Safety Science 134 (2021) 105064 Contents lists available at ScienceDirect Safety Science journal homepage: www.elsevier.com/locate/safety Working conditions and health in Spain during the COVID-19 pandemic: Minding the gap Sergio Salas-Nicás a, Salvador Moncada b, Clara Llorens a, b, c, Albert Navarro a, d, * a Research Group on Psychosocial Risks, Organization of Work and Health (POWAH), Autonomous University of Barcelona (UAB), Cerdanyola del Vallès, Spain b Union Institute of Work, Environment and Health (ISTAS), Reference Centre on Work Organisation and Health, Barcelona, Spain c Sociology Department, Faculty of Sociology and Political Sciences, Autonomous University of Barcelona (UAB) Cerdanyola del Vallès, Spain d Biostatistics Unit, Faculty of Medicine, Autonomous University of Barcelona (UAB), Cerdanyola del Vallès, Spain A R T I C L E I N F O A B S T R A C T Keywords: Background: The emergence of the COVID-19 pandemic has had major consequences in the workplace, both in COVID-19 terms of the number of cases among the working population and the enormous changes made to cope with it. The Working conditions objective of this study is to describe the impact of COVID-19 on the working conditions and health of wage- Occupational health earners in Spain. Inequalties Spain Methods: Cross-sectional study carried out between the end of April and the end of May 2020 among the wage- earning population. Sample included n = 20,328 participants obtained through an online survey. Results: A situation of high-strain was reported by 44.3% of workers, 42.6% were concerned about possible job loss, 75.6% about finding a new job if they lost the present one, 69.7% were worried about salary reduction, 68% about becoming infected at work and 72.3% of being a transmitter of the virus. Among those who regularly went to work, 13.1% did so with symptoms compatible with COVID-19 and 71.2% stated that they had done so without adequate protection measures. 36.7% of workers believe that their health worsened, 41.6% had severe trouble sleeping during the last month, 55.1% were at risk of poor mental health and consumption more than doubled of tranquilisers and opioid analgesics, compared to the pre-pandemic situation. Conclusion: The impact of COVID-19 on the wage-earning population has been enormous, with high exposures to harmful working conditions and very poor health indicators, which, compared to the pre-pandemic situation, means significant deterioration. Important inequalities are observed according to class, gender, age and wage. 1. Introduction The world of employment has been seriously affected, due not only to numbers of cases among the working population, but also to effects To date, the COVID-19 pandemic has affected millions of people derived from the policies and measures decreed by governments to around the world with disastrous consequences for many of them. At the control the pandemic (e.g. “lockdown”, enforced teleworking…) as well time of writing this paper, in mid-September 2020, there were more as by effects deriving from labour management practices, which have than 670,000 confirmed cases in Spain, one of the most badly affected led to the deterioration of working conditions, albeit in a rather uneven countries, more than 4.8 million in Europe and over 30 million manner. worldwide. The COVID-19 pandemic therefore represents a tremendous chal­ Social inequalities in the prevalence of chronic diseases and social lenge for occupational health. Many occupations involve direct contact determinants of health (among them, employment) have greatly influ­ or physical proximity to other people and pose a higher risk of exposure enced the evolution of the pandemic. We may consider the inequalities and infection, while others offer the possibility of working from home, in infection and mortality rates due to COVID-19 as one result of these considerably reducing the risk of contagion. Healthcare professionals inequalities and of previously existing determinants (Bambra et al., are at increased risk of exposure due to their close contact with COVID- 2020). 