Secondary Logo

Journal Logo

Original Studies

Neurologic Complications in Children Hospitalized With Influenza Infections

Prevalence, Risk Factors and Impact on Disease Severity

Solís-García, Gonzalo MD*; Chacón-Pascual, Almudena MD; González Martínez, Felipe MD, PhD*; Miranda Herrero, Mª Concepción MD, PhD; Hernández-Sampelayo, Teresa MD, PhD; Catalán Alonso, Pilar MD§; Rodríguez-Fernández, Rosa MD, PhD*

Author Information
The Pediatric Infectious Disease Journal: September 2020 - Volume 39 - Issue 9 - p 789-793
doi: 10.1097/INF.0000000000002686
  • Free


Influenza virus is one of the main contributors to pediatric respiratory disease, infecting every season 10% to 20% of the world’s population1 and resulting in an estimated 870,000 hospitalizations in children younger than 5 years annually.2 Pediatric and elderly populations bear most of the burden of severe influenza disease3 and children account for the most important part of neurologic complications,1,4 which are associated with higher morbidity and mortality.5

Neurologic manifestations of influenza have been associated with the virus since the 1918 H1N1 pandemic6 and were defined for decades as infrequent complications of the disease.7 However, an apparent increase of incidence and severity associated with the 2009 H1N1 pandemic8,9 has aroused new interest in their epidemiology, pathophysiology and consequences.

The pathogenesis of neurologic events has not been clearly established, both direct invasion of neural cells and inflammatory response seeming to play a role in the process.6 Glycoproteins and glycolipids, which serve as receptors for influenza hemagglutinin proteins, can be found on the surface of most mammalian cells, including brain tissue, and microglia and astrocytes can be infected in vitro by some influenza strains, inducing direct cellular damage and proinflammatory cytokine cascades.10,11

Seizures, both febrile and nonfebrile, are the most frequent of these neurologic complications.4,12 However, Guillain-Barré Syndrome, encephalopathy, stroke and myelitis have also been reported.4,13 The epidemiology and risk factors related to the development of these events have not been clearly established: influenza A has been classically stated as the main neurotropic strain4,8,9,14,15 but other comparative studies find no differences in neurotropism between influenza A and B.12,16 Age younger than 4 years has also been described as a risk factor,5,12 and history of neurologic or neuromuscular disease has also been strongly linked to the risk of neurologic complications.17

The impact of these complications on the morbidity and prognosis of these patients has not been clearly established. Although influenza vaccines are strongly recommended in this group of patients,18 the rates of vaccination continue to be suboptimal19 and both parents and pediatricians fail to recognize these comorbidities as high-risk influenza conditions.20

The aim of this study is to describe neurologic manifestations associated with influenza disease in hospitalized children, analyze clinical and epidemiologic risk factors focusing on prior comorbidities and establish their impact on disease severity.


We conducted a retrospective cohort study and included all hospitalized patients with microbiologic confirmation of influenza disease over 4 epidemic seasons (2015–2018) at Hospital General Universitario Gregorio Marañón, a tertiary-care hospital in Madrid, Spain, which admits in Pediatric wards over 1800 pediatric patients every year. The study was approved by the hospital’s Institutional Review Board, with waived informed consent.

Hospitalized patients younger than 16 years were included if they had at least one positive laboratory test for influenza virus: viral culture, antigen testing or polymerase chain reaction. These laboratory tests are routinely performed during epidemic season when a patient with fever or respiratory symptoms is admitted to the hospital; negative antigen tests are normally confirmed by polymerase chain reaction or viral culture. Positive antigen tests do not need confirmation due to the high specificity of the tests.21 Results were compared with influenza national circulation data from Spain in every epidemic season. There were no exclusion criteria.

