Welcome to the Schizophrenia Resource Centre

Welcome, this website is intended for healthcare professionals in EMEA with an interest in the treatment of schizophrenia. By clicking the link below you are declaring and confirming that you are a healthcare professional

You are here

A systematic review and meta-analysis of the effect of depot antipsychotic frequency on compliance and outcome

Schizophrenia Research, Volume 166, Issue 1-3, August 2015, Pages 178 - 186



Depot antipsychotics are commonly used to improve adherence and clinical outcomes such as relapse and readmission. Dosing regimens vary but are commonly two- and four-weekly. To date, the effect of administration at two-weekly or four-weekly intervals on outcome has not been examined in a meta-analysis.


A systematic review and meta-analysis on whether the frequency of depot antipsychotic administration (e.g., two- vs four-weekly) makes any difference to compliance and outcome.


A systematic search of Medline, EMBASE and PsycInfo for RCTs that compared the frequency of depot administration (e.g., two- vs four-weekly) for an equivalent dose. Outcomes were compliance, psychiatric symptomatology, quality of life, adverse drug reactions (ADRs), patient preference, admission rates, bed-days and costs.


Seven studies from eight papers (n = 3994) were found covering olanzapine, paliperidone, risperidone, haloperidol and fluphenazine enanthate/decanoate with follow-up of up to one year. Meta-analyses were possible for psychotic symptoms and ADRs. There were no differences in psychotic symptoms or quality of life between two- and four-weekly doses. Health service use was not reported. For ADRs, the only significant difference detected was that two-weekly injections were less likely to lead to site pain (RR 0.16, 95% CI 0.07–0.38; 2 studies n = 1667). There were no differences in other ADRs.


There were surprisingly little data on the effect of dosing frequency for an equivalent dose on clinical outcomes. There is a need for long-term studies of a wide range of outcomes including cost-effectiveness. Claims for advantages of new preparations over others require careful evaluation.

Keywords: Depot antipsychotic, LAI, Dosing frequency, Outcomes.

1. Introduction

Psychotropic medications play a key role in the treatment of serious mental illness. However, adherence is often poor resulting in poorer psychosocial outcomes (Gray et al, 2010 and Hill et al, 2010).

Several studies have reported advantages of long-acting injectable (LAI) over oral medication in terms of relapse prevention and adherence. A meta-analysis of 10 randomised trials showed a statistically significant reduction in relapse rates with the use of depots as opposed to orals ( Leucht et al., 2011 ). A further study found that risk of rehospitalisation for patients receiving depots was about one-third of that for patients receiving equivalent oral medications ( Tiihonen et al., 2011 ). Discontinuation rates with orals can reach 74% ( Lieberman et al., 2005 ) compared to around a third for an atypical LAI (Fleischhacker et al, 2003a, Fleischhacker et al, 2003b, Kissling et al, 2005, Keks et al, 2007, and Baker et al, 2012). Long-term benefits may be greater as few studies extend over one year ( Patel and David, 2005 ).

There have been relatively few depot head-to-head comparisons although comparisons with oral medications have shown no convincing advantages for one depot over another (Adams et al, 2001 and Fleischhacker, 2009). Where there have been head-to-head comparisons, no depot appears to have superiority over another, including comparisons between first and second generation antipsychotics (FGA, SGA) (Quraishi and David, 2000, Adams et al, 2001, David et al, 2005, Fricchione Parise et al, 2010, Einarson, 2011, Fleischhacker et al, 2012, and McEvoy et al, 2014).

One area that has not been studied in depth is dosing frequency. Dosing regimens vary but are commonly two- and four-weekly. The effect of administration at two-weekly or four-weekly intervals on subsequent compliance and outcome is unknown. This is topical because many patients are currently being changed from two-weekly depot risperidone to the four-weekly depot of its metabolite, paliperidone on the basis that this benefits patients and the health service. A Cochrane review reported on two studies comparing risperidone depot with paliperidone and found little difference between the two. However it did not expressly investigate dosing frequency, and combined the results of studies with very different follow-up periods (13 and 53 weeks respectively) with no usable intermediate data points. Importantly, it did not investigate other depot psychotropics ( Nussbaum and Stroup, 2012 ).

Reducing dosing frequency may save resources, including time, travel and reduced outreach visits. An industry-sponsored cost analysis of potential savings from changing from a two-weekly to monthly regime estimated an average saving of US$58 per injection avoided, potentially saving US$8.5 million per year ( Dalton et al., 2011 ). In addition, this might give staff more time for other duties, and be more convenient for patients.

Conversely, reduced frequency of contact may mean that opportunities are missed for concurrent non-pharmacological interventions. Comparisons of standard care with either intensive case management or assertive community treatment have consistently shown that greater intensity of contact improves patient compliance and outcomes (Marshall and Lockwood, 2000, Zygmunt et al, 2002, Dieterich et al, 2010, and Guhne et al, 2014). Where compulsory community treatment has shown advantages over treatment as usual, it has been suggested that it is through more frequent contact with mental health clinicians ( Kisely et al., 2013 ).

We therefore undertook a systematic review comparing outcomes of two-weekly versus four-weekly administration on clinical and health service outcomes.

2. Method

The review was registered with PROSPERO, an international database of prospectively registered systematic reviews in health and social care based in the United Kingdom (registration number: CRD42015015764) ( Booth et al., 2012 ). In addition, we followed recommendations for the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement including background, search strategy, methods, results, discussion and conclusions ( Moher et al., 2009 ).

