We summarise a recent study that used real-world data (RWD) to explore the prevalence of adverse effects in patients before and after they started clozapine treatment.
Clozapine is a second-generation antipsychotic drug. It the gold-standard and most effective antipsychotic drug in the management of schizophrenia. However, it is an underutilised medication with only 54% of eligible patients being prescribed it in the UK. One of the primary reasons for this is due to the concerns regarding its side effects. The most common side effects include weight gain and hypoglycaemia which together can result in type II diabetes. A more acute side effect is agranulocytosis – a severe and dangerous lowered white blood cell count.
Adverse drug reactions (ADRs) are troublesome and potentially life-threatening. Improving our ability to detect these can lead to a greater understanding of individual risks and better patient management. Once a drug is on the market, new ADRs are reported by physicians and pharmacists. Current reporting systems have several limitations, including under-reporting of observed ADRs. In recent years, attention has shifted to the use of data obtained within electronic health records (EHRs). This data can provide additional insight into adverse events in real-world settings.
In this study, published in PLOS ONE, researchers used data from de-identified EHRs of three mental health trusts in the UK. Specifically, >50 million documents and over 500,000 patients, 2835 of which were prescribed clozapine. The team explored the prevalence of 33 adverse effects by age, gender, ethnicity, smoking status and admission type, three months before and after the patients started clozapine treatment. They developed a natural language processing pipeline – ADEPt – to detect and validate mention of ADRs in free-text clinical notes. Finally, they compared the results with the prevalence of adverse effects reported in the Side Effects Resource (SIDER).
The researchers identified sedation, fatigue, agitation, dizziness, hypersalivation, weight gain, tachycardia, headache, constipation and confusion as the highest recorded adverse effects. A higher percentage of these were specifically found in the first month of clozapine therapy. They also found a significant association between most of the ADRs and smoking status and hospital admission.
These findings demonstrate the utility of wider ADR extraction beyond clozapine, which can be replicated using other drugs. Most importantly, it also highlights the value of extracting data from free-text clinical documents to help clinicians and researchers predict risks and interventions of drug administration.
Image credit: By rawpixel.com – freepik