19 patients. At the same time they themselves are a potential source of * Corresponding author at: Biostatistics Unit, Faculty of Medicine, Autonomous University of Barcelona (UAB), Avda. Can Domènech S/N, 08193 Cerdanyola del Vallès, Spain. E-mail address: albert.navarro@uab.cat (A. Navarro). https://doi.org/10.1016/j.ssci.2020.105064 Received 25 September 2020; Accepted 20 October 2020 Available online 26 October 2020 0925-7535/© 2020 Published by Elsevier Ltd. S. Salas-Nicás et al. Safety Science 134 (2021) 105064 transmission of the disease to colleagues, patients and relatives (Burdorf Table 1 et al., 2020). Survey questions used in this paper. The challenge involves not only higher qualified health personnel Topic Question Responses Source but also includes more vulnerable workers in manual and less qualified Organisational Since the state of alert Fired / Not renewed Ad-hoc occupations and the most precarious and lowest paid jobs (Sim, 2020). measures and was declared, in what / Working Time For people employed in such occupations, COVID-19 is experienced as a COVID sick situations have you Reduction / Contract synergistic, simultaneous pandemic (syndemic) that interacts with non- leave been? Suspension / communicable diseases and existing social conditions (Bambra et al., AND Supension and Today, what is your Reduction / Mainly 2020). situation? going to work to the At present we still do not know enough about mechanisms of the workpost / pandemic, and it will take some time to generate the information Combination of necessary to permit making decisions based on evidence (Burdorf et al., teleworking and going to work / 2020). Meanwhile, we must continue investigating and acting despite Mainly teleworking / the uncertainties. Publications in the specialised literature so far have COVID-19 sick leave largely been focused on healthcare personnel. This paper, on the other with positive test / hand, presents the main results of the survey “Working conditions, COVID-19 sick leave insecurity and health of workers resident in Spain in the context of with negative test / COVID-19 sick leave COVID-19′′ (COTS, for its acronym in Spanish), whose objective was to for contact with a highlight impacts of the COVID-19 pandemic on the working conditions person with positive and health of the wage-earning population in Spain. We hope that the test / COVID-19 sick results presented here contribute to understanding the effects of this leave for contact with a person with crisis and facilitate the search for measures to face it. negative test Salary covering How often does your Always / Many times ERP16 2. Material and methods basic needs current salary cover / Sometimes / Only the daily basic needs of once / Never 2.1. Study population your home? Job loss insecurity Are you worried about To a very large COPSOQ3 becoming extent / To a large The study population consists of all wage-earners residing in Spain unemployed? extent / Somewhat / who had a job on 14 March 2020 (the day the state of alert began), To a small extent / including people who were subsequently fired, or affected by a tempo­ To a very small extent rary lay-off procedure (hereafter referred to as “ERTE”, from the term Labor market Are you worried about To a very large COPSOQ3 used in spanish, “Expediente de