“Do unto others as…” - Which treatments do psychiatrists prefer? Results from a national survey.

  M Taylor¹, and T Brown². 

¹ Springpark Centre, Glasgow G22 5EU, UK

² Western Infirmary, Glasgow G11 6NT, UK 

Correspondence: Dr Mark Taylor, Springpark Centre, Glasgow G22 5EU, UK .  Email:  mark.taylor@glacomen.scot.nhs.uk

SMJ 2006 52(1): 17-19

Abstract:

We undertook an independent national survey of psychiatrists’ treatment preferences should they become mentally ill. The response rate was 59% from 921 individuals. For psychosis, atypical antipsychotics were generally favoured with risperidone receiving most votes. Psychotherapy and antidepressants were both endorsed as treatments for mild-to-moderate depression, and citalopram; fluoxetine; and venlafaxine were the three preferred antidepressants. ECT received the backing of a large majority of psychiatrists, particularly for severe mood disorder.

Introduction

The Mental Health (Care and Treatment) (Scotland) Act 2003 allows individuals suffering from mental disorder to specify their treatment preferences during future episodes of illness via an Advance Statement.¹ An Advance Statement can only be drawn up by a capable adult, and implies that the individual concerned has some understanding of the risks and benefits of the treatments involved.

We sought to determine the treatment preferences of psychiatrists based in Scotland should they become mentally ill. We argue that this collective real world expert opinion is a powerful form of evidence, and that it complements the narrow rigour of the randomised controlled trials and meta-analytic data by synthesising that data through years of clinical experience.

Method 

The UK Royal College of Psychiatrists (Scottish Division) mailing list of members and fellows who were either employed or participating in continuing professional development in Scotland contained 925 names. These individuals are all experienced psychiatrists, and either fully trained specialists (consultants) or senior trainees. All 925 psychiatrists were sent a standardised questionnaire. All responses were anonymous, but the respondent’s age, gender, seniority, and place of work were recorded.

The questionnaire required the psychiatrist to specify which one antipsychotic medication they would take if they became psychotic; and whether they would prefer psychotherapy (eg cognitive behavioural or interpersonal therapy) or antidepressants (or both) should they suffer from a mild to moderate episode of depression, as well as which one antidepressant they would opt for if they required medication. In both scenarios the psychiatrist could indicate whether their choice of medication was due to the relative efficacy or the side effect profile, or both.

Additionally, psychiatrists were asked if they would consent to electro-convulsive therapy (ECT) and if so, for which clinical condition. Lastly the psychiatrists were asked if they would ever consent to neurosurgery for mental disorder (NMD).

Results

The response rate to the questionnaire was 59%, ie 544 psychiatrists replied from a possible total of 921 individuals, with a legal executor replying that one psychiatrist was dead and hence would not be responding, whilst three retired psychiatrists felt unable to provide answers.

69% of the psychiatrists who responded were consultants (fully trained specialists); with 51% of the total being male. There was an even distribution of age range, with 10% aged between 20-30; 34% between 31 and 40; 32% between 41 and 50; and 24% over 51 years old. Geographic distribution of responses followed staffing patterns across the country, with 50% of all responses being received from the main urban (Glasgow and Edinburgh) regions.  

Table 1: Antipsychotic preferences, and relative reasons for choice 

There was no dramatic difference in pattern of anti-psychotic choice between consultants and other grades of psychiatrist, or between men and women. As can be seen from Table 1, most choices were based on relative side effect profile, rather than relative efficacy with only chlorpromazine, clozapine, and olanzapine having efficacy more frequently specified than side-effect profile.

With regard to treatment for their own mild-to-moderate depression, 34% of psychiatrists preferred only psychotherapy as their treatment, 41% preferred only antidepressant medication, and 25% indicated they would prefer both.

Table 2: Antidepressant preferences, and relative reasons for choice 

Where antidepressant medication had to be rated, neither gender nor seniority affected antidepressant choice. Of the favourite three antidepressants, citalopram was viewed as having best side effect profile; venlafaxine as the most efficacious; with fluoxetine having roughly equal votes for efficacy and side effects, as indicated in Table 2.

85% of all psychiatrists indicated they would accept ECT, with 15% of men and 13% of women stating that they would never consent to ECT. Treatment resistant depression and depressive psychosis were the two most preferred indications for ECT, and most (69%) female psychiatrists indicated they would consent to ECT for post-partum psychosis, whilst male respondents suggested this was not relevant.

The majority of psychiatrists who responded (60%) indicated they could not imagine any clinical scenario where they would consent to neurosurgery for mental disorder (NMD), with 4% not expressing a view.

