Received: 17 September 2020 | Revised: 20 November 2020 | Accepted: 5 December 2020 DOI: 10.1002/da.23129 RESEARCH ARTICLE Thirty‐day suicidal thoughts and behaviors among hospital workers during the first wave of the Spain COVID‐19 outbreak Philippe Mortier1,2 | Gemma Vilagut1,2 | Montse Ferrer1,2,3 | Consol Serra2,4,5 | Juan D. Molina6,7,8,9 | Nieves López‐Fresneña10 | Teresa Puig3,11,12,13 | José M. Pelayo‐Terán14 | José I. Pijoan2,15 | José I. Emparanza2,16 | Meritxell Espuga17 | Nieves Plana2,18 | Ana González‐Pinto9,19 | Rafael M. Ortí‐Lucas9,20 | Alma M. de Salázar21 | Cristina Rius9,22 | Enric Aragonès23,24 | Isabel del Cura‐González25,26,27 | Andrés Aragón‐Peña27,28 | Mireia Campos29 | Mara Parellada9,10 | Aurora Pérez‐Zapata18 | Maria João Forjaz30,31 | Ferran Sanz32,33,34 | Josep M. Haro3,9,35 | Eduard Vieta9,36 | Víctor Pérez‐Solà3,4,9 | Ronald C. Kessler37 | Ronny Bruffaerts38 | Jordi Alonso1,2,33 | the MINDCOVID Working Group 1 Health Services Research Unit, IMIM‐Institut Hospital del Mar d'Investigacions Mèdiques, Barcelona, Spain 2 CIBER Epidemiología y Salud Pública (CIBERESP), Madrid, Spain 3 Universitat Autònoma de Barcelona (UAB), Barcelona, Spain 4 Parc de Salut Mar PSMAR, Barcelona, Spain 5 CiSAL‐Centro de Investigación en Salud Laboral, IMIM/UPF, Barcelona, Spain 6 Villaverde Mental Health Center, Clinical Management Area of Psychiatry and Mental Health, Psychiatric Service, Hospital Universitario 12 de Octubre, Madrid, Spain 7 Research Institute Hospital 12 de Octubre (i+12), Madrid, Spain 8 Faculty of Health Sciences, Francisco de Vitoria University, Madrid, Spain 9 CIBER Salud Mental (CIBERSAM), Madrid, Spain 10 Hospital General Universitario Gregorio Marañón, Madrid, Spain 11 Department of Epidemiology and Public Health, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain 12 Biomedical Research Institute Sant Pau (IIB Sant Pau), Barcelona, Spain 13 CIBER Enfermedades Cardiovasculares (CIBERCV), Madrid, Spain 14 Hospital El Bierzo, León, Spain 15 Hospital Universitario Cruces/OSI EEC, Bilbao, Spain 16 Hospital Universitario Donostia, San Sebastián, Spain 17 Occupational Health Service, Hospital Universitari Vall d'Hebron, Barcelona, Spain 18 Príncipe de Asturias University Hospital, Alcalá de Henares, Madrid, Spain 19 Hospital Universitario Araba‐Santiago, Vitoria‐Gasteiz, Spain 20 Hospital Clínic Universitari, Valencia, Spain 21 UGC Salud Mental, Hospital Universitario Torrecárdenas, Almería, Spain This is an open access article under the terms of the Creative Commons Attribution‐NonCommercial‐NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non‐commercial and no modifications or adaptations are made. © 2020 The Authors. Depression and Anxiety Published by Wiley Periodicals LLC 528 | wileyonlinelibrary.com/journal/da Depression Anxiety. 2021;38:528–544. MORTIER ET AL. | 529 22 Agència de Salut Pública de Barcelona, Barcelona, Spain 23 Institut d'Investigació en Atenció Primària IDIAP Jordi Gol, Barcelona, Spain 24 Atenció Primària Camp de Tarragona, Institut Català de la Salut, Tarragona, Spain 25 Research Unit, Primary Care Management, Madrid Health Service, Madrid, Spain 26 Department of Medical Specialities and Public Health, King Juan Carlos University, Madrid, Spain 27 Fundación Investigación e Innovación Biosanitaria de AP, Comunidad de Madrid, Madrid, Spain 28 Epidemiology Unit, Regional Ministry of Health, Community of Madrid, Madrid, Spain 29 Service of Prevention of Labor Risks, Medical Emergencies System, Generalitat de Catalunya, Barcelona, Spain 30 National Center of Epidemiology, Instituto de Salud Carlos III (ISCIII), Madrid, Spain 31 Health Services Research Network on Chronic Diseases (REDISSEC), Madrid, Spain 32 Research Progamme on Biomedical Informatics (GRIB), Hospital del Mar Medical Research Institute (IMIM), Barcelona, Spain 33 Department of Experimental and Health Sciences, Pompeu Fabra University, Barcelona, Spain 34 Instituto Nacional de Bioinformatica—ELIXIR‐ES, Madrid, Spain 35 Parc Sanitari Sant Joan de Déu, Barcelona, Spain 36 Hospital Clínic, University of Barcelona, IDIBAPS, Barcelona, Spain 37 Department of Health Care Policy, Harvard Medical School, Boston, Massachusetts, USA 38 Center for Public Health Psychiatry, Universitair Psychiatrisch Centrum, KU Leuven, Leuven, Belgium Correspondence Philippe Mortier, IMIM, CIBERESP, IMIM, Abstract PRBB Bldg, Carrer del Doctor Aiguader 88, Background: Healthcare workers are a key occupational group at risk for suicidal 08003 Barcelona, Spain. Email: pmortier@imim.es thoughts and behaviors (STB). We investigated the prevalence and correlates of STB among hospital workers during the first wave of the Spain COVID‐19 outbreak Jordi Alonso, IMIM, CIBERESP, UPF, IMIM, PRBB Bldg, Carrer del Doctor Aiguader 88, (March–July 2020). 08003 Barcelona, Spain. Methods: Data come from the baseline assessment of a cohort of Spanish hospital Email: jalonso@imim.es workers (n = 5450), recruited from 10 hospitals just after the height of the coronavirus Funding information disease 2019 (COVID‐19) outbreak (May 5–July 23, 2020). Web‐based self‐report Fondo de Investigación Sanitaria, Instituto de surveys assessed 30‐day STB, individual characteristics, and potentially modifiable Salud Carlos III (Ministerio de Ciencia e Innovación) /FEDER, Grant/Award Number: contextual factors related to hospital workers' work and financial situation. COV20/00711; ISCIII, Grant/Award Number: Results: Thirty‐day STB prevalence was estimated at 8.4% (4.9% passive ideation Sara Borrell, CD18/00049, PFIS, FI18/00012; FPU, Grant/Award Number: FPU15/05728; only, 3.5% active ideation with or without a plan or attempt). A total of n = 6 Generalitat de Catalunya, professionals attempted suicide in the past 30 days. In adjusted models, 30‐day STB Grant/Award Number: 2017SGR452 remained significantly associated with pre‐pandemic lifetime mood (odds ratio [OR] = 2.92) and anxiety disorder (OR = 1.90). Significant modifiable factors included a perceived lack of coordination, communication, personnel, or supervision at work (population‐attributable risk proportion [PARP] = 50.5%), and financial stress (PARP = 44.1%). Conclusions and Relevance: Thirty‐day STB among hospital workers during the first wave of the Spain COVID‐19 outbreak was high. Hospital preparedness for virus outbreaks should be increased, and strong governmental policy response is needed to increase financial security among hospital workers. KEYWORDS COVID‐19 outbreak, hospital workers, suicidal thoughts and behaviors 530 | MORTIER ET AL. 1 | INTRODUCTION 2 | ME THO D S The coronavirus disesase 2019 (COVID‐19) pandemic has presented 2.1 | Study design, population, and sampling hospital workers with unprecedented challenges in terms of work- load as well as health‐ and work‐related risk and stress exposures. The study design consists in a multicenter, prospective, observational The latter includes exposure to COVID‐19 patients and stress about cohort study of Spanish hospital workers. A convenience sample of getting infected or infecting loved ones, but also moral injury, that is, 10 hospitals from four autonomous communities in Spain (i.e., the psychological distress resulting from actions, or the lack of them, that Basque Country, Castile and Leon, Catalonia, and the Community of violate one's moral or ethical code (Litz et al., 2009). In the current Madrid) agreed to participate. Hospitals were selected to reflect the context, moral injury may result from the lack of hospital pre- geographical and sociodemographic variability in Spain. All partici- paredness for the pandemic, and may lead to traumatic experiences pating hospitals came from regions with high COVID‐19 caseloads. such as having to prioritize care, or seeing patients suffer or die from Here we report on the baseline assessment of the cohort, which COVID‐19 (Greenberg et al., 2020). In line with these concerns, rates consists of de‐identified web‐based self‐report surveys (May 5–July 23, of depression, anxiety, and sleep problems among healthcare work- 2020), conducted soon after the first wave of the COVID‐19 outbreak ers during COVID‐19 outbreaks are high (Muller et al., 2020; Pappa in Spain. In each participating hospital, hospital representatives con- et al., 2020; Vindegaard & Benros, 2020), and those in direct contact tacted all employed hospital workers to participate using the hospitals' with affected patients report posttraumatic stress and psychological administrative email distribution lists (i.e., census sampling). The in- distress (Kisely et al., 2020). These adverse mental health outcomes vitation email included