Background: Many patients who are discharged from the emergency department (ED) with a symptom-based discharge diagnosis (SBD) have post-discharge challenges related to lack of a definitive discharge diagnosis and follow-up plan. Research regarding outcomes for these patients is challenging as there is no well-defined method for identifying patients with a SBD without individual chart review. We describe a method for automated identification of SBDs from ICD-10 codes using the Unified Medical Language System (UMLS) Metathesaurus. The UMLS Metathesaurus is a biomedical thesaurus that organizes various medical vocabularies into unique concepts, then assigns each concept to a semantic relationship built into a Semantic Network.
Methods: We mapped ICD-10 codes from a one-month period of ED discharges at an urban academic ED (annual volume >40,000) to UMLS Concepts and one of 127 possible Semantic Types. For comparison, two physician reviewers then independently manually identified all discharge diagnoses consistent with SBDs within the same dataset. A third physician resolved discrepancies. We calculated inter-rater reliability for manual review and the sensitivity and specificity for our automated process for identifying SBDs against the “gold standard” of manual review.
Results: We mapped 857 unique ICD-10 codes corresponding to 3,642 ED discharges to 10 UMLS semantic types. Over one third (37%, n=1,367) of ED discharges were assigned diagnosis codes that mapped to the “Sign or Symptom” semantic type. Inter-rater reliability for manual review of SBDs was very good (0.87), and 92.9% of diagnoses that were mapped to the semantic type “Sign or Symptom” were confirmed to be SBDs by manual review. Sensitivity and specificity of our automated process for identifying encounters with SBDs were 84.7% and 96.3%, respectively.
Conclusions: Use of our automated process to identify patients discharged with ICD-10 codes that classify into the UMLS “Sign or Symptom” semantic type identified the majority of patients with a SBD. While this method needs refinement to increase capture sensitivity, it has potential to automate an otherwise highly time-consuming process. This novel use of informatics can facilitate further research specific to patients with SBDs and inform improved care delivery for this vulnerable population.