Prevalence and correlates of antipsychotic polypharmacy: a systematic review and meta-regression of global and regional trends from the 1970s to 2009

Abstract

Objective: To assess the prevalence and correlates of antipsychotic polypharmacy (APP) across decades and regions.

Methods: Electronic PubMed/Google Scholar search for studies reporting on APP, published from 1970 to 05/2009. Median rates and interquartile ranges (IQR) were calculated and compared using non-parametric tests. Demographic and clinical variables were tested as correlates of APP in bivariate and meta-regression analyses.

Results: Across 147 studies (1,418,163 participants, 82.9% diagnosed with schizophrenia [IQR=42-100%]), the median APP rate was 19.6% (IQR=12.9-35.0%). Most common combinations included first-generation antipsychotics (FGAs)+second-generation antipsychotics (SGAs) (42.4%, IQR=0.0-71.4%) followed by FGAs+FGAs (19.6%, IQR=0.0-100%) and SGAs+SGAs (1.8%, IQR=0.0-28%). APP rates were not different between decades (1970-1979:28.8%, IQR=7.5-44%; 1980-1989:17.6%, IQR=10.8-38.2; 1990-1999:22.0%, IQR=11-40; 2000-2009:19.2% IQR=14.4-29.9, p=0.78), but between regions, being higher in Asia and Europe than North America, and in Asia than Oceania (p<0.001). APP increased numerically by 34% in North America from the 1980s 12.7%) to 2000s (17.0%) (p=0.94) and decreased significantly by 65% from 1980 (55.5%) to 2000 (19.2%) in Asia (p=0.03), with non-significant changes in Europe. APP was associated with inpatient status (p<0.001), use of FGAs (p<0.0001) and anticholinergics (<0.001), schizophrenia (p=0.01), less antidepressant use (p=0.02), greater LAIs use (p=0.04), shorter follow-up (p=0.001) and cross-sectional vs. longitudinal study design (p=0.03). In a meta-regression, inpatient status (p<0.0001), FGA use (0.046), and schizophrenia diagnosis (p=0.004) independently predicted APP (N=66, R(2)=0.44, p<0.0001).

Conclusions: APP is common with different rates and time trends by region over the last four decades. APP is associated with greater anticholinergic requirement, shorter observation time, greater illness severity and lower antidepressant use.

Figure 1

Quality of Research of Meta-analysis (QUORUM) flow chart.

Figure 2

Antipsychotic polypharmacy rate by decade and geographic region ■p=0.04; ▲Omnibus (p=0.03). Regions were significantly different after post-hoc test using a rank test corrected for multiple comparisons. ^Number of time points included in the analysis. Time point definition: some studies reported in a single article two or more APP rates, examined at different points in time (e.g. 1999 and 2000), constituting two or more different individual samples or “time points” in our database. Total Number of studies: 147. Total number of time points: 196. Studies were classified based on time of data collection and not publication date.