an anonymous link to access the web‐based are well‐known risk factors for suicidal thoughts and behaviors (STB; survey platform (qualtrics.com). Informed consent was obtained from all Franklin et al., 2017). High rates of STB among healthcare profes- participants at the first survey page. Two reminder emails were sent sionals during virus outbreaks can therefore be expected (Gunnell within a 2–4 weeks period after the initial invitation. et al., 2020) especially since this population segment already has The study complies with the principles established by national increased risk for suicidal ideation (Tyssen et al., 2001) and suicide and international regulations, including the Declaration of Helsinki (Dutheil et al., 2019; Hawton et al., 2011) under normal working and the Code of Ethics. The study was approved by the Research conditions. No research to date focused on STB during a virus out- Integrity and Good Scientific Practices Committee of IMIM‐Parc de break in this key occupational group at risk (Salazar de Pablo et al., Salut Mar, Barcelona, Spain (2020/9203/I), and by all participating 2020). In addition, previous studies among hospital workers active centers' institutional review boards (IRBs). during the COVID‐19 pandemic predominantly focused on health- care workers (mostly doctors or nurses; Muller et al., 2020), while many hospital workers not involved in patient care may also be at 2.2 | Measures risk for adverse mental health. Spain was among those countries whose healthcare systems 2.2.1 | STB came under extreme pressure during the first wave of the COVID‐19 pandemic (March–July 2020; Arango, 2020). The Spanish govern- A modified self‐report version of selected items from the Columbia ment declared a state of alarm on March 14, 2020, and between the Suicide Severity Rating Scale (C‐SSRS; Posner et al., 2011), also used beginning of March and mid‐April, more than 2000 new cases were in other large‐scale epidemiological studies (e.g., Nock et al., 2014), reported daily. The healthcare system nearly collapsed during assessed suicidal thoughts and behaviors in the past 30 days, in- April–May due to lack of intensive care unit beds, ventilators, and cluding passive suicidal ideation (“wish you were dead or would go to healthcare personnel (Red Nacional de Vigilancia Epidemiólogica sleep and never wake up”), active suicidal ideation (“have thoughts of [RENAVE], 2020). By the time the situation stabilized in early July, killing yourself”), suicide plans (“think about how you might kill yourself Spain had the eighth highest number of confirmed cases [e.g., taking pills, shooting yourself] or work out a plan of how to kill (i.e., 249,659 on 01/07/2020), and the fifth highest COVID death rate yourself”), and suicide attempt (“make a suicide attempt [i.e., purpo- (i.e., 60.7/100,000 on 01/07/2020) in the world (Roser et al., 2020). sefully hurt yourself with at least some intent to die]”). We present data from the MIND/COVID project (MIND/COVID‐ 19, 2020), a national multiple‐cohort study of the mental health impact of the COVID‐19 pandemic in Spain. We report here on the baseline 2.2.2 | Potentially modifiable contextual factors assessment of the hospital workers cohort, conducted just after the height of the virus outbreak (May 5–July 23, 2020), when demands on Potentially modifiable contextual factors refer to factors that are the Spanish public healthcare system were substantially increased. The related to hospital workers' work and financial situation, that are objectives of the current report are to examine baseline prevalence of relevant with regard to the COVID‐19 outbreak, and that are po- 30‐day STB and to investigate the relationship of potentially modifiable tentially modifiable in the future. contextual factors related to hospital workers' perceived work and fi- Six work‐related factors were assessed: (1) the average weekly nancial situation, with 30‐day STB. hours worked, categorized into 40 h/week or less, 41–50 h/week, MORTIER ET AL. | 531 and 51 h or more per week; (2) the perceived lack of coordination, 2.3 | Data representativeness and quality communication, personnel, or supervision at work, using four 5‐level Likert‐type items ranging from “none of the time” to “all of the time.” A total of 5450 hospital workers participated (response rate = Items were summed and rescaled to a 0.0–4.0 Likert scale (Cronbach 11.8%). It is important to note that the survey view rate (i.e., the α = .858); (3) the perceived frequency of lack of protective equip- proportion of hospital workers that opened the invitation email; ment, using a 5‐level Likert‐type item ranging from “none of the time” Eysenbach, 2004) is unknown, except for one hospital (26.4%), sug- to “all of the time;” (4) the perceived efficiency of the available pro- gesting low survey view rates, and questioning the validity of the tective equipment, using a 4‐level Likert‐type item ranging from response rate as an indicator of data representativeness. Post‐ “sufficient” to “completely insufficient;” (5) having had to make deci- stratification weights were used in all analyses to adjust for potential sions regarding prioritizing care among COVID‐19 patients (assessed nonresponse bias, taking into account sample versus target popula- among medical doctors and nurses only); and (6) having had patient tion differences in age, gender, and profession. Differences in post‐ (s) in care that died from COVID‐19. All items included a specific stratifying variables between our sample and the target population time frame, that is, “since the onset of the virus outbreak in Spain.” were small, suggesting good data representativeness. See Table S1 Two factors related to hospital workers' financial situation were and S2 for more details on response rates and poststratification. To assessed: (1) having suffered a significant loss in personal or familial further assess data representativeness, we compared our observed income due to the COVID‐19 pandemic; and (2) financial stress, COVID‐19 infection rates stratified by Autonomous Community in using two 5‐level Likert‐type items that assessed stress regarding Spain (range 8.4%–21.8%) with official seroprevalence results, and one's financial situation (Dohrenwend et al., 1978) and stress found that they are in very close agreement (see Table S3). regarding job loss or loss of income because of COVID‐19, with For this study, analyses were restricted to n = 5169 (94.8%) that response options ranging from “none” to “very severe.” Items were completed all STB items. An additional n = 5 were excluded because summed and rescaled to 0.0–4.0 Likert scale (Cronbach α = .821). they did identify with neither male or female gender (of those, n = 1 reported 30‐day passive ideation only; n = 2 reported a 30‐day sui- cide attempt). No statistical differences in gender or age were found 2.2.3 | Individual characteristics between those that completed the STB items and those that did not (females 80.8% vs. 82.1%, χ2(1) = 0.315, p = .574; mean age 42.9 vs. Twelve individual characteristics were assessed: (1) age; (2) gender; (3) 42.1, t(5448] = −1.18, p = .240). Median % missingness per variable in marital status; (4) having children in care; (5) self‐reported lifetime mental the analysis sample was 1.4% (see Table S4). Missing data were disorders before the onset of the COVID‐19 outbreak, using Composite handled using multivariate imputation by chained equations International Diagnostic Interview (CIDI) 3.0 adapted screener (van Buuren, 2012). items (Kessler & Üstün, 2004), including mood (i.e., depressive and bipolar disorders), anxiety (i.e., panic, generalized anxiety, and obsessive–compulsive disorders), substance use (i.e., alcohol, illicit drugs, 2.4 | Statistical analysis and prescription drugs with or without prescription), and other disorders; (6) profession, categorized into five categories: medical doctors, nurses, All analyses were conducted with SAS version 9.4, and R version auxiliary nurses, other professions involved in patient care (i.e., midwives; 3.6.2. First, prevalence estimates of 30‐day STB were estimated, dentists or odontologists; pharmaceutical, laboratory, or radiology tech- both total and stratified by individual characteristics, with associated nicians; psychologists, physiotherapists, social workers, patient transport) modified Rao–Scott χ2 tests and Fisher‐Exact tests. Second, we es- and other professions not involved in patient care (i.e., administrative and timated multivariable associations between 30‐day STB and in- management personnel, logistic support [e.g. food, maintenance, sup- dividual