This neurological sub-study of the CCCC Study was designed with the aim to obtain an overview of neurological complications in a large international multicenter cohort of critically ill COVID-19 patients, including recruitment from LMICs as well as HICs. COVID-19 may manifest critical involvement of multiple organ systems in severe disease stage, and neurological complications represent a potentially devastating complication of coronavirus disease 1. In this analysis, a > 12% prevalence of new neurological complications was observed which included ischemic stroke (2.9%), ICH (2.8%), and seizure (2.6%). Myopathy was the most common complication involving the peripheral nervous system.
COVID-19 infection has been shown to have a higher rate of ischemic stroke when compared to other viral infections (influenza A/B), even when adjusting for ICU admission 24. Several recent reports discuss the possible pathophysiology of the innate immune response to COVID-19 infection leading to thromboembolic and in-situ microthromboses. The activation of the inflammasome, as well as disruption of the angiotensin-renin system, can lead to activation of the complement system—resulting in endothelial damage and microthromboses 25–27. Studies have demonstrated that viral infections can lead to seizures, either through direct neural injury, or by decreasing the seizure threshold in a predisposed host 28. Therefore, local neuroinvasion can lead to seizures as well as long-term epilepsy 29. It has been demonstrated that severe or fatal influenza infection can cause seizures, but at a rate of 2.1 percent, slightly lower than this cohort of COVID-19 infection related seizures 30. In our study, the need for ECMO was observed to be an independent risk factor for ICH. This data is in accordance with a recent meta-analysis that revealed COVID-19 patients on ECMO support who developed neurological complications had worse outcomes and higher mortality 31. However, it is possible that ECMO cannulation itself can lead to worse neurological injury 13.
Multiple observational studies have been conducted to evaluate acute respiratory distress syndrome (ARDS) incidence and outcomes. One such study, The Large Observational Study to Understand the Global Impact of Severe Acute Respiratory Failure- LUNG SAFE trial, a multinational trial that included 29,144 patients—of which around 10% developed ARDS 32,33. In that cohort, 23% of ARDS patients required mechanical ventilator support, and it was shown that median length of hospital stay was 17 days, and that of ICU stay was 10 days. Mortality ranged from 35–40%, with higher mortality rates in more severe cases. 32,33. In our cohort, all patients were admitted to a critical care unit, more than 50% of whom required respiratory support (MV and/or ECMO). Additionally, patients who developed neurological complications had a median length ICU stay of 18 days in comparison to 13 days for those who did not develop a neurological complication. Furthermore, our study showed that patients who developed a neurological complication had a mortality of 49.2% in comparison to the 34.8% with no reported neurological complication. Mortality rates depended on the type of neurological complication, with 77.8% of hypoxic ischemic brain injury patients dying in hospital compared with 10% with myopathy. Taken together, this cohort data is in accordance with previous observational studies reflecting ARDS incidence and outcomes with the added benefit of seeing how the development of neurological complications impacts these factors.
Epidemiological studies support the idea that LMIC accrue a bulk of the burden of non-communicable neurological deaths and disability-adjusted life-years (DALYs) 34. This is in accordance with this cohort, which demonstrated the increased incidence of complications per admitted days such as seizures and ICH. More specifically, LMIC have been observed to experience higher COVID-19 case-infection rates compared to HIC, whilst also having limitations of health infrastructure such as decreased number of intensive care beds, hospital beds, as well as availability of ventilators 35. These factors might be an underlying substrate for more advanced and thus severe stage of infection, which is exacerbated by the fact that these countries have higher rates of non-communicable disease comorbidities that include HTN and diabetes 35,36. Taken together, more severe infections and higher prevalence of comorbid conditions associated with higher rates of stroke in LMIC may be the substrate for the higher observed rates of neurological complications. In this study, however, higher complication rates in LMIC were associated with the number of admitted days and not per admission. This suggests that other factors, such as shorter length of stay, may explain the differences that we observed. More research, support, and resources are necessary to understand the reasons for these differences in order to diminish the disproportionate burden that LMIC countries face in response to severe COVID-19 infection and associated neurological complications.
It has also been demonstrated that patients admitted to the emergency department with previous comorbid neurological conditions were more likely to have a more severe form of COVID-19 infection 37. Additionally, hospitalized patients with COVID-19 were more likely to develop neurological conditions, such as encephalopathy, if they were older and had a previous comorbid neurological condition 38. Thus, the development of neurological complications related to COVID-19 infection might be facilitated by an already compromised blood brain barrier, neuronal dysfunction, or preexisting cerebral atrophy—though more studies are needed to identify causality. These factors are exacerbated by the fact that during the pandemic, the accessibility and availability of CT scan for diagnosis was scarce 39. Therefore, patients with neurological comorbidities may benefit from enhanced surveillance of new neurological complications. In particular, low-cost neurodiagnostic tools in resource limited settings are essential for patients from LMICs.
A unique aspect of the CCCC Study is the prospective and international nature of the study—since it allows for an analysis from a diverse patient population spanning many countries (including LMIC), hospital networks, and patient demographics. Additionally, we collected details that allowed for an in-depth examination of the many factors contributing to COVID-19 outcomes and the development of complications. However, this study has some limitations that are important to highlight. For one, two large centers (HIC Kuwait and LMIC Indonesia) comprise more than 1/3rd of the patients in the sub-study. The practices of these two large centers, such as the threshold by which patients are discharged/transferred, may influence the overall results. Additionally, it is likely that treatment practices varied widely between centers and between countries—most notably when comparing HICs to LMICs. These differences are not accounted for in our cohort and might skew results. Finally, some centers had missing data—including those for neurological complications. This may affect data generalizability and interpretation.