Regulación Temporal de Empleo”). insecurity it being difficult for extent / To a large you to find another job extent / Somewhat / 2.2. Design, data collection and sample size if you became To a small extent / unemployed? To a very small extent Observational cross-sectional study. The data come from an online Wage insecurity Are you worried about To a very large COPSOQ3 survey conducted between 29 April and 28 May 2020 (between six and a decrease in your extent / To a large ten weeks after the state of alert was declared in Spain). The sample salary? extent / Somewhat / includes all questionnaires completed by wage-earners, once we had To a small extent / To a very small eliminated those corresponding to persons residing outside Spain (n = extent 21) and those with fewer than 20% of valid responses (n = 82). The final Worring about Are you worried about To a very large Ad-hoc sample consists of n = 20,328 participants. COVID19 the possibility of extent / To a large infection at becoming infected extent / Somewhat / 2.3. Variables work with COVID-19 at To a small extent / work? To a very small extent Various questions were prepared ad hoc for this investigation, others Worring about Are you worried about To a very large Ad-hoc were obtained from the third psychosocial risks survey carried out in spreading the possibility of extent / To a large Spain (ERP-2016)(ISTAS, 2019), from the survey on alcohol and other COVID19 infecting someone extent / Somewhat / with COVID-19? To a small extent / drugs in Spain (EDADES) of the National Drug Plan (Plan Nacional sobre To a very small Drogas, 2018), from the SF-36 (Vilagut et al., 2005) and items corre­ extent sponding to the short version of the Copenhagen Psychosocial Ques­ Working without Since the state of alert Always / Often / Ad-hoc tionnaire (CoPsoQ) (Burr et al., 2019) adapted and validated in Spain protection was declared, do you Sometimes / (COPSOQ-Istas21) (ISTAS, 2020). The variables recorded correspond to against think that you have Seldom/ Never COVID19 had to work without different blocks: a) sociodemographic, b) organisational measures, c) adequate protection working conditions, d) sick leave and sickness presenteeism, e) health measures to avoid and f) drug consumption. The survey questions used in this paper are contagion by COVID- shown in table 1. 19? Working with Since the state of alert No, never / Yes, few Ad-hoc COVID-19 was declared, have you days / Yes, some 2.4. Statistical analysis symptoms gone to work with days / Yes, most days symptoms (fever, / Yes, always Percentages and prevalences of the different items were estimated cough, shortness of with their respective 95% confidence intervals (95% CI). In order to breath or general malaise)? avoid biases in the estimates, we weighted by the distribution of gender General health How do you consider Better / Unaltered / Ad-hoc crossed with age and occupational class (the nine large groups of the your current state of Worse national classification of occupations –CNO-11-) taken from the survey (continued on next page) of the active population (EPA) for the first quarter of 2020, also 2 S. Salas-Nicás et al. Safety Science 134 (2021) 105064 Table 1 (continued ) 6.5% of the participants. The sensitivity analysis consisted of re- Topic Question Responses Source estimating the percentages or global prevalences of all the items of in­ terest, assuming that in reality this group represented only half, that is, health, compared to what you had before 3.25%. the state of alert was declared? 