BRITISH JOURNAL OF PSYCHIATRY %2001), 179, 290^299 REVIEW ARTICLE

Systematic meta-review of depot antipsychotic alal, 1994) including `mirror image' and in- fluential discontinuation trials e.g. Hirsch drugs for people with schizophrenia{{ et aletal, 1973), and concluded that depots were superior to oral drugs in many respects. Similarly, Glazer & Kane 1992) CLIVE E. ADAMS, MARK K. P. FENTON, SEEMA QURAISHI combined several studies comparing the and ANTHONY S. DAVID incidence of tardive dyskinesia for people on depots with those on oral agents. They concluded that depots were no more harm- ful than oral drugs in this respect. Because non-randomised evaluative studies, includ- ing before and after `mirror image') trials, Background Long-acting depot Antipsychotic medication is the mainstay in repeatedly have been shown to over- antipsychotic medicationis a widely used the effective management of schizophrenia. estimate the effect of experimental inter- These drugs reduce symptoms and, when ventions Chalmers et aletal, 1983; Schulz etet treatment for schizophrenia. used as a maintenance treatment, prevent alal, 1995), these have not been included in Aims To synthesise relevant systematic relapse Davis & Andriukaitis, 1986). Cochrane schizophrenia reviews see Translation of this success into clinical Adams & Eisenbruch, 2000; Coutinho etet Cochrane reviews. practice is attenuated by poor compliance alal, 2000; Quraishi & David, 2000aa±±ee;; MethodMethod TheCochraneDatabasewas Weiden & Olfson, 1995), reasons for QuraishiQuraishi et aletal, 2000).,2000). which include adverse effects, level of searched and summary data were insight, severity of illness, complexity of ObjectiveObjective extracted from randomised controlled treatment regimen and the relationship that The aim of this paper is to provide a patients have with mental health practi- clinical trials of depots. synthesis and quantitative summary of the tioners Fenton et aletal, 1997; Kemp & David, findings of Cochrane depot reviews. ResultsResults Standard dose depot vv.. 1997). Long-acting depot antipsychotics placebo resultedin significantlyless were developed in the 1960s, to promote relapse but more movement disorders. compliance in people with chronic psy- METHOD chotic illnesses Simpson, 1984). They Those on depots vv. oral drugs)showed generally consist of an ester of the anti- Search more global change on one outcome psychotic drug in an oily solution injected The search strategy, methods of selection, measure; relapse and adverse effects intramuscularly every 1±6 weeks. Depots quality assessment, data extraction and showed no difference.Comparisons simplify the treatment process by requiring, assimilation within each review are showed no convincing advantages for one for example, the person to attend for injec- published in The Cochrane Library. The.The tion at a specific clinic, thus guaranteeing depot over another. reader is referred to these reviews for the delivery of a known quantity of medi- explicit details. Conclusions Depot antipsychotics are cation Barnes & Curson, 1994; Davis etet alal, 1994). Apart from overcoming covert SelectionSelection safe and effective.They maymayconfer confer a small non-compliance, the pharmacological All reviews of long-acting depot anti- benefitover oral drugs on global outcome. advantages for depots include the avoid- Those for whom depots are most ance of problems associated with absorp- psychotics for schizophrenia were obtained fromfrom The Cochrane Library Issue 1,Issue1, indicated may not be represented. Large tion and hepatic biotransformation. Disadvantages include concerns over 2000) by searching using the term studies are required to discern differences adverse effects, including tardive dyskinesia SCHIZOPHRENIA and scanning the titles in relapse rates andlong-term adverse and those associated with parenteral of completed reviews. The pre-stated effects, and data on satisfaction, qualityofquality of administration per seperse Johnson, 1984). comparisons of interest were of any long- acting depot antipsychotic medication vv.. life and economics. Many clinicians have promoted the use of depots Glazer & Kane, 1992; Gerlach, placebo orplaceboor vv. oral medication and, finally, high-dose depot vv. low-dose depot, for Declaration of interest Funded by 1995; Kane et aletal, 1998), yet patient and people with schizophrenia or schizo- the NHS HealthTechnology Assessment clinician acceptance is variable, with the phrenia-like illnesses. Outcomes of a prioriapriori programme. A.S.D. is participatingin a mode of delivery being a major stumbling block reviewed in companion paper ± interest were intention-to-treat data on trial funded by Janssen-Cilag and has been WalburnWalburn et aletal, 2001, this issue). Figures death, improvement in global functioning, an advisor for them.The Cochrane on the use of depots are sparse but a UK mental state, behaviour, social functioning, Schizophrenia Group has received funding national household survey found that quality of life, carer burden and incidence of attrition and adverse effects. from pharmaceutical companies, the NHS approximately 29% of 390 non-hospitalised patients with psychotic disorders were and DepartmentofDepartment of Health. prescribed depots Foster et aletal, 1996).,1996). Quality assessment Davis undertook meta-analyses, incor- There was no quality assessment of the {{See pp. 300^307, thisissue. porating several methodologies Davis etet primary reviews from which these data

290

Downloaded from https://www.cambridge.org/core. 23 Sep 2021 at 17:23:35, subject to the Cambridge Core terms of use. SYSTEMATIC META-REVIEW OF DEPOT ANTIPSYCHOTICS