2.1. Search strategy

We searched Medline, PsycInfo and EMBASE up till December 2014 using the following text, MeSH or Emtree terms as appropriate: 2 weekly, 2-weekly, 2 weeks, 2-weeks, two weekly, two-weekly, two weeks, two-weeks, once monthly, once-monthly, 4 weekly, 4-weekly, 4 weeks, 4-weeks, four weekly, four-weekly, four weeks, four-weeks, antipsychotic, antipsychotics, neuroleptic, Neuroleptics, Depot, Long Acting Injectable, Long Acting Injectables, LAI. In addition we searched by name all the SGA LAIs and the following most commonly used conventional depots: haloperidol, fluphenazine, flupenthixol and zuclopenthixol.

We inspected titles and abstracts of all papers identified in the electronic searches. The full texts of all randomised control trials (RCTs) were examined for relevance and snowball searches of reference lists. Articles that did not meet the inclusion criteria were cross-referenced for additional potential sources of RCTs.

2.2. Inclusion and exclusion criteria

We included all relevant double or rater-blind RCTs comparing the frequency of depot administration (e.g., two- vs four-weekly) for an equivalent dose determined through published tables ( Andreasen et al., 2010 ). Outcomes were compliance, psychiatric symptomatology, quality of life, adverse drug reactions (ADRs), patient preference, admission rates, bed-days and costs.

We assessed the quality of included studies using the following five criteria of the risk of bias assessment tool, developed by the Cochrane Collaboration to assess possible sources of bias in RCTs: 1. Adequate generation of allocation sequence; 2. Concealment of allocation to conditions; 3. Prevention of knowledge of the allocated intervention to assessors of outcome; 4. Dealing with incomplete outcome data; and 5. Selective reporting of outcomes ( Higgins and Green, 2008 ).

Data were extracted by two independent researchers (SK and ES). All discrepancies during study selection, data extraction, and quality assessment were resolved by re-checking source papers. The other two authors (GR and DS) provided content expertise.

2.3. Statistical analysis

We used Review Manager version 5.2 for Windows, a statistical software package for analysing Cochrane Collaboration systematic reviews. We calculated the mean differences for continuous data where studies used the same scale for each outcome, and the standardised mean difference for data that used different scales. We reported the relative risk (RR) for any dichotomous outcome. Where possible, intention-to-treat analyses were used.

We assessed heterogeneity using the I2 statistic, a measure that does not depend on the number of studies in the meta-analysis and hence has greater power to detect heterogeneity when the number of studies is small. It is calculated using the chi-squared statistic (Q) and its degrees of freedom ( Higgins and Green, 2008 ). An estimate of 50% or greater indicates possible heterogeneity, and scores of 75–100% indicate considerable heterogeneity.

We used the random effects model for all the analyses as we could not definitely exclude between-study variation even in the absence of statistical heterogeneity given the range of medications under review. For any outcomes where there were at least 10 studies, we tested for publication bias using funnel plot asymmetry where low P-values suggest publication bias ( Higgins and Green, 2008 ).

3. Results

We found 220 citations of interest representing 178 papers once duplicates were removed. Of these, 13 full-text papers were potentially relevant and assessed for eligibility. Five papers were excluded for reasons listed in Fig. 1 . This left eight papers from seven studies ( Fig. 1 ). The sum of enrolled patients was 3994 at baseline and 3160 at follow-up. Meta-analyses were possible for five of these studies ( Fig. 1 ).


Fig. 1 PRISMA flow diagram.

Three studies compared two-weekly injections of risperidone with equivalent four-weekly doses of its metabolite, paliperidone. Of the other studies, two compared dosing frequencies of equivalent doses of olanzapine, one of equivalent doses of fluphenazine enanthate and decanoate, and one of equivalent doses of haloperidol and fluphenazine decanoate ( Table 1 ). In one olanzapine study, the results of high and low two-weekly medium doses of olanzapine (300 mg and 150 mg every two weeks) were averaged and compared with a medium monthly dose (405 mg/month) ( Table 1 ). All the studies were of patients meeting DSM (II or IV) or ICD-9 criteria for schizophrenia and/or schizoaffective disorder ( Table 1 ). Two studies ended at 13 weeks (Li et al, 2011 and Pandina et al, 2011), and one study that was reported in 2 papers ended at 8 weeks (Lauriello et al, 2008 and Witte et al, 2012). All other studies lasted between six and 12 months.

Table 1 Summary of included studies.

Author Year Comparison N Outcomes Summary of results
Chouinard G., Annable L., Ross-Chouinard A. 1982 Fluphenazine decanoate (4-weekly) vs fluphenazine enanthate (2-weekly) Randomised = 50

Follow-up = 48

Follow-up decanoate = 24

Follow-up enanthate = 24
Endpoint: 28 weeks. Outcomes measured at 10, 12, 26 and 28 weeks.

Psychiatric symptoms
  • BPRS (endpoint)

ADRs a
  • Use of anti-Parkinsonian medication
  • Treatment emergent symptoms form & ESRS: akathisia, akinesia, dyskinesia, dystonia, dyskinetic movement, hypokinetic factors
  • CGI-S: tardive dyskinesia
Patient inclusion: DSM-II with schizophrenia

BPRS: at 10, 12 and 26 weeks = no difference between groups

At 28 weeks = higher BPRS score in 4-weekly treatment

Anti-Parkinsonian medication: no differences between groups in number of patients requiring antiparkinsonian medication or mean dosage amount

ESRS: At 10 and 12 weeks the 2-weekly group had significantly higher rates of hypokinetic factor and total Parkinsonian symptoms. However there were no differences between groups at 26 and 28 weeks.

