical Journal, 330,1126 - 1126B.

May, C., Allison, G., Chapple, A., Chew-Graham, C., Dixon, C., Gask, L. et al. (2004). Framing the doctor-patient relationship in chronic illness: A comparative study of general practitioners' accounts. Sociology of Health & Illness, 26(2), 135 - 58.

Medawar, C. (1992). Power and dependence. London: Social Audit.

Moncrieff, J. and Crawford, M.J. (2001). British psychiatry in the 20th century: Observations from a psychiatric journal. Social Science & Medicine,53(3), 349 - 56.

Moncrieff, J., Hopker, S. and Thomas, P. (2005). Psychiatry and the pharmaceutical industry. Psychiatric Bulletin, 29,84 -5.

North, D., Davis, P. and Power, A. (1995). Patient responses to benzodiazepine medication: A typology of adaptive repertoires developed by long-term users. Sociology of Health & Illness, 17(5), 632 -50.

Olson, M. and Pincus, H. (1994). Use of benzodiazepines in the community. Archives of Internal Medicine, 154(11), 1235 - 40.

Parr, J.M., Kavanagh, D.J., Young, R.M. and McCafferty, K. (2006). Views of general practitioners and benzodiazepine users on benzodiazepines: A qualitative analysis. Social Science & Medicine, 62(5), 1237 - 44.

Rogers, A. and Pilgrim D. (2001). Mental health policy in Britain. Basingstoke: Palgrave.

Stevenson, F.A., Britten, N., Barry, C.A., Bardley, C.P. and Barber, N. (2002). Perceptions of legitimacy: The influence on medicine taking and prescribing. health: 6(1), 85- 104.

Stimson, G. and Webb, B. (1975). Going to see the doctor: The consultation process in general practice. London: Routledge & Kegan Paul.

Swartz, M.S., Swanson, J., Hiday, V.A., Borum, R., Wagner, R. and Burns, B.J. (1998). Taking the wrong drugs: The role of substance abuse and medication non-compliance in violence among severely mentally ill individuals. Social Psychiatry and Psychiatric Epidemiology, 33(Supp. 1), S75 - 80.

Turner, B.S. (1991). Recent developments in the theory of the body. In M.

Featherstone, M. Hepworth and B.S. Turner (Eds.), The body: Social processes and cultural theory, pp. 1 - 35. London: Sage.

Tyrer, P. (1987). Benefits and risks of benzodiazepines. Proceedings of the Royal Society of Medicines, 114,7 - 11.

Tyrer, P., Harrison-Read, P. and van Horn, E. (1997). Drug treatment in psychiatry. London: Butterworth Heineman.

Author biographies

anne rogers is Professor of the Sociology of Health Care in the Division of Primary Care, School of Medicine, University of Manchester. She has a keen research interest in the Sociology of Mental Health and Illness.

david pilgrim is Clinical Dean, for the Teaching Primary Care Trust for East Lancashire and Honorary Professor at the Universities of Liverpool and Central Lancashire. He is a Clinical Psychologist and has clinical and research interests in many aspects of mental health work.

susie brennan is a medical student at the University of Manchester.

ilyas sulaiman is a medical student at the University of Manchester.

gareth watson is a medical student at the University of Manchester.

carolyn chew-graham is Senior Lecturer in General Practice at the University of Manchester. She has research interests in the areas of professional practice, exploring health professionals' attitudes to their work with patients with mental health problems, back pain and medically unexplained symptoms.