Back

Risk factors for West Nile neuroinvasive disease and mortality in the United States, 2013-2024

Judson, S.; Dowdy, D.

2025-07-21 infectious diseases
10.1101/2025.07.21.25331933 medRxiv
Show abstract

ImportanceWest Nile virus (WNV) is the leading mosquito-borne infection in the United States and can cause West Nile fever (WNF) or West Nile neuroinvasive disease (WNND), resulting in substantial morbidity and mortality. Contemporary risk factor analyses are needed to identify populations at increased risk for severe outcomes and target interventions accordingly. ObjectiveTo assess risk factors for WNND and mortality among adults with WNV infection using a large national cohort from federated real-world data. DesignA retrospective cohort analysis from 2013-2024, consisting of de-identified electronic health record data from the TriNetX Research Network, including adult patients with an ICD-10 diagnosis consistent with WNV infection. Data were accessed from March 28th to April 30th, 2025. SettingThe retrospective cohort included patients from 65 healthcare organizations from across the United States, using the TriNetX Research Network. ParticipantsThe overall initial cohort included 3,064 adult patients with an ICD-10 diagnosis consistent with WNV infection, including 1,328 with WNF and 1,206 with WNND. ExposuresDemographic characteristics (age, sex, race, ethnicity), comorbidities (hematologic malignancy, ischemic heart disease, other forms of heart disease, diabetes, HIV, chronic kidney disease (CKD), liver disease, hypertension, alcohol related disorders, cerebrovascular disease (CEVD), chronic obstructive pulmonary disease, asthma, multiple sclerosis, dementia, rheumatoid arthritis, organ transplant), and medications (immune suppressants, antineoplastics). Main Outcomes and MeasuresThe primary outcomes were the development of WNND and all-cause mortality (30-day, 90-day, and overall mortality). ResultsAmong all patients diagnosed with WNV infection, those with WNND were older (mean age 59 vs 55, p-value <0.0001) and more often male (61% vs 48%, p-value <0.0001). Significant risk factors for WNND included age (adjusted hazard ratio [aHR] 1.11 per decade, 95% confidence interval [CI] 1.06-1.49), male sex (aHR 1.29, 1.15-1.45), CKD (aHR 1.21, 1.002-1.45), CEVD (aHR 1.22, 1.03-1.45), hematologic malignancy (aHR 1.38, 1.09-1.76), immune suppressant use (aHR 1.43, 1.11-1.83), hypertension (aHR 1.18, 1.04-1.34), alcohol related disorders (aHR 1.54, 1.20-1.97), and multiple sclerosis (aHR 2.3, 1.62-3.37). Significant risk factors for mortality were WNND (aHR 2.49 for 30-day mortality, 95% CI 1.37-4.52), age (aHR 1.32 per decade, 95% CI 1.07-1.60), CKD (aHR 2.08, 95% CI 1.01-3.93), and CEVD (aHR 2.00, 95% CI 1.14-3.50). Conclusions and RelevanceRisk factors for WNND broadly reflect impaired immune response and/or central nervous system vulnerability. Patients who developed WNND were at substantially increased risk of death. Targeted prevention strategies and countermeasures for those at greatest risk for WNND could substantially reduce morbidity and mortality. Key PointsO_ST_ABSQuestionC_ST_ABSWhat are risk factors for West Nile virus neuroinvasive disease and mortality in the United States between 2013 and 2024? FindingsIn this retrospective cohort analysis, we found that increased age, chronic kidney disease, and cerebrovascular disease are significantly associated with both West Nile neuroinvasive disease and mortality. Additionally, male sex, hematologic malignancy, immune suppressants, hypertension, alcohol related disorders, and multiple sclerosis were significant risk factors for West Nile neuroinvasive disease. MeaningAn aging population in the United States with increasing rates of comorbidities and immunosuppression faces an increased risk for severe disease from West Nile virus infection.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.