3. Results Sleep problems During the past 4 Always / Many times COPSOQ3 weeks, how often have / Sometimes / A few The characteristics of the COTS sample, unweighted and weighted, you slept badly and times / Never restlessly? and a comparison with EPA data is presented in Table 2. At risk of poor Q1: During the past 4 Always / Most of the SF-36 mental health weeks, how often have time / A good bit of 3.1. Workplace organisational measures taken to address the pandemic you been very the time / Some of nervous? the time / A little of Q2: During the past 4 the time / Never As Table 3 shows, 5.7% of workers lost their jobs during the first weeks, how often have weeks of the state of alert, 3.8% due to dismissal and 1.9% due to the you felt so depressed non-renewal of their contract. One out of four was affected by an ERTE, that nothing could mostly suspension of contract. Just over one in three (37.8%) continued cheer you up? Q3: During the past 4 going to work at their company or institution on a regular basis and weeks, how often have 30.1% carried out telework. you felt calm and peaceful? 3.2. Relevant working conditions during the pandemic Q4: During the past 4 weeks, how often have you felt sad or We may observe in Table 3 that 44.3% of workers suffered high unhappy? strain (high psychological job demands and low job control), a similar Q5: During the past 4 percentage (42.6%) was concerned about losing their job, three out of weeks, how often have four about finding another job if they were to lose the present one, about you been happy? Tranquilizers In the past thirty days, No / Yes, I was EDADES seven out of ten were concerned about the possibility of their wages have you used any taking before and being lowered, becoming infected at work, or transmitting the virus to tranquilizers/sedatives now I take the same others. Among those who regularly went to work at their company or or sleeping pills? dose / Yes, I was institution, 71.2% stated that they had done so at some time without Are included: Lexatín, taking before but orfidal, noctamid, now I have increased trankimazin, rohipnol, the dose or I have Table 2 tranxilium, diacepam, changed to a stronger Characteristics of COTS sample and comparision with EPA. valium, stilnox, one / Yes, I do not COTS Weighted COTS EPA zolpidem, hypnotics, usually consume benzos, benzodiazepine, them but in this % 95CI% % 95CI% % barbiturates, etc. period I have Are not included: Gender valerian, passionflower, Male 42.1 (41.4–42.8) 51.0 (50.0–52.0) 51.8 dormidina, melatonin Female 57.8 (57.1–58.4) 48.9 (47.9–49.9) 48.2 Other 0.1 (0.1–0.2) 0.1 (0.1–0.2) NC Painkillers In the past thirty days, No / Yes, I was EDADES Age (opioids) have you used taking before and 16–34 years old 10.4 (10.0–10.9) 25.6 (24.4–26.7) 26.0 painkillers (opioids)? now I take the same 35–49 years old 44.5 (43.8–45.2) 43.7 (42.7–44.7) 43.8 Are included: Adolonta, dose / Yes, I was 50 years old or more 45.0 (44.4–45.7) 30.7 (30.0–31.5) 30.3 Tramadol, Zaldiar, taking before but Occupational class Enanplus, Tradonal, now I have increased Non manual 56.7 (56.0–57.4) 46.1 (45.1–47.1) 45.0 Algidol, Dolocatil, the dose or I have Manual 43.3 (42.6–44.0) 53.9 (52.9–54.9) 55.0 Codeisan, MST changed to a stronger Occupational group continus, Durogesic, one / Yes, I do not (CNO11) Dolantina, Subuxone, usually consume Directors and managers 0.7 (0.6–0.8) 2.6 (2.1–3.3) 2.3 Oxicodona, Fentanilo, them but in this Scientific and intellectual 25.6 (25.0–26.3) 20.2 (19.5–20.9) 19.3 etc. period I have technicians and Are not included: professionals Paracetamol, Ibuprofen, Technicians and 13.3 (12.9–13.8) 11.3 (10.8–11.9) 11.3 Aspirin, Nolotil, etc. AssociateProfessionals Accounting, administrative 17.1 (16.6–17.6) 12.0 (11.5–12.5) 12.0 ERP16: III Spanish psychosocial risks survey (2016). and other office COPSOQ3: Copenhagen