Long-term Mental Health Morbidity Among Survivors of COVID-19 Critical Illness – A Population-Based Cohort Study

AUTHORS: Shannon M. Fernando, Danial Qureshi, Robert Talarico, Eddy Fan, Daniel I. McIsaac, Simone N. Vigod, Manish M. Sood, Daniel T. Myran, Carol L. Hodgson, Bram Rochwerg, Laveena Munshi, Kirsten M. Fiest, O. Joseph Bienvenu, Dale M. Needham, Daniel Brodie, Niall D. Ferguson, Robert A. Fowler, Deborah J. Cook, Arthur S. Slutsky, Damon C. Scales, Margaret S. Herridge, Peter Tanuseputro, Kwadwo Kyeremanteng BACKGROUND Severe Coronavirus 2019 (COVID-19) is a common cause of critical illness and intensive care unit (ICU) admission, most often due to hypoxemic respiratory failure. Incidence of ICU admission and hospital mortality among COVID-19 patients has varied geographically and across the duration of the pandemic (Tzotzos et al., Crit Care, 2020). The initial focus of research in critical care focused upon treatments for critically ill COVID-19 patients, and adequate use of resources (Alhazzani et al., Intensive Care Med, 2020; Weissman et al., Ann Intern Med, 2020). Since then, there has been great emphasis upon understanding survivorship after COVID-19 critical illness, and the long-term outcomes among COVID-19 ICU survivors (Hosey and Needham, Nat Rev Dis Primers, 2020). Survivors of critical illness are known to have substantial physical morbidity (Herridge et al., N Engl J Med, 2003; Herridge et al., N Engl J Med, 2011), and existing data also shows that these patients are at increased risk of downstream psychiatric morbidity (Wunsch et al., JAMA, 2014; Sivanathan et al., Intensive Care Med, 2019; Olafson et al., Intensive Care Med, 2021), including suicide and self-harm (Fernando et al., BMJ, 2021). Understanding long-term outcomes among survivors of COVID-19 remains an important avenue for future research (Marshall et al., Lancet Infect Dis, 2020). Survivors of COVID-19 critical illness may have experienced invasive critical care interventions, such as invasive mechanical ventilation and extracorporeal life support (Wunsch, Am J Respir Crit Care Med, 2020; Barbaro et al., Lancet, 2020), treatments that have been associated with higher physical and mental health morbidity (Wunsch et al., JAMA, 2014; Hodgson et al., Lancet Respir Med, 2022; Fernando et al., JAMA, 2022). While existing data suggest that survivors of COVID-19 critical illness may experience substantial physical morbidity (Heesakkers et al., JAMA, 2022), data on mental health morbidity is less clear (Sankar et al., Chest, 2022). Furthermore, how mental health morbidity among survivors of COVID-19 critical illness compares to other survivors of critical illness is unknown, and is an important question as the physical morbidity experienced by these patient populations does not appear to differ (Hodgson et al., Am J Respir Crit Care Med, 2022). We seek to investigate the incidence of long-term mental health morbidity in survivors of COVID-19 critical illness, using population-based data from the province of Ontario, and compare this incidence to other survivors of critical illness, as well as non-ICU hospitalized patients with COVID-19. STUDY OBJECTIVES 1. To describe the sociodemographic characteristics of survivors of COVID-19 critical illness, and to examine the incidence of new mental health diagnoses, and self-harm. 2. To examine if COVID-19 critical illness is associated with a higher incidence of mental health diagnoses among survivors, as compared to ICU survivors without COVID-19; 3. To examine the proportion of COVID-19 critical illness survivors using inpatient (i.e., requiring hospitalization for mental health diagnoses) and outpatient (e.g., visiting a psychiatrist as an outpatient, number of visits within 1-year, time to first visit, and costs related to outpatient resource use) mental health services; 4. To investigate the prognostic factors associated with incident downstream mental health diagnoses among survivors of COVID-19 critical illness. METHODS Data Sources and Setting This will be a population-level cohort study using health administrative databases from the province of Ontario in Canada (population 14.6 million). Within Ontario’s single payer healthcare system, all publicly funded healthcare services, physician, hospital, and demographic information for residents are recorded in administrative databases. These datasets are linked using unique encoded identifiers, and analyzed at ICES, an independent, non-profit research institute whose legal status under Ontario’s health information privacy law allows it to collect and analyze healthcare and demographic data, without consent, for health system evaluation and improvement. ICES is funded by an annual grant from the Ontario Ministry of Health and the Ministry of Long-term Care. Patients are linked across provincial databases using their Ontario Health Insurance Plan (OHIP) number, which is unique to each citizen in Ontario. We will link ten databases at ICES, as performed previously (Fernando et al., BMJ, 2021; Fernando et