This site is intended for health professionals only

Switching from typical to atypical antipsychotics

Anne Connolly
BPharm DipClinPharm
Principal Pharmacist Medicines Information

David Taylor
MSc MRPharmS
Chief Pharmacist
Pharmacy Department
Maudsley Hospital

Atypical antipsychotics are the treatment of choice for most patients with psychotic illness. This is because, compared with typical antipsychotics, they cause few or no extrapyramidal side-effects (EPSE)(1) and have a lower risk of causing tardive dyskinesia (TD).(1) In addition, atypical antipsychotics are at least as effective as typical antipsychotics in treating psychosis.(2) Clozapine is more effective than all other antipsychotics in treating refractory schizophrenia.(3) Atypical antipsychotics of course also have adverse effects, such as weight gain and diabetes mellitus.(4,5)

Switching antipsychotics is a common procedure and, when managed well, offers patients significant benefits. This article will provide practical guidance on how to switch antipsychotics, so as to reduce the risks of relapse and withdrawal side-effects.

We conducted a search of the following databases: Medline (from 1966), Embase (from 1980) and Cochrane, in October 2002.

Who to switch?
It is important to have a clear rationale for switching antipsychotics. The two main reasons for switching are intolerance or inefficacy of the current treatment.(6–8) Intolerance to typical antipsychotics is common, particularly because of adverse effects such as EPSE, TD and hyperprolactinaemia.(2) Antipsychotics are not always effective at treating all symptom domains,(2) or may not prevent relapse despite patient compliance.(9) Whatever the reasons for switching, explaining the risks and benefits of the new treatment to patients and carers is important.(6)

Switching may not be beneficial for some patients, such as: those whose risk of symptom exacerbation, particularly aggression and suicidal thoughts, outweighs the benefits of switching; patients who have just recovered from a relapse; patients who do not want to switch; those who are well and have no side-effects from their current medication; patients experiencing stressful life events (eg, new job or relationship changes).(6,8)

Risks of switching
Two main risks when switching antipsychotics are antipsychotic withdrawal reactions and relapse.

The signs and symptoms of antipsychotic withdrawal reactions are presented in Table 1. Many of these (particularly nausea, sweating, vomiting and insomnia) are due to cholinergic hypersensitivity caused by withdrawal from profoundly anticholinergic antipsychotics.(10) This reaction is often called “cholinergic rebound” and can occur one to four days after stopping treatment and lasts one to two weeks.(11) Dyskinesias can occur or worsen when antipsychotics are stopped. They usually disappear after six weeks and are probably caused by oversensitivity of dopamine receptors.  If dyskinesia persists (ie, lasts longer than 6–12 weeks), then tardive dyskinesia or tardive dystonia should be considered.(11) Withdrawal akathisia may also occur within days of stopping a typical antipsychotic and can persist.(12) If akathisia continues the diagnosis may change to tardive akathisia.(13,14)


Cholinergic rebound is managed by gradual withdrawal of both the antipsychotic and any anticholinergic medication.(15,6) Some researchers recommend withdrawal of anticholinergics over three weeks.(7) Unfortunately, withdrawal dystonia and Parkinsonism may need to be managed with anticholinergic medication.(6) Thus patients coming off anticholinergics may need to have their withdrawal halted temporarily. Withdrawal akathisia usually does not respond to anticholinergics, so beta-blockers or a benzodiazepine can be used if necessary.(16)

It is well known that withdrawal of maintenance antipsychotics increases the risk of relapse in schizophrenia.(11,17) What is less well known, however, is whether the abrupt or gradual withdrawal of an antipsychotic affects this risk. One review paper attempted to answer this question by analysing data on 1,210 patients withdrawn from antipsychotics (1,006 abruptly, 204 gradually).(18) A 25% risk of relapse of schizophrenia was found at six weeks for patients abruptly withdrawn from antipsychotics and at 10 weeks for those gradually withdrawn from antipsychotics. The authors say it is uncertain whether gradual withdrawal reduces or postpones relapse and the data should be interpreted with caution. This is because of the wide variability of the studies considered in the review, in areas such as diagnostic criteria and definition of relapse. Given the uncertainty about the link between how antipsychotics are withdrawn and relapse, it would seem prudent to withdraw antipsychotics gradually.