psychosocial questionnarie, third version. employees SF-36: 36-Item short form health survey questionnaire. Hotel and catering workers, 23.3 (22.7–23.9) 21.4 (20.7–22.2) 21.0 EDADES: Spanish survey on alcohol and other drugs. personnel services… Skilled workers in the 0.3 (0.3–0.4) 1.0 (0.7–1.3) 1.0 agricultural sector, respecting the distribution by autonomous communities (Spanish livestock… administrative regions). Skilled workers in the 5.1 (4.8–5.4) 9.8 (9.0–10.5) 10.4 manufacturing industry A sensitivity analysis was performed to examine the effect of a Plant and machinery 2.3 (2.1–2.6) 7.7 (6.8–8.6) 8.1 possible selection bias assuming that more responses would be obtained operators from healthcare workers than expected according to the population Elementary occupations 12.2 (11.8–12.7) 14.0 (13.4–14.7) 14.6 distribution. Unfortunately in Spain the exact number of active health­ Employment contract care professionals is not known, we only have certain data on the public relationship Permanent 80.8 (80.2–81.3) 78.2 (77.3–79.1) 75.0 sector or associations of doctors and nurses. Once the sample obtained Temporary 19.2 (18.7–19.8) 21.8 (20.9–22.7) 25.0 had been weighted, doctors, nurses and nursing assistants accounted for NC: Not considered in the survey. 3 S. Salas-Nicás et al. Safety Science 134 (2021) 105064 Table 3 Organisational measures, working conditions, COVID-19 sick leave and sickness presenteeism, global and by occupational class and salary. Occupational class Salary covering basic needs Global Non manual Manual Sufficent Insufficient % 95%CI % 95%CI % 95%CI % 95%CI % 95%CI Fired 3.8 (3.4–4.2) 2.4 (2.0–2.9) 5.1 (4.4–5.9) NA NA NA NA Not renewed 1.9 (1.7–2.2) 1.3 (1.0–1.6) 2.5 (2.2–3.0) NA NA NA (NA Working Time Reduction (ERTE *) 3.4 (3.0–3.7) 4.1 (3.7–4.6) 2.7 (2.3–3.3) 3.5 (3.1–4.0) 3.5 (3.0–4.2) Contract Suspension (ERTE *) 21.8 (21.0–22.7) 13.1 (12.2–14.0) 29.1 (27.8–30.5) 17.8 (16.9–18.8) 33.4 (31.7–35.2) Supension and Reduction (ERTE *) 1.0 (0.8–1.3) 0.8 (0.6–1.1) 1.2 (0.8–1.6) 1.1 (0.8–1.5) 0.9 (0.6–1.4) Going to work to the workpost 37.8 (36.8–38.8) 25.6 (24.6–26.8) 48.5 (47.0–50.1) 39.1 (37.9–40.3) 40.0 (38.3–41.8) Combination of teleworking and going to work 4.9 (4.5–5.4) 8.4 (7.8–9.2) 2.0 (1.5–2.6) 6.3 (5.7–6.9) 2.8 (2.3–3.5) Mainly teleworking 25.2 (24.4–26.0) 49.9 (48.6–51.2) 4.3 (3.7–4.9) 31.9 (30.8–33.0) 14.4 (13.3–15.6) High strain 44.3 (43.1–45.4) 37.2 (35.9–38.5) 51.0 (49.2–52.8) 39.5 (38.2–40.8) 54.8 (52.7–56.9) Job loss insecurity 42.6 (41.5–43.6) 37.1 (35.8–38.4) 47.4 (45.8–49.1) 37.0 (35.7–38.2) 54.4 (52.5–56.2) Labor market insecurity 75.6 (74.7–76.4) 69.8 (68.7–71.0) 80.8 (79.6–82.1) 71.4 (70.3–72.4) 84.5 (83.0–85.8) Wage insecurity 69.7 (68.7–70.6) 64.4 (63.0–65.7) 74.2 (72.8–75.6) 65.6 (64.4–66.8) 78.3 (76.7–79.8) Worring about COVID19 infection at work 68.0 (67.0–69.0) 62.2 (60.9–63.6) 73.0 (71.6–74.5) 64.9 (63.7–66.1) 74.6 (72.9–76.2) Worring about spreading COVID19 72.3 (71.4–73.3) 69.8 (68.6–71.1) 74.6 (73.2–75.9) 71.2 (70.0–72.3) 76.6 (74.9–78.1) Working without protection against COVID19 $ 71.2 (69.7–72.7) 68.4 (66.0–70.6) 72.9 (70.9–74.8) 68.7 (66.8–70.5) 76.2 (73.6–78.6) COVID-19 sick leave (case or contact) # 6.1 (5.6–6.6) 4.7 (4.2–5.3) 7.3 (6.5–8.1) 5.6 (5.1–6.2) 7.5 (6.5–8.6) COVID-19 sick leave (case) # 3.6 (3.3–4.0) 2.8 (2.5–3.2) 4.3 (3.8–4.9) 3.2 (2.8–3.6) 4.6 (4.0–5.3) Working with COVID-19 symptoms $ 13.1 (12.1–14.2) 12.7 (11.2–14.4) 13.3 (11.9–14.7) 10.6 (9.5–11.7) 18.2 (16.0–20.6) NA Don’t apply. People unemployed at the time of the survey were not asked this question. * Temporary lay-off proceedings. In Spanish, Expediente de Regulación Temporal de Empleo (ERTE). $ Estimated among those who went to