Discussion 

We have undertaken the first independent national survey of psychiatrists’ preferred treatment choices, should they become mentally unwell. In this age of evidence based medicine, treatment choices are based not only on scientific data and expert guidelines, but also personal clinical experience, peer opinion, and marketing influence. We believe the data in this type of study offers a unique perspective on the real world practice of psychiatry, although the results only have a finite shelf life as new treatments emerge.

Before discussing our results we should consider some limitations to this work. Although we have opinion from across a range of ages and localities, with a balance of gender and career grade, it is possible our results are not representative of Scotland or the UK generally. A response rate of 59% is however better than many postal surveys.² The brief standardised questionnaire forced the respondent into specific listed choices, and hence if an option was not listed but preferred this could have acted as a disincentive to completion. Numerous responses contained freehand comment regarding the survey, for example “..absence of a depot (antipsychotic) is a fatal flaw…”

Debate has continued over whether intra-class differences in efficacy exist between atypicals^3,4. In this survey, the first atypical to be licensed in the UK, risperidone, was clearly the preferred antipsychotic. It is worth observing that aripiprazole was only launched some four months before the commencement of this survey. Our results demonstrate that atypicals as a whole command widespread confidence, perhaps reflecting NICE recommendations⁵, and although most choices were based on side effect profile important differences in both efficacy and side-effect profile were felt to exist. Interestingly in a survey of Scottish patients⁶ who had been treated in the mental health services (with 756 replies returned) sulpiride was the preferred antipsychotic medication. Data⁷ on national prescription rates for Scotland in 2003-4, in both primary and secondary care, revealed that only chlorpromazine, risperidone, and olanzapine were prescribed over 20% by volume. Thus there would appear to be differences between what the psychiatrists say and do, although the national data includes family physician or general practitioner prescriptions. Also interesting is that patients would appear to value a relatively non-sedative antipsychotic medication (sulpiride) which was not specifically included on our list although amisulpride is a similar compound.

With regard to the choice of treatment for mild-to-moderate depressive episode, there was a comparatively even split of opinion between psychotherapy and antidepressant medication, and perhaps surprisingly a minority of psychiatrists voted for a combination approach. One study⁸ suggested both generic counselling and antidepressants were effective in primary care for mild-to-moderate depression.

Amongst the preferred antidepressants there appeared an inverse correlation between relatively benign side effects (citalopram) and the perception of added efficacy (venlafaxine). Fluoxetine appeared midway between these two poles, whereas sertraline scored in a similar pattern to citalopram. The Scottish patient survey⁶ listed trazodone as the patients’ preferred antidepressant, and the national prescription rate data⁷ showed that only amitriptyline, citalopram, and fluoxetine were prescribed at over 16% by volume in Scotland during 2003-4. The patients’ preferred antidepressant medication, trazodone, was not on our limited list of choices, and is often employed in part because it provides nocturnal sedation which of course may be desirable during a depressive episode. This may suggest that psychiatrists should consider asking patients specifically if they would prefer a medication which confers nocturnal sedation.

One of the most controversial treatments in psychiatry is ECT, but here an overwhelming majority of expert opinion was favour of its use, particularly for severe mood disorder and post-partum psychosis in women. No data on why it was rejected by 15% of all psychiatrists was available.

Scotland has a specialist centre for neurosurgery for mental disorder. Although it is perhaps difficult to foresee a clinical scenario where all other treatment options have been exhausted, it is possibly surprising that only a minority (36%) of psychiatrists felt they would consent to this low-frequency intervention.

In summary, psychiatrists in Scotland are clearly able to state their own treatment preferences, and genuine differences in both side effect profile and relative efficacy are felt to exist between amongst the listed antipsychotics, and amongst the listed antidepressant medications. Psychotherapy and ECT are both valued as treatment options in mood disorder.

Acknowledgements

Thanks to all the psychiatrists who responded, and to K Addie and A Celini.

References 

1. www.nes.scot.nhs.uk/mha

2. Personal communication, Ben Page, MORI research, UK.

3. Davis J, Chen N, and Glick I. A meta-analysis of the efficacy of second generation antipsychotics.  Arch Gen Psychiatry. 2003; 60: 553-564.

4. Lieberman J, Stroup TS, McEvoy JP, et al. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. NEJM. 2005; 353; 12. 1209-22.

5. www.nice.org.uk (No. 43)

6. Bradstreet S and Norris R, eds. All you need to know? A Scottish survey of people’s experience of psychiatric drugs. www.samh.org.uk 2004.

7. Gold B Prescribing and dispensing in primary healthcare. http://www.isdscotland.org 2004

8. Chilvers C, Dewy M, Fielding K, et al. Counselling versus antidepressants in Primary Care Study Group. BMJ. 2001; 322 (7289): 772-5

Back to February Contents