Background: Guillain-Barré syndrome (GBS) is an outcome of interest in numerous vaccine safety studies, including the vaccines for respiratory syncytial virus (RSV). The positive predictive value (PPV) of a claims-based GBS algorithm commonly used in the United States has been reported as ~75%, but may vary by key patient characteristics such as age and underlying conditions.
Objective: To estimate the PPV of a claims-based GBS algorithm in subgroups relevant to vaccine safety studies.
Methods: GBS was identified among patients aged ≥18 years in 2017 through 2024 via claims data from two US integrated delivery systems. Patients were required to have ≥365 days of medical and pharmacy coverage prior to GBS. Incident GBS was defined as an inpatient ICD-10-CM code G61.0 in the principal position. The index date was the hospital admission date, unless an outpatient visit with G61.0 occurred in the preceding 7 days, in which case the earlier date was used. Incidence was defined as no GBS diagnosis code (G61.0; G65.0, GBS sequelae) in the 365 days prior to index (any position, any setting). From the cohort of potential GBS events, a random sample of 300 is undergoing medical chart abstraction and adjudication using the Brighton Collaboration criteria. PPVs will be calculated overall and by age, sex, year, comorbidity score, risk of severe RSV-associated lower respiratory tract disease, any vaccination in the previous 42 days, and infection in the previous 30 days.
Results: We identified 780 potential GBS episodes among 778 patients. The mean age was 55 yrs (58% aged 18-59 yrs; 31% aged 60-74; 11% aged ≥75); 57% were male, 7% had a recent respiratory infection, 10% had a recent gastrointestinal infection, and 20% had a recent vaccination of any kind. Adjudication has been completed for 71 of the 300 GBS episodes as of Jan 2026. Of these, 49 met Brighton Levels 1-3 (confirmed GBS), for a preliminary overall PPV of 69%. Among the 49 confirmed GBS episodes, a potential etiology was identified for 33 (67%). Of these 33, adjudicators identified 31 (94%) with infection as the potential etiology and 2 (6%) as vaccine.
Conclusions: Of the 71 GBS episodes that have been chart-reviewed, 49 were confirmed, resulting in a preliminary overall PPV of 69%. Chart review will be completed by Spring 2026, enabling reporting of a final overall PPV, subgroup PPVs, and potential etiologies.
View Poster
Bell EJ, Sun JW, Xu S, Wu A, Liu J, Hsu C, Pio J, Canul NM, Garcia C, Perez R, Nolan MB, DeFor TA, Horgan C, Back S, Emerton D, Fuller CC, Round KE, Walsh ST, Cole DV, Rajbhandari R, Vasquez A, Zhang F, Weng J, Maro JC, Platt R, Cocoros NM. Risk-Factor-Stratified Positive Predictive Values of an Algorithm for Guillain-Barré Syndrome. 42nd Annual International Society for Pharmacoepidemiology Conference; 2026 Sep 1; Milan, Italy.