were extracted but the empirical-evidence- sparse see Table 1). Only 111 people have placebo placebo nnˆ51, RR51,RRˆ20.5,20.5, CICIˆ1.3±338;1.3±338; based Cochrane reviews have, in general, been randomised within trials of perphena- NNHNNHˆ3, CI3,CIˆ6.5±1.9).6.5±1.9). been shown to be of higher quality than zine and 117 in bromperidol decanoate others Jadad et aletal, 1998).,1998). studies. On the other hand, 3348people Depot antipsychotics vv. oral.oral have been randomised in trials of fluphena- antipsychotics Data extraction zine decanoate. The duration of studies in the reviews ranged from 2 weeks to 3 years. Death is a rare, inconsistently reported out- Data were extracted from reviews by Most of the included studies employed come. One review presented limited data M.K.P.F. and re-extracted by either operationalised definitions of schizo- and there is no clear effect of either depot C.E.A. or A.S.D. Where disagreement phrenia, which covered several classifi-classifi- or oral antipsychotic nnˆ156, RR156,RRˆ2,2, arose, this was resolved through discussion. cation systems and their revisions. Doses CICIˆ0.19±21). Three reviews present data of depots varied from the very low on global change. Significantly fewer Data analysis 1.25 mg of fluphenazine every 2 weeks) numbers of people allocated to depot All intention-to-treat data were binary out- to the very high 250±1100 mg of fluphena- preparations had no clinically meaningful comes. Risk ratios RRs, random) and their zine weekly±monthly), although analysis change changennˆ127, RR127,RRˆ0.7, CI0.7,CIˆ0.5±0.9;0.5±0.9; 95% confidence intervals CI) were ex- was undertaken only between comparable NNTNNTˆ4,4,CICIˆ2.4±9; see Fig. 2). For out- tracted from the original review and entered dose ranges where appropriate, with most comescomessuchsuch as relapse, study attrition, into RevMan 4.1 http://www.update- falling within British National Formulary needing adjunctive anticholinergic medi- software.com/ccweb/cochrane/revman.htm). ranges British Medical Association & cation and incidence of tardive dyskinesia, These were calculated in preference to odds Royal Pharmaceutical Society of Great no clear differences were demonstrated ratios because they are robust to hetero- Britain, 1999). Reviewers sought clinically betweenbetweenthosethose taking depot and people geneity and more intuitive to clinicians relevant outcomes but only a limited range allocated to oral antipsychotics relapse, BoisselBoissel et aletal, 1999). In turn, where appro- were recorded consistently or presented in a nnˆ844, RR844,RRˆ0.96, CI0.96,CIˆ0.8±1.1).0.8±1.1). priate, summary data from each review usable form. Overall, study attrition was were summated and an overall RR and remarkably low. For example, only about Specific depot antipsychotic the summary 95% CI were calculated. 14% of those randomised to the compari- vv. another depot Where possible, we calculated the numbers sons of one depot with another left the All nine depot reviews contributed to at needed to treat NNT) or harm NNH). For trials early. Trials of oral atypical anti- least one of the outcomes in this compari- a test of heterogeneity we visually inspected psychotic agents have rates of attrition of son. No data were pooled because it was graphs as well as employing the ww22 test.test. 40±60% Thornley & Adams, 1998). impossible to avoid counting data twice: There are some dangers in this overall one review's experimental group was approach. Because it was not possible to another's control. For all outcomes see avoid spurious results by counting partici- Depot antipsychotics vv. placebo.placebo Fig. 3) there were few convincing data that pants twice in the `specific depot vv. other.other depots any real differences exist between depots. depots' comparison, totals are not Understandably, few people have been All data from reviews that compared the produced. Totalling across different randomised within this comparison. Three depot antipsychotic of interest with pharmacological classes of antipsychotics reviews compared depot medication against another, for the outcome of `no important is statistically attractive. Power to demon- placebo bromperidol, fluphenazine and improvement in global functioning' as strate outcomes of interest afforded by haloperidol decanoate). Only one small indexed by Clinical Global Impression summation is increased so that any import- trial within the fluphenazine review CGI) scores Guy, 1976), included the ant effects that the small source trials and reported mortality, with no clear differ- possibility that there were no differences reviews would have missed may be high- ences between groups nnˆ54, RR54,RRˆ5,5, between depots. This also applies to the lighted. Clinically and pharmacologically, CICIˆ0.25±99). One review reported on re- outcome of leaving the study early 25% however, such totalling may not make lapse fluphenazine decanoate vv. placebo),.placebo), attritionattritioninin the experimental groups). The sense. Clinicians who choose to prescribe with the results favouring the active drug outcome of `mental state ± relapse' a specific depot may not be interested in a nnˆ415, RR415,RRˆ0.3, CI0.3,CIˆ0.22±0.4; NNTˆ2,2, showed that zuclopenthixol decanoate summary statistic across all depots. In some CICIˆ1.8±2.6; see Fig. 1). Three reviews pre- was statistically superior to the control circumstances effects could cancel out in an sented data for the numbers leaving the depots largely fluphenazine decanoate) overall summary statistic and causative and studies early. Significantly more people nnˆ296, RR296,RRˆ0.64, CI0.64,CIˆ0.4±0.9, NNTˆ8,8, protective effects could be masked or can- taking depot medication stayed in the CICIˆ5±53), but this could be a function of celled out. Data using summated totals studies than those receiving placebo publication bias see Discussion). must therefore be interpreted with caution. nnˆ152, RR152,RRˆ0.43, CI0.43,CIˆ0.27±0.71). Two reviews reported on the adverse effects of RESULTSRESULTS blurred vision or dry mouth. Curiously, High-dose depot vv.. these symptoms were more frequent in the standard dose, and The original reviews placebo group nnˆ52, RR52,RRˆ0.16,0.16, standard dose vv.lowdose Details of the studies included and excluded CICIˆ0.03±0.8; NNTˆ3, CI3,CIˆ2±9). When There are limited data, but reviews found in specific reviews can be found in the data were reported in the review as `move- no differences between high-dose 250 mg individual Cochrane publications. Overall, ment disorders ± general',statistical signifi- of fluphenazine weekly; 200 mg of flupen- composite data for some compounds were cance was achieved in favour of those taking tixol every 2 weeks) vv. standard-dose