Demographic, laboratory and clinical data, as well as past history, were extracted from patients’ electronic records. Chronic conditions were defined as respiratory, cardiac, neurologic, hematologic, nephrologic, oncologic or metabolic diseases, which required active follow-up by pediatricians. Patient history was also revised in search neurologic events during hospitalization, which was then classified into 4 groups: febrile seizures, nonfebrile seizures, encephalitis and others. The term encephalitis was reserved for those patients meeting criteria from the Consensus Statement of the International Encephalitis Consortium.22 The last group included behavioral changes, focal deficits and paroxysmal movements which Pediatric Neurology specialists did not consider seizures. Status epilepticus was diagnosed when a seizure lasted more than 20 minutes or recurrent seizures happened without clinical recovery and normal mental status between them. Finally, measures of disease severity were captured: mortality, pediatric intensive care unit (PICU) admission, and PICU and hospital length of stay.

Comparison of continuous variables was carried out by using the Student t test or Mann-Whitney U test depending on the normality of distributions. Comparison of categorical variables was analyzed by using the χ2 test or Fisher exact test. Multivariate models were constructed to identify independent predictors of neurologic events, PICU admission, PICU and hospital length of stay, including all variables with previous P < 0.1 in univariate models and their interaction terms. Multivariate logistic regression was used for categorical dependent variables and multifactor ANOVA for continuous dependent variables. The goodness of fit of the final set of predictors was expressed as the R2 of Nagelkerke. Assumption of no multicollinearity was evaluated in all models. Two-sided tests were performed, and P < 0.05 was considered statistically significant Statistical analysis was performed using SPSS version 25.0 SPSS (IBM SPSS Statistics, Armonk, NY) and R statistical software, version 3.5.3: packages used included WRS2 and RMS.


From January 2015 to April 2018, 245 hospitalized patients were included with microbiologic confirmation of influenza disease, from a total number of 5514 hospitalizations (4.4%). The year which registered the highest number of cases was 2016 (94, 38.4%), followed by 2015 (67, 27.3%) and 2018 (49, 20.0%). In total, 70% patients were positive for influenza A and 30.8% for influenza B, with 2 patients showing positive results for both strains. Figure 1 shows detail of strain distribution per epidemic season, compared with national circulations in Spain during those same periods.23 Median age was 21 months [interquartile range (IQR), 6–57) and 57.5% were male. A total of 47.8% had a previous underlying condition. Baseline characteristics are summarized in Table 1. Clinical and analytic features at admission are shown in Table 2.

Table 1.
Table 1.:
Baseline Characteristics
Clinical and Analytical Features at Admission
Proportion of cases which teste positive for influenza A and B per epidemic period, compared with proportion of total cases in Spain over the same epidemic period.

Antiviral treatment with oseltamivir was administered to 86% of patients, with a median of 3 days of symptoms prior to starting treatment. Median hospitalization was 4 days (IQR, 3–6), with a PICU admission rate of 8.9%. Immediate causes for PICU admission where respiratory failure (47.6%), neurologic events (23.8%) and septic shock (14.3%). Of those admitted for a neurologic cause, 80% had mild respiratory disease, and 20% did not have any respiratory symptoms. The median length of PICU stay was 0.6 days (IQR, 0.3–1.5).

Twenty-nine patients (11.8%) developed neurologic events before or during hospitalization. The most frequent were febrile seizures, followed by nonfebrile seizures. Details of neurologic events and are shown in Table 3.

Neurologic Events

Of all seizures, 48% lasted more than 5 minutes, and convulsive status epilepticus occurred in 4 children. Recurrent seizures were observed in 16 patients: 69.6% of all patients who suffered one seizure had a second episode during hospitalization. One patient developed a febrile infection-related epilepsy syndrome.24 In total, 9 patients (39.1% of all suffering neurologic events) required antiepileptic drugs at discharge.

Cerebrospinal fluid (CSF) specimens were obtained from 19 (65%) of patients with neurologic symptoms. Cytochemical analysis results of 18 samples (94.7%) were normal and pleocytosis was noted in one; CSF culture was negative in all 19 samples. EEG recordings were taken in 15 patients with neurologic complications: 11 (73.3%) were normal and the remaining had abnormalities consisting on epileptiform activity. Neuroimaging was obtained in 9 patients: computed tomography scan was performed in 5 cases and brain magnetic resonance imaging in 9. All computed tomography scans were normal, one magnetic resonance imaging identified right hippocampal cortical dysplasia and another one a focal lesion in the right external capsule.

