Despite increased cardiometabolic risk in individuals with mental illness taking antipsychotic medication, metabolic screening practices are often incomplete or inconsistent.

We undertook a systematic search and a PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) meta-analysis of studies examining routine metabolic screening practices in those taking antipsychotics both for patients in psychiatric care before and following implementation of monitoring guidelines.

We identified 48 studies (n=290 534) conducted between 2000 and 2011 in five countries; 25 studies examined predominantly schizophrenia-spectrum disorder populations; 39 studies (n=218 940) examined routine monitoring prior to explicit guidelines; and nine studies (n=71 594) reported post-guideline monitoring. Across 39 studies, routine baseline screening was generally low and above 50% only for blood pressure [69.8%, 95% confidence interval (CI) 50.9–85.8] and triglycerides (59.9%, 95% CI 36.6–81.1). Cholesterol was measured in 41.5% (95% CI 18.0–67.3), glucose in 44.3% (95% CI 36.3–52.4) and weight in 47.9% (95% CI 32.4–63.7). Lipids and glycosylated haemoglobin (HbA1c) were monitored in less than 25%. Rates were similar for schizophrenia patients, in US and UK studies, for in-patients and out-patients. Monitoring was non-significantly higher in case-record versus database studies and in fasting samples. Following local/national guideline implementation, monitoring improved for weight (75.9%, CI 37.3–98.7), blood pressure (75.2%, 95% CI 45.6–95.5), glucose (56.1%, 95% CI 43.4–68.3) and lipids (28.9%, 95% CI 20.3–38.4). Direct head-to-head pre–post-guideline comparison showed a modest but significant (15.4%) increase in glucose testing (p=0.0045).

In routine clinical practice, metabolic monitoring is concerningly low in people prescribed antipsychotic medication. Although guidelines can increase monitoring, most patients still do not receive adequate testing.