New‐onset and persistent neurological and psychiatric sequelae of COVID‐19 compared to influenza: A retrospective cohort study in a large New York City healthcare network

Abstract Objectives Neurological and neuropsychiatric manifestations of post‐acute SARS‐CoV‐2 infection (neuro‐PASC) are common among COVID‐19 survivors, but it is unknown how neuro‐PASC differs from influenza‐related neuro‐sequelae. This study investigated the clinical characteristics of COVID‐19 patients with and without new‐onset neuro‐PASC, and of flu patients with similar symptoms. Methods We retrospectively screened 18,811 COVID‐19 patients and 5772 flu patients between January 2020 and June 2021 for the presence of new‐onset neuro‐sequelae that persisted at least 2 weeks past the date of COVID‐19 or flu diagnosis. Results We observed 388 COVID‐19 patients with neuro‐PASC versus 149 flu patients with neuro‐sequelae. Common neuro‐PASC symptoms were anxiety (30%), depression (27%), dizziness (22%), altered mental status (17%), chronic headaches (17%), and nausea (11%). The average time to neuro‐PASC onset was 138 days, with hospitalized patients reporting earlier onset than non‐hospitalized patients. Neuro‐PASC was associated with female sex and older age (p < 0.05), but not race, ethnicity, most comorbidities, or COVID‐19 disease severity (p > 0.05). Compared to flu patients, COVID‐19 patients were older, exhibited higher incidence of altered mental status, developed symptoms more quickly, and were prescribed psychiatric drugs more often (p < 0.05). Conclusions This study provides additional insights into neuro‐PASC risk factors and differentiates between post‐COVID‐19 and post‐flu neuro‐sequelae.

These symptoms, often referred to as neurological post-acute sequelae of SARS-CoV-2 infection (neuro-PASC), include, but are not limited to, altered mental status (AMS), altered taste and smell, anxiety, ataxia, delirium, depression, dizziness, fatigue, headaches, memory loss, nausea, new-onset post-traumatic stress disorder, seizures, strokes, and tinnitus (Collantes et al., 2021;Graham et al., 2021;Moghimi et al., 2021;Wang et al., 2020). Neuro-PASC may be caused by an immune response to initial infection, by direct viral infection of the central nervous system, or by psychological stressors such as social isolation, fear of illness, stigma, and future uncertainty , Mazza et al., 2020. Neuro-PASC symptoms affect individuals of all age groups, including young children (Ludvigsson, 2021). Emerging data suggests that individuals with mild symptoms from SARS-CoV-2 infection (i.e., not requiring hospitalization) are also susceptible (Townsend et al., 2020;van den Borst et al., 2021).
Neuro-PASC is incompletely understood, and prior studies have largely relied on subjective survey reports, often without detailed clinical data. Moreover, it is unknown whether neuro-PASC symptoms differ from neurological and neuropsychiatric symptoms associated with other respiratory diseases such as influenza, and to what extent these neuro-PASC symptoms are new-onset and persistent in nature.

| Aims of the study
The aim of this study was to characterize neuro-PASC in  patients without prior history of neurological or neuropsychiatric symptoms using detailed real-world data pre-and post-SARS-CoV-2 infection, and to compare with both COVID-19 patients without neuro-PASC and influenza patients with neurological or neuropsychiatric sequelae from the same catchment area and study timeframe. We evaluated clinical and demographic patient variables, neuro-PASC symptoms, timeframes of symptom onset, and patient drug prescriptions associated with neuro-PASC.

| MATERIAL AND METHODS
This study was approved by the Einstein-Montefiore Institutional Review Board with an exemption for informed consent and a HIPAA waiver due to the retrospective and deidentified nature of the data.

