Intended for healthcare professionals

Letters

Prescribing antidepressants in general practice

BMJ 1997; 314 doi: https://doi.org/10.1136/bmj.314.7083.826 (Published 15 March 1997) Cite this as: BMJ 1997;314:826
  1. David Fish, West Midlands Mental Health Education fellow (1993-6)a
  1. a Henley Green Centre, Coventry CV2 1AB
  2. b Three Swans Surgery, Salisbury SP1 1DX
  3. c Cochrane Depression Group, Institute of Health Sciences, University of Oxford, PO Box 777, Oxford OX3 7LF
  4. d Geriatrics and Extended Care, Department of Veterans Affairs Medical Center/Texas Tech University, 6010 Amarillo Boulevard, West Amarillo, TX 79106, USA
  5. e Gwent Community Health NHS Trust, County Hospital (Griffithstown), Pontypool, Gwent NP4 5YA
  6. f Pain Management Clinic, South Cleveland Hospital, Middlesbrough TS4 3BW
  7. g Grampian Health Board, Aberdeen AB9 1RE
  8. h Fitznells Manor Surgery, Ewell, Surrey
  9. i NHS Centre for Reviews and Dissemination, University of York, York YO1 5DD
  10. j Leeds General Infirmary, Leeds LS1 3EX
  11. k Royal Edinburgh Hospital, Edinburgh EH10 5HF
  12. l Dingleton Hospital, Melrose, Roxburghshire
  13. m Hairmyres Hospital, East Kilbride G75 8RG
  14. n West Norwich City Community Team, Norfolk Mental Health Care NHS Trust, Norwich NR6 5BE

    What is an effective dose?

    Editor-What should be the advised dosage of a tricyclic antidepressant for major depression? In his editorial Tony Kendrick quotes a recommended dose of a typical tricyclic antidepressant, amitriptyline, as 125 mg daily.1 This dosage is at the higher end of the range of dosages quoted for all tricyclic antidepressants in the British National Formulary. If followed it would have major implications for the prescribing habits of general practitioners.

    The recommended dosage for amitriptyline for major depression is quoted from a consensus statement published in the BMJ in 1992.2 The source of this recommendation is a double blind placebo controlled trial of amitriptyline among depressed patients in general practice.3 In this trial patients given the active drug received amitriptyline 75 mg daily by the end of the first week, amitriptyline 100 mg during the second week, and amitriptyline 125-175 mg, depending on improvement, during the remaining four weeks of the study. The median dose of amitriptyline given over the six week study period was 125 mg. It is the median tolerated dose. Figure 1 in the paper by Hollyman et al clearly shows a response to treatment in the first two weeks, when the lower dose of amitriptyline was given.3

    Kendrick discusses the results of several investigations into the effectiveness of antidepressants at different dosages. Similar investigations have been used to develop the dose range of every prescribable antidepressant–for example, Kerihuel and Dreyfus reviewed 34 randomised clinical trials of lofepramine.4 In the papers reviewed the concentration of lofepramine giving a clinical response ranged from 105 mg to 210 mg, which is equivalent to 75-150 mg amitriptyline.

    The principle that more severe cases of major depression require higher doses of antidepressants is also valid. It also holds true for newer selective serotonin reuptake inhibitors. For example, a study of prescribing patterns across 23 381 prescriptions shows that a prescription for higher than the initial dose was required in 6.8% of patients taking fluoxetine, 21.1% of patients taking paroxetine, and 51% of patients taking sertraline.5

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.

    More on what is an effective dose

    1. M V Moore, General practitionerb
    1. a Henley Green Centre, Coventry CV2 1AB
    2. b Three Swans Surgery, Salisbury SP1 1DX
    3. c Cochrane Depression Group, Institute of Health Sciences, University of Oxford, PO Box 777, Oxford OX3 7LF
    4. d Geriatrics and Extended Care, Department of Veterans Affairs Medical Center/Texas Tech University, 6010 Amarillo Boulevard, West Amarillo, TX 79106, USA
    5. e Gwent Community Health NHS Trust, County Hospital (Griffithstown), Pontypool, Gwent NP4 5YA
    6. f Pain Management Clinic, South Cleveland Hospital, Middlesbrough TS4 3BW
    7. g Grampian Health Board, Aberdeen AB9 1RE
    8. h Fitznells Manor Surgery, Ewell, Surrey
    9. i NHS Centre for Reviews and Dissemination, University of York, York YO1 5DD
    10. j Leeds General Infirmary, Leeds LS1 3EX
    11. k Royal Edinburgh Hospital, Edinburgh EH10 5HF
    12. l Dingleton Hospital, Melrose, Roxburghshire
    13. m Hairmyres Hospital, East Kilbride G75 8RG
    14. n West Norwich City Community Team, Norfolk Mental Health Care NHS Trust, Norwich NR6 5BE

