Risk Factors for Long COVID Among Individuals with Noninfectious Uveitis in a Large United States Claims Database, 2026, Kumar et al.

[SNT Gatchaman]

Risk Factors for Long COVID Among Individuals with Noninfectious Uveitis in a Large United States Claims Database

Spoiler: Authors

Anika Kumar; Emily Tang; Xinyi Xu; Alythia Vo; Benjamin F Arnold; Nisha R Acharya

PURPOSE
Patients with noninfectious uveitis (NIU) may be more susceptible to complications of COVID-19, including long COVID, yet limited research has evaluated this outcome in NIU populations. This study aimed to assess the prevalence and risk factors of long COVID among individuals with NIU in the United States.

METHODS
A retrospective cohort analysis was conducted using de-identified data from a large healthcare claims database. A time-varying Cox proportional hazard regression analysis was performed to estimate the effects of various risk factors on the development of long COVID. Models were adjusted for use of systemic corticosteroids and other immunosuppressive medications, comorbidities, COVID-19 vaccination status, and demographic variables.

RESULTS
The study population consisted of 63 220 individuals with NIU (mean age (SD) = 62.9 (17.4) years, 60.7% female). There were 621 (0.096%) documented cases of long COVID across 100 105.6 person-years, representing an incidence rate of 6.2 cases of long COVID per 1000 person-years. Exposure to systemic corticosteroids was associated with an average increased risk of long COVID (hazard ratio (HR) = 3.45, 95% CI: 2.68–4.46, p < 0.001). Additionally, COVID-19-related hospitalizations, increased healthcare utilization, and having baseline asthma and mental health disorders were also associated with an average increased risk of long COVID (all p < 0.003). Male sex, topical or local corticosteroid exposure, and COVID-19 vaccination were associated with an average decreased risk of long COVID (all p < 0.001).

CONCLUSION
This large population-based study identified key risk and protective factors for long COVID among patients with NIU. Findings suggest that systemic corticosteroid–induced immune dysregulation may increase vulnerability to long COVID in individuals with severe uveitis.

Web | DOI | Ocular Immunology and Inflammation | Open Access

 

[Hutan]

Might be worth a look. There will probably be issues with the definition of Long Covid of course.

Patients with noninfectious uveitis (NIU) represent a clinically important population in the context of COVID-19–related outcomes. Individuals with NIU are known to be at increased risk of COVID-19 infection and to experience more severe acute disease and related complications, a vulnerability thought to be driven primarily by underlying comorbidities, demographic factors, and the frequent use of systemic immunosuppressive therapies rather than by NIU itself. Patients with NIU are often treated with systemic immunosuppressive medications, such as corticosteroids, disease-modifying anti-rheumatic drugs (DMARDs), and biologics, which may plausibly influence immune responses to viral infection and post-infectious sequelae.

investigation of the disease using insurance claims data in the United States has been made possible with the new International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) diagnosis code, U09.9, which was introduced in October 2021 to represent “Post COVID-19 condition, unspecified.”

Yeah, the data is going to be pretty muddled by people on those immunosuppressants having long term consequences of more severe Covid-19 infections.

 

[Hutan]

Time-varying dichotomous exposures to immunosuppressive medications (systemic corticosteroids (SCs), DMARDs, tumor necrosis factor (TNF)-alpha inhibitors, interleukin (IL)-6 inhibitors, anti-CD20 monoclonal antibodies (anti-CD20), other biologics, and other non-biologic immunosuppressive medications) were extracted from outpatient pharmacy claims data (eAppendix 4), and doses were extracted for all medications prescribed in an individual’s exposure period.

time-varying covariates during the risk period: dichotomous indicator use of SCs, DMARDs, TNFs, and other non-biologic immunosuppressive medications, and average daily prednisone equivalent dose during periods of SC exposure, reported as 10 mg unit change per day,

It's good how they have tried to work out a prednisone equivalent dose for systemic corticosteroid use.

63 220 individuals with NIU

During the risk period, 22 863 (36.2%) individuals were found to have one or more documented prescription fills of an immunosuppressive medication (Table 1). The most common drug class was SCs (filled by 32.3% of the overall cohort), followed by DMARDs (filled by 5.4%), TNF-α inhibitors (filled by 2.9%), and IL-6 inhibitors, anti-CD20 monoclonal antibodies, and other biologics (each filled by less than 0.1%).

Additionally, any exposure to a SC was associated with an increased risk of long COVID (HR = 3.45, 95% CI: 2.68–4.46, p < 0.001), while a decreased risk of long COVID was associated with exposure to DMARDS (HR = 0.64, 95% CI: 0.43–0.96, p = 0.03) and treatment with topical or local corticosteroids, a proxy of active NIU and disease severity (HR = 0.62, 95% CI: 0.52–0.73, p < 0.001).

Having one or more COVID-19 hospitalizations was associated with a greater risk of long COVID (HR = 5.73, 95% CI: 4.47–7.35, p < 0.001), while having one or more COVID-19 vaccination records was associated with a decreased risk of long COVID (HR = 0.56, 95% CI: 0.48–0.66, p < 0.001). Increased healthcare utilization measured by number of ambulatory visits were associated with a mildly increased risk of long COVID (HR = 1.004, 95% CI: 1.001–1.006, p = 0.003).

The baseline comorbidities that were significantly associated with an increased risk of long COVID in the adjusted models were asthma (HR = 1.46, 95% CI: 1.17–1.82, p < 0.001) and mental health disorders (HR = 1.31, 95% CI: 1.08–1.59, p = 0.007) (Table 3, Figure 1). Reported hazard ratios summarize effects over the study period.

The association of increased health care utilisation with long covid gave an HR of 1.004. That's nothing and really should not have been mentioned.

So, systemic corticosteroid use had an HR of Long Covid of 3.45. But, one or more Covid-19 hospitalisations had an HR of 5.73 which is huge. It seems likely that there is some interactions between systemic corticosteroid use and severe acute Covid-19 there. And that high HR of long Covid for hospitalisations strongly suggests that a lot of the Long Covid cases are not ME/CFS.

People using DMARDS had a lower risk of Long Covid (HR 0.58). Perhaps that's worth looking into. People on TNF-a inhibitors had a HR of Long Covid of 0.88, perhaps also interesting.