| Data source
All data originated from observational databases in the Montefiore Health System as described in previous reports Hoogenboom, Pham, et al., 2021;Lu, 2021;Lu et al., 2022). Briefly, patient data between January 1, 2020, and June 9, 2021, were standardized to the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) version 6, which uses standard vocabulary concepts, allowing for the systematic analysis of disparate observational databases such as electronic medical records and disease classification systems (e.g., ICD-10, SNOWMED, LOINC) (Hripcsak et al., 2015). Cohort building and searches of vocabulary concepts were performed using ATLAS, a web-based tool developed by the Observational Health Data Sciences and Informatics (OHDSI) community that enables navigation of patient-level, observational data in the CDM format. OMOP concept IDs used to define neuro-PASC symptoms are presented in Table S1.
Data was subsequently imported into an SQLite database (www. sqlite.org) and queried using the DB Browser (version 3.12.2).

| Study population and inclusion criteria
All individuals included in this report were patients of the Montefiore Health System, one of the largest healthcare systems in New York City with more than 180 primary and specialty care locations in the Bronx environs serving a large racially and ethnically diverse population (Hoogenboom, Pham, et al., 2021;Wadhera et al., 2020). Three groups of patients were studied: (1) COVID-19 patients with neuro-PASC symptoms, (2) COVID-19 patients without neuro-PASC symptoms, and (3) Flu patients with neurological and neuropsychiatric symptoms.
For the COVID-19 cohort, all individuals who tested positive for SARS-CoV-2 infection (laboratory confirmed by real-time PCR test) were included. Patients were subsequently screened for the presence of new-onset neurological and neuropsychiatric symptoms that persisted at least 2 weeks past the date of COVID-19 diagnosis, the standard period during which SARS-CoV-2 has been found to replicate (Lamontagne et al., 2021), to exclude transient, hyperacute effects.
Patients were included in the flu cohort if diagnosed with influenza (by probe detection of Influenza Virus A or B RNA) in the same hospital system and within the same timeframe, but without a positive COVID-19 diagnosis.

| Demographics and clinical variables
Collected data included demographics, comorbidities, and vital signs.
Demographic data included age, sex, race, and ethnicity. Race/ethnicity were based on patient self-identification and categorized as American Indian, Asian, Black, Pacific Islander, White, or Other (comprising patients selecting multiple groups or some other race). Chronic comorbidities included hypertension, diabetes, congestive heart failure (CHF), chronic kidney disease (CKD), coronary artery disease (CAD), chronic obstructive pulmonary disease (COPD), and asthma. COVID-19 disease severity was defined by hospitalization status (i.e., outpatient vs. inpatient general floor or ICU), which is indicative of need for escalated care. Vitals and laboratory values were collected at admission and included body mass index (BMI), pulse oximetry, oral temperature, lactate dehydrogenase (LDH), C-reactive protein (CRP), D-dimer, blood urea nitrogen (BUN), lymphocytes, and leukocyte measurements.

| Statistical analysis
Statistical analyses were performed using R programming language (version 4.0.2), RStudio (version 1.3.1056), Microsoft Excel for Mac (version 16.52), and Excel add-in program Analysis ToolPak. We used descriptive statistics to report patient demographic characteristics, including mean � standard deviation (e.g., age, BMI, pulse oximetry).

| Neuro-PASC symptoms
The most common neuro-PASC symptoms in the COVID-19 cohort were anxiety disorders (30% of neuro-PASC patients), followed by F I G U R E 1 Inclusion/exclusion criteria for neuro-PASC and control cohorts IOSIFESCU ET AL.
-3 of 12 T A B L E 1 Demographics, comorbidities, and laboratory data at admission of COVID-19 patients with and without neuro-PASC, and flu patients with neurological and neuropsychiatric symptoms  Flu patients displayed significantly more delayed symptom onset, averaging 238 days (p < 0.05; Figure 3c).

