Cookie policy: This site uses cookies (small files stored on your computer) to simplify and improve your experience of this website. Cookies are small text files stored on the device you are using to access this website. For more information please take a look at our terms and conditions. Some parts of the site may not work properly if you choose not to accept cookies.


Subscribe or Register

Existing user? Login


Prescribing in mental illness: will short-term gains lead to long-term costs?

By Celia Feetam and Chris Hawley

The performance of many NHS staff is measured, in part at least, by the kornilovdream/dreamstime.comsavings they can achieve on selected budgets, medicines expenditure included. Getting better value from the NHS is in everybody’s interest because it is the public purse — our money — that is being spent. Specific prescribing strategies have been proposed across a variety of therapeutic areas to reduce costs.1–3 Some, if implemented, might be powerful in reducing expenditure on medicines in certain areas, but with respect to the treatment of mental illness, several can be contested and could in fact lead to increased long-term costs as well as prove detrimental to patient care.


The National Institute for Health and Clinical Excellence advocates “patient centred care”4,5 and encourages us to offer choice.5,6 Patients are told by NICE to expect it.7,8 This is particularly the case in the treatment of schizophrenia, where the choice of antipsychotic should be influenced by a patient’s preferences, considering the relative potential of individual antipsychotics to cause certain side effects such as extrapyramidal side effects, metabolic effects (eg, weight gain, glucose dysregulation) and hyperprolactinaemia.6 A central proposal by some is to substitute generic risperidone for a branded atypical antipsychotic1,3 but this is not readily reconciled with policy directives emphasising patient choice and the individualisation of treatment.

This strategy rests on the assumption that for each patient, considered individually, all or any of the antipsychotics would be equivalent with respect to effectiveness, safety, tolerability and adherence. However, given that there are significant differences in the side effect profiles of antipsychotics such an assumption is open to challenge. For example, symptom control may be maintained if the switch is made correctly and the patient responds to risperidone but what would be the cost of relapse if this were not the case? Relapse can also occur if the patient fails to adhere to the treatment, for example, due to sexual dysfunction attributable to risperidone as a result of hyperprolactinaemia.9 The long-term costs of a pituitary adenoma, osteoporosis or breast cancer — all acknowledged as potential risks of chronic hyperprolactinaemia — are also worthy of consideration.

Medicines in some other therapy areas, such as statins, are so pharmacologically similar as to be readily interchangeable but atypical antipsychotics are not. Treatment selection has to be according to clinical profile and context. For example, aripiprazole, being non-sedative, is not good for a distressed patient, olanzapine is not suitable for a patient already exhibiting a propensity to obesity, and risperidone cannot be a preferred choice in women of reproductive potential because the resultant hyperprolactinaemia can lead to infertility. It has been said that that “substitution is well accepted by patients”.10 Our experience in psychiatry suggests otherwise. Antipsychotic switching in schizophrenia can be a delicate, sometimes precarious, matter.

Typical antipsychotics

Another worrying cost-saving strategy recently proposed within some NHS organisations is endorsement of a typical antipsychotic, such as chlorpromazine, as first-line treatment for schizophrenia. We consider such a suggestion unconscionable, paying no heed as to why chlopromazine and similar treatments so rapidly passed out of routine use in the 1990s. Many studies have shown a high incidence of extrapyramidal symptoms and tardive dyskinesia associated with typical antipsychotics. Muscettola et al reported 29.4 per cent experiencing acute extrapyramidal symptoms and 18.3 per cent showing early signs of tardive dyskinesia.11 Similarly, Jeste et al showed a high prevalence of tardive dyskinesia after only a year’s treatment with low-dose typicals (the median chlorpromazine equivalent dose being 68.4mg/day). Prevalence was higher in those who had received previous antipsychotics for more than 30 days than in those who were neuroleptic naive (36.9 per cent vs 22.3 per cent).12 The incidence of tardive dyskinesia in older people (mean age 65 years) treated with typical antipsychotics has been shown to be 60 per cent after three years13 and in younger people (mean age 29 years) 19 per cent after four years and 26 per cent after six years.14 Chlorpromazine in particular has been associated with hepatitis,15 profound hypotension and an overall tolerability no better, and possibly worse, than clozapine.16 In addition, people prescribed a phenothiazine such as chlorpromazine require high factor sun protection in the summer because they are likely to suffer photosensitivity.17 This incurs further cost.