Dyskinetic movements: no differences between groups. However, there was a significant increase in the mean total score for dyskinetic movements from 26 to 28 weeks in the 4-weekly group.

CGI-S: no differences between groups
Fleishhacker W., Gopal S., Lane R., Gassmann-Mayer C., Lim P., Hough D., Remmerie B., Eerdekens M.

(initially presented as a poster in 2009)
2012 Paliperidone palmitate (50 mg eq. on days 1 and 8 and flexible dosing once monthly–4-weekly) vs risperidone LAI (2-weekly injections of 25 mg on days 8 and 22 and flexible dosing from day 36 with allowed oral supplementation) Randomised = 749

Follow-up = 339

Primary efficacy, total = 570

Primary efficacy, paliperidone = 288

Primary efficacy, risperidone = 282

Secondary efficacy, total = 674

Secondary efficacy, paliperidone = 343

Secondary efficacy, risperidone = 331
Endpoint: 372 days

Psychiatric symptoms (LOCF)
  • PANSS (score change)
  • PSP scale
  • CGI-S

  • Use of anti-Parkinsonian medication
  • TEAEs: hyperkinesia, akathisia, dyskinesia, extrapyramidal symptoms, cardiovascular events, evaluation of injection site
Patient inclusion: DSM-IV with schizophrenia

PANSS: higher scores in 4-weekly group

PSP: no differences between groups

CGI-S: reduction of symptoms was greater in 2-weekly group

TEAEs: proportion slightly greater in serious TEAES in 4-weekly group

Extrapyramidal symptoms: similar between groups (both low incidence)

Median plasma levels of paliperidone (the active metabolite in both arms) were higher for the risperidone arm (20 ng/mL) than the paliperidone arm (7.5 ng/mL) by day 64 and remained so for the remainder of the study.
Kane J., Detke H., Naber D., Sethuraman G., Lin D., Bergstrom R., McDonnell D. 2010 Olanzapine 150 mg/2 weeks (low), 405 mg/4 weeks (medium), 300 mg/2 week (high) against a very low reference dose of olanzapine 45 mg/4 weeks and oral olanzapine Randomised = 1065

Follow-up = 1063

low = 140

medium = 318

high = 141

very low = 144

oral = 322
Endpoint: 24 weeks. Outcomes were measured weekly for the first 12 weeks and fortnightly for the second 12 weeks

Psychiatric symptoms (ITT, LOCF)
  • BPRS (score change)
  • PANSS (score change)
  • CGI-S

  • TEAEs: extrapyramidal symptoms, akathisia, insomnia
Patient inclusion: DSM-IV with schizophrenia

The 4 week regimen and pooled 2-week regime had similar efficacies with no significant differences in exacerbation rates
Kissling W., Moller H., Walter K., Wittmann B., Krueger R., Trenk D. 1985 Haloperidol decanoate (4-weekly) vs fluphenazine decanoate (2-weekly) Randomised = 54

Follow-up = 31

Follow-up, haloperidol = 22

Follow-up, fluphenazine = 9
Endpoint: day 170. All outcomes measured on days 1–4, 9 and every week thereafter until end of study

Psychiatric symptoms a
  • BPRS (endpoint)
  • IMPS
  • PD-S & Bf-S

ADRs a
  • Use of anti-Parkinsonian medication
Patient inclusion: ICD-9 patients with schizophrenic and schizoaffective psychosis

BPRS: no differences between groups

Self rating level (PD-S, Bf-S): no differences between groups

Anti-Parkinson agents: much higher consumption in 2-weekly group

Extra-pyramidal side effects: 2-weekly group mean values were higher during the first three months

Mean plasma levels in both arms reached steady state by 90 days
Pandina G., Lane R., Srihari S., Gassmann-Mayer C., Hough D., Remmerie B., Simpson G. 2011 Paliperidone palmitate (on days 1 and 8 and flexible dosing once monthly on days 36 and 64 — 4-weekly) vs risperidone LAI (On days 8, 22, 36, 50, 64, 78 — 2-weekly) Randomised = 1220

Follow-up = 927

Follow-up, PP = 456

Follow-up, RIS-LAI = 471
Endpoint: 13 weeks. Outcomes recorded on days 1, 4, 15, 22, 36, 64, and 92

Psychiatric symptoms a
  • PANSS (score change)
  • PSP
  • CGI-S

  • TEAEs: hyperkinesia, akathisia, dyskinesia, extrapyramidal symptoms, cardiovascular events, evaluation of injection site, prolactin levels, weight gain
Patient inclusion: DSM-IV with schizophrenia

PANSS: difference in change in mean score between the two groups was 0.4 in favour of 4-weekly regime

Difference in change in the PANSS total score was 1.2 in favour of 4-weekly regime

EPS rating scales: no clinically relevant differences between groups

Use of anti-EPS medication: no clinically relevant differences between groups.

