This study is not about these drugs, and that needs saying before anything else [1]. It measures whether people with mental illness receive the routine monitoring that diabetes care depends on — blood tests, eye screening, foot checks. It is here because of what this desk has published about the same population — including the prescribing gap documented in four drugs in one class, and two showed nothing.
Across 49 studies and 5.5 million people, those with a diagnosed mental disorder were less likely to receive recommended monitoring: an odds ratio of 0.81 (95% CI 0.70–0.94) overall, and the same direction for HbA1c testing (0.81), retinal screening (0.77), cholesterol measurement (0.83) and foot examination (0.85). The gaps are modest and consistent, with several upper bounds sitting close to parity.
For someone buying through a telehealth seller the connection tightens rather than loosens. A cash-pay prescription arrives without the monitoring infrastructure that a diabetes clinic would provide, and the people most likely to be routed toward that model are often the people already receiving less attention elsewhere. What patients report about the variability of clinical support is in what patients said when asked, and what happens when support is absent in every drug worked better with support attached.
The pooled figures come with the usual caveat for a meta-analysis spanning four continents: health systems differ enormously, and an odds ratio averaged across Asia, North America, Europe and Australasia describes none of them precisely. Most included studies were rated high quality, and all 49 are observational.
The practical translation is a question to ask rather than a number to use. If you or someone you are buying for has a psychiatric diagnosis, the monitoring that should accompany a metabolic drug is the thing statistically most likely to be missing — which makes it worth asking a seller what follow-up they actually provide, alongside the access questions in who the price filtered out.