Physical health

Excess physical morbidity and mortality in those with schizophrenia is well documented, cardiovascular disease and diabetes in particular.18 It is well established that some atypical antipsychotics are associated with adverse metabolic effects, such as weight gain, dyslipidaemia and glucose dysregulation, further increasing the risk of coronary heart disease and diabetes and consequently the costs of treatment. Barnett et al describe the increased costs, both direct and indirect, of antipsychotic-induced diabetes and coronary heart disease in people with schizophrenia.19 A projected risk analysis of metabolic data using modified Stern and Framingham models showed that relative to treatment with olanzapine, risperidone or quetiapine, aripiprazole treatment was predicted to result in fewer new cases of diabetes over 7.5 years and a lower incidence of coronary heart disease over 10 years. Cost projections suggest that compared with the other three aforementioned antipsychotics, aripiprazole may be associated with long-term cost savings to the UK health service with 23.4 per 1,000 fewer new cases of diabetes over 7.5 years (saving £37,261,293) and 3.7 per 1,000 fewer CHD events over 10 years (saving £7,506,770). This analysis is not without its limitations but it does illustrate that judicious choice of treatment, rather than the cheapest one, has the potential to reduce the health burden for people with schizophrenia as well as cost to the health economy.


Schizophrenia is typically a chronic condition requiring management over several decades or even a life-time. Medicines adherence is notoriously poor in schizophrenia20 and is a leading cause of relapse.21 Patients have almost complete control over the prescribed medicine option because they have the ultimate sanction — not to take it. During the initial years of treatment a patient may typically cycle between several antipsychotics, eventually settling on the one that is the least unsatisfactory.

Reduced prescribing of modified release formulations, another proposed cost-saving strategy, may also impair adherence by increasing the number of doses required as well as the intensity of side effects.

Cost of relapse

Antipsychotics at any price are extremely high value medicines. The cost of a relapse of schizophrenia requiring hospital admission has been estimated at £25,00022 so any constraint that compromises the clinician’s flexibility entails a high-stake gamble.

The optimal management of schizophrenia involves doing everything possible to squeeze that last few per cent of clinical effectiveness out of available treatments. A typical mental health trust might be responsible for 2,000 people with schizophrenia. Even a 1 per cent extra relapse rate per year requiring hospital admission could cost an additional £500,000. Thus one must contextualise the costs of schizophrenia; the major costs of managing this distressing illness are those of human resources, not medicines acquisition.

Bipolar disorder

In mental health services the cost of medicines is far outweighed by other health and social service costs. In 2009/10 over 60 per cent of the annual NHS cost of managing bipolar disorder in the UK was accounted for by hospital admissions and day-care attendances, with only 7.4 per cent of the annual total spent on primary care prescribing.23 In 2010 there were 1,048 admission for mania in England24 with an average length of stay of about 37 days, amounting to an annual total of 39,582 bed-days and costing approximately £12m. A Markov model analysis demonstrated that quetiapine plus a mood stabiliser was more cost-effective than a mood stabiliser alone in the maintenance treatment of bipolar I disorder:25 adjunctive quetiapine was associated with a 56 per cent reduction in acute mood events, a 29 per cent reduction in associated hospital admissions, 4 per cent improvement in quality-adjusted life year gains and 5 per cent lower total direct costs compared with a mood stabiliser, either lithium or valproate, prescribed as monotherapy. A subsequent review by the Centre for Evidence Based Medicine found the methodology of this analysis to be valid and well presented and the conclusions to be valid and robust.26


In 2007, the King’s Fund estimated that medicines make up only 1 per cent of the total cost of managing depression in the UK.27 In 2010 there were 21,775 hospital admissions for depression in England with the average length of stay 40 days resulting in 783,972 bed-days25 at a cost of around £235m. If savings can be made on admissions, length of stay and other NHS costs by using a more cost-effective antidepressant despite a higher acquisition cost,5 this may be more cost-efficient in the longer term than choosing a cheap treatment.