characteristics. Logistic regression was used for all plies], research‐only personnel); (7) the frequency of direct exposure to multivariable analysis. Third, we estimated the multivariable asso- COVID‐19 patients during professional activity, using one 5‐level Likert ciation between 30‐day STB and each potentially modifiable con- type item, ranging from “none of the time” to “all of the time;” (8) changes textual factor separately, adjusting for all individual characteristics. in assigned functions, team, or working location, categorized into having Fourth, we identified the subset of individual characteristics and changed to a specific COVID‐19‐related work location (e.g., emergency potentially modifiable contextual factors that best explain STB in room, COVID ward, fever clinic, intensive care unit, quarantine center, multivariable models, using the lasso shrinkage method (Hastie et al., field hospital), having changed of team or assigned functions, and all 2009), optimizing the Bayesian Information Criterion. Variance was others; (9) the frequency of working at home during the COVID‐19 estimated using the Taylor series linearization method taking into outbreak, using a 6‐level Likert item, ranging from “never” to “always;” account post‐stratification and within‐hospital clustering of data. (10) COVID‐19 infection history, categorized into having been hospita- Potential deviations from a continuous linear effect in the logit were lized for COVID‐19, having had a positive COVID‐19 test or medical assessed using likelihood ratio tests comparing full categorical versus diagnosis not requiring hospitalization, and all others; (11) having been in continuous variable specifications. All analyses were adjusted for isolation of quarantine because of exposure to COVID‐19‐infected per- hospital membership and time of survey. Fifth, population‐ son(s); and (12) having close ones infected with COVID‐19. attributable risk proportions (PARP; Krysinska & Martin, 2009), and 532 | MORTIER ET AL. associated bootstrap percentile confidence intervals (500 replica- significantly associated with pre‐existing mood and anxiety disorders tions), were simulated based on individuals' predicted probabilities (OR range 1.63–5.05), the perceived lack of coordination, commu- estimated by the multivariable logistic regression equations (Nock nication, personnel, or supervision at work (OR range 1.35–1.85), et al., 2015). PARP provide estimates of the proportions of STB that and the degree of financial stress (OR range 1.28–1.50), with OR could potentially be attributed to specific predictor variables as- generally being higher for active ideation, plan, or attempt. suming a causal pathway between these predictor variables and STB. Approximately 50.5% of 30‐day STB was potentially attributable to a perceived lack of coordination, communication, personnel, or supervision at work (Figure 1), and approximately 44.1% of STB was 3 | RESULTS potentially attributable to financial stress. Interventions that com- pletely eliminate both these risk factors could potentially eliminate Prevalence of 30‐day STB was estimated at 8.4% (Table 1). Of those, 73.3% of any 30‐day STB (61.5%–61.6% of 30‐day passive or active more than half (4.9% in total) were passive suicidal ideation only. The ideation only, and 81.3% of active ideation with plan or attempt). others (3.5%) consisted of active suicidal ideation without plan or PARP for pre‐existing mental disorders were 35.8% for any 30‐day attempt (0.8%) and with plan or attempt (2.7%). A total of n = 6 STB, 24.8% for passive ideation only, and 47.7%–49.5% for active reported a 30‐day suicide attempt. Thirty‐day STB was substantially ideation, plan, or attempt. Additionally eliminating pre‐existing elevated among those with pre‐existing mood and anxiety disorders mental disorders could increase reductions in 30‐day STB to 83.5% (any STB range 13.5%–22.3%; plan or attempt range 4.8%–9.6%) and (71.2% for passive ideation only, and 80.1%–90.4% for active idea- among those with a hospitalization for COVID‐19 (any STB = 12.6%; tion, plan, or attempt). plan or attempt = 8.1%). STB was also elevated among those aged 18–29, non‐married respondents, those without children in care, those that had frequent exposure to COVID‐19 patients, those that 4 | DISCUSS ION changed to specific COVID‐19‐related work locations, and those that never worked at home. Any STB ranged from 6.2% to 12.0% across Approximately 8.4% of hospital workers had 30‐day STB during the hospitals (Table S1). first wave of the Spain COVID‐19 pandemic. Of those, about 4 out of In multivariable analysis of individual characteristics (Table 2), STB 10 had active suicidal ideation (3.5% in total), including 6 hospital remained strongly associated with pre‐existing mood and anxiety dis- workers with a 30‐day suicide attempt. Substantial proportions of orders (any STB OR range 2.12–3.22; plan or attempt OR range 30‐day STB were associated with perceived lack in communication, 2.62–4.64). Detailed analysis of age showed that odds for STB increased coordination, personnel, or supervision (50.5%) and with financial slightly between age 30 and 49, and declined for those aged 50 or more. stress (44.1%). Those with female gender, married or having children in care all had A major strength of this study is the large probability sample of lower odds for any STB (OR range 0.62–0.80). Female gender and being hospital workers, opposed to the high amount of small samples used married was also protective for plan or attempt (OR range = 0.40–0.68). in mental health research among healthcare professionals during No clear associations were found between STB and professional status. virus outbreaks (median n = 333 [IQR 131–769]; Kisely et al., 2020). The frequency of direct exposure to COVID‐19 patients was positively This enabled us to provide detailed analyses of all presumed pre- associated with any STB (OR = 1.11), and passive suicidal ideation only cursor states in the suicidal spectrum (Ribeiro et al., 2016). Despite (OR = 1.17). Those working more frequently at home had generally the large sample size, few cases of 30‐day suicide attempt were lower odds for STB (OR range 0.78–0.82). No significant associations included, precluding detailed analysis of this outcome. An important were found with COVID‐19 infection history and with having been limitation of our study is the low survey response rate (11.8%), in line isolated or quarantined. with the low and declining survey response rates among healthcare Next, we analyzed each of the potentially modifiable contextual professionals worldwide (Cho et al., 2013). While nonresponse bias factors separately, each time adjusting for individual characteristics may affect prevalence estimates in either direction, the expected bias (Table 3). The perceived lack of coordination, communication, per- in associations is substantially lower (Amaya & Presser, 2016). We sonnel, or supervision at work as well as the degree of financial adjusted all estimates for potential nonresponse bias using post- stress were consistently associated with all 30‐day STB outcomes stratification weights. Differences in poststratifying variables be- (OR range 1.18–1.87). Perceived frequency of lack of protective tween our sample and the census population were small, and material (OR = 1.18), and significant loss in personal or familial in- COVID‐19 infection rates in our sample were very similar to official come (OR = 1.35) were also associated with any STB, but only with prevalence estimates among healthcare professionals. Two other passive ideation in the detailed analyses. Having to make decisions limitations of this study are worth mentioning: (1) only those hospital regarding prioritizing care among COVID‐19 patients was only as- workers with institutional email addresses were eligible, which ex- sociated with active suicidal ideation with plan or attempt cludes hospital workers with jobs that are potentially subcontracted (OR = 1.57). to external services (e.g., food and cleaning services), and those who In multivariable models including the variables selected after lost their job by time of survey invitation; and (2) although we ad- lasso shrinkage (Table 4), 30‐day STB outcomes remained justed all analyses for time of survey, we did not investigate how TABLE 1 Thirty‐day prevalence of suicidal thoughts and behaviors (STB), total, and stratified by individual characteristics (n = 5164) Passive ideation Active ideation without plan Active ideation with plan MORTIER Any STB (n = 395) only (n = 243) or attempt (n = 41) or attempt (n = 111) ET AL. n % (SE) % (SE) % (SE) % (SE) % (SE) Total 8.4 (0.8) 4.9 (0.5) 0.8 (0.2) 2.7 (0.3) Age 50 years or more 1730 37.0 (2.1) 7.6 (0.6) 4.7 (0.5) 0.1 (0.0) 2.9 (0.5) 30–49 years 2553 47.7 (1.1) 7.5 (1.1) 4.3 (0.6) 1.0 (0.3) 2.2 (0.3) 18–29 years 881 15.3 (1.6) 12.7 (1.8) 7.2 (1.3) 2.1 (0.6) 3.3 (0.6) 2 2 2 Modified Rao–Scott (R–S) χ test or Fisher‐ R–S χ (2) = 12.59, p = .002* R–S χ (2) = 8.63, p = .013* F‐E p < .001* R–S χ2(2) = 2.66, p = .265 Exact (F‐E) test Gender Male 992 22.7 (1.3) 8.5 (1.0) 3.1 (0.5) 1.3 (0.5) 4.1 (0.5) Female 4172 77.3 (1.3) 8.3 (0.8) 5.4 (0.6) 0.7 (0.1) 2.2 (0.3) 2 2 2 Modified Rao–Scott (R–S) χ test or Fisher‐ R–S χ (1) = 0.11, p = .738 R–S χ (1) = 10.73, p = .001* F‐E p = .425 R–S χ2(1) = 10.82, p = .001* Exact (F‐E) test Marital status Single, divorced, or legally separated, or 2753 52.1 (2.0) 10.7 (1.2) 6.1 (1.0) 1.3 (0.3) 3.3 (0.4) widowed Married 2411 47.9 (2.0) 5.8 (0.3) 3.6 (0.3) 0.3 (0.1) 1.9 (0.5) Modified Rao–Scott (R–S) χ2 test or Fisher‐ R–S χ2(1) = 21.46, p < .001* R–S χ2(1) = 4.71, p = .030* F‐E p < .001* R–S χ2(1) = 4.25, p = .039* Exact (F‐E) test Having children in care Children in care 1974 38.7 (0.8) 5.8 (0.7) 3.5 (0.3) 0.3 (0.1) 2.0 (0.5) No children in care 3190 61.3 (0.8) 10.0 (1.0) 5.8 (0.7) 1.1 (0.2) 3.1 (0.4) Modified Rao–Scott (R–S) χ2 test or Fisher‐ R–S χ2(1) = 15.79, p < .001* R–S χ2(1) = 10.20, p = .001* F‐E p = .002* R–S χ2(1) = 2.57, p = .109 Exact (F‐E) test Lifetime mental disorders before onset COVID‐19 outbreak Lifetime mood disorder 568 12.1 (0.7) 22.3 (1.8) 9.9 (1.9) 2.8 (0.6) 9.6 (1.3) No lifetime mood disorder 4596 87.9 (0.7) 6.4 (0.6) 4.2 (0.3) 0.6 (0.1) 1.7 (0.2) 2 2 2 2 Modified Rao–Scott (R–S) χ test or Fisher‐ R–S χ (1) = 71.74, p < .001* R–S χ (1) = 9.67, p = .002* R–S χ (1) = 15.41, p < .001* R–S χ2(1) = 29.57, p < .001* Exact (F‐E) test Lifetime anxiety disorder 1893 37.9 (1.1) 13.5 (1.3) 7.2 (0.8) 1.6 (0.4) 4.8 (0.5) No lifetime anxiety disorder 3271 62.1 (1.1) 5.2 (0.6) 3.5 (0.4) 0.4 (0.1) 1.4 (0.3) | (Continues) 533 534 TABLE 1 (Continued) | Passive ideation Active ideation without plan Active ideation with plan Any STB (n = 395) only (n = 243) or attempt (n = 41) or attempt (n = 111) n % (SE) % (SE) % (SE) % (SE) % (SE) 2 2 2 2 Modified Rao–Scott (R–S) χ test or Fisher‐ R–S χ (1) = 50.69, p < .001* R–S χ (1) = 31.63, p < .001* R–S χ (1) = 9.38, p = .002* R–S χ2(1) = 25.02, p < .001* Exact (F‐E) test Lifetime substance use disorder 67 1.4 (0.2) 9.6 (3.8) 1.5 (1.4) 0.0 (0.0) 8.1 (3.0) No lifetime substance use disorder 5097 98.6 (0.2) 8.3 (0.8) 4.9 (0.5) 0.8 (0.2) 2.6 (0.3) Modified Rao–Scott (R–S) χ2 test or Fisher‐ F‐E p = .353 F‐E p = .375 – F‐E p = .003* Exact (F‐E) test Other lifetime mental disorder 135 2.4 (0.3) 15.1 (3.4) 5.2 (2.2) 0.5 (0.5) 9.5 (3.1) No other lifetime mental disorder 5029 97.6 (0.3) 8.2 (0.8) 4.9 (0.5) 0.8 (0.2) 2.5 (0.3) Modified Rao–Scott (R–S) χ2 test or Fisher‐ R–S χ2(1) = 3.22, p = .073 R–S χ2(1) = 0.02, p = .899 F‐E p = 1.000 R–S χ2(1) = 4.22, p = .040* Exact (F‐E) test Number of lifetime mental disorders Two or more 454 9.9 (0.8) 25.1 (1.6) 10.8 (2.3) 3.2 (0.7) 11.2 (2.1) Exactly one 1710 33.1 (1.1) 9.2 (0.9) 5.5 (0.6) 0.9 (0.2) 2.8 (0.4) Zero 3000 57.0 (1.2) 4.9 (0.6) 3.5 (0.4) 0.4 (0.1) 1.1 (0.3) 2 2 2 2 Modified Rao–Scott (R–S) χ test or Fisher‐ R–S χ (2) = 118.39, p < .001* R–S χ (2) = 16.09, p < .001* R–S χ (2) = 23.76, p < .001* R–S χ2(2) = 32.86, p < .001* Exact (F‐E) test Profession Medical doctor 1372 20.8 (0.4) 7.2 (1.7) 4.0 (1.0) 1.0 (0.4) 2.2 (0.5) Nurse 1635 30.6 (0.6) 8.3 (1.0) 5.2 (0.5) 0.8 (0.2) 2.2 (0.5) Auxiliary nurse 681 18.8 (1.6) 10.9 (1.8) 5.9 (1.3) 1.5 (0.6) 3.5 (0.7) Other profession involved in patient care 568 12.8 (0.7) 7.3 (0.9) 3.6 (0.8) 0.4 (0.3) 3.3 (0.7) Other profession not involved in 908 17.0 (1.6) 7.9 (1.2) 5.2 (1.0) 0.2 (0.1) 2.5 (0.6) patient care Modified Rao–Scott (R–S) χ2 test or Fisher‐ R–S χ2(4) = 5.48, p = .241 R–S χ2(4) = 4.16, p = .385 F‐E p = .083 R–S χ2(4) = 4.20, p = .380 Exact (F‐E) test Frequency of direct exposure to COVID‐19 patients All of the time 1478 29.3 (2.6) 10.4 (0.7) 6.6 (0.6) 1.1 (0.3) 2.6 (0.5) Most of the time 918 18.1 (1.0) 9.8 (2.0) 5.6 (1.5) 1.0 (0.3) 3.2 (0.5) Some of the time 1351 26.3 (1.3) 7.9 (1.2) 4.0 (0.4) 1.0 (0.5) 2.8 (0.6) A little of the time 707 13.6 (0.8) 5.9 (1.3) 3.6 (1.2) 0.3 (0.2) 2.0 (0.7) None of the time 710 12.8 (2.1) 5.3 (0.7) 3.0 (0.9) 0.1 (0.1) 2.3 (0.5) MORTIER ET AL. TABLE 1 (Continued) MORTIER Passive ideation Active ideation without plan Active ideation with plan Any STB (n = 395) only (n = 243) or attempt (n = 41) or attempt (n = 111) ET AL. n % (SE) % (SE) % (SE) % (SE) % (SE) Modified Rao–Scott (R–S) χ2 test or Fisher‐ R–S χ2(4) = 13.38, p = .010* R–S χ2(4) = 10.86, p = .028* F‐E p = .072 F‐E p = .363 Exact (F‐E) test Changes at work Changed to specific COVID‐19‐related 1679 31.9 (2.2) 10.7 (1.1) 6.6 (0.7) 1.5 (0.2) 2.7 (0.4) work location Changed of team or assigned functionsb 1316 25.5 (1.0) 8.5 (1.2) 4.9 (1.0) 0.7 (0.2) 2.9 (0.4) No changes in assigned functions, team, or 2169 42.6 (1.6) 6.5 (0.6) 3.6 (0.4) 0.4 (0.2) 2.5 (0.4) work location Modified Rao–Scott (R–S) χ2 test or Fisher‐ R–S χ2(2) = 19.86, p < .001* R–S χ2(2) = 11.82, p = .003* F‐E p < .001* R–S χ2(2) = 0.74, p = .692 Exact (F‐E) test Frequency of working at home Always 195 2.5 (0.8) 3.3 (1.0) 1.9 (0.8) 0.0 (0.0) 1.5 (0.8) Almost always 158 2.3 (0.6) 5.3 (2.1) 3.8 (1.8) 1.1 (1.2) 0.5 (0.5) Often 266 4.6 (0.6) 2.4 (1.1) 2.0 (1.0) 0.1 (0.1) 0.2 (0.2) Sometimes 452 8.1 (1.1) 7.0 (1.3) 2.6 (0.7) 0.9 (0.5) 3.5 (0.9) Seldom 471 8.3 (0.7) 7.4 (1.7) 3.5 (0.6) 1.0 (0.5) 2.9 (1.3) Never 3622 74.2 (2.9) 9.2 (1.0) 5.6 (0.6) 0.8 (0.2) 2.8 (0.4) Modified Rao–Scott (R–S) χ2 test or Fisher‐ R–S χ2(5) = 20.13, p = .001*a R–S χ2(5) = 28.61, p < .001*a R–S χ2(4) = 1.31, p = .859a R–S χ2(5) = 9.36, p = .095a Exact (F‐E) test COVID‐19 infection history Having been hospitalized for COVID‐19 55 1.3 (0.2) 12.6 (4.8) 4.5 (2.4) 0.0 (0.0) 8.1 (4.6) Positive COVID‐19 test or medical 845 16.6 (2.0) 10.0 (2.6) 5.9 (1.8) 1.0 (0.5) 3.2 (0.6) COVID‐19 diagnosisc None of the above 4264 82.0 (2.2) 7.9 (0.6) 4.7 (0.4) 0.8 (0.2) 2.5 (0.3) 2 Modified Rao–Scott (R–S) χ test or Fisher‐ F‐E p = .058 F‐E p = .556 F‐E p = .789 F‐E p = .004* Exact (F‐E) test Isolation or quarantine because of COVID‐19 Having been isolated or quarantined 1258 23.3 (1.3) 8.7 (1.7) 4.6 (1.0) 1.1 (0.4) 2.9 (0.5) Not having been isolated or quarantined 3906 76.7 (1.3) 8.3 (0.6) 5.0 (0.5) 0.7 (0.1) 2.6 (0.2) 2 2 2 Modified Rao–Scott (R–S) χ test or Fisher‐ R–S χ (1) = 0.08, p = .782 R–S χ (1) = 0.13, p = .722 F‐E p = .856 R–S χ2(1) = 0.44, p = .509 Exact (F‐E) test | Close ones infected with COVID‐19 535 (Continues) 536 | MORTIER ET AL. potentially modifiable contextual factors (e.g., hospital preparedness) Active ideation with plan changed over time. To optimize intervention planning efforts, future R–S χ2(2) = 3.91, p = .141 or attempt (n = 111) research should provide more qualitative insight into these con- textual factors (e.g., operationalizing hospital preparedness with clear definitions), and investigate changes over time in these factors in response to viral outbreaks. 3.5 (0.7) 2.8 (0.4) 1.9 (0.3) % (SE) To our knowledge, we present the first study that estimated STB among hospital workers during a virus outbreak. Over the past century, Spain consistently had relatively low rates of suicide (range Active ideation without plan 3.5–9.0/100,000 for the years 1906–2016; Alfonso‐Sánchez et al., 2020), with 12‐month suicidal ideation and plans among Spanish adults also being consistently low, that is, 0.7%–0.9%, and or attempt (n = 41) 0.2%–0.4%, respectively (Bernal et al., 2007; Kovess‐Masfety et al., 2011; Miret et al., 2014). Against this, we found a 30‐day prevalence F‐E p = .273 0.5 (0.2) 0.9 (0.2) 0.8 (0.4) of 3.5% for active ideation, plans, or attempts, suggesting that STB % (SE) among hospital workers during the COVID‐19 outbreak is at least three times higher—but likely much higher—than in the general Spanish population before the COVID‐19 outbreak. One previous study among 375 neurosurgeons recruited through snowball sam- R–S χ2(2) = 0.53, p = .766 The category “changed of team or assigned functions” excludes those that changed to a specific COVID‐19‐related work location. pling found a prevalence of 5.1% of active suicidal ideation during the The category “positive COVID‐19 test or medical COVID‐19 diagnosis” excludes those having been hospitalized for COVID‐19. COVID‐19 pandemic (Sharif et al., 2020), somewhat higher than our Passive ideation estimate (3.5%). No previous studies report on STB prevalence only (n = 243) The category “other family, friends, or others” excludes having a partner, children, or parents infected with COVID‐19. among Spanish hospital or healthcare workers under normal working Fisher‐Exact test could not be estimated due to computational limitations; the modified Rao–Scott χ2 test is shown. 