Why diagnosis codes alone fall short for long COVID surveillance
A global pharmaceutical company examined about 24.1 million patients with both closed medical claims and HealthVerity Notes from October 2021 through September 2025 to see how many long COVID (PASC) patients diagnosis codes alone miss, since long COVID is often documented in clinical notes rather than formal codes. Combining ICD-10 code U09.9 with keyword searches ("long COVID," "post-COVID," "PASC") in clinical notes identified 287,000 total patients versus 242,000 found via diagnosis codes alone — a roughly 19% increase (45,000 additional patients) from notes. Diagnosis coding plateaued at about 75% of identified cases, overlap between the two sources stayed minimal (4-6%), and by 2025 only about 10% or fewer of note-identified patients had a corresponding diagnosis code; adults 45+ made up most identified cases.
The customer is anonymised in HealthVerity’s own case study — not withheld by us.
HealthVerity is a real-world-data marketplace aggregating claims, EHR and lab data from 75+ sources via a proprietary patient-matching layer.
This sits in the Data Providers stage of the AI use-case lifecycle: Named examples from the real-world-data, claims and competitive-intelligence providers pharma AI systems are trained on, benchmarked against, or run inside of.
Published by HealthVerity on its own site — read the original case study on blog.healthverity.com ↗
Related use cases
- 01Symphony Health and dentsu health partner to efficiently reach rare disease patients
- 02Differentiating TYK2 inhibitors with real-world evidence in psoriasis
- 03How HealthVerity Audiences delivered 26.9x targeting precision for a rare kidney disease campaign
- 04Reducing targeting noise by supplementing claims with oncology labs: Metastatic breast cancer case study
- 05Accurately targeting chronic liver disease patients for specialty therapy
- 06Accurately reaching patients with rare hematologic disease