Zika virus infection and risk of Guillain-Barré syndrome: A meta-analysis

Abstract

Objective: Findings from studies of the association between Zika virus (ZIKV) infection and Guillain-Barré syndrome (GBS) are inconsistent. I conducted a systematic review and meta-analysis to clarify the nature of this association.

Methods: I searched PubMed, Scopus, Cochrane, CINAHL, Web of Science, Scielo, and DOAJ for case report, ecological, and analytic studies with "Zika" and "Guillain-Barré syndrome" as keywords, published up to July 1st^th 2018. I evaluated if ZIKV infection status influenced the diagnosis of GBS (detection bias) in case-report and analytic studies; assessed if changes in weekly number of cases of ZIKV infection during outbreaks were followed by changes in number of GBS cases 1-8 weeks later; gauged the likelihood of selection, confounding, information, sparse data, and time-dependent bias (i.e. when ZIKV infection was ascertained after GBS onset) in analytic studies; and calculated the average ZIKV-GBS odds ratio (OR) in studies without time-dependent bias.

Results: In case reports, ZIKV infection prevalence in GBS cases was 2.4 to 25 times higher than expected. Changes in the number of ZIKV-infection cases during outbreaks were not consequentially followed by changes in the number of GBS cases (OR: 1.01; 95% CI: 0.99-1.03). Major biases were likely in all but one analytic study, which showed a non-significant ZIKV-GBS association. The average ZIKV-GBS OR in studies without time-dependent bias was 1.57 (95% CI: 0.86-2.86).

Interpretation: These findings indicate the available evidence is insufficient to claim ZIKV infection causes GBS. Therefore, stakeholders may want to reconsider current ZIKV-GBS public health and patient care recommendations.

Keywords: Disease outbreaks; Guillain-Barre syndrome; Systematic review; Zika virus.