Median plasma levels in both arms reached steady state by approximately 15 days and remained at around 20 ng/mL for the remainder of the study.
Li H., Rui Q., Ning X., Gu N. 2011 Paliperidone palmitate (4-weekly) vs risperidone LAI (2-weekly) Randomised = 452

Follow-up = 350
Endpoint: 13 weeks. Outcomes recorded at weeks 1, 5, 9 and 13

Psychiatric symptoms (ITT, LOCF)
  • PANSS (score change)
  • PSP
  • CGI-S

  • Use of anti-Parkinsonian medication
  • TEAEs: akathisia, tremor, insomnia, extrapyramidal symptoms, cardiovascular events, evaluation of injection site, prolactin levels, weight gain
Patient inclusion: DSM-IV with schizophrenia

PANSS: two-weekly had significantly greater improvement using mean differences but not least square means on per-protocol analyses. PANSS ITT results were restricted to least square means

TEAEs: similar in both groups

Although plasma levels were not measured, the study used the same doses as Pandina where they were (see above).
Sibling studies Lauriello J., Lambert T., Andersen S., Lin D., Taylor C., Mcdonnell D. 2008 Olanzapine: 210 mg/2 weeks

300 mg/2 weeks, 405 mg/4 weeks & placebo/2 weeks
Randomised = 404

Follow-up = 402

Primary outcome:

210 mg/2 weeks = 106

300 mg/2 weeks = 98

405 mg/4 weeks = 100

Placebo = 98
Endpoint: 8 week follow-up

Psychiatric symptoms:(Lauriello a , Witte: ITT, LOCF)
  • PANSS (Score change)
  • PANSS derived BPRS
  • CGI-S
  • QLS (endpoint scores)
  • SF-36 (endpoint scores)

ADRs: (Lauriello a , Witte: ITT, LOCF)
  • TEAEs: changes in glucose, akathisia, extrapyramidal symptoms, dyskinesia
Patient inclusion: DSM-IV with schizophrenia

The response profile when compared to placebo of both the 210 mg/2 weeks and 405/4 weeks regimes was similar for PANSS, BPRS, CGI-S, QLS and SF-36 total score and subscores with 2 exceptions
  • In QLS interpersonal relations the 2-weekly regime was significantly better to placebo whilst the 4-weekly regime was not
  • In SF-36 mental component scores, the 4-weekly regime was significantly better to placebo whilst the 2-weekly regime was not

Improvements in PANSS scores correlated with QOL & SF-36 scores with no difference between LAI regimes

There was no difference in TAES between patients on 210 mg/2 weeks and 405/4 weeks

Olanzapine plasma levels did not reach steady-state by 8 weeks — no data were presented about any differences between the treatment groups
Witte M., Case M., Schuh K., Ascher-Svanum H. 2012 Randomised = 404

Follow-up = 355

Primary outcome:

210 mg/2 weeks = 93

300 mg/2 weeks = 87

405 mg/4 weeks = 92

Placebo = 83

a Did not state whether ITT analyses of LOCF were used.

BPRS: Brief Psychiatric Rating Scale.

ESRS: Extrapyramidal Symptom Rating Scale.

CGI-S: Clinical Global Impression — Severity of illness score.

PANSS: Positive and Negative Syndrome Scale.

PSP: Personal and Social Performance Scale.

TEAS: Treatment-Emergent Adverse Events.

IMPS: Inpatient Multidimensional Psychiatric Scale.

PD-S: Paulhus Deception Scale and Bf-S: Befindlichkeits-Skala — self rating mood scale.

DOTES: dosage record and treatment emergent symptom scale.

QLS: Quality Of Life Scale.

Study quality was moderate on the risk of bias assessment tool ( Table 2 ). Generation of the random allocation sequence was adequate in three studies, whilst in the other four it was unclear. Adequate allocation concealment was unclear in five studies, high in one and low in another. Six studies were described as double-blinded and all of these had adequate descriptions of the blinding process including measures to ensure that participants (and researchers) could not guess their treatment arm through the use of oral and/or depot placebo. The final study was only rater-blinded.

Table 2 Risk of bias table.

  Random sequence generation (selection bias) Allocation concealment (selection bias) Blinding of participants, personnel (performance bias) Blinding of outcome assessment (detection bias) Incomplete outcome data (attrition bias) Selective outcome reporting (reporting bias) Other sources of bias
Chouinard et al. (1982) Unclear Unclear Low Low Low Unclear Unclear
Fleischhacker et al. (2012) Low Unclear Low Low High High High a
Kane et al. (2010) Unclear Unclear Low Low Low Unclear High a
Kissling et al. (1985) Low Unclear Low Low Low Unclear High a
Lauriello et al. (2008) b Unclear Unclear Low Low Low Unclear High a
Li et al. (2011) Unclear High High Low Low Low High a
Pandina et al. (2011) Low Low Low Low Low Low High a

a Pharmaceutical company sponsored.

b Lauriello et al includes Witte et al (2012) .

Three studies had less than 10% loss at follow-up (Chouinard et al, 1982, Lauriello et al, 2008, Kane et al, 2010, and Witte et al, 2012). In two others, the follow-up rate was 75% (Li et al, 2011 and Pandina et al, 2011). Another paper had a follow-up of 55% ( Kissling et al., 1985 ), whilst the final study had a follow-up of only 45% ( Fleischhacker et al., 2012 ). Five of the papers reported both per-protocol and intention-to-treat (ITT) analyses but in most circumstances gave greater prominence to the former for their primary outcomes of clinical response. ITT analyses were largely restricted to the assessment of secondary outcomes such awards. In the case of the study with the lowest follow-up, the ITT results for primary outcomes were not given but stated to be similar to the per-protocol analyses. This study was therefore rated at high risk for selective reporting ( Table 2 ).