Switching technique
Two common techniques when switching antipsychotics are abrupt switching or cross-tapering.(6,8,15)

Abrupt switching, where the typical antipsychotic is stopped abruptly and the atypical gradually started, has some advantages. These include establishing the new treatment quickly, a low risk of interactions between the two antipsychotics and a simple medication regimen during the switch. However, as discussed earlier, the risk of antipsychotic withdrawal effects is higher using this approach. Abrupt switching means stopping the oral typical antipsychotic on one day and starting the oral atypical the next. If the patient is on a typical depot then the injection is stopped abruptly and the atypical started when the next injection is to be given. This approach is suitable for patients experiencing a life-threatening side-effect, such as neuroleptic malignant syndrome (NMS) or clozapine-induced neutropenia, or when switching from a typical depot. A variation on this approach would be if the typical antipsychotic were gradually withdrawn and stopped then the atypical started. This method is recommended when switching to clozapine for reasons discussed below.

Cross-tapering is a method by which the typical antipsychotic is gradually withdrawn while the dose of the atypical antipsychotic is gradually increased.  With this method the new medication is started quickly, minimising the risks of relapse and antipsychotic withdrawal effects. Unfortunately, co-prescribed drugs can interact during the cross-taper and medication regimens are relatively more complicated. One author recommended a dose reduction of the typical antipsychotic of 30–50% every three to seven days while titrating the atypical to a therapeutic dose.(8) This method of switching is recommended when the priorities are treatment of symptoms, speed of initiation of the new medication and avoidance of antipsychotic withdrawal effects.

Article continues below this sponsored advert
Cogora InRead Image
Explore the latest advances in clinical care at events delivered by renowned experts from CofE

Initiation and switching atypical antipsychotics

There are two studies that make recommendations for switching to amisulpride. One study describes a consensus meeting about the practical use of amisulpride.(19) The consensus group suggested cross-tapering over one to four weeks as the best method for switching to amisulpride. A second study also suggests a cross-tapering method over up to four weeks.(12) Both recommendations are based on knowledge of amisulpride’s pharmacology and clinical experience of the switching process. Amisulpride has minimal adverse effects and interactions that are important when switching, which makes this a relatively low risk procedure.

The recommended method of switching to clozapine is to gradually and completely withdraw the typical antipsychotic before gradual initiation of clozapine.(20) This is largely because some antipsychotics are associated with agranulocytosis and coadministration with clozapine may increase the risk of this life-threatening side-effect.(15) This method of switching is also important because of the danger of other pharmacokinetic and pharmacodynamic interactions.

Occasionally it is not possible to completely withdraw the typical antipsychotic before initiating clozapine. This could be because of the patient’s mental state or if they are receiving depot medication. If cross-tapering is necessary then monitoring of blood pressure, neutrophils and electro-encephalogram are needed.(15) Clozapine also causes profound sedation, hypotension and seizures, effects that can be magnified by typical antipsychotics. Clozapine is metabolised by the cytochrome P450 enzymes 2D6 and 1A2,(21) increasing the risk of pharmacokinetic interactions with typical antipsychotics metabolised in the same way.

Two industry-funded studies have investigated methods of switching to olanzapine. The earliest study openly randomised patients (n=209) on typical antipsychotics and risperidone to one of four switching groups: method one, abrupt withdrawal of current antipsychotic then immediate olanzapine 10mg initiation; method two, abrupt withdrawal of current antipsychotic then gradual initiation of olanzapine (one week on placebo then olanzapine 5mg then 10mg); method three, gradual withdrawal of current antipsychotic (50% of previous dose for one week then stop) plus immediate olanzapine 10mg initiation; method four, gradual withdrawal and then gradual initiation of olanzapine (as method two).(22) All four methods were as effective as each other, although the authors suggest method three was the most favourable option because it showed the earliest improvements in efficacy and a low incidence of adverse effects.