      Editor-I think that there is some misconception about the effective doses of tricyclic antidepressants in primary care. The consensus guidelines state that doses between 125 mg and 150 mg have been shown to be effective and doses of 75 mg have not.1 This statement is widely interpreted to mean that 125 mg or more is the recommended dose and that the minimum effective dose is 125 mg. Much has been made of general practitioners failure to prescribe at recommended doses.2 3 What is the evidence that this interpretation of the guidelines is justified?

      In none of the three studies cited in the consensus guidelines was there any attempt at dose titration. Amitriptyline was compared with either placebo or substantially lower doses. In one of these studies, a double blind placebo controlled trial in general practice, the final dose of amitriptyline achieved in the treatment group was a median of 125 mg and a mean of 119 mg.4 In other words, at least half of the treatment group were receiving doses of less than those recommended. Doses based on how many tablets were returned were lower still: 113 mg and 111 mg at four and six weeks respectively. The same study data reported elsewhere show that significant differences between the study groups were present after only two weeks, when treated patients were taking only 100 mg amitriptyline. Indeed, the authors concluded that their study had not provided evidence about the optimum dose of amitriptyline.4

      In conclusion, I suggest that it is unjustified to assume that doses of tricyclic antidepressants below 125 mg of amitriptyline or equivalent are not effective in primary care. I would echo the sentiments of Tony Kendrick, who questions whether general practitioners are really wrong to prescribe at lower doses.5 I welcome the news of further research into effectiveness of low dose tricyclic antidepressants in primary care.5

      References

      1. 1.
      2. 2.
      3. 3.
      4. 4.
      5. 5.

      Systematic review of all pertinent trials is required to establish guidelines

      1. David Gill, Joint editor (david.gill{at}psychiatry.oxford.ac.uk)c
      1. a Henley Green Centre, Coventry CV2 1AB
      2. b Three Swans Surgery, Salisbury SP1 1DX
      3. c Cochrane Depression Group, Institute of Health Sciences, University of Oxford, PO Box 777, Oxford OX3 7LF
      4. d Geriatrics and Extended Care, Department of Veterans Affairs Medical Center/Texas Tech University, 6010 Amarillo Boulevard, West Amarillo, TX 79106, USA
      5. e Gwent Community Health NHS Trust, County Hospital (Griffithstown), Pontypool, Gwent NP4 5YA
      6. f Pain Management Clinic, South Cleveland Hospital, Middlesbrough TS4 3BW
      7. g Grampian Health Board, Aberdeen AB9 1RE
      8. h Fitznells Manor Surgery, Ewell, Surrey
      9. i NHS Centre for Reviews and Dissemination, University of York, York YO1 5DD
      10. j Leeds General Infirmary, Leeds LS1 3EX
      11. k Royal Edinburgh Hospital, Edinburgh EH10 5HF
      12. l Dingleton Hospital, Melrose, Roxburghshire
      13. m Hairmyres Hospital, East Kilbride G75 8RG
      14. n West Norwich City Community Team, Norfolk Mental Health Care NHS Trust, Norwich NR6 5BE

        Editor-“Watchful waiting for minor depression, full dose treatment for major depression” is a fairly catchy subtitle to Tony Kendrick's editorial,1 but it is questionable how far it is based on evidence.

        Depression is continuously, not dichotomously distributed in the population, with no point of rarity between major depression and other forms. Intrinsically, non-major depression is likely to respond to antidepressants to some extent if major depression does. Such a view gains no support from evidence about the treatment of other continuously distributed health problems such as pain or hypertension, and the trials quoted have conflicting results.

        If patients with mild depression benefit from antidepressants they may indeed be responding to a “placebo with side effects,” but other options desired by patients such as counselling seem to be ineffective and are likely to be more expensive. If general practitioners believe that they have to do something they know that the hypnotic and anxiolytic properties of antidepressants are often helpful in the mixed neurotic states seen in primary care, irrespective of any direct effect on mood.