Clinical and epidemiologic factors were compared between the patients who did and did not develop neurologic complications of influenza infection. No individual factor was proved to be predictive of neurologic events. However, in a logistic regression model which included sex, age, influenza strain and previous underlying condition, a previous underlying condition had a greater risk of developing a neurologic complication [odds ratio (OR), 4.55; confidence interval (CI) 95%, 1.23–16.81) (Table 4).

Risk Factors for Neurologic Events and PICU Admission

The severity of infection was measured in terms of PICU admissions and PICU length of stay. Patients with a neurologic event had a higher risk of PICU admission (relative risk, 2.8; CI 95%, 1.1–7.0). Multivariate logistic regression models were also designed for this purpose: male sex (OR, 3.21; CI 95%, 1.22–8.33), influenza B virus (OR, 2.82; CI 95%, 1.14–7.14) and neurologic events (OR, 3.34; CI 95%, 1.10–10.19) were found to be risk factors for PICU admission. In a bifactorial ANOVA analysis, the PICU length of stay was influenced by influenza strain (P = 0.003), neurologic complications (P = 0.02) and interactions between them and underlying conditions (Pinteraction < 0.05).


Our study describes the epidemiology and risk factors associated with neurologic complications of influenza disease in a cohort of hospitalized pediatric patients and evaluates their impact on morbidity and disease severity.

Influenza-related neurologic symptoms are not as prevalent as respiratory manifestations in pediatric populations but have been increasingly reported and have been stated to be the third cause of hospitalization after respiratory and constitutional symptoms.25 These studies report a prevalence of neurologic complications in those hospitalizations of 7% to 11%.5,8,12 Our study found a similar but slightly higher prevalence (12%) which may be explained by the high rate of underlying chronic conditions in our cohort (49.8%), which are one of the risk factors associated with complications and which can be related to our hospital being reference for a wide range of these patients. Indeed, chronic conditions were one of the factors associated with neurologic complications in our multivariate models. This supports the recommendation of generalized influenza vaccination in this group of patients18 and highlights the need of a more active approach to vaccination in chronic patients, whose caregivers often lack information regarding such high risk.20 Education of families and healthcare providers is thus key to prevent and recognize severe infections.

Interestingly, almost a quarter of the seizures registered lasted more than 10 minutes, another quarter of patients had electroencephalographic abnormalities, and recurrent seizures during hospitalization were very common (69.6% of all seizures). Despite most of those patients having had fever, influenza-related seizures were thus clinically different from simple febrile seizures, which by definition last less than 10 minutes and are do not recur within 24 hours or within the same febrile illness.26 These findings support the idea of a different mechanism between febrile seizures and influenza-related seizures, even when both are preceded by fever. Febrile seizures are believed to be caused by a temperature-induced dynamic reduction of mutant surface gamma-aminobutyric acid receptors,27 whereas influenza-related seizures can be caused by infection of brain tissue which causes cellular damage and inflammatory cytokine cascades.10,11

Although in our study, there were not healthy subjects acting as a control group, it is of note that most CSF samples and imaging studies did not encounter any abnormal findings. Organic involvement in neurologic complications has widely been described15,24 but similarly to our study, the most frequent events are seizures4,12 which even though may recur and increase severity of disease, do not normally lead to organic central nervous system damage.

According to our results, neurologic events seem to be linked to influenza disease severity. We found that the rate of PICU admission was higher in patients who had developed a neurologic event, along with those of male sex and those who had an influenza B strain. Although chronic conditions were not a significant factor in our PICU admission model, they did influence PICU length of stay in the bifactorial ANOVA model. Those findings are in line with previous publications: a US population study found that most influenza-related deaths occurred in patients with a preexisting chronic condition28 and that the most common was a neurologic disorder. Other studies have suggested that PICU admissions and mortality are related to neurologic events, especially in patients with prior neurologic disease.12