work to his/her company or institution. # Confirmed or suspected. adequate protection measures against COVID-19. that their health status worsened during the pandemic, two in five (41.6%) often or almost always slept poorly during the last four weeks, and more than half (55.1%) were classified as at risk of poor mental 3.3. Sick leave and sickness presenteeism health. Tranquilizer use went from 9.5% pre-pandemic to 21.5%, and one in three workers who were already consuming either increased the Still looking at Table 3, 6.1% of workers had at least one sick leave dose or switched to a stronger drug. The consumption of opioid anal­ episode related to COVID-19, of which 3.6% were due to being a case gesics doubled, from 8.9% to 18.6%. (suspected or confirmed), while 13.1% of workers who went to work regularly during the state of alert, reported having done so at some time with symptoms compatible with COVID-19. 3.5. Working conditions and health-related inequalities 3.4. Health and consumption of selected pharmaceutical drugs Results by occupational class and salary are shown in Tables 3 and 5. Regarding organisational measures, Table 3, it appears that occupa­ Table 5 shows that over one in three (36.7%) wage earners affirm tional class discriminates better than just salary, so we see for example Table 4 Organisational measures, working conditions, COVID-19 sick leave and sickness presenteeism, by gender and age. Gender * Age Male Female 16–34 years old 35–49 years old 50 years old or more % 95%CI % 95%CI % 95%CI % 95%CI % 95%CI Fired 4.3 (3.6–5.1) 3.3 (2.9–3.7) 6.6 (5.3–8.2) 3.3 (2.8–3.8) 2.2 (1.8–2.6) Not renewed 1.8 (1.4–2.2) 2.1 (1.8–2.4) 3.4 (2.7–4.3) 1.7 (1.4–2.0) 1.0 (0.8–1.3) Working Time Reduction (ERTE #) 4.0 (3.4–4.6) 2.8 (2.4–3.2) 3.3 (2.5–4.3) 3.9 (3.5–4.4) 2.7 (2.3–3.1) Contract Suspension (ERTE #) 23.5 (22.1–24.9) 20.1 (19.2–21.0) 26.9 (24.5–29.5) 21.8 (20.7–22.8) 17.6 (16.6–18.6) Supension and Reduction (ERTE #) 1.3 (0.9–1.9) 0.6 (0.5–0.8) 1.7 (1.1–2.8) 0.9 (0.6–1.2) 0.5 (0.4–0.7) Going to work to the workpost 41.7 (40.1–43.3) 33.8 (32.7–34.9) 35.7 (33.0–38.6) 38.5 (37.2–39.8) 38.5 (37.3–39.8) Combination of teleworking and going to work 4.8 (4.1–5.6) 5.0 (4.5–5.5) 4.1 (3.0–5.6) 5.1 (4.6–5.7) 5.3 (4.8–5.8) Mainly teleworking 20.8 (19.6–22.1) 29.7 (28.7–30.7) 22.4 (20.2–24.8) 26.4 (25.3–27.5) 25.8 (24.8–26.8) High strain 40.8 (39.0–42.6) 47.8 (46.5–49.0) 42.9 (39.6–46.2) 46.1 (44.6–47.5) 42.7 (41.3–44.1) Job loss insecurity 44.1 (42.4–45.8) 40.9 (39.7–42.1) 52.2 (49.1–55.3) 44.1 (42.7–45.4) 32.7 (31.5–34.0) Labor market insecurity 78.3 (77.0–79.6) 72.8 (71.7–73.8) 80.5 (78.1–82.8) 78.1 (76.9–79.2) 68.2 (67.0–69.3) Wage insecurity 70.4 (68.8–72.0) 68.9 (67.8–70.0) 70.4 (67.5–73.2) 70.0 (68.7–71.1) 68.7 (67.5–69.9) Worring about COVID19 infection at work 66.6 (65.0–68.2) 69.5 (68.4–70.5) 66.5 (63.5–69.4) 69.1 (67.9–70.3) 67.7 (66.5–68.9) Worring about spreading COVID19 71.8 (70.3–73.3) 73.0 (71.9–74.0) 73.6 (70.9–76.1) 73.5 (72.3–74.6) 69.7 (68.5–70.9) Working without protection against COVID19 $ 70.1 (67.8–72.3) 72.6 (70.7–74.3) 74.9 (70.4–78.8) 72.3 (70.4–74.2) 66.7 (64.6–68.8) COVID-19 sick leave (case or contact) & 5.1 (4.5–5.8) 7.1 (6.5–7.8) 6.1 (5.0–7.5) 5.7 (5.1–6.4) 6.6 (6.0–7.2) COVID-19 sick leave (case) & 2.9 (2.4–3.4) 4.4 (3.9–4.9) 3.1 (2.4–4.0) 3.6 (3.1–4.1) 4.1 (3.7–4.6) Working with COVID-19 symptoms $ 9.9 (8.5–11.5) 17.2 (15.8–18.8) 14.0 (11.2–17.4) 12.9 (11.5–14.4) 12.5 (11.2–13.9) * Gender “Other” was not considered due to its low sample size. # Temporary lay-off proceedings. In Spanish, Expediente de Regulación Temporal de Empleo (ERTE). $ Estimated among those who went to work to his/her company or institution. & Confirmed or suspected. 