        The various guidelines advocating high doses for major depression in general practice are based on evidence,2 but, again, general practitioners know that starting with a low dose is the only way that many patients will take tricyclic antidepressants, and many patients will have recovered or stopped treatment long before their general practitioner has been able to increase the dose anywhere near 150 mg or 225 mg daily. Selective serotonin reuptake inhibitors, another option, are more expensive, lack dose flexibility, and have not been shown to be superior in practice.

        Only a systematic review of all pertinent trials from all countries, published and unpublished, can answer these questions. This is one of the main objectives of the recently formed Cochrane Collaborative Review Group for Depression and Neurosis, which welcomes any help towards this aim. Premature guideline development should be avoided.

        References

        1. 1.
        2. 2.

        Low dose tricyclic antidepressants are effective in treating major depression

        1. Robert S Tan, Associate chief of staffd
        1. a Henley Green Centre, Coventry CV2 1AB
        2. b Three Swans Surgery, Salisbury SP1 1DX
        3. c Cochrane Depression Group, Institute of Health Sciences, University of Oxford, PO Box 777, Oxford OX3 7LF
        4. d Geriatrics and Extended Care, Department of Veterans Affairs Medical Center/Texas Tech University, 6010 Amarillo Boulevard, West Amarillo, TX 79106, USA
        5. e Gwent Community Health NHS Trust, County Hospital (Griffithstown), Pontypool, Gwent NP4 5YA
        6. f Pain Management Clinic, South Cleveland Hospital, Middlesbrough TS4 3BW
        7. g Grampian Health Board, Aberdeen AB9 1RE
        8. h Fitznells Manor Surgery, Ewell, Surrey
        9. i NHS Centre for Reviews and Dissemination, University of York, York YO1 5DD
        10. j Leeds General Infirmary, Leeds LS1 3EX
        11. k Royal Edinburgh Hospital, Edinburgh EH10 5HF
        12. l Dingleton Hospital, Melrose, Roxburghshire
        13. m Hairmyres Hospital, East Kilbride G75 8RG
        14. n West Norwich City Community Team, Norfolk Mental Health Care NHS Trust, Norwich NR6 5BE

          Editor-T M McDonald and colleagues commented that tricyclic antidepressants used in low doses were unlikely to be of benefit in treating major depression.1 I disagree on the basis of work that I did with colleagues at Hammersmith Hospital.2 We studied the effect of low dose lofepramine (a tricyclic antidepressant) in 46 depressed elderly people and found that one third of the recommended dose was sufficient to improve major depression. We also found that placebo was as effective in people with mild depression (P=0.04, 95% confidence interval 0.4 to 7.9). A similar study of another tricyclic antidepressant was conducted in America with low dose doxepin in 24 elderly patients.3 Low doses of doxepin (10-20 mg) were also effective in significantly reducing depressive symptoms.

          I believe that the biological half lives of tricyclic antidepressants do not correlate well with the physiological half lives. This might explain why low doses of these agents are effective. Currently, therapeutic ranges have been developed for amitriptyline and imipramine. In my experience the clinical effects of tricyclic antidepressants do not always parallel serum concentrations. MacDonald and colleagues found that 72% of subjects were given subtherapeutic dosages,1 but whether this group improved with the lowered dosages is not known.

          However, I concur that low doses of tricyclic antidepressants are safe as this was also observed in our study of low dose lofepramine.2 This is especially important in ambulatory elderly patients as they may not be able to tolerate higher dosages.4

          Tony Kendrick argues in his editorial that watchful waiting is sufficient for minor depression.5 We also found placebo to be as effective for people with minor depression. An attentive ear is often more effective than drugs. However, the statement that a full dose is warranted for major depression is challenged by the results of our study.2

          References

          1. 1.
          2. 2.
          3. 3.
          4. 4.
          5. 5.