We did not find neurotropic differences between influenza A and B, although B infections did have a higher risk of PICU admission. Most previous studies have suggested that influenza A is the main neurotropic strain,4,8,9,14,15 and others are in line with our findings, showing no differences between A and B.12,16 However, neurologic complications in influenza B, such as encephalitis, have also been described29–31 and their prevalence might be underestimated.32 Our higher rate of PICU admission in influenza B infections does not seem to be related to neurologic complications but to slightly more severe general or respiratory symptoms. Although this was not the aim of the study and should be carefully interpreted, other studies have pointed out similar results, with influenza B being responsible for up to 33%–38% of influenza-related pediatric deaths in the United States when accounting for only 26% of total circulating virus.28,33

Our study has some limitations. As any retrospective study, conclusions should be carefully interpreted as selection, and information bias may have been present. Our hospital is a tertiary-care center with a high prevalence of chronic patients, which may also have affected our results and overestimated morbidity. We did not have a control group with healthy subjects. Thus, more studies are needed, especially prospective and controlled, to support recommendations and highlight the significance and consequences of those neurologic events.

In conclusion, we have found that a significant proportion of influenza-related hospitalized patients can suffer neurologic complications that seizures are the most common of them and that patients with a first seizure may be at higher risk of suffering recurrent ore prolonged seizures during hospitalization. We also found that the severity of complications is variable and that diagnostic tests unreliably distinguish severity and long-term outcomes. Previous underlying conditions pose the greatest risk to a neurologic event and that those neurologic events increase disease severity and seem to play an important role in morbidity associated to the infection.


We hereby acknowledge that the material is original research, the manuscript has been reviewed and approved by each author, and that no portion of the manuscript has been published or is under consideration for publication elsewhere.


