AI Tool Finds More Long COVID Cases Than Codes - Report - MDSpire
Advertisement
AI Tool Finds More Long COVID Cases Than Codes
Nearly 90% of patients who met algorithmic criteria for postacute sequelae of SARS-CoV-2 infection had at least 1 chronic or potentially chronic condition requiring ongoing clinical management.
Clinical Report: AI Tool Finds More Long COVID Cases Than Codes
Overview
A study utilizing an AI-enabled electronic health record phenotyping approach identified 16% of COVID-19 patients as having postacute sequelae of SARS-CoV-2 infection (PASC), significantly higher than traditional diagnostic coding methods. The algorithm demonstrated consistent performance across diverse populations.
Background
Long COVID, or postacute sequelae of SARS-CoV-2 infection (PASC), is a chronic condition affecting many individuals after COVID-19. Accurate identification and monitoring of PASC are critical for effective healthcare planning and resource allocation. Traditional diagnostic coding has been shown to capture only a small fraction of affected individuals.
Data Highlights
Region
PASC Prevalence
New England
19%
Southeast Texas
20%
Southern California
23%
Western Pennsylvania
14%
Key Findings
The AI phenotyping algorithm identified PASC in 16% of COVID-19 patients.
Regional prevalence rates varied, with Southern California showing the highest at 23%.
Patients with PASC were generally older, had higher comorbidity burdens, and were more likely to be female.
Only 4% of the associated ICD-10 codes were classified as acute conditions.
Systemic manifestations were the most common PASC symptoms, accounting for 23% to 25% of cases across regions.
Clinical Implications
Clinicians should be aware of the high prevalence of chronic conditions among patients with PASC.
Conclusion
Continued efforts are needed to improve the accuracy of PASC detection and management.