          Prescription does not mean that major depression is being treated

          1. Imad M Ali, Senior registrar in psychiatrye
          1. a Henley Green Centre, Coventry CV2 1AB
          2. b Three Swans Surgery, Salisbury SP1 1DX
          3. c Cochrane Depression Group, Institute of Health Sciences, University of Oxford, PO Box 777, Oxford OX3 7LF
          4. d Geriatrics and Extended Care, Department of Veterans Affairs Medical Center/Texas Tech University, 6010 Amarillo Boulevard, West Amarillo, TX 79106, USA
          5. e Gwent Community Health NHS Trust, County Hospital (Griffithstown), Pontypool, Gwent NP4 5YA
          6. f Pain Management Clinic, South Cleveland Hospital, Middlesbrough TS4 3BW
          7. g Grampian Health Board, Aberdeen AB9 1RE
          8. h Fitznells Manor Surgery, Ewell, Surrey
          9. i NHS Centre for Reviews and Dissemination, University of York, York YO1 5DD
          10. j Leeds General Infirmary, Leeds LS1 3EX
          11. k Royal Edinburgh Hospital, Edinburgh EH10 5HF
          12. l Dingleton Hospital, Melrose, Roxburghshire
          13. m Hairmyres Hospital, East Kilbride G75 8RG
          14. n West Norwich City Community Team, Norfolk Mental Health Care NHS Trust, Norwich NR6 5BE

            Editor-The study of John Donoghue and colleagues raises the question why general practitioners continue to prescribe low dose tricyclic antidepressants.1

            They cite my study of people taking long term antidepressants, which showed that 62% of the 177 responders to a postal survey reported moderate to severe depressive symptoms at follow up; prescription of low dose tricyclic antidepressants was widespread even among this group.2 Patients frequently consulted their general practitioner with psychological complaints, which tends to refute the notion that low doses continue to be prescribed because of a belief in their efficacy. Furthermore, the suggestion that general practitioners avoid raising the dose because of side effects cannot be the whole explanation because the logical solution would be simply to prescribe an alternative antidepressant.

            Another explanation is that the general practitioner may have revised his or her original diagnosis. The diagnostic process in primary care takes place over several consultations with the accumulation of information which helps to gain a more complete formulation of the patient's problems. Factors becoming apparent later, such as marital and financial difficulties, social adversity, life events, and even a poor response to drug treatment, could influence the general practitioner to revise the original diagnosis to one of dysthymia, adjustment disorder, or social problems. This may induce a sense of therapeutic nihilism, in which symptoms are perceived as being understandable and appropriate in the given psychosocial circumstances and therefore as unresponsive to antidepressants.

            Of course some patients initially diagnosed as being depressed by their general practitioner may not have met standard criteria for major depression in the first place and therefore antidepressants would not be indicated. The problem, however, lies with patients who do meet criteria for major depression, which is treatable. Helping general practitioners to identify these patients and recognise that antidepressants are effective even when the depression is understandable has rightly been one of the aims of the Defeat Depression Campaign.

            Low dose tricyclic antidepressants probably continue to be prescribed because depression in primary care comprises a heterogeneous group of major and minor depression, adjustment and anxiety disorders, and dysthymia, in which low dose tricyclic antidepressants are prescribed for insomnia and anxiety while clarification of the diagnosis is awaited. Although major depression may be included in a general practitioner's differential diagnosis, it may be discarded later. The important thing is that patients with a diagnosis of depression from a general practitioner (without a concurrent standardised psychiatric assessment) may be flawed because the concept of depression in primary care is broad and the use of antidepressants in itself does not mean that major depression is being treated.

            References

            1. 1.
            2. 2.

            Tricyclic antidepressants are also used for relief of chronic pain

            1. K Milligan, Clinical directorf
            1. a Henley Green Centre, Coventry CV2 1AB
            2. b Three Swans Surgery, Salisbury SP1 1DX
            3. c Cochrane Depression Group, Institute of Health Sciences, University of Oxford, PO Box 777, Oxford OX3 7LF
            4. d Geriatrics and Extended Care, Department of Veterans Affairs Medical Center/Texas Tech University, 6010 Amarillo Boulevard, West Amarillo, TX 79106, USA
            5. e Gwent Community Health NHS Trust, County Hospital (Griffithstown), Pontypool, Gwent NP4 5YA
            6. f Pain Management Clinic, South Cleveland Hospital, Middlesbrough TS4 3BW
            7. g Grampian Health Board, Aberdeen AB9 1RE
            8. h Fitznells Manor Surgery, Ewell, Surrey
            9. i NHS Centre for Reviews and Dissemination, University of York, York YO1 5DD
            10. j Leeds General Infirmary, Leeds LS1 3EX
            11. k Royal Edinburgh Hospital, Edinburgh EH10 5HF
            12. l Dingleton Hospital, Melrose, Roxburghshire
            13. m Hairmyres Hospital, East Kilbride G75 8RG
            14. n West Norwich City Community Team, Norfolk Mental Health Care NHS Trust, Norwich NR6 5BE