The only clinical outcomes that were reported were psychiatric symptomatology, quality of life and ADRs ( Table 1 ). Four studies presented data on plasma levels ( Table 1 ). Effects on compliance, health services use or costs were not reported, and neither was patient preference. In terms of other sources of bias, six of the seven studies were funded by pharmaceutical companies and/or carried out by their employees.

3.1. Psychiatric symptoms and quality of life

Studies reported on a wide range of endpoint and change from baseline scores ( Table 1 ). The most commonly used were the Brief Psychiatric Rating Scales (BPRS), the Positive and Negative Syndrome Scale (PANSS) and the Clinical Global Impression — Severity scale (CGI-S). Quality of life ratings were based on the Quality of Life Scale (QLS) and SF-36.

The results of individual studies showed very little difference in outcome between the two-weekly and four-weekly regimes in all comparisons and at all time points ( Table 1 ) with the exception of one where the improvement in PANSS was greater with risperidone ( Li et al., 2011 ). However, the difference was of unclear clinical significance ( Fig. 1 ).

Given the wide range of outcomes and follow-up periods, meta-analyses were only possible for a limited number of measures at two to six month follow-up. As it was not possible to combine the results from different scales, we only calculated mean differences. These analyses showed no difference in psychiatric symptomatology between dosing regimens irrespective of the agents compared ( Fig. 2 ). One further study reported changes in PANSS but the follow-up period extended to 53 weeks, with no reported data from intermediate time points ( Fleischhacker et al., 2012 ). Including this study in the meta-analysis did not change the results.


Fig. 2 Psychiatric symptoms at two to six months.

Two-weekly risperidone and four-weekly paliperidone have the same active metabolite (paliperidone). Two studies examined the effect of dosing on its bioavailability (Pandina et al, 2011 and Fleischhacker et al, 2012). The first found that at a presumed equivalent dose, depot paliperidone achieved suboptimal bioavailability (see the section below on serum levels for more details) ( Table 1 ). In the subsequent study, a higher dose of paliperidone was therefore used. A sensitivity analysis of excluding the study with suboptimal bioavailability made no difference to the outcomes.

3.2. Adverse drug reactions

The main ADRs reported were extra-pyramidal symptoms and glycaemic changes assessed through standardised instruments, clinician ratings or investigations ( Table 1 ). Measures included the Extrapyramidal Symptom Rating Scale (ESRS), Barnes Akathisia Rating Scale, and Abnormal Involuntary Movements Scale (AIMS). Meta-analyses were only possible for some comparisons (Fig 3 and Fig 4). The only significant difference was that two-weekly injections were less likely to lead to site pain ( Fig. 3 ). There were no differences in Parkinsonian symptoms, dyskinesia, akathisia, or diabetes between regimes at either two to six month follow-up ( Fig. 3 ) or beyond six months ( Fig. 4 ). Anticholinergic drug use was also the same at both time points (Fig 3 and Fig 4). The study that could not be included in the meta-analyses did report greater extra-pyramidal consequences in the two-weekly group ( Table 1 ).


Fig. 3 Adverse drug reaction at two to six months.


Fig. 4 Adverse drug reactions at six to 12 months.

3.3. Plasma levels

Four studies had data on plasma levels ( Table 1 ). In all but one, a study on olanzapine, plasma levels reached steady state before the study's end. In the case of the olanzapine study, the effect was unclear as no data on differences between the treatment groups were presented, the authors merely commenting that the possible benefit across all subjects could have been greater if steady state had been reached (Lauriello et al, 2008 and Witte et al, 2012). Direct comparisons of levels were only possible in the two studies where the active metabolite, paliperidone, was the same between the two and four weekly treatment groups. In the first, median plasma levels were consistently higher in the 2-weekly compared with the 4-weekly arm (20 v 7.5 ng/mL respectively) ( Fleischhacker et al., 2012 ). The latter would correspond to a central D2-receptor occupancy of approximately 60%, which is at the lower end of the range required for efficacy ( Fleischhacker et al., 2012 ). A higher dose of paliperidone was therefore used in the subsequent study resulting in median levels in both arms of around 20 ng/mL ( Pandina et al., 2011 ). A further study used the same higher paliperidone dose although bioavailability was not measured ( Li et al., 2011 ).

3.4. Publication bias

We could not test for publication bias as there were insufficient studies for any of the outcomes.

4. Discussion

There were surprisingly little data from RCTs on the effect of dosing frequency for an equivalent dose on clinical outcomes (n = 7). Indeed, studies comparing depots often ensured that both active treatments were given at similar time intervals. This removes any consideration of the effects of dosing frequency on outcomes. On the limited evidence from these meta-analyses, two-weekly or four-weekly injections do not lead to notable differences in clinical outcomes. If anything, two-weekly risperidone injections were less likely to lead to site pain than monthly paliperidone. Additionally, in one study, risperidone produced greater improvement in PANSS scores ( Li et al., 2011 ). There were no RCT data on cost-effectiveness, the literature being based on industry-funded economic modelling (Dalton et al, 2011, Mehnert et al, 2012, and Zeidler et al, 2013).

As clinical outcomes do not appear to support one regimen over the other in either FGA or SGA agents, choice should be governed by patient convenience and preference. This however, needs to be balanced against the importance of patient engagement and facilitated access to non-pharmacological interventions to enhance recovery.

Further controlled trials evaluating both two- and four-weekly doses are indicated to explore the clinical impact of four-weekly relative to two-weekly dosing. These should be long-term, include cost-effectiveness, and take into account the wider costs and benefits of reducing, or increasing, frequency of patient contact.