291

Downloaded from https://www.cambridge.org/core. 23 Sep 2021 at 17:23:35, subject to the Cambridge Core terms of use. ADAMS ET AL Global improvement ;CGI) Mental state ;BPRS, CPRS) Behaviour ;NOSIE) Leaving the study early Use of additional medication Side-effects Death Leaving the study early Relapse Use of additional medication Mental state ;BPRS, CPRS, HRSD) Side-effects ;SAS) Global improvement ;CGI) Mental state ;BPRS, CPRS) Behaviour ;NOSIE) Leaving the study early Use of additional medication Side-effects Global effect ;CGI) Leaving the study early Mental state ;use of additional medication) Side-effects ;DOTES) Outcomes Death Leaving the study early Relapse Use of additional medication Mental state ;BPRS, CPRS, HRSD) Side-effects ;SAS) Global effect ;CGI) Leaving the study early Mental state ;use of additional medication) Side-effects ;DOTES) Outcomes 39)39) 58) 58) ˆ nn ˆ 1963) 1963) nn 16)16) ˆ ˆ nn 10) 10) nn 70) 70) ˆ ˆ nn 48)48) nn 36) 36) ˆ ˆ nn nn 23)23) 93)93) ˆ ˆ nn 30) 30) 19) 19) nn 27) 27) ˆ ˆ nn ˆ nn 17)17) nn ˆ nn 184)184) ˆ nn 396) 396) ˆ nn 23)23) 142)142) 359) 359) 139) 139) ˆ ˆ ˆ nn 10) 75)75) ˆ 10) nn nn nn ˆ ˆ nn nn 96)96) 48)48) ˆ ˆ nn nn 6.25^400 mg/2^5 weeks ; ^ ek 24weeks ; 2^4 weeks ;2^4 ;; range2.5^60mg/day; 200 mg/4 weeks to 600 mg/2 weeks ; 64^400 mg ; 6.25 mg) standard^high dose 25^1100 mg/2^4 weeks ; 05 g;05 mg ; 10^50 mg ;10^50 ;; ^ ek 23weeks ; 2^3 weeks ;2^3 6.25^400 mg/2^5 weeks ; range2.5^60mg/day; 200 mg/4 weeks to 600 mg/2 weeks ; 64^400 mg ; 6.25 mg) standard^high dose 25^1100 mg/2^4 weeks ; 9. Haloperidol decanoate: mean dose 109.4 mg, range 15^900 mg/ 8. Fluspirilene decanoate: 3^20 mg/week ; 7. Fluphenazine enanthate: dose range 2.5^387.5 mg/2^4 weeks 6. Fluphenazine hydrochloride ;oral): mean dose 18.9 mg, dose 5. Flupentixol: dose range 30^40 mg/2^4 weeks ; 4. Clopenthixol decanoate: mean dose 220 mg/3^4 weeks, range 2. Bromperidol decanoate: mean dose 242 mg/month, range 3. Chlorpromazine: dose range 50^100 mg/day ; 1. Fluphenazine decanoate: mean dose 51.73 mg, range ;low 1.25^ 6. Pipothiazine: mean dose 100 mg/month ; 5. Penfluridol: mean dose 20 mg/week ; 4. Haloperidol decanoate: mean dose 151mg/injection ; 3. Flupenthixol decanoate: mean dose 25 mg/3 weeks, range 2. Clopenthixol decanoate: dose range 50^600 mg/2^4 weeks 1. Flupentixol decanoate: dose range 9 mg/2^3 weeks to 300 mg/ . Paeo;.lcb ; 4. Placebo ;4.Placebo 9. Haloperidol decanoate: mean dose 109.4 mg, range 15^900 mg/ 8. Fluspirilene decanoate: 3^20 mg/week ; 3. Haloperidol decanoate: mean dose 119 mg/month ; 7. Fluphenazine enanthate: dose range 2.5^387.5 mg/2^4 weeks 2. Fluphenazine decanoate: dose range 16.7^300 mg/month ; 6. Fluphenazine hydrochloride ;oral): mean dose 18.9 mg, dose 1. Bromperidol decanoate: dose range 50^242 mg/month ; 5. dose Flupentixol: range 30^40 mg/2^4 weeks ; 4. decanoate: mean Clopenthixol dose 220 mg/3^4 weeks, range 2. Bromperidol decanoate: mean dose 242 mg/month, range 3. Chlorpromazine: dose range 50^100 mg/day ; 1. Fluphenazine decanoate: mean dose 51.73 mg, range ;low 1.25^ 6. Pipothiazine: mean dose 100 mg/month ; 5. Penfluridol: mean dose 20 mg/week ; 4. Haloperidol decanoate: mean dose 151mg/injection ; 3. Flupenthixol decanoate: mean dose 25 mg/3 weeks, range 2. Clopenthixol decanoate: dose range 50^600 mg/2^4 weeks 1. Flupentixol decanoate: dose range 9 mg/2^3 weeks to 300 mg/ 3. Haloperidol decanoate: mean dose 119 mg/month ; 2. Fluphenazine decanoate: dose range 16.7^300 mg/month ; 1. Bromperidol decanoate: dose range 50^242 mg/month ; 14. Trifluoperazine: dose 10 mg/day ; 12. Pipotiazine palmitate: mean dose 88.3 mg, dose range 13. Placebo ; 1Pmzd:ds m/a^ek ag1^0g; range10^60mg 8mg/day^week, dose 11.Pimozide: 10. Penfluridol: dose range 20^160 mg ; 14. Trifluoperazine: dose 10 mg/day ; 12. Pipotiazine palmitate: mean dose 88.3 mg, dose range 13. Placebo ; 11.Pimozide:dose8mg/day^week,range10^60mg; 10. Penfluridol: dose range 20^160 mg ; Intervention Intervention 2^392^39 55 1054 F 1054 F 193 F, unknown 193 F, unknown 42 F 42 F 44 44 44 55 M;55M; 373 M; 38M11 M; 1318 M;1318 373 M; 3348 3348 615 615 117117 44 44 44 ˆ ˆ nn ˆ nn nn 48,48, 4, 4, 15,15, schizoaffective disorder schizoaffective disorder years in1trialin1trial years ˆ ˆ ˆ Age: range 24^70 years Setting: community and in-patient Gender: NN Duration of illness: range Diagnosis: schizophrenia and Age: range 17^79 years Setting: community and in-patient Diagnosis: schizophrenia Duration of illness: 1^29 years NN Gender: Age: range 24^70 years Setting: community and in-patient Gender: Duration of illness: range Setting: community and in-patients Diagnosis: schizophrenia and Participants Diagnosis: schizophrenia Age: range 20^65 years Gender: NN Age: range 17^79 years Setting: community and in-patient Diagnosis: schizophrenia Duration of illness: 1^29 years Gender: Setting: community and in-patients Participants Diagnosis: schizophrenia Age: range 20^65 years Gender: blinding blinding method in5studies method in5studies Allocation: all 48 studies randomised Blinding: varying degrees of double Duration: 2 weeks^2 years Two studies used a crossover Allocation: all 15 studies randomised Blinding: double, no further details Duration: 8 weeks^2 years Informed consent from participants Allocation: all 4 studies randomised Blinding: double, no further details Duration: 6 months^1 year Allocation: all 48 studies randomised Blinding: varying degrees of double Duration: 2 weeks^2 years Two studies used a crossover Allocation: all 15 studies randomised Blinding: double, no further details Duration: 8 weeks^2 years Informed consent from participants Allocation: all 4 studies randomised Blinding: double, no further details Duration: 6 months^1 year Methods Summary of included reviews Summary of included reviews decanoatedecanoate decanoatedecanoate depot depot Fluphenazine Flupentixol Bromperidol a ble 1a l e 1 Ta b l e 1Tab Review Fluphenazine Flupentixol Bromperidol Review Methods