              Editor-I consider the paper by T M MacDonald and colleagues and the accompanying editorial by Tony Kendrick to be misleading because they criticise general practitioners for prescribing low dose tricyclic antidepressants, claiming that there will be no therapeutic effect.1 2 This may be the case in the treatment of clinical depression, but it must be emphasised that many patients suffering chronic pain and mild depression are also prescribed these drugs for their therapeutic effect on neuropathic pain and their evident benefit on night time sedation. Any therapeutic action on mood is a bonus but is not the primary goal of treatment. Most of this group of patients will titrate the dose of drug against side effects (dry mouth and daytime sedation) and pain modification. Many patients will use these drugs long term after the drugs have been correctly titrated to minimise the above side effects and it has been explained to them that they are not addictive. Also, the newer (cleaner) antidepressant drugs seem to be less efficacious than the tricyclic antidepressants in this respect.

              Most pain management clinics prescribe low dose tricyclic antidepressants on a regular basis, and newly referred patients are now commonly already taking these drugs as part of the management of their chronic pain. I think that it would be a retrograde step if general practitioners took away the wrong message from the above articles1 2 and increased the drug dose for the wrong reasons or mistakenly switched to newer drugs.

              References

              1. 1.
              2. 2.

              Prescribing rates vary widely between practices

              1. Ann F Bisset, Senior registrar, department of public health medicineg,
              2. J P Reid, Medical prescribing adviserg
              1. a Henley Green Centre, Coventry CV2 1AB
              2. b Three Swans Surgery, Salisbury SP1 1DX
              3. c Cochrane Depression Group, Institute of Health Sciences, University of Oxford, PO Box 777, Oxford OX3 7LF
              4. d Geriatrics and Extended Care, Department of Veterans Affairs Medical Center/Texas Tech University, 6010 Amarillo Boulevard, West Amarillo, TX 79106, USA
              5. e Gwent Community Health NHS Trust, County Hospital (Griffithstown), Pontypool, Gwent NP4 5YA
              6. f Pain Management Clinic, South Cleveland Hospital, Middlesbrough TS4 3BW
              7. g Grampian Health Board, Aberdeen AB9 1RE
              8. h Fitznells Manor Surgery, Ewell, Surrey
              9. i NHS Centre for Reviews and Dissemination, University of York, York YO1 5DD
              10. j Leeds General Infirmary, Leeds LS1 3EX
              11. k Royal Edinburgh Hospital, Edinburgh EH10 5HF
              12. l Dingleton Hospital, Melrose, Roxburghshire
              13. m Hairmyres Hospital, East Kilbride G75 8RG
              14. n West Norwich City Community Team, Norfolk Mental Health Care NHS Trust, Norwich NR6 5BE

                Editor-The information on general practitioners' prescribing is interesting, but John Donoghue and colleagues give no information on the variation in prescribing between practices.1

                Grampian region had a population of 532 500 in 1994, with 340 general practitioners in 89 practices. Analysis of the prescribing rates per 100 patients by each practice for psychiatric drugs varied from 2.63 to 15.38 for hypnotics and tranquillisers and from 3.50 to 14.84 for antidepressants, whereas prescribing of antipsychotic drugs ranged from 0.36 to 3.81. These ranges were based on the number of drugs prescribed, and they therefore overestimate the numbers of patients treated when several drugs are given to one patient or when frequent prescriptions are given (perhaps because of perceived risk of suicide). Prescribing rates bore no correlation to the proportion of patients in the practice who lived in deprived areas (as measured by Carstairs deprivation category for postcode) or to the rate of referrals to inpatient or outpatient psychiatry. Only three of the 12 practices with the highest prescribing rates for antidepressants were in the city of Aberdeen: the others were in small rural and fishing villages.

                More research is needed on reasons for such a wide variation in the prescribing practice of general practitioners.

                References

                1. 1.

