A ten-year retrospective evaluation of acute flaccid myelitis at 5 pediatric centers in the United States, 2005–2014

Background Acute flaccid myelitis (AFM) is a severe illness similar to paralytic poliomyelitis. It is unclear how frequently AFM occurred in U.S. children after poliovirus elimination. In 2014, an AFM cluster was identified in Colorado, prompting passive US surveillance that yielded 120 AFM cases of unconfirmed etiology. Subsequently, increased reports were received in 2016 and 2018. To help inform investigations on causality of the recent AFM outbreaks, our objective was to determine how frequently AFM had occurred before 2014, and if 2014 cases had different characteristics. Methods We conducted a retrospective study covering 2005–2014 at 5 pediatric centers in 3 U.S. regions. Possible AFM cases aged ≤18 years were identified by searching discharge ICD-9 codes and spinal cord MRI reports (>37,000). Neuroradiologists assessed MR images, and medical charts were reviewed; possible cases were classified as AFM, not AFM, or indeterminate. Results At 5 sites combined, 26 AFM cases were identified from 2005–2013 (average annual number, 3 [2.4 cases/100,000 pediatric hospitalizations]) and 18 from 2014 (12.6 cases/100,000 hospitalizations; Poisson exact p<0.0001). A cluster of 13 cases was identified in September–October 2014 (temporal scan p = 0.0001). No other temporal or seasonal trend was observed. Compared with cases from January 2005–July 2014 (n = 29), cases from August–December 2014 (n = 15) were younger (p = 0.002), more frequently had a preceding respiratory/febrile illness (p = 0.03), had only upper extremities involved (p = 0.008), and had upper extremity monoplegia (p = 0.03). The cases had higher WBC counts in cerebrospinal fluid (p = 0.013). Conclusion Our data support emergence of AFM in 2014 in the United States, and those cases demonstrated distinctive features compared with preceding sporadic cases.