Relationship between Guillain-Barré syndrome, influenza-related hospitalizations, and influenza vaccine coverage

Some studies reported an increased risk of Guillain-Barré syndrome (GBS) within six weeks of influenza vaccination. It has also been suggested that this finding could have been confounded by influenza illnesses. We explored the complex relationship between influenza illness, influenza vaccination, and GBS, from an ecologic perspective using nationally representative data. We also studied seasonal patterns for GBS hospitalizations. Monthly hospitalization data (2000-2009) for GBS, and pneumonia and influenza (P&I) in the Nationwide Inpatient Sample were included. Seasonal influenza vaccination coverage for 2004-2005 through the 2008-2009 influenza seasons (August-May) was estimated from the National Health Interview Survey data. GBS seasonality was determined using Poisson regression. GBS and P&I temporal clusters were identified using scan statistics. The association between P&I and GBS hospitalizations in the same month (concurrent) or in the following month (lagged) were determined using negative binomial regression. Vaccine coverage increased over the years (from 19.7% during 2004-2005 to 35.5% during 2008-2009 season) but GBS hospitalization did not follow a similar pattern. Overall, a significant correlation between monthly P&I and GBS hospitalizations was observed (Spearman's correlation coefficient=0.7016, p<0.0001). A significant (p=0.001) cluster of P&I hospitalizations during December 2004-March 2005 overlapped a significant (p=0.001) cluster of GBS hospitalizations during January 2005-February 2005. After accounting for effects of monthly vaccine coverage and age, P&I hospitalization was significantly associated (p<0.0001) with GBS hospitalization in the concurrent month but not with GBS hospitalization in the following month. Monthly vaccine coverage was not associated with GBS hospitalization in adjusted models (both concurrent and lagged). GBS hospitalizations demonstrated a seasonal pattern with winter months having higher rates compared to the month of June. P&I hospitalization rates were significantly correlated with hospitalization rates for GBS. Vaccine coverage did not significantly affect the rates of GBS hospitalization at the population level.