5.5 (1.5) 4.8 (0.5) 4.6 (0.5) conditions. Few previous studies outside of Spain report on suicidal % (SE) ideation among physicians under normal working conditions, and found 12‐month estimates in the range 2.6%–14.3% (Fridner et al., 2009; Hem et al., 2000), and 30‐day estimates of 4.3% (Brooks et al., R–S χ2(2) = 2.59, p = .273 2018; Loas et al., 2018). However, these studies were restricted to female physicians or report data from countries with higher suicide Any STB (n = 395) rates (Sweden, Norway, Belgium) than Spain. To the extent that STB among hospital workers during the COVID‐19 outbreak was effectively increased, and given the asso- 9.5 (1.9) 8.5 (0.7) 7.3 (0.7) % (SE) ciations between STB and future suicide (Ribeiro et al., 2016), our study may warn for an expected increase in suicides among Spanish hospital workers. We identified three important STB correlates that 14.4 (2.0) 60.3 (1.3) 25.3 (1.9) provide empirical evidence for suicide intervention preventions. % (SE) First, a large proportion of 30‐day STB (50.5%) was associated with a perceived inadequate response to the COVID‐19 outbreak in terms of coordination, supervision, communication, and available 3175 1259 730 healthcare personnel. To the extent that this perception reflects the Abbreviation: COVID‐19, coronavirus disease 2019. n objective hospitals' response, this finding underscores the need of Modified Rao–Scott (R–S) χ2 test or Fisher‐ increasing hospital preparedness for virus outbreaks (European Centre for Disease Prevention and Control, 2020). But this finding also highlights the importance of societal efforts to reduce the Other family, friends, or othersd spread of the virus to avoid healthcare systems being overwhelmed, *Statistically significant (α = .05). Partner, children, or parents and to avoid adverse mental health among hospital workers facing a (Continued) sudden lack of hospital preparedness. Second, considerable propor- None of the above tions of 30‐day STB (44.1%) were associated with financial stress, Exact (F‐E) test including stress related to potential income or job loss due to the COVID‐19 virus pandemic. Spain was among those countries hit hard TABLE 1 by the 2007–2008 economic crisis, with subsequent cuts in health- care expenditure (Thomson et al., 2013). This may exacerbate fears among Spanish hospital workers related to the future economic b d a c TABLE 2 Multivariable associations between 30‐day suicidal thoughts and behaviors (STB) and individual characteristics (n = 5164) Passive ideation Active ideation without plan Active ideation with plan or MORTIER % (SE) or Med Any STB (n = 395) only (n = 243) or attempt (n = 41) attempt (n = 111) ET AL. n (SE) (IQR) OR (95% CI)a OR (95% CI)a OR (95% CI)a,b OR (95% CI)a Age Spline 18–29 years 0.92 (0.83–1.02) 0.94 (0.84–1.05) 0.84 (0.74–0.95)* 0.95 (0.82–1.09) Spline 30–49 years 1.03 (1.00–1.06)* 1.03 (0.99–1.07) 1.02 (0.96–1.08) 1.05 (1.01–1.10)* Spline 50+ years 0.93 (0.88–0.97)* 0.94 (0.88–1.01) 0.77 (0.62–0.95)* 0.91 (0.85–0.97)* Gender—female (vs. male) 4167 77.3 (1.3) 0.80 (0.67–0.97)* 1.46 (1.05–2.04)* 0.42 (0.23–0.78)* 0.40 (0.28–0.58)* Marital status—married (vs. single, divorced, or 2411 47.9 (2.0) 0.75 (0.59–0.95)* 0.82 (0.54–1.24) 0.69 (0.31–1.58) 0.68 (0.46–0.99)* legally separated, or widowed) Children in care (vs. no children in care) 1974 38.7 (0.8) 0.62 (0.43–0.89)* 0.62 (0.42–0.91)* 0.46 (0.20–1.06) 0.63 (0.37–1.10) Lifetime mood disorder before onset COVID‐19 568 12.1 (0.7) 3.22 (2.67–3.89)* 2.34 (1.65–3.30)* 5.03 (2.69–9.38)* 4.64 (3.00–7.18)* outbreak Lifetime anxiety disorder before onset COVID‐19 1893 37.9 (1.1) 2.12 (1.59–2.83)* 1.82 (1.41–2.36)* 2.64 (1.41–4.93)* 2.62 (1.49–4.62)* outbreak Lifetime substance use disorder before onset 67 1.4 (0.2) 0.62 (0.30–1.30) 0.21 (0.04–1.26) 0.24 (0.02–2.97) 1.34 (0.68–2.63) COVID‐19 outbreak Other lifetime mental disorder before onset 135 2.4 (0.3) 1.63 (0.96–2.77) 1.03 (0.38–2.80) 0.56 (0.09–3.44) 3.71 (1.90–7.23)* COVID‐19 outbreak Profession Medical doctor 1372 20.8 (0.4) 0.78 (0.48–1.27) 0.64 (0.37–1.13) 2.60 (0.64–10.58) 1.00 (0.43–2.31) Nurse 1635 30.6 (0.6) 0.66 (0.43–1.00) 0.54 (0.34–0.87)* 1.94 (0.46–8.27) 0.86 (0.45–1.65) Auxiliary nurse 681 18.8 (1.6) 0.80 (0.56–1.14) 0.59 (0.38–0.91)* 3.78 (0.89–16.04) 1.23 (0.62–2.46) Other professions involved in patient care 568 12.8 (0.7) 0.67 (0.49–0.93)* 0.54 (0.28–1.05) 1.05 (0.22–5.10) 1.01 (0.52–1.96) Other profession not involved in patient care 908 17.0 (1.6) (ref) (ref) (ref) (ref) Frequency of direct exposure to COVID‐19 patients 1.9 (0.1) (0.9‐3.1) 1.11 (1.01–1.23)* 1.17 (1.04–1.31)* 0.98 (0.73–1.30) 1.04 (0.76–1.44) (scaled 0–4) Changes at work Changed to specific COVID‐19‐related work 1679 31.9 (2.2) 1.27 (0.88–1.85) 1.40 (0.92–2.12) 1.68 (0.76–3.73) 0.91 (0.48–1.75) location Changed of team or assigned functionsc 1316 25.5 (1.0) 1.17 (0.93–1.48) 1.22 (0.85–1.75) 1.24 (0.54–2.81) 1.09 (0.67–1.76) No changes in assigned functions, team or work 2169 42.6 (1.6) (ref) (ref) (ref) (ref) location Frequency of working at home (scaled 0–5) 0.0 (0.2) (0.0–0.1) 0.82 (0.75–0.90)* 0.82 (0.70–0.96)* 1.07 (0.80–1.44) 0.78 (0.64–0.94)* | (Continues) 537 538 TABLE 2 (Continued) | Passive ideation Active ideation without plan Active ideation with plan or % (SE) or Med Any STB (n = 395) only (n = 243) or attempt (n = 41) attempt (n = 111) n (SE) (IQR) OR (95% CI)a OR (95% CI)a OR (95% CI)a,b OR (95% CI)a COVID‐19 infection history Having been hospitalized for COVID‐19 55 1.3 (0.2) 1.43 (0.46–4.44) 1.05 (0.33–3.38) 1.37 (0.10–18.32) 2.30 (0.71–7.42) Positive COVID‐19 test or medical COVID‐19 845 16.6 (2.0) 1.32 (0.84–2.07) 1.44 (0.78–2.63) 0.94 (0.40–2.20) 1.24 (0.75–2.05) diagnosisd None of the above 4264 82.0 (2.2) (ref) (ref) (ref) (ref) Having been isolated or quarantined because of 1258 23.3 (1.3) 0.80 (0.63–1.01) 0.71 (0.49‐1.04) 1.30 (0.63–2.68) 0.78 (0.50–1.23) COVID‐19 Close ones infected with COVID‐19 Partner, children, or parents 730 14.4 (2.0) 1.32 (0.80–2.18) 1.28 (0.65–2.51) 0.68 (0.24–1.94) 1.72 (0.80–3.68) e Other family, friends or others 3175 60.3 (1.3) 1.26 (0.94–1.69) 1.15 (0.84–1.58) 1.02 (0.51–2.01) 1.60 (0.81–3.17) None of the above 1259 25.3 (1.9) (ref) (ref) (ref) (ref) Abbreviations: COVID‐19, coronavirus disease 2019; IQR, interquartile range. *Statistically significant (α = .05). a All analyses adjust for time of survey (weeks), hospital membership, and all predictors shown in the rows. b Due to data sparseness, the model was estimated using Penalized Maximum Likelihood Estimation (Firth‐type Estimation). c The category “changed of team or assigned functions” excludes those that changed to a specific COVID‐19‐related work location. d The category “positive COVID‐19 test or medical COVID‐19 diagnosis” excludes those having been hospitalized for COVID‐19. e The category “other family, friends or others” excludes having a partner, children, or parents infected with COVID‐19. MORTIER ET AL. TABLE 3 Multivariable associations between 30‐day suicidal thoughts and behaviors (STB) and potentially modifiable contextual factors (n = 5164) MORTIER Passive ideation Active ideation without plan Active ideation with plan or ET AL. % (SE) or Med Any STB (n = 395) only (n = 243) or attempt (n = 41) attempt (n = 111) n (SE) (IQR) OR (95% CI)a OR (95% CI)a OR (95% CI)a,b OR (95% CI)a Work‐related Average weekly hours worked factors 51 h or more 775 13.4 (1.1) 0.78 (0.49–1.25) 0.80 (0.42–1.55) 1.12 (0.50–2.50) 0.61 (0.32–1.17) 41–50 h 1354 25.6 (2.2) 0.78 (0.71–0.84)* 0.84 (0.72–0.99)* 0.75 (0.38–1.45) 0.61 (0.43–0.87)* 40 h or less 3035 61.0 (3.0) (ref) (ref) (ref) (ref) Perceived lack of communication, coordination, 1.9 (0.1) (1.0–2.5) 1.48 (1.33–1.64)* 1.36 (1.20–1.53)* 1.87 (1.37–2.56)* 1.67 (1.44–1.94)* personnel, or supervision (scaled 0.0–4.0)c Perceived lack of coordination (scaled 0–4) 1.6 (0.1) (0.8–2.4) 1.36 (1.18–1.57)* 1.27 (1.08–1.49)* 1.53 (1.16–2.03)* 1.54 (1.27–1.86)* Perceived lack of communication (scaled 0–4) 1.5 (0.1) (0.7–2.3) 1.37 (1.22–1.53)* 1.22 (1.07–1.40)* 1.76 (1.35–2.31)* 1.59 (1.30–1.94)* Perceived lack of personnel (scaled 0–4) 1.5 (0.2) (0.3–2.6) 1.22 (1.11–1.33)* 1.18 (1.10–1.27)* 1.49 (1.16–1.91)* 1.24 (1.02–1.50)* Perceived lack of supervision (scaled 0–4) 1.1 (0.1) (0.0–2.1) 1.28 (1.22–1.34)* 1.24 (1.13–1.36)* 1.39 (1.11–1.75)* 1.34 (1.21–1.49)* Perceived frequency of lack of protective 1.8 (0.1) (1.1–2.6) 1.18 (1.05–1.34)* 1.21 (1.07–1.36)* 1.23 (0.92–1.64) 1.08 (0.90–1.30) equipment (scaled 0–4) Perceived inefficiency of protective equipment 0.3 (0.0) (0.0–1.1) 1.06 (0.85–1.33) 1.04 (0.79–1.38) 1.05 (0.79–1.39) 1.05 (0.80–1.37) (scaled 0–3) Having to make decisions regarding prioritizing 858 14.9 (1.3) 1.18 (0.79–1.75) 1.07 (0.67–1.70) 0.96 (0.43–2.16) 1.57 (1.07–2.32)* care among COVID‐19 patients Having patient(s) in care that died from 1993 37.8 (2.7) 1.07 (0.85–1.36) 1.46 (0.79–2.71) 0.99 (0.50–1.99) 0.64 (0.40–1.02) COVID‐19 Financial factors Significant loss in personal or familial income 1058 20.4 (0.9) 1.32 (1.03–1.69)* 1.43 (1.05–1.95)* 0.97 (0.51–1.84) 1.13 (0.65–1.95) Financial stress (scaled 0.0–4.0)d,e 0.7 (0.0) (0.0–1.7) 1.39 (1.31–1.48)* 1.32 (1.16–1.51)* 1.14 (0.90–1.46) 1.60 (1.28–2.00)* Stress related to job or income loss due to 0.2 (0.0) (0.0–1.4) 1.27 (1.18–1.37)* 1.21 (1.04–1.42)* 1.01 (0.81–1.25) 1.44 (1.16–1.79)* COVID‐19 (scaled 0–4)e Stress related to financial situation (scaled 0–4)e 0.6 (0.0) (0.0–1.6) 1.38 (1.31–1.45)* 1.32 (1.21–1.45)* 1.27 (1.01–1.61)* 1.52 (1.30–1.78)* Abbreviations: COVID‐19, coronavirus disease 2019; IQR, interquartile range. *Statistically significant (α = .05). a Each cell represents a separate regression model, each time adjusting for time of survey (weeks), hospital membership, and all individual characteristics (i.e., age, gender, marital status, having children in care, lifetime mental disorders before onset COVID‐19 outbreak, profession, frequency of direct exposure to COVID‐19 patients, changes in assigned functions, team or work location, frequency of working at home, COVID‐19 infection history, having been isolated or quarantined related to COVID‐19, and COVID‐19 infection of close ones). b Due to data sparseness, the model was estimated using Penalized Maximum Likelihood Estimation (Firth‐type Estimation). c The Perceived lack of communication, coordination, personnel, or supervision scale is created using the four 5‐level Likert‐type items shown in the rows below (see Section 2.2). d The Financial stress scale is created using the two 5‐level Likert‐type items shown in the rows below (see Section 2.2). e Likelihood ratio tests for linearity of effect were significant, and models including polynomials suggest a better model fit when including a quadratic term. For ease of interpretability, we present here the models including the linear term only; models including the quadratic term are presented in Table S5. | 539 540 TABLE 4 Final multivariable models predicting 30‐day suicidal thoughts and behaviors (STB; n = 5164) | Any Passive ideation Active ideation without Active ideation with plan % (SE) or Med STB (n = 395) only (n = 243) plan or attempt (n = 41) or attempt (n = 111) n (SE) (IQR) OR (95% CI)a OR (95% CI)a OR (95% CI)a,b OR (95% CI)a Individual characteristics Age Spline 18‐29 years 0.91 (0.82–1.01) 0.94 (0.84–1.04) 0.81 (0.71–0.92)* 0.94 (0.82–1.07) Spline 30‐49 years 1.04 (1.01–1.07)* 1.03 (0.99–1.08) 1.02 (0.96–1.08) 1.06 (1.03–1.10)* Spline 50+ years 0.93 (0.89–0.98)* 0.95 (0.88–1.03) 0.76 (0.61–0.96)* 0.92 (0.86–0.98)* Gender—female (vs. male) 4167 77.3 (1.3) 0.82 (0.69–0.98)* 1.52 (1.04–2.22)* 0.45 (0.24–0.83)* 0.38 (0.29–0.51)* Marital status—married (vs. single, divorced, 2411 47.9 (2.0) 0.73 (0.55–0.96)* 0.78 (0.48–1.25) 0.71 (0.29–1.74) 0.65 (0.44–0.96)* or legally separated, or widowed) Children in care (vs. no children in care) 1974 38.7 (0.8) 0.61 (0.42–0.89)* 0.60 (0.41–0.87)* 0.56 (0.23–1.38) 0.61 (0.35–1.06) Lifetime mood disorder before onset 568 12.1 (0.7) 2.92 (2.40–3.54)* 2.09 (1.46–2.99)* 5.05 (2.68–9.52)* 4.21 (2.70–6.58)* COVID‐19 outbreak Lifetime anxiety disorder before onset 1893 37.9 (1.1) 1.90 (1.40–2.58)* 1.63 (1.24–2.14)* 2.24 (1.17–4.29)* 2.39 (1.30–4.40)* COVID‐19 outbreak Frequency of direct exposure to COVID‐19 1.9 (0.1) (0.9–3.1) 1.05 (0.95–1.16) 1.10 (1.02–1.19)* 1.07 (0.82–1.39) 0.95 (0.71–1.25) patients (scaled 0–4) Frequency of working at home (scaled 0–5) 0.0 (0.2) (0.0–0.1) 0.84 (0.73–0.96)* 0.84 (0.75–0.95)* 0.94 (0.69–1.26) 0.78 (0.60–1.00)* Potentially modifiable Perceived lack of communication, 1.9 (0.1) (1.0–2.5) 1.44 (1.27–1.63)* 1.35 (1.13–1.61)* 1.85 (1.34–2.56)* 1.59 (1.34–1.90)* contextual factors coordination, personnel, or supervision (scaled 0.0–4.0) Perceived inefficiency of protective 0.3 (0.0) (0.0–1.1) 0.89 (0.73–1.08) 0.90 (0.71–1.15) 0.86 (0.64–1.15) 0.83 (0.62–1.11) equipment (scaled 0–3) Financial stress (scaled 0.0–4.0)c 0.7 (0.0) (0.0–1.7) 1.34 (1.26–1.43)* 1.28 (1.14–1.44)* 1.10 (0.86–1.41) 1.50 (1.24–1.83)* Note: Individual characteristics and potentially modifiable contextual factors included in the final multivariable models are selected using the lasso shrinkage method, optimizing the Bayesian Information Criterion. Abbreviations: COVID‐19, coronavirus disease 2019; IQR, interquartile range. *Statistically significant (α = .05). a All analyses adjust time of survey (weeks), hospital membership, and all predictors shown in the rows. b Due to data sparseness, the model was estimated using Penalized Maximum Likelihood Estimation (Firth‐type Estimation). c Likelihood ratio tests for linearity of effect were significant, and models including polynomials suggest a better model fit when including a quadratic term. For ease of interpretability, we present here the models including the linear term only; models including the quadratic term are presented in Table S6. MORTIER ET AL. MORTIER ET AL. | 541 F I G U R E 1 Multivariable PARPs (All analyses adjust for time of survey [weeks], hospital membership, age, gender, marital status, having children in care, frequency of direct exposure to COVID‐19 patients, frequency of working at home, perceived inefficiency of protective equipment, and the predictors shown in the figure.) for 30‐day suicidal thoughts and behaviors (STB; n = 5164). Individual characteristics and potentially modifiable contextual factors included in the final multivariable models are selected using the lasso shrinkage method, optimizing the Bayesian Information Criterion. Variables included for PARP analysis are potentially modifiable contextual factors, and lifetime mental disorders before the onset of the COVID‐19 outbreak; only PARPs that are statistically different from zero are shown. See Tables S7 and S8 for more details on PARP analysis. PARP of individual risk factors do not sum to 100% because of the multifactorial etiology of STB, that is, one risk factor can be part of multiple risk factor constellations leading to STB. COVID‐19, coronavirus disease 2019; PARP, population‐attributable risk proportion; STB, suicidal thoughts and behavior impact of the COVID‐19 pandemic, and points to the need of in- financial situation. The COVID‐19 pandemic has revealed how de- creasing financial security for hospital workers active during the pendent we are as a society on a well‐functioning healthcare system. COVID‐19 outbreak. Third, our results suggest that those hospital Improving future mental health and promoting fair financial and workers with a history of mental disorder are at increased risk for working conditions among hospital workers should therefore be an STB during high‐stress periods such as a virus outbreak, in line with a absolute priority. stress‐diathesis etiological model of suicide (van Heeringen & Mann, 2014). Importantly, a recent meta‐analysis documented an absolute lack of research on effective interventions for mental disorders and The MIND C OVID W or k i ng Gr oup is STB among healthcare personnel, both on the individual‐ and on the formed by organizational level (Petrie et al., 2019). Our study points to the need for adequate mental health support for hospital workers active Itxaso Alayo, Jordi Alonso, Manuel Alonso, Mar Álvarez‐Villalba, during virus outbreaks (Vieta et al., 2020). When suicidal ideation is Benedikt Amann, Franco F. Amigo, Gerard Anmella, Andrés Aragón, present, suicide prevention guidelines recommend additional support Núria Aragonès, Enric Aragonès, Ana Isabel Arizón, Angel Asunsolo, for healthcare workers such as crisis helplines or online interventions Alfons Ayora, Laura Ballester, Puri Barbas, Josep Basora, Elena Bere- (Gunnell et al., 2020; Moutier, 2020; Wasserman et al., 2020), which ciartua, Inés Bravo, Ignasi Bolíbar, Xavier Bonfill, Ronny Bruffaerts, should take into account specific experiences such as moral injury Alberto Cotillas‐Rodero, Paula Cristóbal‐Narváez, Andrés Cuartero, and COVID‐19‐related traumatic