                Prescribing is short term and follow up poor

                1. Tim Johnson, General practitioner registrarh,
                2. Richard Martin, General practitionerh,
                3. Jonathan Morrell, General practitionerh
                1. a Henley Green Centre, Coventry CV2 1AB
                2. b Three Swans Surgery, Salisbury SP1 1DX
                3. c Cochrane Depression Group, Institute of Health Sciences, University of Oxford, PO Box 777, Oxford OX3 7LF
                4. d Geriatrics and Extended Care, Department of Veterans Affairs Medical Center/Texas Tech University, 6010 Amarillo Boulevard, West Amarillo, TX 79106, USA
                5. e Gwent Community Health NHS Trust, County Hospital (Griffithstown), Pontypool, Gwent NP4 5YA
                6. f Pain Management Clinic, South Cleveland Hospital, Middlesbrough TS4 3BW
                7. g Grampian Health Board, Aberdeen AB9 1RE
                8. h Fitznells Manor Surgery, Ewell, Surrey
                9. i NHS Centre for Reviews and Dissemination, University of York, York YO1 5DD
                10. j Leeds General Infirmary, Leeds LS1 3EX
                11. k Royal Edinburgh Hospital, Edinburgh EH10 5HF
                12. l Dingleton Hospital, Melrose, Roxburghshire
                13. m Hairmyres Hospital, East Kilbride G75 8RG
                14. n West Norwich City Community Team, Norfolk Mental Health Care NHS Trust, Norwich NR6 5BE

                  Editor-In their study of antidepressant prescribing in general practice John Donoghue and colleagues found that selective serotonin reuptake inhibitors were more likely than older tricyclic antidepressants to be prescribed at recommended doses.1 However, effective treatment of major depression requires antidepressants to be continued for at least four to six months after clinical improvement because treatment for an inadequate amount of time is more likely to result in relapse.2

                  We performed a retrospective review of antidepressant prescribing in two fully computerised, paperless, general practices in Ewell, Surrey. The combined list size is 11 400 patients. All consultations and prescriptions are entered directly on to a record linked practice computer by each doctor. We performed a computer search of all patients who had a diagnosis of depression and who were prescribed an antidepressant between 1 January 1995 and 31 June 1995. Patients prescribed an antidepressant in the previous year were excluded as we wished to examine newly treated patients. We reviewed the notes manually one year after the initial prescription to determine the duration of treatment, the number of follow up examinations by the general practitioner, the rate of relapse, and the reason for stopping treatment.

                  We identified 78 patients (annual incidence 1.4%), of whom 47 (60%) were prescribed a serotonin reuptake inhibitor and 31 (40%) a tricyclic antidepressant (including four patients who were given lofepramine). The average age and sex distribution of the patients was similar for both serotonin inhibitors (45.3 years, 37 women (79%)) and tricyclic antidepressants (48.6 years, 25 women (80%)). The median length of treatment with a serotonin reuptake inhibitor was only four weeks compared with six weeks for tricyclic antidepressants. Fourteen patients (30%) initially prescribed a serotonin inhibitor had a relapse after stopping treatment compared with five patients (16%) initially prescribed a tricyclic antidepressant. The average number of general practitioner follow up examinations was 2.0 for serotonin inhibitors and 3.4 for tricyclic antidepressants. In most cases (44 patients (56%)) no reason for stopping treatment was recorded, and most patients had only one follow up consultation or none at all (19 (24%) and 24 (31%) patients, respectively).

                  We found that both selective serotonin inhibitors and tricyclic antidepressants were given for short durations for the treatment of depression and that follow up by general practitioners was poor. In this comparatively small study we were unable to control for confounding by severity or prescribing history. However, we agree with Donoghue and colleagues that research is urgently required to determine the clinical and economic costs of such prescribing.

                  References

                  1. 1.
                  2. 2.

                  Costs should have been considered

                  1. Simon Gilbody, MRC training fellow in health services researchi,
                  2. Trevor Sheldon, Directori,
                  3. Fujian Song, Research fellowi,
                  4. Allan House, Consultantj
                  1. a Henley Green Centre, Coventry CV2 1AB
                  2. b Three Swans Surgery, Salisbury SP1 1DX
                  3. c Cochrane Depression Group, Institute of Health Sciences, University of Oxford, PO Box 777, Oxford OX3 7LF
                  4. d Geriatrics and Extended Care, Department of Veterans Affairs Medical Center/Texas Tech University, 6010 Amarillo Boulevard, West Amarillo, TX 79106, USA
                  5. e Gwent Community Health NHS Trust, County Hospital (Griffithstown), Pontypool, Gwent NP4 5YA
                  6. f Pain Management Clinic, South Cleveland Hospital, Middlesbrough TS4 3BW
                  7. g Grampian Health Board, Aberdeen AB9 1RE
                  8. h Fitznells Manor Surgery, Ewell, Surrey
                  9. i NHS Centre for Reviews and Dissemination, University of York, York YO1 5DD
                  10. j Leeds General Infirmary, Leeds LS1 3EX
                  11. k Royal Edinburgh Hospital, Edinburgh EH10 5HF
                  12. l Dingleton Hospital, Melrose, Roxburghshire
                  13. m Hairmyres Hospital, East Kilbride G75 8RG
                  14. n West Norwich City Community Team, Norfolk Mental Health Care NHS Trust, Norwich NR6 5BE