Even if four-weekly regimes prove to be beneficial, this does not necessarily mean changing to more expensive preparations given some FGA antipsychotics are available in monthly preparations. Emerging literature suggests SGA agents may not have clear-cut advantages over conventional agents. In one comparison of depot risperidone and zuclopenthixol, there were no significant differences in treatment discontinuation, psychotic symptoms or tolerability including ADRs, lipid and glycaemic profile and Body Mass Index (BMI) ( Fricchione Parise et al., 2010 ). In another study, use of paliperidone palmitate or haloperidol decanoate did not result in statistically significant differences in efficacy ( McEvoy et al., 2014 ). Although haloperidol decanoate was associated with more akathisia, paliperidone palmitate was associated with more weight gain and greater increases in serum prolactin ( McEvoy et al., 2014 ). In addition, if the choice of medication is restricted to SGAs, a monthly regime does not necessarily mean a change of agent. Olanzapine can be given two- or four-weekly (Chue and Chue, 2012 and Case et al, 2013) and there is the suggestion that risperidone can be used monthly in stabilised patients. In a small open-label study (n = 87), once-monthly risperidone depot maintained the baseline status over one year of stable patients with schizophrenia or schizoaffective disorder, with no unexpected safety and tolerability findings ( Gharabawi et al., 2007 ).

There are several limitations to this study. The most obvious is that we could only find seven studies and conduct meta-analyses for the results of five. The maximum number of studies in any meta-analysis was three. On the other hand, because of the size of the studies, the number of subjects included in meta-analyses was as high as 1714. It is therefore unlikely that the lack of difference between two- and four-weekly regimes was solely due to type 2 error. A further study averaged the results of high and low two-weekly doses of olanzapine (300 mg and 150 mg every two weeks) with a medium monthly dose (405 mg/month) ( Kane et al., 2010 ). This was noted in the narrative review and, importantly, the results did not contribute to the meta-analyses.

We compared a range of different agents. Whilst bias is possible, studies reporting on the same agents found that the bioavailability was mostly similar. Removal of a study with suboptimal bioavailability of one agent in a sensitivity analysis had no effect upon outcomes.

Some of our results showed heterogeneity and it was not possible to explore this further using sensitivity analyses as there were insufficient studies for this to be appropriate, or because sensitivity analyses made no difference to the result. Accordingly, we used a random-effects model throughout to incorporate heterogeneity into our analyses ( Higgins and Green, 2008 ). However, although we have tried to minimize the effect, results with heterogeneity should still be treated with caution.

In conclusion, long-term studies are needed of the effect of the frequency of depot antipsychotic administration on a wide range of outcomes including cost-effectiveness. Claims for advantages of new preparations over others require careful evaluation.

Role of funding source

This project was supported by the University of Queensland Summer Research programme. This funding source provided funds for a research assistant only.


Author SK designed the study and wrote the protocol. Literature searches and analyses were completed by SK and ES.

Statistical analyses were undertaken by SK and DS. The first draft of the manuscript was written by SK. All authors edited and contributed to drafts of the manuscript. The final draft of the manuscript was approved by all authors.

Conflict of interest

The authors have none to disclose.


This project was supported by the University of Queensland Summer Research programme.