292

Downloaded from https://www.cambridge.org/core. 23 Sep 2021 at 17:23:35, subject to the Cambridge Core terms of use. SYSTEMATIC META-REVIEW OF DEPOT ANTIPSYCHOTICS ContinuedContinued Krawiecka Scale, MADRS) UKU) Krawiecka Scale, MADRS) UKU) Death Global impression ;CGI) Leaving the study early Use of additional medication Side-effects Death Global impression ;CGI) Mental state ;BPRS, CPRS, depression, Behaviour ;Wing Ward Scale) Leaving the study early Use of additional medication Side-effects ;AIMS, Bordeleau Scale, SAS, Global effect ;CGI) Leaving the study early Side-effects ;UKU) Death Global impression ;CGI) Leaving the study early Use of additional medication Side-effects Global effect ;CGI) Mental state ;BPRS) Leaving the study early Use of additional medication Side-effects ;Bordeleau Scale, HRSD) Death Global impression ;CGI) Mental state ;BPRS, CPRS, depression, Behaviour ;Wing Ward Scale) Leaving the study early Use of additional medication Side-effects ;AIMS, Bordeleau Scale, SAS, Global effect ;CGI) Leaving the study early Side-effects ;UKU) Global effect ;CGI) Mental state ;BPRS) Leaving the study early Use of additional medication Side-effects ;Bordeleau Scale, HRSD) 23) 23) 20)20) ˆ nn ˆ 26)26) 31)31) nn ˆ ˆ nn nn 20)20) ˆ nn 15) 15) ˆ nn 11)11) ˆ nn 39)39) ˆ nn 24) 87)87) 238) 71)71) 42)42) 284)284) 279)279) 24) 238) 111)111) 125) 13)13) 160)160) 125) ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ ˆ nn nn nn nn nn nn nn nn nn nn nn ;; ;; ;; ;; ;; ;; ;; ;; ;; ;; ;; 3. Perphenazine enanthate: mean dose 108.5 mg/2 weeks 2. Clopenthixol decanoate: dose range 50^800 mg/2 weeks 1. Perphenazine decanoate: dose range 20^600 mg/2 weeks 6. Zuclopenthixol decanoate: mean dose 284 mg/month ; 4. Pipotiazine palmitate: dose range 50^125 mg/month ; .Paeo;.Placebo ; 5. Placebo ;5. 3. ;oral): Haloperidol dose not specified ; 2. Fluphenazine decanoate: dose range 2.5^300 mg/2^4 weeks 1. Haloperidol decanoate: dose range 15^900 mg/2^4 weeks 5. Pipotiazine undecylenate: mean dose 103.8 mg/2 weeks 4. Fluphenazine enanthate: mean dose 7.55 mg/week ; 2. Chlorpromazine: dose range 370^720 mg/day ; 1. Fluspirilene decanoate: dose range 1^23 mg/1^2 weeks 3. Fluphenazine decanoate: dose range 25^150 mg/2^3 weeks 5. Pipotiazine palmitate: dose range 25^250 mg/2^4 weeks 4. Fluspirilene decanoate: dose range 1^14 mg/week ; 3. decanoate: dose Fluphenazine range 2.5^500 mg/2^4 weeks 2. Chlorpromazine: mean dose 388 mg/day ; 1. Fluphenazine enanthate: dose range 3.5^387.5 mg/2^4 weeks 3. Perphenazine enanthate: mean dose 108.5 mg/2 weeks 2. Clopenthixol decanoate: dose range 50^800 mg/2 weeks 1. Perphenazine decanoate: dose range 20^600 mg/2 weeks 6. Zuclopenthixol decanoate: mean dose 284 mg/month ; 4. Pipotiazine palmitate: dose range 50^125 mg/month ; 3. ;oral): Haloperidol dose not specified ; 2. Fluphenazine decanoate: dose range 2.5^300 mg/2^4 weeks 1. Haloperidol decanoate: dose range 15^900 mg/2^4 weeks 5. Pipotiazine undecylenate: mean dose 103.8 mg/2 weeks 4. Fluphenazine enanthate: mean dose 7.55 mg/week ; 2. Chlorpromazine: dose range 370^720 mg/day ; 1. Fluspirilene decanoate: dose range 1^23 mg/1^2 weeks 3. Fluphenazine decanoate: dose range 25^150 mg/2^3 weeks 5. Pipotiazine palmitate: dose range 25^250 mg/2^4 weeks 4. Fluspirilene decanoate: dose range 1^14 mg/week ; 3. decanoate: dose Fluphenazine range 2.5^500 mg/2^4 weeks 2. Chlorpromazine: mean dose 388 mg/day ; 1. Fluphenazine enanthate: dose range 3.5^387.5 mg/2^4 weeks ^5yas^5years 2^25 years2^25 7 15F 175 F175 87 F 87 F 137 F 137 F 55 60 years60years 44 44 44 2years 44 2years 44 264 M;264M; 98 M;98M; 135 M; 135 M; 451451 455455 236236 290 290 44 44 44 ˆ ˆ ˆ ˆ nn nn nn nn 2, 7,7, 2, 11, 14, 11, 14, psychosis schizoaffective disorder hospitalised psychosis schizoaffective disorder hospitalised ˆ ˆ ˆ ˆ Setting: community and in-patient Age: range 18 to Gender:Gender: NN Diagnosis: schizophrenia or acute Duration of illness: Setting: community and in-patient Diagnosis: schizophrenia or Duration of illness: 0^38 years NN Age: range 18^66 years Gender:Gender: Setting: in-patient and community Diagnosis: schizophrenia Duration of illness: 2^39 years NN Age: range 16^80 years Gender:Gender: Age: range 17^65 years Setting: community and in-patient NN Diagnosis: schizophrenia Duration of illness: acute to Setting: community and in-patient Age: range 18 to Diagnosis: schizophrenia or acute Duration of illness: Setting: community and in-patient Diagnosis: schizophrenia or Duration of illness: 0^38 years g:rne86 years range18^66 Age: Setting: in-patient and community Diagnosis: schizophrenia Duration of illness: 2^39 years g:rne68 years range16^80 Age: Age: range 17^65 years Setting: community and in-patient Diagnosis: schizophrenia Duration of illness: acute to Allocation: both studies randomised Blinding: double, no further details Duration: range 6 weeks^6 months Allocation: all 11 studies randomised Blinding: double, no further details Duration: 16^60 weeks Blinding: double, no further details Duration: 4 weeks^6 months Allocation: all 14 studies randomised Blinding: double, no further details Duration: 2 weeks^1 year One study used a crossover design Allocation: both studies randomised Blinding: double, no further details Duration: range 6 weeks^6 months Allocation: all 11 studies randomised Blinding: double, no further details Duration: 16^60 weeks Allocation: all 7 studies randomised Blinding: double, no further details Duration: 4 weeks^6 months Allocation: all 14 studies randomised Blinding: double, no further details Duration: 2 weeks^1 year One study used a crossover design depot decanoatedecanoate enanthate depot enanthate Perphenazine HaloperidolHaloperidol Fluspirilene Fluphenazine Perphenazine Fluspirilene Allocation: all 7 studies randomised Fluphenazine

293293

Downloaded from https://www.cambridge.org/core. 23 Sep 2021 at 17:23:35, subject to the Cambridge Core terms of use. ADAMS ET AL ,1962); SAS, ,1962); SAS, et al et al ,1967);CGI,Clinical ,1967); CGI,Clinical et al et al DOTES) DOTES) Death Global impression ;CGI) Relapse Leaving the study early Use of additional medication Side-effects ;UKU) Discharge Global impression ;CGI) Mental state ;BPRS, HRSD) Leaving the study early Use of additional medication Side-effects ;AIMS, Bordeleau Scale, Outcomes Death Global impression ;CGI) Relapse Leaving the study early Use of additional medication Side-effects ;UKU) Discharge Global impression ;CGI) Mental state ;BPRS, HRSD) Leaving the study early Use of additional medication Side-effects ;AIMS, Bordeleau Scale, Outcomes 365) 365) ˆ nn ohm16) odla cl ;Bordeleau Scale Bordeleau l & Gorham,1962); ohm16) odla Scale ;Bordeleau l Bordeleau & Gorham,1962); 87) 87) ˆ nn 198)198) 87) 87) 21) 21) ˆ ˆ ˆ nn nn nn 17)17) , 1978); Observation Scale Nurses' NOSIE, for In-patient ;Honigfeld Evaluation 13) , 1978); Observation Scale Nurses' NOSIE, for In-patient ;Honigfeld Evaluation 13) ˆ nn ˆ nn et al et al 85) 28)28) 48)48) 85) ˆ ˆ ˆ nn nn nn ;; ;; ^ ek 24weeks ; 2^4 weeks ;2^4 ;; 4. Perphenazine enanthate: dose range 20^600 mg/2 weeks .aoeio decanoate: dose3.Haloperidol range 39^200 mg/4 weeks 1. Zuclopenthixol decanoate: dose range 100^600 mg/ 2. Flupentixol palmitate: dose range 25^300 mg/4 weeks 6. Oral antipsychotics ;various) ; 5. Haloperidol decanoate ; 4. Fluspirilene ; 3. Fluphenazine enanthate ; 2. Fluphenazine decanoate ; 1. Pipotiazine palmitate and undecylenate ; Intervention 4. Perphenazine enanthate: dose range 20^600 mg/2 weeks 3.Haloperidol decanoate: dose range 39^200 mg/4 weeks 1. Zuclopenthixol decanoate: dose range 100^600 mg/ 2. Flupentixol palmitate: dose range 25^300 mg/4 weeks 6. Oral antipsychotics ;various) ; 5. Haloperidol decanoate ; 4. Fluspirilene ; 3. Fluphenazine enanthate ; 2. Fluphenazine decanoate ; 1. Pipotiazine palmitate and undecylenate ; Intervention 3^34 years 3^34 years , 1987); Wing Ward Scale ; Wing, 1961). , 1987); Wing Ward Scale ; Wing, 1961). 55 , 1978); DOTES, Dosage Record and Treatment Emergent Symptom Scale ;Guy,1976); Hamilton HRSD, Rating Scale for Depression ;Hamilton, , 1978); DOTES, Dosage Record and Treatment Emergent Symptom Scale ;Guy,1976); HRSD, Hamilton Rating Scale for Depression ;Hamilton, et aletal 2years 2years 0 25F 205 F205 71 F 71 F et al et al 44 44 44 380 M; 380 M; 9 ;9 M; 197 M;197 771771 332 332 44 44 ˆ ˆ nn nn 4,4, 14,14, ˆ ˆ NN Setting: community and in-patient Diagnosis: schizophrenia Duration of illness: Age: range 20^65 years Gender: g:rne86 years range18^69 Age: Gender: Informed consent given in 2 studies Setting: community and in-patient NN Diagnosis: schizophrenia Duration of illness: range Participants Setting: community and in-patient Diagnosis: schizophrenia Duration of illness: Age: range 20^65 years Gender: g:rne86 years range18^69 Age: Gender: Informed consent given in 2 studies Setting: community and in-patient Diagnosis: schizophrenia Duration of illness: range Participants , 1977); Depression Rating Montgomery^Ðsberg Scale MADRS, ;Montgomery , 1977); MADRS, Montgomery^Ðsberg Depression Rating Scale ;Montgomery et al et al blinding blinding blinding blinding Allocation: all 4 studies randomised Blinding: varying degrees of double Duration: 12 weeks^1 year Methods Allocation: all 14 studies randomised Blinding: varying degrees of double Duration:range11weeks^3years Allocation: all 4 studies randomised Blinding: varying degrees of double Duration: 12 weeks^1 year Allocation: all 14 studies randomised Blinding: varying degrees of double Duration:range11weeks^3years Methods , of number participants; M, males; Scale BPRS, F, Movement Brief Rating ;Guy,1976); females; Involuntary Psychiatric Scale AIMS, Abnormal ;Overal , of number participants; M, males; Scale BPRS, F, Movement Brief Rating ;Guy,1976); females; Scale Involuntary Psychiatric AIMS, Abnormal ;Overal nn Continued Continued depot depot depot depot , number of trials; ,numberoftrials; Zuclopenthixol Simpson ^Angus Scale ;Simpson & Angus, 1970); UKU, UKU Side Effects Rating Scale ;Lingjaerde 1960); Krawiecka Scale ;Manchester Scale) ;Krawiecka PipotiazinePipotiazine Global Impression ;Guy,1976); CPRS, Comprehensive Psychopathological Rating Scale ;Ðsberg able 1a l e 1 Ta b l e 1Tab Review NN Zuclopenthixol Simpson^Angus Scale ;Simpson & Angus, 1970); UKU UKU, Side Effects Rating Scale ;Lingjaerde 1960); Scale Krawiecka ;Manchester Scale) ;Krawiecka Global Impression ;Guy,1976); CPRS, Comprehensive Psychopathological Rating Scale ;Ðsberg Review