experiences. Concha de Paz, Isabel del Cura‐González, Maria Jesús del Yerro, Joke De Vocht, Domingo Díaz, Joan Domènech‐Abella, José Luís Domingo, José I. Emparanza, Mireia Espallargues, Meritxell Espuga, Patricia 5 | C O N CL U S I O N Estevan‐Burdeus, Mireia Félez‐Nobrega, M. Isabel Fernández, Tania Fernández, Montse Ferrer, Yolanda Ferreres, Giovanna Fico, María This study provides the first empirical evidence for a potential in- João Forjaz, Rosa García‐Barranco, Carles García‐Ribera, J. Manuel crease of STB among hospital workers during virus outbreaks, and García‐Torrecillas, Araceli Garrido‐Barral, Elisa Gil, María Giola‐Insigna, suggests important associations with a lack in hospital preparedness, Marta Gómez, Javier Gómez, Ana González‐Pinto, Josep Maria Haro, financial stress, and pre‐existing adverse mental health. Future re- Margarita Hernando, Milagros Iriberri, Leontien Jansen, Núria Jiménez, search should confirm our findings by including a non‐pandemic Xavi Jiménez, Ronald C. Kessler, Amparo Larrauri, Fernando León‐ control condition, and by using a prospective study design including Vázquez, Mayte López‐Atanes, Nieves López‐Fresneña, Carmen López‐ objective markers of hospital functioning and hospital workers' Rodríguez, Juan A. López‐Rodríguez, Germán López‐Cortacans, Alba 542 | MORTIER ET AL. Marcos, Jesús Martín, Vicente Martín, Mercedes Martínez‐Cortés, the manuscript. All authors reviewed the initial draft and made critical Raquel Martínez‐Martínez, Alma D. Martínez de Salázar, Isabel Martínez, contributions to the interpretation of the data and approved the manu- Marco Marzola, Nelva Mata, Josep María Molina, Juan D. Molina, Emilia script. The corresponding authors attest that all listed authors meet Molinero, Philippe Mortier, Carmen Muñoz‐Ruipérez, Andrea Murru, authorship criteria and that no others meeting the criteria have been Lydia Navarro, Beatriz Olaya, Jorge Olmedo‐Galindo, Rafael M. omitted. Ortí‐Lucas, Rafael Padrós, Meritxell Pallejà, Raúl Parra, Julio Pascual, José María Pelayo‐Terán, Rosa Pla, Nieves Plana, Coro Pérez‐Aznar, Beatriz DATA A VAILABILITY STA TEMENT Pérez‐Gómez, Víctor Pérez‐Solà, Aurora Pérez‐Zapata, José Ignacio The de‐identified participant data as well as the study protocol and Pijoan, Elena Polentinos‐Castro, Beatriz Puértolas, María Teresa Puig, statistical analysis plan used for this study are available upon rea- Álex Quílez, María Jesús Quintana, Antonio Quiroga, David Rentero, sonable request from the corresponding authors (Philippe Mortier and Cristina Rey, Cristina Rius, Carmen Rodríguez‐Blázquez, M. José Jordi Alonso) as long as the main objective of the data sharing request Rojas‐Giraldo, Yamina Romero‐Barzola, Gabriel Rubio, Pedro Ruiz, is replicating the analysis and findings as reported in this paper. Mercedes Rumayor, Margarita Sáenz, Jesús Sánchez, Ignacio Sánchez‐ Arcilla, Ferran Sanz, Consol Serra, Victòria Serra‐Sutton, Manuela OR C ID Serrano, Sílvia Solà, Sara Solera, Miguel Soto, Alejandra Tarragó, Nativi- Philippe Mortier https://orcid.org/0000-0003-2113-6241 dad Tolosa, Mireia Vázquez, Margarita Viciola, Eduard Vieta, Gemma Juan D. Molina https://orcid.org/0000-0001-8561-8130 Vilagut, Wouter Voorspoels, Sara Yago‐González, Jesús Yáñez‐Sánchez, Ronald C. Kessler https://orcid.org/0000-0003-4831-2305 Yolanda Zapico, Luís María Zorita, Iñaki Zorrilla, and Saioa L. Zurbano. Jordi Alonso https://orcid.org/0000-0001-8627-9636 A C K N O W L E D GM E N T S REFER E NC ES The authors would like to sincerely thank Puri Barbas and Franco Alfonso‐Sánchez, J. L., Martin‐Moreno, J. M., Martinez, I. M., & Martinez, A. A. Amigo for the management of the project as well as all hospital (2020). Epidemiological study and cost analysis of suicide in Spain: Over 100 years of evolution. Archives of Suicide Research, 24(suppl 2), workers that participated in the study. This study was supported by S356–S369. https://doi.org/10.1080/13811118.2019.1612802 Fondo de Investigación Sanitaria, Instituto de Salud Carlos III Amaya, A., & Presser, S. (2016). Nonresponse bias for univariate and (Ministerio de Ciencia e Innovación)/FEDER (COV20/00711); ISCIII (Sara multivariate estimates of social activities and roles. Public Opinion Borrell, CD18/00049) (PM); FPU (FPU15/05728) (LB); ISCIII (PFIS, Quarterly, 81(1), nfw037. https://doi.org/10.1093/poq/nfw037 Arango, C. (2020). Lessons learned from the coronavirus health crisis in FI18/00012) (BP); Generalitat de Catalunya (2017SGR452). Madrid, Spain: How COVID‐19 has changed our lives in the last 2 weeks. Biological Psychiatry, 88(7), e33–e34. https://doi.org/10. C O NF L IC T O F IN T E R ES T S T A T E M EN T 1016/j.biopsych.2020.04.003 Eduard Vieta reports personal fees from Abbott, personal fees from Bernal, M., Haro, J. M., Bernert, S., Brugha, T., de Graaf, R., Bruffaerts, R., Allergan, personal fees from Angelini, grants from Novartis, grants Lépine, J. P., de Girolamo, G., Vilagut, G., Gasquet, I., Torres, J. V., Kovess, V., Heider, D., Neeleman, J., Kessler, R., & Alonso, J. (2007). from Ferrer, grants and personal fees from Janssen, personal fees Risk factors for suicidality in Europe: Results from the ESEMED from Lundbeck, personal fees from Sage, personal fees from Sanofi, study. Journal of Affective Disorders, 101(1–3), 27–34. https://doi. outside the submitted work. Juan D. Molina reports personal fees org/10.1016/j.jad.2006.09.018 from Janssen, personal fees and nonfinancial support from Brooks, E., Gendel, M. H., Early, S. R., & Gundersen, D. C. (2018). When doctors struggle: Current stressors and evaluation recommendations Otsuka, personal fees and nonfinancial support from Lundbeck, for physicians contemplating suicide. Archives of Suicide Research, 22(4), personal fees from Angelini, personal fees and nonfinancial support 519–528. https://doi.org/10.1080/13811118.2017.1372827 from Accord, outside the submitted work. In the past 3 years, Ronald Cho, Y. I., Johnson, T. P., & VanGeest, J. B. (2013). Enhancing surveys of C. Kessler was a consultant for Datastat, Inc., Sage Pharmaceuticals, health care professionals. Evaluation & the Health Professions, 36(3), and Takeda. All other authors reported no conflict of interest. 382–407. https://doi.org/10.1177/0163278713496425 Dohrenwend, B. S., Krasnoff, L., Askenasy, A. R., & Dohrenwend, B. P. (1978). Exemplification of a method for scaling life events: the Peri AUTHOR CONTRIBUTIONS Life Events Scale. Journal of Health and Social Behavior, 19(2), Philippe Mortier, Gemma Vilagut, and Jordi Alonso reviewed the litera- 205–229. https://doi.org/10.2307/2136536 ture. Philippe Mortier, Gemma Vilagut, Montse Ferrer, Jordi Alonso, Dutheil, F., Aubert, C., Pereira, B., Dambrun, M., Moustafa, F., Mermillod, M., Enric Aragonès, Víctor Pérez‐Solà, Josep M. Haro, Ronald C. Kessler, and Baker, J. S., Trousselard, M., Lesage, F. X., & Navel, V. (2019). Suicide among physicians and health‐care workers: A systematic review and Ronny Bruffaerts conceived and designed the study. Consol Serra, meta‐analysis. PLOS One, 14(12), e0226361. https://doi.org/10.1371/ Juan D. Molina, Nieves López‐Fresneña, Teresa Puig, José M. journal.pone.0226361 Pelayo‐Terán, José I. Pijoan, José I. Emparanza, Meritxell Espuga, Nieves European Centre for Disease Prevention and Control. (2020). Preparedness Plana, Ana González‐Pinto, Consol Serra, Enric Aragonès, Isabel del for COVID‐19. Retrieved from https://www.ecdc.europa.eu/en/covid- 19/preparedness-and-response Cura‐González, Andrés Aragón‐Peña, Mireia Campos, Aurora Pérez‐ Eysenbach, G. (2004). Improving the quality of web surveys: The Zapata, Eduard Vieta, and Víctor Pérez‐Solà acquired the data. Philippe Checklist for Reporting Results of Internet E‐Surveys (CHERRIES). Mortier and Gemma Vilagut cleaned and analyzed the data. Philippe Journal of Medical Internet Research, 6(3), e34. https://doi.org/10. Mortier, Gemma Vilagut, and Jordi Alonso drafted the initial version of 2196/jmir.6.3.e34 MORTIER ET AL. | 543 Franklin, J. C., Ribeiro, J. D., Fox, K. R., Bentley, K. H., Kleiman, E. M., Moutier Christine (2020). Suicide Prevention in the COVID‐19 Era. JAMA Huang, X., Musacchio, K. M., Jaroszewski, A. C., Chang, B. P., & Psychiatry. http://dx.doi.org/10.1001/jamapsychiatry.2020.3746 Nock, M. K. (2017). Risk factors for suicidal thoughts and behaviors: Muller, A. E., Hafstad, E. V., Himmels, J. P. W., Smedslund, G., Flottorp, S., A meta‐analysis of 50 years of research. Psychological Bulletin, Stensland, S. Ø., Stroobants, S., Van de Velde, S., & Vist, G. E. (2020). 