Manya Prasad, Ettore Beghi, K. Prasad, Fan Kee Hoo, Tom Solomon, Andrea Sylvia Winkler, Arpna Srivastava, Tarun Dua, Ajeet Kumar, Kiran T. Thakur, Benedict D Michael, Kolappa K, Ericka L. Fink, Alla Guekht, Sanjeev Misra, Amit Agarwal, Sherry H.-Y. Chou, Amir Kheradmand, Divya M. Radhakrishnan, Carlos A. Pardo, Erich Schmutzhard, Abanti Das, and Greta K. Wood
SummaryObjectiveTo summarize the frequency of neurological manifestations reported in COVID-19 patients and investigate the association of these manifestations with disease severity and mortality.DesignSystematic review and meta-analysisEligibility criteriaStudies enrolling consecutive COVID-19 patients (probable or confirmed) presenting with neurological manifestations.Data sourcesPubMed, Medline, Cochrane library, clinicaltrials.gov and EMBASE from 31st December 2019 to 15th December 2020.Data extraction and analysisTwo authors independently screened titles and abstracts retrieved by literature search. Risk of bias was examined using Joanna Briggs Institute (JBI) scale. A random-effects meta-analysis was performed, and pooled prevalence and 95% Confidence Intervals (CI) were calculated for neurological manifestations. Odds ratio (OR) and 95%CI were calculated to determine the association of neurological manifestations with disease severity and mortality. Presence of heterogeneity was assessed using I-square, meta-regression, and subgroup analyses. Statistical analyses were conducted in R version 3.6.2.ResultsOf 2,455 citations, 350 studies were included in this review, providing data on 145,634 COVID-19 patients, 89% of whom were hospitalized. Forty-one neurological manifestations (24 symptoms and 17 diagnoses) were identified. Pooled prevalence of the most common neurological symptoms included: fatigue (32%), myalgia (20%), taste impairment (21%), smell impairment (19%) and headache (13%). A low risk of bias was observed in 85% of studies; studies with higher risk of bias yielded higher prevalence estimates. Stroke was the most common neurological diagnosis (pooled prevalence-2%). In COVID-19 patients aged >60, the pooled prevalence of acute confusion/delirium was 34% and the presence of any neurological manifestations in this age group was associated with mortality (OR 1.80; 95%CI 1.11 to 2.91).ConclusionsUp to one-third of COVID-19 patients analysed in this review experienced at least one neurological manifestation. One in 50 patients experienced stroke. In those over 60, more than one-third had acute confusion/delirium; the presence of neurological manifestations in this group was associated with near doubling of mortality. Results must be interpreted keeping in view the limitations of observational studies and associated bias.Systematic review registrationPROSPERO CRD42020181867.What is already known on this topicThe frequency of neurological manifestations including fatigue, myalgia, taste and smell impairments, headache and dizziness in COVID-19 patients has been reported in a few systematic reviews and meta-analyses. However, considerable heterogeneity has been observed in terms of methodological quality of the studies, severity of the disease, mean age and hospitalization status of the patients. The evidence regarding the frequency of neurological diagnoses including stroke, encephalitis, Guillain Barré syndrome (GBS) is also limited to case reports and case series and no data exists thus far on the pooled prevalence estimates for neurological diagnoses in COVID-19 patients.What this study addsTo the best of the authors’ knowledge, this is the largest systematic review and meta-analysis to date (including 350 studies with data on 145,634 cases) summarizing the evidence on the frequency of the full spectrum of neurological manifestations in COVID-19 patients in the overall, young and elderly populations. For the first time, our review reports the pooled prevalence of stroke in COVID-19 patients. Risk of bias, old age and disease severity were potential determinants of the frequency and nature of neurological manifestations as well as its association with mortality. Our review also highlights the need to develop reporting standards for studies describing the frequency of clinical features. Moreover, we note that this will be the first systematic review and meta-analysis on this subject to include studies reported in all languages.