More recently, a randomised open-label study investigated two techniques for switching Asian patients from mainly typical antipsychotics to olanzapine.(23) Method one was by abrupt switching (n=54), where the patient’s antipsychotic was stopped abruptly and then olanzapine 10mg started. Method two was by cross-tapering (n=54), where olanzapine 10mg was started and then the patient’s current antipsychotic was withdrawn over two weeks. Both methods of switching were equally successful (74.1% of abrupt switch group; 67.9% of cross-taper group). Success was defined as completing six weeks of olanzapine treatment without worsening of symptoms and EPSE. Interestingly, patients in both groups reported sedation, insomnia and headache as side-effects. This could have been a result of the abrupt and relatively fast (over one week) withdrawal of the previous antipsychotic.

Overall these two studies show that a variety of methods are suitable for switching to olanzapine. Cross-tapering, by gradual withdrawal of the typical antipsychotic (for reasons discussed earlier) and immediate initiation of olanzapine, is recommended. This is because initiation of olanzapine at an effective dose did not cause side-effects in the above two studies. However, some patients (eg, women, the elderly and nonsmokers) may require a lower starting dose.(24) Olanzapine is sedating and can, less commonly, cause postural hypotension making pharmacodynamic interactions with typical antipsychotics a possibility.

One review article discusses methods of switching to quetiapine but does not specifically state a preferred option.(25) It refers to a poster presentation where abrupt stopping of the patient’s current antipsychotic and then gradual initiation of quetiapine was well tolerated. However, until we have more robust data about how to switch to quetiapine, a cross-tapering regimen is recommended (because of the reasons outlined earlier). Quetiapine can cause moderate hypotension and is also sedative, so is always initiated slowly. When cross-tapering quetiapine with a typical antipsychotic, consideration of any possible additive hypotensive effects of the typical antipsychotic is important.

Risperidone Oral formulation
Opinions differ as to how switching to oral risperidone should be managed. An international consensus group recommended abrupt switching, while the USA consensus group recommended cross-tapering.(26) A variety of methods of switching to risperidone have been studied. Abrupt withdrawal of oral and depot typical antipsychotic and then immediate initiation of oral risperidone can be a successful method of switching.(26,27) Gradual withdrawal of typical and then initiation of oral risperidone is the preferred method of switching of one reviewer.(28) However, for the reasons outlined earlier in the article, cross-tapering is the recommended method of switching. Risperidone is usually initiated slowly twice daily over three days, because it can cause hypotension.

Long-acting injection
Risperidone long-acting injection (RLAI) is the first atypical antipsychotic available in a long-acting formulation. Switching from oral and depot typical antipsychotics to RLAI can be complex. This is because RLAI is not released for the first three weeks after injection and does not reach steady-state plasma levels until just before the fourth injection.(20)

  • Switching from oral typical antipsychotics to RLAI. There are two recommended methods of switching from oral typical antipsychotics to RLAI: the first, switch to oral risperidone for a few days treatment and, if tolerated, give RLAI and then stop oral risperidone after three to four weeks; the second, give RLAI and then withdraw oral typical antipsychotic gradually after three to four weeks. The advantage of the first method is that patients are pre-exposed to risperidone, so tolerability can be assessed before giving a long-acting preparation.
  • Switching from depot typical antipsychotics to RLAI. There are a number of possibilities when switching from depot typical antipsychotics to RLAI. The manufacturer suggests various options: option one, giving RLAI one week before last typical depot due; option two, giving RLAI when next depot is given; option three, switching patient to oral risperidone, assessing tolerability, then switching to RLAI.

No published studies about switching to sertindole were found, although the manufacturer recommends initiation of sertindole after stopping previous antipsychotic treatment. This is because sertindole can cause hypotension and prolong cardiac QT interval, possibly causing life-threatening arrhythmias (torsades de pointes). Electrocardiogram monitoring is mandatory with sertindole: before starting treatment; after three weeks of treatment (when sertindole has reached steady state); before and then three weeks after any dose change; after the addition of an interacting medication. Typical antipsychotics can also prolong the QT interval, making coadministration dangerous. In addition, serum sertindole levels are significantly elevated by inhibitors of cytochrome p450 3A and 2D6 or drugs metabolised through the same hepatic enzymes. This includes many psychotropic drugs.