294

Downloaded from https://www.cambridge.org/core. 23 Sep 2021 at 17:23:35, subject to the Cambridge Core terms of use. SYSTEMATIC META-REVIEW OF DEPOT ANTIPSYCHOTICS

DepotDepot Placebo RRRR WeightWeight RRRR StudyStudy ;;n/N)) ;;n/N)) ;95% CI random) ;%) ;95% CI random)

01Mental state: General ^ relapse Fluphenazine decanoate 38/210 122/205122/205 100.0100.00.30 ;0.22^0.41) Subtotal ;95% CI) 38/21038/210 122/205122/205 ^ 100.0100.00.30 ;0.22^0.41)

Test for overall effect zzˆ777.55, PP550.00001

02Leaving the study early Bromperidol decanoate 2/10 5/105/10 12.512.50.40 ;0.10^1.60) Fluphenazine decanoate 9/45 17/4517/45 49.949.90.53 ;0.26^1.06) Haloperidol decanoate 5/20 16/2216/22 37.537.50.34 ;0.15^0.77) Subtotal ;95% CI) 16/7516/75 38/7738/77 ^ 100.00.43 ;0.27^0.71) Test for heterogeneity ww22ˆ0.65, d.f.d.f.0.65, ˆ2,2, PPˆ0.720.72 Test for overall effect zzˆ773.33, PPˆ0.0009

03Side-effects: 1. Dry mouth/blurred vision Bromperidol decanoate 0/103/10 3 33.10.14 ;0.01^2.45) Haloperidol decanoate 1/166/16 66.966.90.17 ;0.02^1.23) Subtotal ;95% CI) 1/269/26 100.00.16 ;0.03^0.81) Test for heterogeneity ww22ˆ0.01, d.f.ˆ1,1, PPˆ0.930.93 Test for overall effect zzˆ772.21,2.21, PPˆ0.03

04Side-effects: 2. Movement disorders ^ general

Fluphenazine decanoate 8/23 0/280/28 3 100.0100.020.54 ;1.25^337.96) Subtotal ;95% CI) 8/238/23 0/280/28 ^3 100.0100.020.54 ;1.25^337.96) Test for heterogeneity ww22ˆ0.0, d.f.ˆ00 Test for overall effect zzˆ2.12,2.12, PPˆ0.03

0.010.1 110 1 10 100100 Favours depot Favours placebo

Fig. 11Fig. Depot antipsychoticvv. placebo depot: all outcomes.

12.5±50 mg of fluphenazine every 2 weeks; a depot is most indicated were included, depot 21% vv. 49% over the same time 40 mg of flupentixol biweekly) depot anti- however, is less certain. It would be period) can be interpreted as a positive psychotics for global outcome, mental problematic to recruit those who are reluc- outcome, assuming that those who left state, adverse effects or attrition. The tant with a prescription for oral antipsy- early were unlikely to be well. Data from estimates of effect all had wide confidence chotics into any clinical trial. The reviews within this comparison suggest that the intervals. Within the standard dose vv. low.low mostly comprised those who were stable adverse effects of blurred vision or dry dose comparison, most data were available on oral medication. Some participants mouth are not good indicators of anti- for the outcome of relapse nnˆ638). Pooled whose course of illness had not been helped psychotic activity because they are more data across three phenothiazine prepara- previously by a variety of medications were frequent in the placebo group. Unsurpris- tions flupentixol, fluphenazine decanoate included, but it is unclear whether these ingly, drugs such as fluphenazine decanoate and enanthate) suggest that the standard people were non-compliant or unresponsive are associated with movement disorders. dose 12.5±5012.5±50 mg every 2 weeks) is more to treatment. Studies that compared people We have calculated that, on average, be- effective than the low doses 1.25±25 mg who were stable on oral medication and tween two and seven people have to be every 2 weeks) RRˆ2.5, CI2.5,CIˆ1.1±5.9;1.1±5.9; then were randomised to receive either given depot for one person to suffer sig- NNTNNTˆ7, CI7,CIˆ5±12). Although no clear depot or inactive placebo, such that the nificant general movement disorders differences were demonstrated between comparison group are undergoing discon- NNHNNHˆ3, CI3,CIˆ2±6.5), which is admittedly the standard dose and low dose on global tinuation of treatment, were not included a crude index. functioning, attrition and adverse effects in this overview. movement disorders), data are limited. Depot antipsychotics Depot antipsychotics vv. oral antipsychotics DISCUSSION vv. placebo depots An underlying assumption in psychiatric Currently, it would be difficult to undertake therapeutics is that people with serious Generalisability a trial comparing placebo to neuroleptic mental illnesses may not take oral medi- This overview collates a great deal of trial depotdepotinin the treatment of schizophrenia, cation reliably, resulting in relapse. If this data. All trial populations were slightly given the availability of effective treat- assumption is correct, then the comparison different and this clinical heterogeneity ments. Even with the limited data of relapse rates should demonstrate an may mean that at least some participants, nnˆ415), fluphenazine decanoate clearly advantage for those on depots vv. oral drugs. treatment regimens and circumstances reduces relapse between 12 weeks and 2 Although the advantage on one outcome should resemble those seen in everyday years NNTˆ2). That significantly more measure in favour of depots was statisti- practice. Whether those patients for whom people stayed in the study if allocated to cally significant global improvement:

295295

Downloaded from https://www.cambridge.org/core. 23 Sep 2021 at 17:23:35, subject to the Cambridge Core terms of use. ADAMS ET AL

DepotDepot Oral RR Weight RR Study ;;n/N)) ;;n/Nn/N)) ;95% CI random) ;%) ;95% CI random)

01 DeathDeath

Fluphenazine decanoate 2/781/78 3 100.02.00 ;0.19^21.61) Subtotal ;95% CI) 2/781/78 ^3 100.02.00 ;0.19^21.61) Test for overall effect zzˆ0.57, PPˆ0.60.6 02Global functioning: No important global change Fluphenazine decanoate 22/3834/36 67.60.61 ;0.46^0.81) Fluphenazine enanthate 5/167/15 6.60.67 ;0.27^1.66) Haloperidol decanoate 8/119/11 25.90.89 ;0.56^1.40) Subtotal ;95% CI) 35/6535/65 50/6250/62 ^ 100.0100.00.68 ;0.54^0.86) Test for heterogeneity ww22ˆ1.85, d.f.d.f.1.85, ˆ2,2, PPˆ0.40 Test for overall effect zzˆ773.27, PPˆ0.001 03Mental state: General ^ relapse Fluphenazine decanoate 129/339 142/345142/345 92.90.92 ;0.77^1.11) Fluspirilene decanoate 2/20 2/202/20 0.91.00 ;0.16^6.42) Pipotiazine palmitate 15/61 10/6310/63 6.21.55 ;0.76^3.18) Subtotal ;95% CI) 146/420146/420154/428 100.00.96 ;0.80^1.14) ^ Test for heterogeneity ww22ˆ1.88, d.f.ˆ2,2, PPˆ0.39 Test for overall effect zzˆ770.50,0.50, PPˆ0.60.6

04Leaving the study early Flupentixol decanoate3/30 1/301/30 3 1.23.00 ;0.33^27.24) Fluphenazine decanoate 85/298 77/31077/310 82.882.81.15 ;0.88^1.50) Fluspirilene decanoate 2/20 2/202/20 1.71.00 ;0.16^6.42) Pipotiazine palmitate 16/8515/81 14.31.02 ;0.54^1.92) Subtotal ;95% CI) 106/433106/433 95/44195/441 ^ 100.0100.01.14 ;0.90^1.45) Test for heterogeneity ww22ˆ0.89, d.f.d.f.0.89, ˆ3, PPˆ0.830.83 Test for overall effect zzˆ1.06, PPˆ0.30.3 05Side-effects: 1. Movement disorders ^ general ^ needing anticholinergic medication medication Flupentixol decanoate 19/3016/30 13.41.19 ;0.77^1.83) Fluphenazine decanoate 54/75 54/8054/80 32.51.07 ;0.87^1.31)

Fluspirilene decanoate 19/20 14/2014/20 21.821.81.36 ;1.00^1.84) Haloperidol decanoate 3/111/11 3 0.73.00 ;0.37^24.58) Pipotiazine palmitate 42/61 49/6349/63 31.60.89 ;0.71^1.10) Subtotal ;95% CI) 137/197137/197 134/204134/204 ^ 100.0100.01.08 ;0.90^1.30) Test for heterogeneity ww22ˆ6.46, d.f.ˆ4,4, PPˆ0.17 Test for overall effect zzˆ0.87,0.87, PPˆ0.40.4 06Side-effects: 2. Movement disorders ^ tardive dyskinesia Fluphenazine decanoate 9/7216/76 81.20.59 ;0.28^1.26) Pipotiazine palmitate 3/61 3/633/63 18.81.03 ;0.22^4.92) Subtotal ;95% CI) 12/13319/139 ^ 100.00.66 ;0.33^1.30) Test for heterogeneity ww22ˆ0.39, d.f.ˆ1,1, PPˆ0.53 Test for overall effect zzˆ771.21,1.21, PPˆ0.20.2

0.10.2 151 51010 Favours depot Favours oral

Fig. 22Fig. Depot antipsychoticvv. oral antipsychotic: all outcomes.

NNTNNTˆ4, CI4,CIˆ2±9), other important out- Specific depot antipsychotic the comparator drugs were high and, comes such as relapse, attrition and adverse vv. another depot pharmacologically,there are no grounds to effects were not. Reviews involving over suspect any superiority. On the other hand, 800 participants did not demonstrate a Many of these comparisons can be seen as being one of the newest preparations, it has statistically significant difference between fulfilling the need to market a new not been used as the comparator depot in depots and oral medications RRˆ0.9,0.9, substance rather than answering any rele- any other trial Gilbody & Song, 2000). CICIˆ0.8±1.2) in terms of relapse, despite vant clinical question. No differences were By the same token, this may explain, to good statistical power. It could be argued seen on any global measures of change. some extent, the poor results for fluphena- that those participating in trials were All nine reviews reported data on relapse. zine compounds, at least when it comes to reasonably compliant with oral medi- One found a statistically significant result adverse effects. These compounds have cations so that the demonstration of anyany in favour of zuclopenthixol decanoate been used more than any other as the advantages to depot and absence of dis- NNTNNTˆ8,CI8,CIˆ5±53). Unlike the other control drugs and these data may be the advantages) is noteworthy. Trials suggest depots, this finding in favour of zuclo- summation of a publication/reporting bias.bias. that adverse effects, reported as the proxy penthixol was consistent across the out- outcome of `needing additional anti- comes of leaving the study early and High-dose depot cholinergic medication', occur in about needing additional anticholinergic drugs. vv. standard dose, and two-thirds of people on antipsychotics, It is feasible that zuclopenthixol decanoate standard dose vv.lowdose whether administered by depot or given is indeed a better depot in terms of the out- Data from trials support the clinical orally.orally. comescomesmeasured,measured, although relapse rates in impression that there is no clear advantage

296

Downloaded from https://www.cambridge.org/core. 23 Sep 2021 at 17:23:35, subject to the Cambridge Core terms of use. SYSTEMATIC META-REVIEW OF DEPOT ANTIPSYCHOTICS

Named Depot Control depot RRRR RRRR StudyStudy ;;n/Nn/N)) ;;n/N)) ;95% CI random) ;95% CI random)