NICE recommends a generic serotonin reuptake inhibitor first-line for the treatment of moderate to severe depression because SSRIs have a favourable risk-benefit ratio and generics are cheaper.5 The guideline also states that SSRIs are as equally effective as other antidepressants although this has recently been contested.28 Cipriani et al concluded that clinically important differences in both efficacy and acceptability do, in fact, exist between commonly prescribed antidepressants.28 Efficacy and acceptability are important considerations when defining care pathways to treat more people more effectively. The findings by Cipriani et al have been reflected by a naturalistic study conducted in a community mental health clinic in Bedfordshire.29 The “top four” antidepressants, as ranked by Cipriani et al (sertraline, escitalopram, venlafaxine and mirtazapine), were associated with the highest discharge rates and the greatest reductions in suicidal behaviour; with sertraline having the most significant effect in reducing suicidal behaviour and escitalopram the highest discharge rate.29

NICE performed a decision tree cost-effectiveness analysis of antidepressants30 using the efficacy data from Cipriani28 and found sertraline to be the most cost-effective SSRI followed by escitalopram If the response to the first SSRI is inadequate after three to four weeks, adherence is confirmed and increasing the dose is not appropriate or acceptable, NICE recommends a second SSRI be prescribed.5 Many organisations recommend citalopram and fluoxetine as first-or second-line choices. Escitalopram and sertraline have not only been shown to be more cost-effective than both of these,30 escitalopram also appears to be less toxic and has been associated with a lower seizure rate following acute overdose than citalopram.31 Furthermore, another recommended cost saving strategy, to switch patients maintained on escitalpram to citalopram,1 may in fact increase costs. A study in the US has shown that switching patients already stabilised and doing well on escitalopram to another SSRI for non-medical reasons incurs greater costs rather than reducing them.32

Although the cost-saving strategies proposed have many strengths we dispute that these generalise well to the treatment of mental illness. Short-term gains can easily be achieved by choosing a cheap medicine, but at what cost to patient care, long-term outcomes and the health economy overall? It is insufficient to consider only acquisition costs: the chronic and complicated nature of these conditions demands that the wider perspective of cost effectiveness is given greater emphasis. We need more detailed, independently validated, cost-effectiveness analyses of the available options for treating mental illness. Until more such data are available we urge decision-makers to proceed perspicaciously in awareness that medicine acquisition costs form only one small part of the overall cost of treating mental illness.


Declarations of interest

Celia Feetam has received honoraria from Astra Zeneca, BMS/Otsuka, Janssen, Lilly and Lundbeck for speaker meetings and attending advisory boards. Chris Hawley has received honoraria from, BMS/Otsuka, Lilly and Lundbeck for speaker meetings.

About the authors

Celia Feetam is specialist mental health pharmacist in the department of psychiatric pharmacy, Aston University, Birmingham, and Chris Hawley is consultant psychiatrist at Hertfordshire Partnership Foundation Trust and the University of Hertfordshire


