FC 07-09Hazard of complex regional pain syndrome (CRPS) following HPV vaccination among adolescents in the United States

We estimated the association between adolescent vaccination and incident complex regional pain syndrome (CRPS) in adolescent girls in the U.S. 

We used insurance claims to identify claims for CRPS and adolescent vaccination, including HPV, tetanus-diphtheria-acellular pertussis (Tdap) and meningococcal conjugate vaccine (MenACWY), using diagnosis and procedure codes. We identified 11-year-old girls between 2006-2014 without a prior history of adolescent vaccination or CRPS, and followed them from the 11th birthday until CRPS diagnosis, loss to follow-up, or December 31, 2014. Time-dependent Cox models estimated hazard ratios (HRs) comparing CRPS hazard following recent vaccination (30 days) to CRPS hazard following prior (>30 days) or no vaccination. HRs were adjusted for history of physical trauma in the year preceding follow-up or an instance of trauma during follow-up. We then estimated time-dependent HRs for CRPS following recent co-administration of common vaccine combinations. Finally, we identified common health diagnoses received by girls in the sample over follow-up, and estimated time-dependent HRs for CRPS comparing girls with each diagnosis to girls without. 

We identified 563 CRPS cases among 1,232,572 girls (incidence rate: 20.6/100,000 person-years). CRPS hazard was not significantly elevated following recent HPV (adjusted HR [aHR]: 1.41, 95% CI: 0.83, 2.40), Tdap (aHR: 1.20, 95% CI: 0.64, 2.59), or MenACWY (aHR: 1.21, 95% CI: 0.57, 2.55) vaccination. Comparing recent or non-recent vaccination to no vaccination, Tdap (HR: 1.59, 95% CI: 1.34, 1.89) and MenACWY (HR: 1.70, 95% CI: 1.43, 2.02) vaccination were associated with CRPS in crude analysis, but were not associated after adjusting for trauma (Tdap - aHR:1.09, 95% CI: 0.91, 1.29; MenACWY - aHR: 1.19, 95% CI: 1.00, 1.42). CRPS hazard comparing concomitant HPV vaccination to HPV vaccination alone was not significantly elevated (aHR: 2.30, 95% CI: 0.80, 6.56). Girls with lower limb injuries had the greatest CRPS hazard compared to girls without (HR: 12.4, 95% CI: 10.4, 14.7). Girls with anxiety disorders had a threefold hazard of CRPS compared to girls without (HR: 3.12, 95% CI: 2.41, 4.04). Common pediatric illnesses (e.g. asthma, respiratory infections, allergies) were positively associated with CRPS in bivariate analyses.

Adolescent CRPS is rare in the U.S., and adolescent vaccination was not significantly associated with CRPS hazard. Crude HRs for CRPS following vaccination were confounded by physical trauma, a known CRPS risk factor. Future research should consider health-related risk factors for CRPS in adolescents, particularly injuries, and inflammatory and mental illnesses.