  • Adams et al., 2001 C.E. Adams, M.K. Fenton, S. Quraishi, A.S. David. Systematic meta-review of depot antipsychotic drugs for people with schizophrenia. Br. J. Psychiatry. 2001;179:290-299 Crossref
  • Andreasen et al., 2010 N.C. Andreasen, M. Pressler, P. Nopoulos, D. Miller, B.C. Ho. Antipsychotic dose equivalents and dose-years: a standardized method for comparing exposure to different drugs. Biol. Psychiatry. 2010;67(3):255-262 Crossref
  • Baker et al., 2012 R. Baker, W. Fleischhacker, R. Sanchez, P. Perry, N. Jin, B. Johnson, R.A. Forbes, R.D. McQuade, W.H. Carson, J.M. Kane. Long-term safety and tolerability of once-monthly aripiprazole intramuscular depot for maintenance treatment in schizophrenia. Int. J. Neuropsychopharmacol.. 2012;15:111
  • Booth et al., 2012 A. Booth, M. Clarke, G. Dooley, D. Ghersi, D. Moher, M. Petticrew, L. Stewart. The nuts and bolts of PROSPERO: an international prospective register of systematic reviews. Syst. Rev.. 2012;1:2 10.1186/2046-4053-1-2 (Feb 9) Crossref
  • Case et al., 2013 M.G. Case, J.L. Karagianis, K.J. Schuh, V.P. Hoffmann. Real-world use patterns of olanzapine long-acting injection in the United States: comparison to the recommended dosing strategy. Curr. Med. Res. Opin.. 2013;29(11):1533-1537 Crossref
  • Chouinard et al., 1982 G. Chouinard, L. Annable, A. Ross-Chouinard. Fluphenazine enanthate and fluphenazine decanoate in the treatment of schizophrenic outpatients: extrapyramidal symptoms and therapeutic effect. Am. J. Psychiatry. 1982;139(3):312-318
  • Chue and Chue, 2012 P. Chue, J. Chue. A review of olanzapine pamoate. Expert. Opin. Pharmacother.. 2012;13(11):1661-1670 Crossref
  • Dalton et al., 2011 A. Dalton, T. Lambert, R. Schrover, J. Hertel, D.K. Smith. The cost associated with administering risperidone long-acting injections in the Australian community. BMC Health Serv. Res.. 2011;11:236 Crossref
  • David et al., 2005 A. David, C.E. Adams, M. Eisenbruch, S. Quraishi, J. Rathbone. Depot fluphenazine decanoate and enanthate for schizophrenia. Cochrane Database Syst. Rev.. 2005;1:CD000307
  • Dieterich et al., 2010 M. Dieterich, C.B. Irving, B. Park, M. Marshall. Intensive case management for severe mental illness. Cochrane Database Syst. Rev.. 2010;10:Cd007906
  • Einarson, 2011 T. Einarson. Comparing the effectiveness of paliperidone palmitate versus olanzapine pamoate for relapse prevention in schizophrenia: post HOC indirect analysis using published placebo-controlled studies. Value Health. 2011;14(7):A560 Crossref
  • Fleischhacker, 2009 W.W. Fleischhacker. Second-generation antipsychotic long-acting injections: systematic review. Br. J. Psychiatry Suppl.. 2009;52:S29-S36 Crossref
  • Fleischhacker et al., 2003a W.W. Fleischhacker, M. Eerdekens, K. Karcher, G. Remington, P.M. Llorca, W. Chrzanowski, S. Martin, O. Gefvert. Treatment of schizophrenia with long-acting injectable risperidone: a 12-month open-label trial of the first long-acting second-generation antipsychotic. J. Clin. Psychiatry. 2003;64(10):1250-1257 Crossref
  • Fleischhacker et al., 2003b W.W. Fleischhacker, M.A. Oehl, M. Hummer. Factors influencing compliance in schizophrenia patients. J. Clin. Psychiatry. 2003;64(Suppl. 16):10-13
  • Fleischhacker et al., 2012 W.W. Fleischhacker, S. Gopal, R. Lane, C. Gassmann-Mayer, P. Lim, D. Hough, B. Remmerie, M. Eerdekens. A randomized trial of paliperidone palmitate and risperidone long-acting injectable in schizophrenia. Int. J. Neuropsychopharmacol.. 2012;15(1):107-118 Crossref
  • Fricchione Parise et al., 2010 V. Fricchione Parise, G. Balletta, G. Manna. Risperidone depot versus zuclopenthixol depot for schizophrenia and schizophrenia-like psychoses: real world outcomes. Eur. Neuropsychopharmacol.. 2010;20:S472-S473
  • Gharabawi et al., 2007 G.M. Gharabawi, N.C. Gearhart, R.A. Lasser, R.A. Mahmoud, Y. Zhu, E. Mannaert, I. Naessens, C.A. Bossie, M. Kujawa, G.M. Simpson. Maintenance therapy with once-monthly administration of long-acting injectable risperidone in patients with schizophrenia or schizoaffective disorder: a pilot study of an extended dosing interval. Ann. Gen. Psychiatry. 2007;6:3 10.1186/1744-859X-6-3 Crossref
  • Gray et al., 2010 R. Gray, J. White, M. Schulz, C. Abderhalden. Enhancing medication adherence in people with schizophrenia: an international programme of research. Int. J. Ment. Health Nurs.. 2010;19(1):36-44 Crossref
  • Guhne et al., 2014 U. Guhne, S. Weinmann, K. Arnold, T. Becker, S.G. Riedel-Heller. S3 guideline on psychosocial therapies in severe mental illness: evidence and recommendations. Eur. Arch. Psychiatry Clin. Neurosci.. 2014;265(3):173-188
  • Higgins and Green, 2008 J.P. Higgins, S. Green. Cochrane Handbook for Systematic Reviews of Interventions. (Wiley Online Library, 2008)
  • Hill et al., 2010 M. Hill, N. Crumlish, P. Whitty, M. Clarke, S. Browne, M. Kamali, A. Kinsella, J.L. Waddington, C. Larkin, E. O'Callaghan. Nonadherence to medication four years after a first episode of psychosis and associated risk factors. Psychiatr. Serv.. 2010;61(2):189-192
  • Kane et al., 2010 J.M. Kane, H.C. Detke, D. Naber, G. Sethuraman, D.Y. Lin, R.F. Bergstrom, D. McDonnell. Olanzapine long-acting injection: a 24-week, randomized, double-blind trial of maintenance treatment in patients with schizophrenia. Am. J. Psychiatry. 2010;167(2):181-189 Crossref