  1.  Moon J.  Flett  AS,  Godman BB,  Grosso  AM, Wierzbicki AS. Getting better value from the NHS drugs budget BMJ 2011; 342:30-2.
  2. Department of Health. Health care: medicine use and procurement :top tips – efficiency schemes in provider settings.
  3. QIPP medicines use and procurement work stream: mental health medicines use efficiency schemes. September 2011.
  4. National Institute for Health and Clinical Excellence. The management of bipolar disorder in adults, children and adolescents, in primary and secondary care quick reference guide. CG38 July 2006.
  5.  National Institute for Health and Clinical Excellence. Depression: treatment and management of depression in adults, including adults with a chronic physical health problem quick reference guide. CG90&91 October 2009.
  6. National Institute for Health and Clinical Excellence. Core interventions in the treatment and management of schizophrenia in primary and secondary care (update) quick reference guide CG 82:March 2009.
  7. National Institute for Health and Clinical Excellence. Schizophrenia (update): understanding NICE guidance CG 82: March 2009
  8. National Institute for Health and Clinical Excellence. Bipolar disorder: understanding NICE guidance CG38: July 2006.
  9. Risperidal. Summary of product characteristics. Available at (accessed on 30 October 2011).
  10. Holt RI. Medical causes and consequences of hyperprolactinaemia. A context for psychiatrists. Journal of Psychopharmacology. 2008;22(2 Suppl):98-103.
  11. Muscettola G. Barbato G, Pampallona S, Casiello M, Bollini P. Extrapyramidal syndromes in neuroleptic-treated patients: prevalence, risk factors, and association with tardive dyskinesia. Journal of Clinical Psychopharmacology 1999;19(3):203-8.
  12. Jeste DV, Lacro JP, Palmer B,  Rockwell E, Harris MJ, Caligiuri MP. Incidence of tardive dyskinesia in early stages of low-dose treatment With typical neuroleptics in older patients. American Journal of Psychiatry 1999;156:309-11.
  13. Jeste DV,Gilbert PL, McAdams LA, Harris MJ. Considering neuroleptic maintenance and taper on a continuum: need for individual rather than dogmatic approach.  Archives of General Psychiatry 1995;52(3):209-12.
  14.  Kane, J.M., Tardive dyskinesia: epidemiological and clinical presentation. In: Bloom DE, Kupfer DJ (Editorss.), Psychopharmacology: the fourth generation of progress. Raven Press, New York, 1995.pp. 1485–96.
  15. Salhab AS,  Dujovn CA. Toxicity of promazine and chlorpromazine to isolated rat hepatocytes and its modification by liposome entrapment. Pharmacology 1986;33:311-21.
  16. Phillips MR, Li X, Li K, Jiang H, Wu C, Duan N, Niu Y, Lieberman JA Clozapine v chlorpromazine in treatment-naive, first-episode schizophrenia: 9-year outcomes of a randomised clinical trial. British Jounral of Psychiatry 2011 bjp.bp.110.081471; published ahead of print February 3, 2011.
  17. Largactil summary of product characteristics. Available at (accessed 30 October 2011).
  18. Correll CU. Estimated prevalence and relative risk factors of modifiable cardiovascular disease risk factors in schizophrenia and bipolar disorder. CNS Spectr. 2007;12(Suppl 17):12-20,35.
  19. Barnett HA, Millar HL, Loze JY, l’Italien GJ,  van Baardewijk M. Knapp M. UK cost-consequence analysis of aripiprazole in schizophrenia:diabetes and coronary heart disease risk projections (STAR study) European Archives of Psychiatry and Clinical Neuroscience 2009;259:239–47.
  20. Feetam CL, Roberts H. Medicine taking behaviour in schizophrenia (Part 1).  Progress in Neurology and Psychiatry 2010;14.
  21. Morken G, Widen JH, Grawe RW. Non-adherence to antipsychotic medication, relapse and rehospitalisation in recent-onset schizophrenia. BMC Psychiatry 2008.
  22. Munro J,  Osborne S,  Dearden L,  Pascoe K, Gauthier A, Price M. Hospital treatment and management in relapse of schizophrenia in the UK: associated costs. .  The Psychiatrist Online 2011;35:95-100.
  23. Young AH, Rigney U, Shaw S, Emmas C, Thompson JM. Annual cost of managing bipolar disorder to the UK healthcare system. Journal of Affective Disorders 2011;133(3):450-6
  24. Accessed 30/10/11
  25. Fajutrao L, Paulsson B, Liu S, Locklear J, Cost-effectiveness of quetiapine plus mood stabilizers compared with mood stabilisers alone in the maintenance therapy of bipolar I disorder:Results of a markov model analysis. Clinical Therapeutics 2009;31:1456-68.
  26. Centre for Reviews and Dissemination. NHS Economic Evaluation Database (NHS EED).
  27. McCrone P, Dhanasiri S, Patel A, Knapp M, Lawton-Smit S.  Paying the price: The cost of mental health care in England to 2026.
  28. Cipriani A, Furukawa TA, Salanti G et al. Comparative efficacy and acceptability of 12 new-generation antidepressants:a multiple-treatments meta-analysis.  Lancet. 2009 373(9665):746-58.
  29. Agius M, Gardner J, Liu K, Zaman R . An audit to compare discharge rates and suicidality between antideprerssant monotherapies prescribed for unipolar depression. Psychiatria Danubina, 2010;22:350–3.
  30. National Institute for Health and Clinical Excellence. Depression: treatment and management of depression in adults, including adults with a chronic physical health problem. CG 90 & 91:October 2009.
  31. Yilmaz Z, Ceschi A, Rauber-Lüthy C et al. Escitalopram causes fewer seizures in human overdose than citalopram. Clinical Toxicology (Phila). 2010 Feb 20. [Epub ahead of print]
  32. Wu EQ, Ben-Hamadi R, Yu AP et al. Healthcare utilization and costs incurred by patients with major depression after being switched from escitalopram to another SSRI for non-medical reasons. Journal of Medical Economics 2010;13:314-23.

    Citation: The Pharmaceutical Journal URI: 11096709

    Have your say

    For commenting, please login or register as a user and agree to our Community Guidelines. You will be re-directed back to this page where you will have the ability to comment.

    Recommended from Pharmaceutical Press

    • Print
    • Share
    • Comment
    • Save
    • Print Friendly Version of this pagePrint Get a PDF version of this webpagePDF

    Jobs you might like

    Newsletter Sign-up

    Want to keep up with the latest news, comment and CPD articles in pharmacy and science? Subscribe to our free alerts.