| DISCUSSION
This study characterized new-onset neurological and neuropsychiatric sequelae in COVID-19 patients at least 2 weeks after COVID-19 diagnosis using real-world EHR data from a diverse population in the Bronx. Comparisons were made with COVID-19 patients without neuro-PASC as well as flu patients with neurological and neuropsychiatric sequelae from the same period and catchment area. The most common neuro-PASC symptoms identified were anxiety disorders, depression, dizziness, fatigue, altered mental status, chronic headaches, and nausea, in line with those reported previously (Carfì et al., 2020;Garrigues et al., 2020;Mao et al., 2020;Townsend et al., 2020). However, most prior studies were surveys and did not include detailed EHR data (Davis et al., 2021;Garrigues et al., 2020;Moghimi et al., 2021). Our analysis also excluded patients with preexisting neurological and neuropsychiatric symptoms and considered only new-onset cases, whereas most previous studies have included individuals with such pre-existing conditions (Carfì et al., 2020;Lamontagne et al., 2021). We further excluded hyperacute neuropsychiatric symptoms within the first 2 weeks after COVID-19 diagnosis, whereas some prior investigations did not Tenforde et al., 2020).
Onset of neuro-PASC symptoms was earlier in COVID-19 patients who required hospitalization. It is possible that severe COVID-19 triggers sudden symptom onset, while mild COVID-19 is associated with a more gradual neuro-PASC symptom development. This finding suggests that screening for neuro-PASC should start as early as possible, especially among those previously hospitalized for COVID-19.
Female sex was associated with higher incidence of neuro-PASC, even though women have lower risk of severe COVID-19 outcomes than men . Sex differences in psychiatric symptom A major finding is that presence of neuro-PASC symptoms was not associated with COVID-19 severity, when the latter was measured by either hospitalization status, pulse oximetry values (indicating need for supplemental oxygen), oral temperature (indicating fever), or comorbidities linked with worse COVID-19 outcomes. This result is further supported by several smaller studies (Henry et al., 2020;Lin et al., 2021;Mazza et al., 2020;Sahu et al., 2020) and by laboratory measures observed herein (in particular LDH, CRP, and D-dimer), which were less extreme for

F I G U R E 3 (Continued)
IOSIFESCU ET AL.
Taken together, these findings suggest that development of neuro-PASC symptoms is not limited to hospitalized patients with severe COVID-19 and supports prior results of multiple smaller investigations (Townsend et al., 2020;van den Borst et al., 2021;Garrigues et al., 2020). An important clinical implication is that screening for neuro-PASC should be done for all individuals post SARS-CoV-2 infection, regardless of hospitalization.
Increased usage of neuropsychiatric drugs after discharge was associated with inpatient status/ICU hospitalization, female sex, and older age. The latter two associations are expected, as older and female patients also developed neuro-PASC symptoms more frequently. Nevertheless, given the lack of association between COVID-19 disease severity and the incidence of neuro-PASC, the relationship between hospitalization and prescribed drugs warrants further investigation.
We compared neuropsychiatric sequelae of COVID-19 patients with those of flu patients over the same period and in the same catchment area, enabled by the use of EHR. COVID-19 patients with neuro-PASC were markedly older, a difference that can be attributed to how COVID-19 differentially affects the older population.
Compared to flu patients, COVID-19 patients with relevant sequelae had lower incidence of COPD/asthma, suggesting that COPD/asthma is a major risk factor for flu but less so for COVID-19. Other comorbidities did not contribute to differences in incidence of new neuropsychiatric symptoms.
Neuropsychiatric symptoms in both cohorts were similar, except that COVID-19 patients experienced higher incidence of altered mental status and trending higher incidence of fatigue.
Consequently, future studies should include quantitative measures of fatigue scores.
Drugs used to treat neuropsychiatric conditions were prescribed significantly more to COVID-19 patients than to flu patients following their respective diagnoses, potentially indicating that

F I G U R E 3 (Continued)
neuro-PASC is considerably more severe than sequelae of other respiratory conditions (Taquet et al., 2021 In conclusion, this study provides additional insights on neuro-PASC associations with key clinical variables, which may be used to support at-risk patients, develop effective interventions and screening methods, and understand the potential for future neurological and psychiatric sequelae related to COVID-19.

CONFLICT OF INTEREST
None.

AUTHOR CONTRIBUTION
Andrei L. Iosifescu

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.

F I G U R E 4 (Continued)
10 of 12 -IOSIFESCU ET AL.