01 DeathDeath Fluphenazine decanoate 1/19 0/19 3 3.00 ;0.13^69.32) Haloperidol decanoate 0/45 1/521/52 3 0.38 ;0.02^9.20) Perphenazine decanoate 0/85 1/871/87 3 0.34 ;0.01^8.26) Zuclopenthixol decanoate 2/123 0/113 " 4.60 ;0.22^94.74) 02Global functioning: No important improvement Bromperidol decanoate 3/15 2/16 1.60 ;0.31^8.29) Fluphenazine decanoate 38/55 32/5532/55 1.19 ;0.89^1.58) Fluspirilene decanoate 2/25 5/255/25 3 0.40 ;0.09^1.87) Perphenazine decanoate 50/85 45/8745/87 1.14 ;0.87^1.49) Pipotiazine palmitate 86/92 91/95 0.98 ;0.91^1.05) 03Mental state: General ^ relapse Bromperidol decanoate 9/33 2/342/34 " 4.64 ;1.08^19.87) Flupentixol decanoate 39/131 34/149 1.30 ;0.88^1.94) Fluphenazine decanoate 79/446 83/45283/452 0.96 ;0.73^1.27) Fluphenazine enanthate 7/42 5/475/47 1.57 ;0.54^4.57) Fluspirilene decanoate 6/75 10/6510/65 0.52 ;0.20^1.35) Haloperidol decanoate 26/155 23/162 1.18 ;0.71^1.98) Perphenazine decanoate 37/85 29/87 1.31 ;0.89^1.92) Pipotiazine palmitate palmitate 41/212 39/205 1.02 ;0.69^1.51) Zuclopenthixol decanoate 33/153 48/143 0.64 ;0.44^0.94) 04Leaving the study early Bromperidol decanoate 10/48 5/49 2.04 ;0.75^5.53) Flupentixol decanoate 27/89 24/9924/99 1.25 ;0.78^2.00) Fluphenazine decanoate 112/464 108/480108/480 1.07 ;0.85^1.35) Fluphenazine enanthate 8/57 17/6217/62 0.51 ;0.24^1.09) Fluspirilene decanoate 6/886/88 10/7810/78 0.53 ;0.20^1.40) Haloperidol decanoate 34/187 33/18433/184 1.01 ;0.66^1.56) Perphenazine decanoate 37/8537/85 29/87 1.31 ;0.89^1.92) Pipotiazine palmitate palmitate 74/23174/231 54/22454/224 1.33 ;0.99^1.79) Zuclopenthixol decanoate 36/17136/171 49/16149/161 0.69 ;0.48^1.00) 05Side-effects: Movement disorders ^ general ^ needing anticholinergic medication medication Bromperidol decanoate 24/48 31/49 0.79 ;0.55^1.13) Flupentixol decanoate 45/92 56/10156/101 0.88 ;0.67^1.16) Fluphenazine decanoate 239/362 192/365192/365 1.26 ;1.11^1.42) Fluphenazine enanthate 28/53 23/6923/69 1.58 ;1.04^2.41) Fluspirilene decanoate 22/88 36/7836/78 0.54 ;0.35^0.84) Haloperidol decanoate 73/124 80/133 0.98 ;0.80^1.20) Perphenazine decanoate 82/8582/85 75/87 1.12 ;1.02^1.23) Pipotiazine palmitate palmitate 97/191 95/17995/179 0.96 ;0.79^1.16) Zuclopenthixol decanoate 100/153 112/143 0.83 ;0.72^0.96) 0.10.1 0.20.2 151 51010 Favours named depot Favours control

Fig. 33Fig. Specific depot antipsychoticvv. control depot: all outcomes, no summations.

in the use of high-dose depot preparations evidence that one depot is clearly better any one of these widely used compounds introduced for treatment-resistant cases, than another, and none that high or ultra- within a large, long, simple and clinically and that ultralow doses are little more than low doses have advantages. relevant randomised trial Hotopf et aletal,, placebo.placebo. Direct data on economic outcomes, 1999). Further research should, ideally, quality of life and satisfaction were not focus on those living outside of hospital Limitations found. Such outcomes were scarcely consid- in community settings, whose non-adherence Many outcomes, stated by trialists to ered in randomised trials from the 1960s to to treatment and follow-up is thought to have been recorded, were lost owing to early 1980s. A review of what limited contribute to relapses in their condition. poor reporting. Modern trialists recom- evidence there is relating to satisfaction Such studies are, by their very nature, dif- mendmendthatthat all outcome measures should with depot antipsychotics suggests that ficult to perform. Those designing evalua- be reported Begg et aletal, 1996). Data from patients on depots are, on average, reason- tive studies of depots in the future, often poorly reported, small trials of ably satisfied see companion paper ± including `atypical' compounds, should limited generalisability, when taken WalburnWalburn et aletal, 2001, this issue). learn from the limitations and strengths together with larger trials, support the seen in depot trial design over the past value of depot antipsychotic preparations. three decades. Such studies would have This complements information from less Future studies to be of longer duration than the majority methodologically rigorous studies Davis Clinicians and recipients of care could conducted to date, in order to capture a et aletal, 1994). There is little convincing still benefit from thorough evaluation of sufficient number of relapses. Long-term

297

Downloaded from https://www.cambridge.org/core. 23 Sep 2021 at 17:23:35, subject to the Cambridge Core terms of use. ADAMS ET AL

trials specifically designed to examine out- comes such as tardive dyskinesia also are CLINICAL IMPLICATIONS required. The definition of relapse requires careful consideration and would need to && Depot neuroleptic medication is an effective maintenance therapy for be operationalised. Obtaining useful cost- schizophrenia. effectiveness data and data on quality of life, satisfaction, disability, etc. is a && There may be a slight therapeutic advantage ;and no obvious disadvantage) of research priority. depot over oral medication, but the evidence is weak. weak.

&& There are few advantages of one depot over another. REFERENCES LIMITATIONS Adams, C. E. & Eisenbruch, M. %2000)%2000) DepotDepot fluphenazine for schizophrenia. Cochrane Library,issue3. && Data on patients for whom depots are most indicated are lacking. Oxford: Update Software.

Ðsberg, M., Montgomery, S., Perris, C., et aletal %1978)%1978) && Patient satisfaction, cost-effectiveness and other outcomes have not been studied The comprehensive psychopathological scale. ActaActa in controlled trials. Psychiatrica Scandinavica Supplementum,, 271271, 5^27.,5^27.

Barnes,T. R. & Curson, D. A.%19 %1994) 94) Long-term depot && There may be unforeseen methodological problems in meta-analyses of several antipsychotics. A risk^benefit assessment. Drug Safety,, systematic reviews. 1010, 464^479.,464^479.

Begg, C., Cho, M., Eastwood, S., et aletal %19 9 6) Improving the quality of reporting of randomised controlled trials.The CONSORTstatement. Journal of the American Medical Association,, 276276,,637^639. 637^639.

Boissel, J. P., Cucherat, M., Li,W., et aletal %1999)%1999) TheThe CLIVE E. ADAMS, MRCPsych, MARKMARKK. K. P.P.FENTON,RMN,Cochrane FENTON, RMN, Cochrane Schizophrenia Group, Summertown problem of therapeutic efficacy indices. 3.Comparison of Pavilion, Oxford; SEEMA QURAISHI, MSc, ANTHONY S. DAVID, FRCPsych, Institute of Psychiatry & Guy's, the indices and their use. TherapieThe¨ rapie,, 5454,405^411., 405^411. King's and St Thomas' School of Medicine, London

Bordeleau, J. M., Albert, J. M., Killer, J., et aletal %1967)%1967) CorrespCorrespondence:ondence: Professor Anthony S.David,InstituteS.David, Institute of PsychPsychiatryiatry & GKT School of Medicine, MeMedication¨ dication antiparkinsonienne et bilan extrapyramidal. Denmark Hill,London SE5 8AF,UK.Tel: 020 7848 0138; Fax: 020 7848 0572; e-mail: EEtude¨ tude du trihexyphenidyl. Canadian Psychiatric Association Journal,, 1212, 585^595.,585^595. a.david@@iop.kcl.ac.uk

British Medical Association & Royal First received 6 September 2000, final revision 5 February 2001, accepted 8 February 2001) Pharmaceutical Society of Great Britain%19 %1999) 9 9) British National Formulary. London & Wellingford: BMJ Books & Pharmaceutical Press.

