TY - JOUR
T1 - Association between discontinuation due to withdrawal of consent and use of long-acting injectable antipsychotics
T2 - A meta-analysis of randomized trials for schizophrenia
AU - Kishi, Taro
AU - Sakuma, Kenji
AU - Okuya, Makoto
AU - Iwata, Nakao
N1 - Publisher Copyright:
© 2020 Elsevier Ltd
PY - 2021/1
Y1 - 2021/1
N2 - We investigated the association between discontinuation due to withdrawal of consent (DWC) in schizophrenia trials and the use of long-acting injectable antipsychotics (LAI-APs). In two categorical meta-analyses of randomized controlled trials, we compared DWC: individual and pooled LAI-APs vs. (1) placebo and (2) oral antipsychotics (OAPs). We also performed conducted a single-group meta-analysis to calculate the average DWC and a meta-regression analysis to examine the association between the results of the meta-analyses and factors related to study design, treatment, and patients. We identified 52 studies (total adult patients = 18,675, LAI-APs = 12,613, placebo = 2,083, and OAPs = 3979; median study duration = 32 weeks). DWC was higher for LAI-aripiprazole than for the placebo [risk ratio (95% confidence interval) = 1.70 (1.23–2.39)]. Neither pooled nor individual LAI-APs differed from the placebo for fluphenazine, olanzapine, paliperidone, and risperidone or from the OAPs for aripiprazole, fluphenazine, haloperidol, olanzapine, paliperidone, risperidone, and zuclopenthixol. The average DWC of each LAI-AP was as follows: LAI-aripiprazole = 10.98%, LAI-fluphenazine = 7.65%, LAI-flupenthixol = 3.33%, LAI-haloperidol = 6.71%, LAI-olanzapine = 10.50%, LAI-paliperidone = 10.38%, LAI-perphenazine = 7.06%, LAI-risperidone = 10.39%, LAI-zuclopenthixol = 4.45%, pooled LAI-APs = 9.88%, and placebo = 11.17%. Meta-regression analysis demonstrated that publication year (β = 0.019), percentage of males (β = 0.020), and mean age (β = 0.045) were associated with an average DWC for pooled LAI-APs. Study duration (β = −0.028), percentage of males (β = 0.078), and patient status (β = −0.847) were associated with an average DWC for LAI-aripiprazole. Presence of a placebo arm (β = 1.596) was associated with an average DWC for LAI-fluphenazine. LAI-AP use was unlikely to be associated with DWC. Although the LAI-aripiprazole had a higher DWC than did the placebo, its average DWC was similar to other those of LAI-APs.
AB - We investigated the association between discontinuation due to withdrawal of consent (DWC) in schizophrenia trials and the use of long-acting injectable antipsychotics (LAI-APs). In two categorical meta-analyses of randomized controlled trials, we compared DWC: individual and pooled LAI-APs vs. (1) placebo and (2) oral antipsychotics (OAPs). We also performed conducted a single-group meta-analysis to calculate the average DWC and a meta-regression analysis to examine the association between the results of the meta-analyses and factors related to study design, treatment, and patients. We identified 52 studies (total adult patients = 18,675, LAI-APs = 12,613, placebo = 2,083, and OAPs = 3979; median study duration = 32 weeks). DWC was higher for LAI-aripiprazole than for the placebo [risk ratio (95% confidence interval) = 1.70 (1.23–2.39)]. Neither pooled nor individual LAI-APs differed from the placebo for fluphenazine, olanzapine, paliperidone, and risperidone or from the OAPs for aripiprazole, fluphenazine, haloperidol, olanzapine, paliperidone, risperidone, and zuclopenthixol. The average DWC of each LAI-AP was as follows: LAI-aripiprazole = 10.98%, LAI-fluphenazine = 7.65%, LAI-flupenthixol = 3.33%, LAI-haloperidol = 6.71%, LAI-olanzapine = 10.50%, LAI-paliperidone = 10.38%, LAI-perphenazine = 7.06%, LAI-risperidone = 10.39%, LAI-zuclopenthixol = 4.45%, pooled LAI-APs = 9.88%, and placebo = 11.17%. Meta-regression analysis demonstrated that publication year (β = 0.019), percentage of males (β = 0.020), and mean age (β = 0.045) were associated with an average DWC for pooled LAI-APs. Study duration (β = −0.028), percentage of males (β = 0.078), and patient status (β = −0.847) were associated with an average DWC for LAI-aripiprazole. Presence of a placebo arm (β = 1.596) was associated with an average DWC for LAI-fluphenazine. LAI-AP use was unlikely to be associated with DWC. Although the LAI-aripiprazole had a higher DWC than did the placebo, its average DWC was similar to other those of LAI-APs.
UR - http://www.scopus.com/inward/record.url?scp=85092908156&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85092908156&partnerID=8YFLogxK
U2 - 10.1016/j.jpsychires.2020.10.009
DO - 10.1016/j.jpsychires.2020.10.009
M3 - Article
C2 - 33096355
AN - SCOPUS:85092908156
SN - 0022-3956
VL - 132
SP - 144
EP - 150
JO - Journal of Psychiatric Research
JF - Journal of Psychiatric Research
ER -