Elsevier

Schizophrenia Research

Volume 39, Issue 1, 23 August 1999, Pages 1-16
Schizophrenia Research

The treatment of tardive dyskinesia—a systematic review and meta-analysis

https://doi.org/10.1016/S0920-9964(99)00021-3Get rights and content

Abstract

This systematic review aimed to collate randomized controlled trials (RCTs) of various interventions used to treat tardive dyskinesia (TD) and, where appropriate, to combine the data for meta-analysis. Clinical trials were identified by electronic searches, handsearches and contact with principal investigators. Data were extracted independently by two reviewers, for outcomes related to improvement, deterioration, side-effects and drop out rates. Data were pooled using the Mantel–Haenzel Odds Ratio (fixed effect model). For treatments that had significant effects, the number needed to treat (NNT) was calculated. From 296 controlled clinical trials, data were extracted from 47 trials. For most interventions, we could identify no RCT-derived evidence of efficacy. A meta-analysis showed that baclofen, deanol and diazepam were no more effective than a placebo. Single RCTs demonstrated a lack of evidence of any effect for bromocriptine, ceruletide, clonidine, estrogen, gamma linolenic acid, hydergine, lecithin, lithium, progabide, seligiline and tetrahydroisoxazolopyridinol. The meta-analysis found that five interventions were effective: l-dopa, oxypertine, sodium valproate, tiapride and vitamin E; neuroleptic reduction was marginally significant. Data from single RCTs revealed that insulin, alpha methyl dopa and reserpine were more effective than a placebo. There was a significantly increased risk of adverse events associated with baclofen, deanol, l-dopa, oxypertine and reserpine. Meta-analysis of the impact of placebo (n=485) showed that 37.3% of participants showed an improvement. Interpretation of this systematic review requires caution as the individual trials identified tended to have small sample sizes. For many compounds, data from only one trial were available, and where meta-analyses were possible, these were based on a small number of trials. Despite these concerns, the review facilitated the interpretation of the large and diverse range of treatments used for TD. Clinical recommendations for the treatment of TD are made, based on the availability of RCT-derived evidence, the strength of that evidence and the presence of adverse effects.

Introduction

Since the 1950s, neuroleptic medication has been extensively used to treat people with chronic mental illnesses such as schizophrenia. These drugs can effectively control positive symptoms such as hallucinations, thought disorder and delusions and, when used as maintenance therapy, can reduce the risk of relapse of psychosis (Schooler and Keith, 1993). Neuroleptic medication, however, has been associated with a wide range of adverse effects such as tardive dyskinesia (TD), a movement disorder characterized by abnormal, repetitive and involuntary movements. TD tends to have an insidious onset, followed by a prolonged but fluctuating course (American Psychiatric Association, 1992). Every year, 4–5% of those who continuously use these drugs begin to show signs of the disorder (American Psychiatric Association, 1992), which can result in considerable social and physical disability (Barnes and Edwards, 1993).

There are few conditions in modern psychiatry that appear to be so consistently resistant to effective interventions as the treatment of neuroleptic-induced tardive dyskinesia. That so many different compounds have been used to treat TD over the last three decades is testament to our ineffectiveness in this area. To choose the best interventions, clinicians are usually guided by reviews in journal articles and textbooks. Over the last 25 years, reviewers have regularly attempted to collate the literature concerning the treatment of TD. Early reviews provided narrative summaries of trials (Kazamatsuri et al., 1972, Mackay and Sheppard, 1982), whereas more recent reviews have compared treatments quantitatively (American Psychiatric Association, 1992, Egan et al., 1997, Jeste and Wyatt, 1982, Jeste et al., 1988)

In recent years, systematic reviews and meta-analyses have been recognized as important tools in guiding clinicians and researchers (Chalmers and Altman, 1995, Kraemer and Pruyn, 1990, Sackett et al., 1997). Systematic reviews should have clear and reproducible methods sections. The rules for inclusion and exclusion of studies should be explicit and as bias-free as possible (Egger et al., 1997), and every effort should be made to identify and extract data from eligible trials.

Those who conduct systematic reviews and meta-analyses also have to decide on the type of trials to include. Randomized controlled trials (RCTs), because of their ability to reduce the risk of bias and confounding (Sackett and Wennberg, 1997, World Health Organization, 1991), are widely regarded as the most stringent research design available to assess the efficacy of an intervention. In the development of evidence-based practice guidelines, it is not unusual for RCTs to be given more weight than other types of controlled trials (e.g. pre-post design) and the combination of a systematic review and meta-analysis of RCT-derived data is accorded the greatest weight (Guyatt et al., 1995).

