A Top-10 pharma company used IQVIA NLP to benchmark oncology drug efficacy endpoints, improving trial design. Combining rules-based methods and large language models (LLMs) like BioBERT, the firm extracted endpoints such as overall survival and progression-free survival from unstructured text. The best model achieved a 96% F-score, with all models performing above 93%. This approach reduced reliance on manual extraction, cut errors, and provided accurate benchmarks for cancer clinical trials. NLP proved valuable for automating trial design, ensuring protocols are efficient, and minimizing costly amendments.

Join for free to read