Table 4 Example of a generated discharge summary using prompt chaining.
From: Automated generation of discharge summaries: leveraging large language models with clinical data
Generated text | |
|---|---|
I have reviewed and adjusted the two sections again to ensure that no content is duplicated. Here are the two sections: **History and findings** The 65-year-old male patient was admitted to our clinic on 30.06.2033 with acute pancreatitis. The medical history revealed previous treatment for colon cancer, which was operated on 03/2033, as well as a colon polyp initial diagnosis on 02/2033. The patient also suffered from arterial hypertension and diabetes mellitus since 02/2032, for which he has been insulin-dependent since 03/2032 **Therapy and course** The patient underwent an exploratory laparotomy, left pancreatic resection with splenectomy and cholecystectomy on 04/07/2033. The operation was without complications. Following the operation, the patient received pain therapy with opioids p.o. and i.v. as well as regional anesthesia (including epidural anesthesia (PDA)). The wound healed without irritation and the patient could be mobilized as before the operation. The patient’s diet was built up in good time and adequate bowel transit was achieved from 08.07.2033 I hope that these are the corrected sections! |