4 S. Salas-Nicás et al. Safety Science 134 (2021) 105064 Table 5 Health and selected medical drugs consumption, global and by occupational class and salary. Occupational class Salary covering basic needs Global Non manual Manual Sufficent Insufficient % 95%CI % 95%CI % 95%CI % 95%CI % 95%CI Positive change in general health 5.2 (4.7–5.7) 5.6 (5.0–6.3) 4.8 (4.1–5.6) 5.0 (4.5–5.6) 5.2 (4.3–6.3) Negative change in general health 36.7 (35.7–37.6) 37.5 (36.2–38.8) 36.2 (34.7–37.6) 33.6 (32.5–34.9) 43.4 (41.6–45.2) Sleep problems: Sometimes 30.6 (29.6–31.5) 30.8 (29.6–32.0) 30.4 (28.9–31.8) 31.2 (30.0–32.3) 29.5 (27.7–31.3) Sleep problems: Many times / Always 41.6 (40.7–42.6) 40.9 (39.6–42.2) 42.5 (41.0–44.0) 36.9 (35.7–38.1) 51.0 (49.2–52.9) At risk of poor mental health * 55.1 (54.0–56.1) 52.8 (51.5–54.1) 57.1 (55.4–58.7) 48.7 (47.4–50.0) 67.2 (65.3–69.0) Tranquilizers: same dose 6.1 (5.7–6.5) 5.6 (5.3–6.2) 6.2 (5.6–6.8) 5.7 (5.3–6.2) 6.8 (6.1–7.6) Tranquilizers: increased dose 3.4 (3.1–3.8) 3.3 (2.9–3.7) 3.5 (3.1–4.0) 2.7 (2.4–3.1) 5.0 (4.4–5.6) Tranquilizers: new consumer 12.0 (11.4–12.6) 11.6 (10.8–12.4) 12.5 (11.6–13.5) 10.5 (9.8–11.2) 14.8 (13.6–16.0) Painkillers (opioids): same dose 6.6 (6.2–7.1) 5.3 (4.9–5.9) 7.7 (7.0–8.4) 5.8 (5.3–6.3) 8.6 (7.7–9.5) Painkillers (opioids): increased dose 2.3 (2.0–2.6) 1.8 (1.5–2.2) 2.7 (2.3–3.1) 1.7 (1.4–2.0) 3.6 (3.0–4.2) Painkillers (opioids): new consumer 9.7 (9.1–10.3) 8.0 (7.3–8.7) 11.1 (10.3–12.0) 8.3 (7.7–9.0) 12.4 (11.3–13.5) * To define “at risk of poor mental health” we used the SF-36 cut-off ≤ 56. that more manual workers lost their job than non-manual (7.6% vs. inequalities in terms of organisational measures, working conditions and 3.7%), more were affected by ERTEs of contract suspension (30.3% vs. sick leave and presenteeism, although smaller than those related to 13.9%), many more had to regularly go to work in person (48.5% vs. occupational class or salary (Table 4). The differences according to 25.6%) and they teleworked much less (6.3% vs. 58.3%). On the other gender are observed above all in health (Table 6): almost 10% more hand, when we look at working conditions, it seems that the differences women than men report their health worsened, 16.1% more women according to salary or occupational class are very similar: the most often or almost always have sleep problems, and 17% more women are harmful exposures occur systematically in manual workers and among at risk of poor mental health. Consumption of tranquilizers is much those whose wages never or almost never cover their basic needs. For higher in women (15.4% vs. 8.8% of new consumers and 28.4% vs. example, differences for manual/non-manual and sufficient/insufficient 15.8% of all consumers), and the same occurs with the consumption of salary are approximately 14% for the case of high job stress, and over opioids (22.2% consumers among women, 15.2% among men). 10% for the three items of job insecurity (job loss, find another job, Regarding age, the loss of employment among young people stands out wages). In addition, manual and low-wage workers more often took sick (10% for those under 35, compared to 5% for those between 35 and 49 leave due to COVID-19. On the other hand, while occupational class years old and 3.2% among those over 50), young people are also the does not discriminate in terms of going to work with symptoms, salary most affected by ERTEs of contract suspension and, in accordance with evidently does so (18.2% among those with insufficient wages vs. 10.6% the above, those most exposed to high insecurity about losing their job among those with sufficient wages). Table 5 presents results stratified in (more than half of