                    Editor-The survey of John Donoghue and colleagues shows a 133% rise in prescriptions for specific serotonin reuptake inhibitors between 1993 and 1995.1 If generalisable this represents a major shift in prescribing, the economic implications of which are ignored in the accompanying editorial by Tony Kendrick,2 who is unable to clinically justify this trend.

                    The largest absolute difference in total discontinuation rates between specific serotonin reuptake inhibitors and tricyclic antidepressants in a meta-analysis is 2.8%3 (rather than 5-10%, as quoted by Kendrick2). The number of patients who need to be treated with specific serotonin reuptake inhibitors instead of tricyclic antidepressants to prevent one discontinuation is 38. The clinical importance of a small difference in discontinuation rates is hard to judge and has been little studied.

                    If, however, small increases in compliance are regarded as sufficiently important to achieve, then the cost effectiveness of alternative strategies should be evaluated,4 rather than presume that this is best addressed by a major shift in prescribing.

                    For example, the finding of Robert G Priest and colleagues that most of the public believes antidepressants to be addictive5 is a possible contributory factor to overall treatment discontinuation rates of over 30%, irrespective of what is prescribed. A more cost effective strategy than such a change in prescribing might be to give better information to patients starting antidepressant treatment in order to allay their fears.

                    It is important that well intentioned national initiatives (such as the Defeat Depression Campaign) designed to extend quality health care to a wider number of people are not used opportunistically by commercial interests to promote a particular product. We were surprised that the fact that Hiram Wildgust, one of Donoghue's colleagues,1 is an employee of Lilly Industries was not acknowledged as a conflict of interest.

                    References

                    1. 1.
                    2. 2.
                    3. 3.
                    4. 4.
                    5. 5.

                    Shared care of people with major mental illness

                    1. Valerie Murray, Research fellowk,
                    2. Helen Walker, Research fellowl,
                    3. Caroline Mitchell, Consultant psychiatristm,
                    4. Anthony J Pelosi, Consultant psychiatristm
                    1. a Henley Green Centre, Coventry CV2 1AB
                    2. b Three Swans Surgery, Salisbury SP1 1DX
                    3. c Cochrane Depression Group, Institute of Health Sciences, University of Oxford, PO Box 777, Oxford OX3 7LF
                    4. d Geriatrics and Extended Care, Department of Veterans Affairs Medical Center/Texas Tech University, 6010 Amarillo Boulevard, West Amarillo, TX 79106, USA
                    5. e Gwent Community Health NHS Trust, County Hospital (Griffithstown), Pontypool, Gwent NP4 5YA
                    6. f Pain Management Clinic, South Cleveland Hospital, Middlesbrough TS4 3BW
                    7. g Grampian Health Board, Aberdeen AB9 1RE
                    8. h Fitznells Manor Surgery, Ewell, Surrey
                    9. i NHS Centre for Reviews and Dissemination, University of York, York YO1 5DD
                    10. j Leeds General Infirmary, Leeds LS1 3EX
                    11. k Royal Edinburgh Hospital, Edinburgh EH10 5HF
                    12. l Dingleton Hospital, Melrose, Roxburghshire
                    13. m Hairmyres Hospital, East Kilbride G75 8RG
                    14. n West Norwich City Community Team, Norfolk Mental Health Care NHS Trust, Norwich NR6 5BE

                      Editor-Tony Kendrick and Tom Burns (a general practitioner and a psychiatrist) argue that family doctors should take back the sole continuing care of stable and compliant patients with psychotic dis- orders.1 However, we stand by our assertion that it is the job of multidisciplinary community teams to oversee the care of almost everybody with a major mental disorder,2 although always in collaboration with general practitioners. This debate is at least a healthy sign amid concerns about the failure of community care and in the light of recent guidance from the General Medical Services Committee which could greatly limit general practitioners' role in the management of major mental illness.3