  • Keks et al., 2007 N.A. Keks, M. Ingham, A. Khan, K. Karcher. Long-acting injectable risperidone v. olanzapine tablets for schizophrenia or schizoaffective disorder. Randomised, controlled, open-label study. Br. J. Psychiatry. 2007;191(2):131-139 Crossref
  • Kisely et al., 2013 S. Kisely, N. Preston, J. Xiao, D. Lawrence, S. Louise, E. Crowe. Reducing all-cause mortality among patients with psychiatric disorders: a population-based study. CMAJ. 2013;185(1):E50-E56 Crossref
  • Kissling et al., 1985 W. Kissling, H. Möller, K. Walter, B. Wittmann, R. Krueger, D. Trenk. Double-blind comparison of haloperidol decanoate and fluphenazine decanoate effectiveness, side-effects, dosage and serum levels during a six months' treatment for relapse prevention. Pharmacopsychiatry. 1985;18(3):240-245 Crossref
  • Kissling et al., 2005 W. Kissling, S. Heres, K. Lloyd, E. Sacchetti, P. Bouhours, R. Medori, P.M. Llorca. Direct transition to long-acting risperidone — analysis of long-term efficacy. J. Psychopharmacol.. 2005;19(5 Suppl.):15-21 Crossref
  • Lauriello et al., 2008 J. Lauriello, T. Lambert, S. Andersen, D. Lin, C.C. Taylor, D. McDonnell. An 8-week, double-blind, randomized, placebo-controlled study of olanzapine long-acting injection in acutely ill patients with schizophrenia. J. Clin. Psychiatry. 2008;69(5):790-799 Crossref
  • Leucht et al., 2011 C. Leucht, S. Heres, J.M. Kane, W. Kissling, J.M. Davis, S. Leucht. Oral versus depot antipsychotic drugs for schizophrenia—a critical systematic review and meta-analysis of randomised long-term trials. Schizophr. Res.. 2011;127(1):83-92 Crossref
  • Li et al., 2011 H. Li, Q. Rui, X. Ning, H. Xu, N. Gu. A comparative study of paliperidone palmitate and risperidone long-acting injectable therapy in schizophrenia. Prog. Neuropsychopharmacol. Biol. Psychiatry. 2011;35(4):1002-1008 Crossref
  • Lieberman et al., 2005 J.A. Lieberman, T.S. Stroup, J.P. McEvoy, M.S. Swartz, R.A. Rosenheck, D.O. Perkins, R.S. Keefe, S.M. Davis, C.E. Davis, B.D. Lebowitz, J. Severe, J.K. Hsiao. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. N. Engl. J. Med.. 2005;353(12):1209-1223 Crossref
  • Marshall and Lockwood, 2000 M. Marshall, A. Lockwood. Assertive community treatment for people with severe mental disorders. Cochrane Database Syst. Rev.. 2000;2:CD001089
  • McEvoy et al., 2014 J.P. McEvoy, M. Byerly, R.M. Hamer, R. Dominik, M.S. Swartz, R.A. Rosenheck, N. Ray, J.S. Lamberti, P.F. Buckley, T.M. Wilkins, T.S. Stroup. Effectiveness of paliperidone palmitate vs haloperidol decanoate for maintenance treatment of schizophrenia: a randomized clinical trial. JAMA. 2014;311(19):1978-1987 Crossref
  • Mehnert et al., 2012 A. Mehnert, D. Nicholl, H. Pudas, M. Martin, A. McGuire. Cost effectiveness of paliperidone palmitate versus risperidone long-acting injectable and olanzapine pamoate for the treatment of patients with schizophrenia in Sweden. J. Med. Econ.. 2012;15(5):844-861 Crossref
  • Moher et al., 2009 D. Moher, A. Liberati, J. Tetzlaff, D.G. Altman. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann. Intern. Med.. 2009;151(4):264-269 (w269) Crossref
  • Nussbaum and Stroup, 2012 A.M. Nussbaum, T.S. Stroup. Paliperidone palmitate for schizophrenia. Cochrane Database Syst. Rev.. 2012;6:CD008296
  • Pandina et al., 2011 G. Pandina, R. Lane, S. Gopal, C. Gassmann-Mayer, D. Hough, B. Remmerie, G. Simpson. A double-blind study of paliperidone palmitate and risperidone long-acting injectable in adults with schizophrenia. Prog. Neuropsychopharmacol. Biol. Psychiatry. 2011;35(1):218-226 Crossref
  • Patel and David, 2005 M.X. Patel, A.S. David. Why aren't depot antipsychotics prescribed more often and what can be done about it?. Adv. Psychiatr. Treat.. 2005;11(3):203-211 Crossref
  • Quraishi and David, 2000 S. Quraishi, A. David. Depot perphenazine decanoate and enanthate for schizophrenia. Cochrane Database Syst. Rev.. 2000;2:CD001717
  • Tiihonen et al., 2011 J. Tiihonen, J. Haukka, M. Taylor, P.M. Haddad, M.X. Patel, P. Korhonen. A nationwide cohort study of oral and depot antipsychotics after first hospitalization for schizophrenia. Am. J. Psychiatry. 2011;168(6):603-609 Crossref
  • Witte et al., 2012 M.M. Witte, M.G. Case, K.J. Schuh, H. Ascher-Svanum. Effects of olanzapine long-acting injection on levels of functioning among acutely ill patients with schizophrenia. Curr. Med. Res. Opin.. 2012;28(3):315-323 Crossref
  • Zeidler et al., 2013 J. Zeidler, J. Mahlich, W. Greiner, S. Heres. Cost effectiveness of paliperidone palmitate for the treatment of schizophrenia in Germany. Appl. Health Econ. Health Policy. 2013;11(5):509-521 Crossref
  • Zygmunt et al., 2002 A. Zygmunt, M. Olfson, C.A. Boyer, D. Mechanic. Interventions to improve medication adherence in schizophrenia. Am. J. Psychiatry. 2002;159(10):1653-1664 Crossref


a The University of Queensland School of Medicine, Qld, Australia

b Metro South Health Service, Woolloongabba, Qld, Australia

c Griffith Institute of Health, Griffith University, Qld, Australia

d School of Medicine, James Cook University, Qld, Australia

Corresponding author at: School of Medicine, The University of Queensland, Level 4, Building 1, Princess Alexandra Hospital, Ipswich Road, Woolloongabba, Qld 4102, Australia. Tel.: + 61 3176 6438.