As far as we are aware there is no published information on how to switch to ziprasidone. Ziprasidone has a low incidence of sedation and hypotension but can, like most antipsychotics, cause a prolonged QT interval.(29) Because of its effect on QT, ideally the typical antipsychotics should be withdrawn before ziprasidone is started. As when switching all antipsychotics, additive pharmacodynamic and pharmacokinetic interactions must be considered.

No published studies about how to switch to zotepine were found. Zotepine can cause profound sedation and moderate hypotension, and because there is no specific recommendation for switching cross-tapering is recommended for the reasons discussed earlier. As always, consideration of any interactions of zotepine with the antipsychotic being withdrawn is important.

Atypical antipsychotics offer benefits to patients through improved tolerability and, in the case of clozapine, improved efficacy. For patients to achieve these benefits the switching process must be discussed, initiated and monitored carefully. Prescribers need to be aware of the risks of switching (antipsychotic withdrawal effects, relapse and drug interactions). When these risks are anticipated and managed, a successful switch from typical to atypical is likely.


  1. Stanniland C, Taylor D. Drug Saf 2000;22(3):195-214.
  2. National Institute for Clinical Excellence. Technology Appraisal Guidance, number 43. Guidance on the use of newer (atypical) antipsychotic drugs for the treatment of schizophrenia. London: NICE; 2002.
  3. Taylor D, Duncan-McConnell D. J Psychopharmacol 2000;14(4):409-18.
  4. Taylor D, McAskill R. Acta Psychiatr Scand 2000;101:416-32.
  5. Mir S, Taylor D. Int Clin Psychopharmacol 2001;16:63-74.
  6. Weiden PJ, Aquila R, Dalheim L, Standard JM. J Clin Psychiatry 1997;58 Suppl 10:63-72.
  7. Masand PS, Berry SL. Ann Pharmacother 2000;34:200-7.
  8. Peuskens J. Int Clin Psychopharmacol 2000;15 Suppl 4:S15-9.
  9. Csernansky JG, Mahmoud R, Brenner R. N Engl J Med 2002;346:16-22.
  10. Luchins DJ, Freed WJ, Wyatt RJ. Am J Psychiatry 1980;137:1395-8.
  11. Gardos G, Cole JO, Tarsy D. Am J Psychiatry 1978;135:1321-4.
  12. Burns T, Chabannes JP, Demyttenaere K. Curr Med Res Opin 2002;18(4):201-8.
  13. Dufresne R, Wagner RL. J Clin Psychiatry 1988;49:435-8.
  14. Lang AE. Mov Disord 1994;9:188-92.
  15. Taylor D. CNS Drugs 1997;8(4):285-92.
  16. Fleischhacker WW, Roth SD, Kane JM. J Clin Psychopharmacol 1990;10:12-21.
  17. Wistedt B. Acta Psychiatr Scand 1981;64:65-84.
  18. Viguera AC, Baldessarini RJ, Hegarty JD, et al. Arch Gen Psychiatry 1997;54:49-55.
  19. Lecrubier Y, Azorin M, Bottai T, et al. Neuropsychobiology 2001;44:41-6.
  20. Summary of product characteristics, Clozaril, Risperdal Consta. Datapharm Communications Ltd.
  21. Taylor D. Br J Psychiatry 1997;171:109-12.
  22. Kinon BJ, Basson BR, Gilmore JA, et al. J Clin Psychiatry 2000;61:833-40.
  23. Lee CT, Conde BJL, Mazlan M, et al. J Clin Psychiatry 2002;63:569-76.
  24. Fulton B, Goa KL. Drugs 1997;53:281-98.
  25. Cutler AJ, Goldstein JM, Tumas JA. Clin Ther 2002;24(2):209-22.
  26. Kirov GK, Murray RM, Seth RV, Feeney S. Acta Psychiatr Scand 1997;95:439-43.
  27. Desai NM, Huq Z, Martin SD, McDonald G. Adv Ther 1999;16(2):78-88.
  28. Borison RL. Clin Ther 1996;18(4):592-607.
  29. Glassman AH, Bigger JT. Am J Psychiatry 2001;158:1774-82.

Be in the know
Subscribe to Hospital Pharmacy Europe newsletter and magazine