(20) Using Healthcare Claims to Evaluate Long Haul COVID-19 Disease

Publication Number: 351

Background: While many ongoing studies are investigating the real-world effect of available COVID-19 vaccines on morbidity and mortality, one aspect of COVID-19 vaccine effectiveness that has received less attention is the effect of the vaccines on long haul COVID-19 disease after breakthrough infections. Research into this clinical condition has been hampered, until very recently, by the lack of an ICD-10 code to denote the presence of long haul COVID-19 disease in electronic healthcare databases. In October 2021, a new ICD-10 code (U09.9) was introduced for long-haul COVID-19.

Objectives: We sought to evaluate the prevalence of the new long haul COVID-19 diagnosis code, and describe characteristics of patients who receive it within a large U.S. healthcare claims database.

Methods: Beginning in October 2021, and extending through the most recent data available (December 15, 2021), we identified enrolled patients who had a claim bearing the ICD-10 code U09.9, corresponding to long haul COVID-19. For these patients, we also identified any other diagnosis codes received up to 1 month before and 2.5 months after their initial long haul COVID-19 diagnosis. We then ranked these diagnoses by prevalence among the long haul COVID-19 population.

Results: A total of 8,223 patients received a long haul COVID-19 diagnosis code during this 2.5 month window. The majority (58%) of these patients were female and the majority (69%) were below 55 years of age. The most common co-occurring diagnoses were other COVID-19-related diagnoses, such as personal history of or exposure to coronavirus. Other commonly recorded conditions included essential hypertension (38%), unspecified hyperlipidemia (19%), fatigue (19%), dyspnea (16%), Type II diabetes mellitus (16%), hypoxemia (15%), gastro-esophageal reflux (14%), anxiety (14%), and chest pain (13%). Many other diagnoses related to cardiovascular problems, weakness/fatigue, and respiratory issues were also present.

Conclusions: This exploratory analysis provides insight into early usage of the code for long haul COVID-19 in a large US claims database. The population explored in this analysis is highly heterogeneous, as it may include patients with recent coronavirus infection as well as patients who were infected early in the pandemic and have had ongoing symptoms. Subsequent analyses will explore the phenotypic patterns associated with this code, including disease trajectory and seek to elucidate the time course from infection to the long haul designation.