Advances Towards Automating Safety Data Entry: New Solution Builds On Optical Character Recognition To Extract SAE Form Data
Advances Towards Automating Safety Data Entry: New Solution Builds On Optical Character Recognition To Extract SAE Form Data
9 PagesSafety data reporting for adverse events (AEs) and serious adverse events (SAEs) is a costly, labor-intensive part of pharmacovigilance (PV), with data entry consuming 45% of team time and each case requiring 1.5–3 hours. Errors remain high—up to 80% of extracted data can be incorrect even after QC. With PV spending projected to reach $14.95B by 2028, efficiencies are critical. Rising AE volumes—an 84% increase in FDA reports from 2014–2020 and a 93% jump in EMA’s 2021 cases (driven by COVID-19 vaccines)—make manual models unsustainable, highlighting the urgent need for automation and new approaches.