al., JAMA, 2022) at the individual patient level, from January 1, 2020, through March 31, 2022. Data on illness severity (Multiple Organ Dysfunction Score [MODS]) and co-interventions will be obtained from the Critical Care Information System (CCIS). CCIS provides near-real time information on every patient admitted to a level 2 (designation for those requiring increased observation, those “stepping down” from higher levels of care, or those requiring monitoring and support for an organ system) or level 3 (designation for those requiring advanced respiratory support alone, or monitoring and support for two or more organ systems) critical care unit in Ontario’s acute care hospitals. The system captures data on bed availability, critical care service utilization and patient outcomes. This provides consistent and reliable information on the utilization of critical care resources across the province. The system provides an important medium for monitoring and managing the province’s critical care resources more effectively, and for highlighting opportunities to implement quality improvement initiatives at individual hospitals and across Local Health Integration Networks (LHINs). Data contained in ICES are full and complete, with the exception of emigration from Ontario, which represents approximately 0.5% of patients per year. Patients The entire study period (including outcome ascertainment) will be from January 1, 2020 to September 30, 2022. We will include consecutive adult patients (≥ 18 years of age), with an index intensive care unit (ICU) discharge in Ontario from January 1, 2020, through March 30, 2022, with a diagnosis of COVID-19, and who survived to hospital discharge. For patients with multiple ICU admissions, we will randomly select one admission per patient during the accrual period. We will identify ICU admission through the use of previously validated algorithms from the Canadian Institute for Health Information Discharge Abstract Database (Scales et al., J Clin Epidemiol, 2006). Since routine SARS-CoV-2 testing has been done in Ontario hospitals during the study period, patients with COVID-19 will be identified by a positive polymerase chain reaction (PCR) test for SARS-CoV-2, linked within 14 days to an index admission where the most responsible diagnosis is COVID-19 (using International Classification of Diseases, Version 10 [ICD-10] codes U071 and U072), as performed previously (McNaughton et al., CMAJ, 2022). We will not exclusively rely on PCR testing, due to concerns surrounding incidental positive cases that may be found during ICU admission, particularly in 2022 with the Omicron variant. The primary control group will be adult patients admitted to an ICU with pulmonary infection, but without a positive PCR test for SARS-CoV-2 linked to the index admission, and surviving to hospital discharge. Pulmonary infection will be identified using validated ICD-10 coding for either pneumonia or influenza (J09-J18; Skull et al., Epidemiol Infect, 2010). We will also include additional control groups: 1) Adult patients admitted to the ICU during the study period for any cause, without a positive PCR test for SARS-CoV-2 linked to the index admission, and surviving to hospital discharge; 2) Adult patients admitted to hospital with a positive PCR test for SARS-CoV-2 linked to the index admission, but without ICU admission, and surviving to hospital discharge; and 3) Adult patients with an outpatient positive PCR test for SARS-CoV-2, and not requiring hospital admission within 30 days prior to or after the date of the positive test. We will identify important patient characteristics at the time of the index admission, including age, sex, Charlson comorbidity index (CCI), date of admission, and the number of hospital admissions in the previous year. We will calculate duration of ICU and hospital length of stay from admission and discharge dates. We will obtain neighbourhood income (categorized into quintiles), rurality, and area-level measures of essential worker and visible minority volume through postal code conversion files based on Statistics Canada census data. Recent immigrant status will be captured from the Immigration, Refugees and Citizenship Canada (IRCC) database, which includes all immigration records for people landing in Ontario from 1985 onwards. We will also capture history of pre-existing mental health diagnoses that occurred in the 5 years prior to the index admission, through the use of ICD-10 codes and whether patients had any outpatient mental health visits with a primary care provider or psychiatrist in the previous year (Fernando et al., JAMA, 2022). We will also capture the Charlson Comorbidity Index in the 5 years prior to index admission. Finally, we will record life support interventions received during hospital admission, including invasive mechanical ventilation (delivered through an endotracheal or tracheostomy tube), non-invasive mechanical ventilation (by facemask), renal replacement therapy, trac

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.