143(2), 187–232. https://doi.org/10.1037/bul0000084 The mental health impact of the covid‐19 pandemic on healthcare Fridner, A., Belkic, K., Marini, M., Minucci, D., Pavan, L., & Schenck‐ workers, and interventions to help them: A rapid systematic review. Gustafsson, K. (2009). Survey on recent suicidal ideation among female Psychiatry Research, 293(August), 113441. https://doi.org/10.1016/j. university hospital physicians in Sweden and Italy (the HOUPE study): psychres.2020.113441 Cross‐sectional associations with work stressors. Gender Medicine, 6(1), Nock, M. K., Borges, G., & Ono, Y. (Eds.). (2012). Suicide: Global 314–328. https://doi.org/10.1016/j.genm.2009.04.006 perspectives from the WHO World Mental Health Surveys. Gunnell, D., Appleby, L., Arensman, E., Hawton, K., John, A., Kapur, N., … Cambridge University Press. Yip, P. S. (2020). Suicide risk and prevention during the COVID‐19 Nock, M. K., Stein, M. B., Heeringa, S. G., Ursano, R. J., Colpe, L. J., pandemic. The Lancet Psychiatry, 7(6), 468–471. https://doi.org/10. Fullerton, C. S., Hwang, I., Naifeh, J. A., Sampson, N. A., 1016/S2215-0366(20)30171-1 Schoenbaum, M., Zaslavsky, A. M., & Kessler, R. C. (2014). Prevalence Hastie, T., Tibshirani, R., & Friedman, J. (2009). Springer series in statistics. The and correlates of suicidal behavior among soldiers. JAMA Psychiatry, elements of statistical learning—Data mining, inference, and prediction. 71(5), 514. https://doi.org/10.1001/jamapsychiatry.2014.30 Springer. https://doi.org/10.1007/b94608 Pappa, S., Ntella, V., Giannakas, T., Giannakoulis, V. G., Papoutsi, E., & Hawton, K., Agerbo, E., Simkin, S., Platt, B., & Mellanby, R. J. (2011). Katsaounou, P. (2020). Prevalence of depression, anxiety, and insomnia Risk of suicide in medical and related occupational groups: A among healthcare workers during the COVID‐19 pandemic: A national study based on Danish case population‐based registers. systematic review and meta‐analysis. Brain, Behavior, and Immunity, Journal of Affective Disorders, 134(1–3), 320–326. https://doi.org/ 88(January), 901–907. https://doi.org/10.1016/j.bbi.2020.05.026 10.1016/j.jad.2011.05.044 Petrie, K., Crawford, J., Baker, S. T. E., Dean, K., Robinson, J., Veness, B. G., Hem, E., Grønvold, N., Aasland, O., & Ekeberg, Ø. (2000). The Randall, J., McGorry, P., Christensen, H., & Harvey, S. B. (2019). prevalence of suicidal ideation and suicidal attempts among Interventions to reduce symptoms of common mental disorders and Norwegian physicians. Results from a cross‐sectional survey of a suicidal ideation in physicians: A systematic review and meta‐ nationwide sample. European Psychiatry, 15(3), 183–189. https:// analysis. The Lancet Psychiatry, 6(3), 225–234. https://doi.org/10. doi.org/10.1016/S0924-9338(00)00227-3 1016/S2215-0366(18)30509-1 Kessler, R. C., & Üstün, T. B. (2004). The World Mental Health (WMH) Posner, K., Brown, G. K., Stanley, B., Brent, D. A., Yershova, K. V., Survey Initiative version of the World Health Organization (WHO) Oquendo, M. A., Currier, G. W., Melvin, G. A., Greenhill, L., Shen, S., & Composite International Diagnostic Interview (CIDI). International Mann, J. J. (2011). The Columbia—Suicide Severity Rating Scale: Initial Journal of Methods in Psychiatric Research, 13(2), 93–121. https://doi. validity and internal consistency findings from three multisite studies org/10.1002/mpr.168 with adolescents and adults. American Journal of Psychiatry, 168(12), Kisely, S., Warren, N., McMahon, L., Dalais, C., Henry, I., & Siskind, D. 1266–1277. https://doi.org/10.1176/appi.ajp.2011.10111704 (2020). Occurrence, prevention, and management of the Red Nacional de Vigilancia Epidemiólogica (RENAVE). Ministerio de Sanidad, psychological effects of emerging virus outbreaks on healthcare España. (2020). Retrieved from https://www.mscbs.gob.es/ workers: rapid review and meta‐analysis. BMJ, 369, m1642. https:// profesionales/saludPublica/ccayes/alertasActual/nCov-China/home.htm doi.org/10.1136/bmj.m1642 Ribeiro, J. D., Franklin, J. C., Fox, K. R., Bentley, K. H., Kleiman, E. M., Kovess‐Masfety, V., Boyd, A., Haro, J. M., Bruffaerts, R., Villagut, G., Chang, B. P., & Nock, M. K. (2016). Self‐injurious thoughts and Lépine, J. P., Gasquet, I., & Alonso, J. (2011). High and low suicidality behaviors as risk factors for future suicide ideation, attempts, and in Europe: A fine‐grained comparison of France and Spain within the death: A meta‐analysis of longitudinal studies. Psychological Medicine, ESEMeD surveys. Journal of Affective Disorders, 133(1–2), 247–256. 46(2), 225–236. https://doi.org/10.1017/S0033291715001804 https://doi.org/10.1016/j.jad.2011.04.014 Roser, M., Ritchie, H., Ortiz‐Ospina, E., & Hasell, J. (2020). Coronavirus Krysinska, K., & Martin, G. (2009). The struggle to prevent and evaluate: Pandemic (COVID‐19). Retrieved from https://ourworldindata.org/ Application of population attributable risk and preventive fraction coronavirus to suicide prevention research. Suicide and Life‐Threatening Behavior, Salazar de Pablo, G., Vaquerizo‐Serrano, J., Catalan, A., Arango, C., 39(5), 548–557. https://doi.org/10.1521/suli.2009.39.5.548 Moreno, C., Ferre, F., Shin, J. I., Sullivan, S., Brondino, N., Solmi, M., & Litz, B. T., Stein, N., Delaney, E., Lebowitz, L., Nash, W. P., Silva, C., & Fusar‐Poli, P. (2020). Impact of coronavirus syndromes on physical Maguen, S. (2009). Moral injury and moral repair in war veterans: A and mental health of health care workers: Systematic review and preliminary model and intervention strategy. Clinical Psychology meta‐analysis. Journal of Affective Disorders, 275, 48–57. https://doi. Review, 29(8), 695–706. https://doi.org/10.1016/j.cpr.2009.07.003 org/10.1016/j.jad.2020.06.022 Loas, G., Lefebvre, G., Rotsaert, M., & Englert, Y. (2018). Relationships Sharif, S., Amin, F., Hafiz, M., Benzel, E., Peev, N., & Dahlan, R. H., … World between anhedonia, suicidal ideation and suicide attempts in a large Spinal Column Society Executive Board. (2020). COVID 19— sample of physicians. PLOS One, 13(3), 1–23. https://doi.org/10. Depression and Neurosurgeons. World Neurosurgery, 21(1), 1–9. 1371/journal.pone.0193619 https://doi.org/10.1016/j.wneu.2020.06.007 MIND/COVID‐19: Mental health Impact and NeeDs associated with Thomson, S., Figueras, J., Evetovits, T., Jowett, M., Mladovsky, P., COVID‐19: A comprehensive national evaluation in Spain (COV20/ Maresso, A., … Kluge, H. (2013). Economic Crisis, Health Systems 00711). (2020). Retrieved from http://eu-isciii.es/covidfundinspain/ and Health in Europe Impact and implications for policy. European mind-covid-19-mental-health-impact-and-needs-associated-with- Observatory on Health Systems and Policies Series IV–16. covid-19-a-comprehensive-national-evaluation-in-spain/ Tyssen, R., Vaglum, P., Grønvold, N. T., & Ekeberg, Ø. (2001). Suicidal Miret, M., Caballero, F. F., Huerta‐Ramírez, R., Moneta, M. V., Olaya, B., ideation among medical students and young physicians: a Chatterji, S., Haro, J. M., & Ayuso‐Mateos, J. L. (2014). Factors nationwide and prospective study of prevalence and predictors. associated with suicidal ideation and attempts in Spain for different Journal of Affective Disorders, 64(1), 69–79. https://doi.org/10. age groups. Prevalence before and after the onset of the economic 1016/S0165-0327(00)00205-6 crisis. Journal of Affective Disorders, 163, 1–9. https://doi.org/10. van Buuren, S. (2012). Flexible imputation of missing data. Chapman & Hall. 1016/j.jad.2014.03.045 https://doi.org/10.1201/b11826 544 | MORTIER ET AL. van Heeringen, K., & Mann, J. J. (2014). The neurobiology of suicide. SU PP O RT I N G I N FO RM ATI ON The Lancet Psychiatry, 1(1), 63–72. https://doi.org/10.1016/ Additional Supporting Information may be found online in the S2215-0366(14)70220-2 supporting information tab for this article. Vieta, E., Pérez, V., & Arango, C. (2020). Psychiatry in the aftermath of COVID‐19. Revista de Psiquiatría y Salud Mental, 13(2), 105–110. https://doi.org/10.1016/j.rpsm.2020.04.004 Vindegaard, N., & Benros, M. E. (2020). COVID‐19 pandemic and mental How to cite this article: Mortier P, Vilagut G, Ferrer M, et al. health consequences: Systematic review of the current evidence. Brain, Behavior, and Immunity, 89, 531–542. https://doi.org/10.1016/ Thirty‐day suicidal thoughts and behaviors among hospital j.bbi.2020.05.048 workers during the first wave of the Spain COVID‐19 Wasserman, D., Iosue, M., Wuestefeld, A., & Carli, V. (2020). Adaptation outbreak. Depression Anxiety. 2020;38:528–544. of evidence‐based suicide prevention strategies during and after the https://doi.org/10.1002/da.23129 COVID‐19 pandemic. World Psychiatry, 19(3), 294–306. https://doi. org/10.1002/wps.20801