                      No matter who takes primary responsibility, the monitoring of seriously mentally ill patients should lead to adequate recognition of their needs and the appropriate responses. Several studies, including our own, indicate that routine psychiatric services often fail in this task.2 4 Our subjects who had returned to the sole care of their general practitioner also had important unidentified needs, although Kendrick and Burns are right to criticise our failure to study the important group of psychotic patients who have never been in touch with specialists.1

                      Continuing care of patients with major mental diseases is dominated by the fact that many do not seek help when their condition deteriorates. This is why we now insist that even patients whose condition is stable require systematic review, usually in their own homes. Our work in Lanarkshire and an ongoing investigation by one of us (HW) in the Scottish borders suggests that a rolling survey of all patients with identified psychotic disorders can be undertaken without prohibitive additional costs. Several standardised schedules have been developed which could guide this clinical process.5 Further research is required to show whether this routine needs assessment will help specialists and general practitioners to prioritise mental health care according to need rather than demand. This will probably lead to increased input to the least vocal and most vulnerable psychiatric patients–that is, those with schizophrenia, manic depressive illness, and other brain disorders. This will inevitably divert services from patients with milder acute conditions, but it is for each area to establish comprehensive services within available resources. Potential shortfalls in healthcare provision will not be avoided by leaving general practitioners to provide all community care once patients recover from the acute phase of major psychotic illness.

                      References

                      1. 1.
                      2. 2.
                      3. 3.
                      4. 4.
                      5. 5.

                      People may become psychologically dependent on antidepressants

                      1. D B Double, Consultant psychiatristn
                      1. a Henley Green Centre, Coventry CV2 1AB
                      2. b Three Swans Surgery, Salisbury SP1 1DX
                      3. c Cochrane Depression Group, Institute of Health Sciences, University of Oxford, PO Box 777, Oxford OX3 7LF
                      4. d Geriatrics and Extended Care, Department of Veterans Affairs Medical Center/Texas Tech University, 6010 Amarillo Boulevard, West Amarillo, TX 79106, USA
                      5. e Gwent Community Health NHS Trust, County Hospital (Griffithstown), Pontypool, Gwent NP4 5YA
                      6. f Pain Management Clinic, South Cleveland Hospital, Middlesbrough TS4 3BW
                      7. g Grampian Health Board, Aberdeen AB9 1RE
                      8. h Fitznells Manor Surgery, Ewell, Surrey
                      9. i NHS Centre for Reviews and Dissemination, University of York, York YO1 5DD
                      10. j Leeds General Infirmary, Leeds LS1 3EX
                      11. k Royal Edinburgh Hospital, Edinburgh EH10 5HF
                      12. l Dingleton Hospital, Melrose, Roxburghshire
                      13. m Hairmyres Hospital, East Kilbride G75 8RG
                      14. n West Norwich City Community Team, Norfolk Mental Health Care NHS Trust, Norwich NR6 5BE

                        Editor-Robert G Priest and colleagues advocate educating patients that discontinuing antidepressant treatment will not be a problem but remarkably do not cite any evidence to support their recommendation.1 They also complain that many lay people regard antidepressants as addictive. They suggest that people may be extrapolating from what they have heard about benzodiazopines. This may be, but it is also common sense to believe that discontinuing taking a drug that is thought to improve mood may be difficult. I think that the general public understands this issue better than the Royal Colleges of Psychiatrists and General Practitioners, which are responsible for the Defeat Depression Campaign.

                        Of course what Priest and colleagues mean is that there is little evidence of physical dependence caused by antidepressants, but this is not what they say. There are, however, case reports of a withdrawal syndrome.2 Clinical experience is that it can be difficult to withdraw treatment with antidepressants for various reasons. The general public might reasonably expect psychiatrists specialising in disorders of the mind to recognise psychological dependence, base their advice on clinical experience, and use their common sense.

                        Randomised controlled trials of discontinuation of antidepressant treatment have a relapse rate varying from 92%3 to 36%4 in the placebo group. Relapse rate is significantly reduced by continuing antidepressant treatment. Some patients therefore do maintain their therapeutic gains when antidepressants are withdrawn, but the relapse rate is not insubstantial and seems to support the general public's commonsense view rather than the Defeat Depression Campaign's purist scientific statement. Perhaps the public needs to be suspicious of the motives of a campaign that encourages them to seek medical treatment and also tries to help doctors recognise depression. Patronising misinformation is not constructive.

                        References

                        1. 1.
                        2. 2.
                        3. 3.
                        4. 4.
                        View Abstract