It has been difficult to compare the large number of different interventions used to treat TD because: (1) many different scales have been used to measure TD, (2) statistical techniques used to assess the impact of the intervention have varied widely, and (3) there has been a lack of uniformity in the reporting of results. Also, many trials have used small sample sizes, leading to a higher probability of Type II error. Meta-analysis uses a uniform method of extracting data from different studies and, by combining the results of several studies, increases the power to identify true clinical effects.

In response to these issues, the Cochrane Collaboration, an international network of clinicians and researchers, was developed to prepare, maintain and disseminate comprehensive and systematic reviews of the effects of all types of health care (Chalmers et al., 1992, Chalmers et al., 1997). The Cochrane Collaboration focuses on state-of-the-art systematic reviews and meta-analysis based on RCT-derived evidence. These reviews are available in CD-ROM format (Cochrane Collaboration, 1998). Over the last two years, the authors have undertaken systematic reviews and meta-analyses under the auspices of the Cochrane Collaboration, which has examined the various treatments of TD. In this paper, we will summarize the essential features of these reviews and compare the efficacy and safety of the many interventions that have been used to treat TD. Our main goal is to provide high-quality objective data on the treatment of TD to health-care providers, consumers and those involved in the development of evidence-based clinical practice guidelines. In addition, we hope that this systematic review and meta-analysis may help researchers identify deficits in our knowledge base about the treatment of TD that may be suitable for future research.

Section snippets

Selection of trials

Several electronic databases (Armstrong, 1993) (Biological Abstracts, Cochrane Controlled Trial Register, EMBASE, LILACS, MEDLINE, and PsycLIT) were systematically searched from January 1966 to September 1997 (details of the search strategies available from the authors). Once identified, each RCT was sought as a citation on the SCISEARCH database in order to identify more potential RCTs (Armstrong, 1993). The reference list of each RCT found was checked for more potential RCTs.

Inclusion criteria

Both authors

Trial allocation

After the removal of duplicates, the search strategies generated 3830 references. The title and abstract of these references were reviewed in order to identify controlled clinical trials. Both authors read the full text of all potential controlled clinical trials, resulting in the identification of 296 trials. These trials, which involved 114 different intervention, were then systematically allocated to one of three categories: Included, Excluded, Awaiting Assessment. Following this process, 47

Discussion

Of the 24 different interventions analyzed, only seven were identified by this systematic review as effective treatments for TD. Table 2 lists interventions according to whether or not we were able to locate RCT-derived evidence. Those with RCT-derived evidence were further subdivided according to the “strength” of the evidence base (meta-analysis based on three or more trials listed under “reasonable evidence base”) and efficacy compared to placebo (equivalent versus superior to placebo).

Acknowledgements

The authors are indebted to the many trialists who assisted this project. We also acknowledge the support of Clive Adams, Geoff Davies, Jon Deeks, Robyn Hayes, Jair Mari, Kirsten Mason, Carmel Meir, Leanne Roberts and Rochelle Seifas for assistance with this project.

References (86)

  • J Ananth et al.

    Baclofen in the treatment of tardive dyskinesia

    Curr. Ther. Res. Clin. E

    (1987)
  • P Armitage

    Br. J. Clin. Pharmacol.

    (1991)
  • C.J Armstrong

    World Databases in Medicine

    (1993)
  • T.R.E Barnes et al.

    The side-effects of antipsychotic drugs. I. CNS and neuromuscular effects

  • C Begg et al.

    Improving the quality of reporting of randomized controlled trials: the CONSORT statement

    J. Am. Med. Assoc.

    (1996)
  • J Browne et al.

    The use of clonidine in the treatment of neuroleptic-induced tardive dyskinesia

    J. Clin. Psychopharmacol.

    (1986)
  • M Burner et al.

    Preliminary observations on the therapeutic action of progabide in tardive dyskinesia

    Brain Dysfunc.

    (1989)
  • O.J.S Buruma et al.

    Tiapride in the treatment of tardive dyskinesia

    Acta Neurol. Scand.

    (1982)
  • D Casey et al.

    Tardive dyskinesia: what is the long-term outcome?

  • I Chalmers et al.

    Getting to grips with Archie Cochrane's agenda

    Br. Med. J.

    (1992)
  • I Chalmers et al.

    Systematic Reviews

    (1995)
  • I Chalmers et al.

    The Cochrane Collaboration

  • G Chatellier et al.

    The number needed to treat: a clinically useful nomogram in its proper context

    Br. Med. J.

    (1996)
  • G Chouinard

    Effects of risperidone in tardive dyskinesia: An analysis of the Canadian multicenter risperidone study

    J. Clin. Psychopharmacol.

    (1995)
  • The Cochrane Library (database on disk and CD ROM)

    (1998)
  • I.B Cookson

    The effects of a 50% reduction of cis(z)-flupenthixol decanoate in chronic schizophrenic patients maintained on a high dose regime

    Int. Clin. Psychopharmacol.

    (1987)
  • J.G Csernansky et al.