1. Sellers SA, Hagan RS, Hayden FG, et al. The hidden burden of influenza: a review of the extra-pulmonary complications of influenza infection. Influenza Other Respir Viruses. 2017;11:372–393.
2. Lafond KE, Nair H, Rasooly MH, et al. Global role and burden of influenza in pediatric respiratory hospitalizations, 1982-2012: a systematic analysis. PLoS Med. 2016;13:e1001977
3. Zhou H, Thompson WW, Viboud CG, et al. Hospitalizations associated with influenza and respiratory syncytial virus in the United States, 1993-2008. Clin Infect Dis. 2012;54:1427–1436.
4. Cárdenas G, Soto-Hernández JL, Díaz-Alba A, et al. Neurological events related to influenza A (H1N1) pdm09. Influenza Other Respir Viruses. 2014;8:339–346.
5. Britton PN, Blyth CC, Macartney K, et al.; Australian Childhood Encephalitis (ACE) Study Investigators, Influenza Complications Alert Network (FluCAN) Investigators, and Paediatric Active Enhanced Disease Surveillance (PAEDS) Network. The spectrum and burden of influenza-associated neurological disease in children: combined encephalitis and influenza sentinel site surveillance from Australia, 2013-2015. Clin Infect Dis. 2017;65:653–660.
6. Davis LE, Koster F, Cawthon A. Neurologic aspects of influenza viruses. Handb Clin Neurol. 2014;123:619–645.
7. Neurological complications of influenza. Br Med J. 1970;1:248–249
8. Ekstrand JJ, Herbener A, Rawlings J, et al. Heightened neurologic complications in children with pandemic H1N1 influenza. Ann Neurol. 2010;68:762–766.
9. Okumura A, Tsuji T, Kubota T, et al. Acute encephalopathy with 2009 pandemic flu: comparison with seasonal flu. Brain Dev. 2012;34:13–19.
10. Kuiken T, Taubenberger JK. Pathology of human influenza revisited. Vaccine. 2008;26(suppl 4):D59–D66.
11. Ng YP, Lee SM, Cheung TK, et al. Avian influenza H5N1 virus induces cytopathy and proinflammatory cytokine responses in human astrocytic and neuronal cell lines. Neuroscience. 2010;168:613–623.
12. Newland JG, Laurich VM, Rosenquist AW, et al. Neurologic complications in children hospitalized with influenza: characteristics, incidence, and risk factors. J Pediatr. 2007;150:306–310.
13. Kondrich J, Rosenthal M. Influenza in children. Curr Opin Pediatr. 2017;29:297–302.
14. Chiu SS, Tse CY, Lau YL, et al. Influenza A infection is an important cause of febrile seizures. Pediatrics. 2001;108:E63.
15. Tsai JP, Baker AJ. Influenza-associated neurological complications. Neurocrit Care. 2013;18:118–130.
16. Tran D, Vaudry W, Moore D, et al. Hospitalization for Influenza A Versus B. Pediatrics. 2016;138: e20154643
17. Mistry RD, Fischer JB, Prasad PA, et al. Severe complications in influenza-like illnesses. Pediatrics. 2014;134:e684–e690.
18. AAP Committee on Infectious Diseases. Recommendations for prevention and control of influenza in children, 2018-2019. Pediatrics. 2018;142:e20182367
19. Yang L, Peng J, Deng J, et al. Vaccination status of children with epilepsy or cerebral palsy in Hunan rural area and a relative KAP survey of vaccinators. Front Pediatr. 2019;7:84.
20. Smith M, Peacock G, Uyeki TM, et al. Influenza vaccination in children with neurologic or neurodevelopmental disorders. Vaccine. 2015;33:2322–2327.
21. Merckx J, Wali R, Schiller I, et al. Diagnostic accuracy of novel and traditional rapid tests for influenza infection compared with reverse transcriptase polymerase chain reaction: a systematic review and meta-analysis. Ann Intern Med. 2017;167:394–409.
22. Venkatesan A, Tunkel AR, Bloch KC, et al.; International Encephalitis Consortium. Case definitions, diagnostic algorithms, and priorities in encephalitis: consensus statement of the international encephalitis consortium. Clin Infect Dis. 2013;57:1114–1128.
23. Instituto de Salud Carlos III. Informe de Vigilancia de la Gripe en España [internet]. Temporada 2018-19. Sistema de Vigilancia de la Gripe en España.Available at: Accessed January 2, 2020.
24. Lee HF, Chi CS. Febrile infection-related epilepsy syndrome (FIRES): therapeutic complications, long-term neurological and neuroimaging follow-up. Seizure. 2018;56:53–59.
25. Arístegui Fernández J, González Pérez-Yarza E, Mellado Peña MJ, et al. Child hospital admissions associated with influenza virus infection in 6 Spanish cities (2014-2016). An Pediatr (Barc). 2019;90:86–93
26. Patel AD, Vidaurre J. Complex febrile seizures: a practical guide to evaluation and treatment. J Child Neurol. 2013;28:762–767.
27. Kang JQ, Shen W, Macdonald RL. Why does fever trigger febrile seizures? GABAA receptor gamma2 subunit mutations associated with idiopathic generalized epilepsies have temperature-dependent trafficking deficiencies. J Neurosci. 2006;26:2590–2597.
28. Shang M, Blanton L, Brammer L, et al. Influenza-Associated pediatric deaths in the United States, 2010-2016. Pediatrics. 2018;141: e20172918.
29. Popescu CP, Florescu SA, Lupulescu E, et al. Neurologic complications of influenza B virus infection in adults, Romania. Emerg Infect Dis. 2017;23:574–581.
30. Piet E, Tattevin P, Mailles A, et al. Influenza B meningoencephalitis. Med Mal Infect. 2017;47:435–436.
31. Straumanis JP, Tapia MD, King JC. Influenza B infection associated with encephalitis: treatment with oseltamivir. Pediatr Infect Dis J. 2002;21:173–175.
32. Lin CH, Huang YC, Chiu CH, et al. Neurologic manifestations in children with influenza B virus infection. Pediatr Infect Dis J. 2006;25:1081–1083.
33. Centers for Disease Control and Prevention (CDC). Influenza-associated pediatric deaths--United States, September 2010-August 2011. MMWR Morb Mortal Wkly Rep. 2011;60:1233–1238.

influenza; seizures; encephalopathy

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.