Chalmers,Chalmers,T. T. C., Celano, P., Sacks, H. S., et aletal %19 83) Bias in treatment assignment in controlled clinical trials. Guy, W.%19 %1976) 76) ECDEU Assessment Manual for Consensus Conference in Siena, Italy. EuropeanEuropean New England Journal of Medicine,, 309309,1358^1361.,1358^1361. Psychopharmacology. Revised DHEW Pub. ADM). Neuropsychopharmacology,, 88,55^66. Coutinho, E., Fenton, M. & Quraishi, S. %2000) Rockville, MD: National Institute for Mental Health. Kemp, R. A. & David, A. S.%19 %1997) 9 7) Insight andandInsight Zuclopenthixol decanoate for schizophrenia and other Hamilton, M.%19 %1960) 6 0) A rating scale for depression. compliance. InTreatment Compliance and the Therapeutic serious mental illnesses. Cochrane Library,issue2.,issue2. Journal of Neurology,Neurosurgery and Psychiatry,, 2323,, Alliance ed. B. Blackwell), pp. 61^84. Newark, NJ: Oxford: Update Software. 56^62.56^62. Harwood Academic. Davis, J. M. & Andriukaitis, S.%19 %1986) 8 6) The natural Hirsch, S. R., Gaind, R., Rohde, P.P.D., D., et aletal %1973)%1973) Krawiecka, M., Goldberg, D. & Vaughn, M.%19 %1977) 7 7) course of schizophrenia and effective maintenance drug Outpatient maintenance of chronic schizophrenic Standardised psychiatric assessment scale for chronic treatment.treatment. Journal of Clinical Psychopharmacology,, 66,, patients with long-acting fluphenazine: double-blind psychiatric patients. Acta Psychiatrica Scandinavica,, 3636,, 2S^10S.2S^10S. placebo controlled trial. British Medical Journal,, ii,, 25^31.

__ , Matalon, L.,Watanabe, M. D., et aletal %19 94)94)%19 DepotDepot 633^637. antipsychotic drugs. Place in therapy. Drugs,, 4747, 741^773. Lingjaerde, O., Ahlfors,U.Ahlfors, U. G., Bech, P., et aletal %19 887) 7) Honigfeld,G.,Gillis,Honigfeld, G., Gillis, R. D. & Klett,C.Klett, C. J.%19 %1962) 62) NOSIE-NOSIE- The UKU side effects rating scale. Acta Psychiatrica Fenton,W. S., Blyler,C. R. & Heinssen, R. K.%19 %1997) 9 7) 30: a treatment sensitive ward behavior scale. Scandinavica,, 7676 suppl. 334),334),1^100. 1^100. Determinants of medication compliance in Psychological Reports,, 1010,799^812., 799^812. schizophrenia: empirical and clinical findings. Montgomery, S. A., Taylor, P. & Montgomery, D. Hotopf, M., Churchill, R. & Lewis, G.%1999) Schizophrenia Bulletin,, 2323,,637^651. 637^651. %1978)%19 78) Development of a schizophrenia scale sensitive to Pragmatic randomised controlled trials in psychiatry. change. Neuropharmacology,, 78,1061^1062. Foster, K., Meltzer,Meltzer,H., H., Gill, B., et aletal %19 9 6) AdultsAdults British Journal of Psychiatry,, 175175,217^223., 217^223. with a psychotic disorder living in the community. OPCS Overall, J. E. & Gorham, D. R.%19 %1962) 62) The Brief Jadad, A. R., Cook, D. J., Jones, A., et aletal %19 9 8) Surveys of Psychiatric Morbidity in Great Britain,Report8.,Report8. Psychiatric Rating Scale. Psychological Reports,, 1010,, Methodology and reports of systematic reviews and London: HMSO. 799^812.799^812. meta-analyses: a comparison of Cochrane reviews with Gerlach,J.%1995)Gerlach, J. %19 95) Depot neuroleptics in relapse articles published in paper-based journals. Journal of the Quraishi, S. & David, A.%2000 aa)) Depot flupenthixol prevention: advantages and disadvantages. InternationalInternational American Medical Association,, 280, 278^280.,278^280. decanoate for schizophrenia or other similar psychotic Clinical Psychopharmacology,, 99 suppl.5),17^20. suppl. 5), 17^20. disorders. Cochrane Library,issue3.Oxford:Update Johnson, D. A.%19 %1984) 8 4) Observations on the use of long- Software.Software. Gilbody,S.Gilbody, S. M. & Song, F. %2000)%2000) Publication bias and acting depot neuroleptic injections in the maintenance the integrity of psychiatry research. Psychological therapy of schizophrenia. Journal of Clinical Psychiatry,, 4545,, __ && __ %2000%2000bb)) Depot fluspirilene for schizophrenia. MedicineMedicine,, 3030, 253^258. 1313^21. ^ 21. Cochrane Library,issue3.Oxford:UpdateSoftware.

Glazer,W. M. & Kane, J. M.%19 %1992) 92) Depot neuroleptic Kane, J. M., Aguglia, E., Altamura, A. C., et aletal __ && __ %2000%2000cc)) Depot haloperidol decanoate for therapy: an underutilized treatment option. Journal of %19 9 8)8)%19 Guidelines for depot antipsychotic treatment in schizophrenia. Cochrane Library,issue3.Oxford:Update Clinical Psychiatry,, 5353, 426^433.,426^433. schizophrenia. European Neuropsychopharmacology Software.Software.

298

Downloaded from https://www.cambridge.org/core. 23 Sep 2021 at 17:23:35, subject to the Cambridge Core terms of use. SYSTEMATIC META-REVIEW OF DEPOT ANTIPSYCHOTICS

__ && __ %2000dd)) Depot perphenazine decanoate and treatment effects in controlled trials. Journal of the Walburn,Walburn,J.,Gray,R.,Gournay,K., J., Gray, R., Gournay, K., et aletal %2001)%2001) enanthate for schizophrenia. Cochrane Library,issue3. American Medical Association,, 273273,408^412., 408^412. Systematic review of patient and nurse attitudes to Oxford: Update Software. depot antipsychotic medication. British Journal of Simpson, G. M.%19 %1984) 8 4) AAbriefhistoryofdepot brief history of depot PsychiatryPsychiatry,, 179, 300^307.,300^307. __ && __ %2000ee)) Depot pipothiazine palmitate and undecylenate for schizophrenia. Cochrane Library,issue3.,issue3. neuroleptics. Journal of Clinical Psychiatry,, 4545,3^4., 3^4. Oxford: Update Software. Simpson, M. & Angus, J.W. %1970)%1970) A rating scale for Weiden,P.J.&Olfson,M.%1995)Weiden,P.J.&Olfson,M.%19 95) Cost of relapse in __ ,, __ &Adams,C.E.%2000)&Adams,C.E.%2000) Depot bromperidol extrapyramidal side effects. Acta Psychiatrica schizophrenia. Schizophrenia Bulletin,, 2121, 419^429.,419^429. decanoate for schizophrenia. Cochrane Library,issue3.,issue3. Scandinavica,, 212212,11^19. Oxford: Update Software.

Schulz, K. F., Chalmers, I., Hayes, R. J., et aletal %19 995) 5) Thornley, B. & Adams, C.%19 %1998) 9 8) Content and quality Wing,J.K.%1961)Wing, J. K. %19 61) A simple and reliable subclassification Empirical evidence of bias. Dimensions of of 2000 controlled trials in schizophrenia over 50 years. of chronic schizophrenia. Journal of Mental Science,, 107107,, methodological quality associated with estimates of British Medical Journal,, 317317,,1181^1184. 1181^118 4. 862^876.

299299

Downloaded from https://www.cambridge.org/core. 23 Sep 2021 at 17:23:35, subject to the Cambridge Core terms of use.