    The effect of benzodiazepines on tardive dyskinesia symptoms

    J. Clin. Psychopharmacol.

    (1988)
  • L.M Dabiri et al.

    Effectiveness of vitamin E for treatment of long-term tardive dyskinesia

    Am. J. Psychiatry

    (1994)
  • G Davey-Smith et al.

    Meta-analysis: Unresolved issues and future developments

    Br. Med. J.

    (1998)
  • C de Montigny et al.

    Ineffectiveness of deanol in tardive dyskinesia: A placebo controlled study

    Psychopharmacology

    (1979)
  • M.F Egan et al.

    Treatment of tardive dyskinesia with vitamin E

    Am. J. Psychiatry

    (1992)
  • M.F Egan et al.

    Treatment of tardive dyskinesia

    Schizophr. Bull.

    (1997)
  • M Egger et al.

    Bias in meta-analysis detected by a simple, graphical test

    Br. Med. J.

    (1997)
  • M Egger et al.

    Meta-analysis: Spurious precision? Meta-analysis of observational studies

    Br. Med. J.

    (1998)
  • M Egger et al.

    Meta-analysis: Bias in location and selection of studies

    Br. Med. J.

    (1998)
  • A.M Elkashef et al.

    Vitamin E in the treatment of tardive dyskinesia

    Am. J. Psychiatry

    (1990)
  • G.G Fisk et al.

    The effect of sodium valproate on tardive dyskinesia—Revisited

    Br. J. Psychiatry

    (1987)
  • J.L Fleiss

    The statistical basis of meta-analysis

    Stat. Meth. Med. Res.

    (1993)
  • A Frances et al.

    The Expert Consensus Guidelines Series: treatment of schizophrenia

    J Clin Psychiatry

    (1996)
  • H.L Freeman et al.

    A controlled trial of oxypertine in tardive dyskinesia

    Int. Pharmacopsychiatr.

    (1980)
  • J George et al.

    Double blind controlled trial of deanol in tardive dyskinesia

    Aust. NZ J. Psychiatry

    (1981)
  • J Gerlach et al.

    Effect of baclofen on tardive dyskinesia

    Psychopharmacology

    (1978)
  • P.L Gilbert et al.

    Neuroleptic withdrawal in schizophrenia patients: a review of the literature

    Arch. Gen. Psychiatry

    (1995)
  • Cited by (82)

    • Drug-Induced Extrapyramidal Syndromes: Implications for Contemporary Practice

      2016, Psychiatric Clinics of North America
      Citation Excerpt :

      However, complete and permanent reversibility beyond the withdrawal period may be uncommon.100,101 In a metaanalysis, Soares and McGrath102 reported that 37.3% of patients assigned to placebo across studies showed some improvement in TD, but concluded that evidence was insufficient to support antipsychotic drug cessation or dose reduction in view of the risk for psychotic relapse. Data on the prevalence of TD during continued treatment with antipsychotics have been inconclusive, with some studies showing an increase and others a decrease or no change at all.96

    • Tardive dyskinesia is associated with greater cognitive impairment in schizophrenia

      2013, Progress in Neuro-Psychopharmacology and Biological Psychiatry
      Citation Excerpt :

      TD is characterized by involuntary, hyperkinetic, abnormal movements (Correll and Schenk, 2008; de Leon, 2007; Remington, 2007), the presence of which is associated with poor quality of life, non-adherence with medications, and increased medical morbidity and mortality (Ballesteros et al., 2000; Browne et al., 1996; Youssef and Waddington, 1987). The pathophysiology of TD is not well understood and treatments lack efficacy, therefore prevention and early recognition may help improve treatment outcome (Remington, 2007; Soares and McGrath, 1999). A number of risk factors have been associated with increased risk of developing TD.

    • Management of Psychotropic Medication Side Effects in Children and Adolescents

      2012, Child and Adolescent Psychiatric Clinics of North America
      Citation Excerpt :

      Also, clinicians should be aware of both exacerbation of movements due to withdrawal of the antipsychotic or rebound due to switching to an antipsychotic with lower affinity for receptors.63 Neuroleptic-induced EPS can be treated with anticholinergic agents, benzodiazepines, β-blockers, vitamin E, switching antipsychotic agents, benzodiazepines, amantadine, or naloxone.64–66 Metabolic syndrome was first identified nearly 50 years ago when it was described by the World Health Organization and the National Cholesterol Education Program Adult Treatment Panel.

    View all citing articles on Scopus

    This project was presented at the 9th Biennial Winter Workshop for Schizophrenia (Davos, Switzerland) with the travel expenses funding by FAPESP (Fundação de Amparo à Pesquisa Estado de São Paulo). The project was supported by CAPES (Ministry of Education, Brazil) and the Queensland Health (Australia).

    View full text