those aged 16–34 years, almost 20% more than those terms of health and drug consumption. In this case, except for mental over 50), (Table 4). Regarding health, Table 6, the differences are health and consumption of opioid analgesics, where the situation is moderate, generally those over 50 being in a better situation. Regarding worse for manual workers, occupational class does not seem to be drug consumption, although older participants consume the most, the associated; salary however is: those with insufficient salary present relative increase in numbers of young people consuming during the almost 10% more deterioration in general health (43.4% vs 33.6%), pandemic is clearly superior, especially in the case of tranquilizers: the 14.1% more problems sleeping often or almost always (51% vs 36.9%), number of consumers aged 16–34 years is 3.5 times higher than the pre- 18.5% a greater risk of poor mental health (67.2% vs 48.7%) and their pandemic figure, while among those aged 35–49 years it has multiplied consumption of tranquilisers and opioid analgesics is consistently higher by 2.4 and among those over 50 year by 1.8. than those with a sufficient salary, both among new consumers (14.8% in tranquilizers and 12.4% in analgesics vs 10.5% and 8.3% respec­ tively) and among all consumers (26.6% and 24.6% for tranquilizers and 3.6. Sensitivity analysis analgesics, vs 18.9% and 15.8%). Tables 4 and 6 show the results by gender and age. There are gender If doctors, nurses and nursing assistants represented half of what they represent in our sample (that is, 3.25% instead of 6.5%), the point Table 6 Health and selected medical drugs consumption by gender and age. Gender * Age Male Female 16–34 years old 35–49 years old 50 years old or more % 95%CI % 95%CI % 95%CI % 95%CI % 95%CI Positive change in general health 4.9 (4.2–5.7) 5.5 (5.0–6.1) 7.4 (6.0–9.0) 4.8 (4.2–5.4) 3.9 (3.4–4.3) Negative change in general health 31.9 (30.4–33.5) 41.6 (40.5–42.7) 37.9 (35.1–40.7) 38.6 (37.3–39.8) 32.9 (31.7–34.0) Sleep problems: Sometimes 32.1 (30.6–33.7) 28.9 (27.9–30.0) 30.5 (27.9–33.2) 30.5 (29.4–31.8) 30.6 (29.4–31.8) Sleep problems: Many times / Always 33.8 (32.2–35.4) 49.9 (48.7–51.0) 44.8 (41.9–47.7) 42.2 (40.9–43.5) 38.2 (37.0–39.4) At risk of poor mental health # 46.8 (45.1–48.5) 63.8 (62.7–64.9) 58.0 (55.0–61.0) 56.8 (55.5–58.1) 49.7 (48.4–51.0) Tranquilizers: same dose 4.8 (4.3–5.4) 7.3 (6.8–7.9) 2.6 (1.9–3.5) 5.5 (4.9–6.1) 9.8 (9.1–10.6) Tranquilizers: increased dose 2.2 (1.8–2.7) 4.7 (4.3–5.1) 2.1 (1.5–3.0) 3.6 (3.2–4.1) 4.2 (3.7–4.7) Tranquilizers: new consumer 8.8 (8.0–9.7) 15.4 (14.5–16.2) 11.9 (10.4–13.7) 12.7 (11.9–13.6) 11.2 (10.4–12.0) Painkillers (opioids): same dose 5.3 (4.8–6.0) 8.0 (7.4–8.6) 4.4 (3.5–5.5) 6.8 (6.3–7.5) 8.2 (7.5–9.0) Painkillers (opioids): increased dose 1.4 (1.0–1.8) 3.3 (2.9–3.7) 1.7 (1.1–2.4) 2.5 (2.2–3.0) 2.5 (2.1–2.9) Painkillers (opioids): new consumer 8.5 (7.7–9.4) 10.9 (10.3–11.7) 9.4 (8.0–11.0) 10.4 (9.7–11.2) 8.9 (8.2–9.7) * Gender “Other” was not considered due to its low sample size. # To define “at risk of poor mental health” we used the SF-36 cut-off ≤ 56. 5 S. Salas-Nicás et al. Safety Science 134 (2021) 105064 estimates obtained would vary, at most, by 1.8% for the percentage of to a stronger drug; on the other hand, in the ERP-2016 it was estimated workers who consider that they had to work without adequate protec­ that the percentage of subjects who reported often or almost always tion measures (it would be 69.4% instead of 71.2%), followed by a sleeping badly in the last month was 12.3% (Salas-Nicás et al., 2020) and decrease of 1.1% of workers who would have gone to work regularly those who were at risk of poor mental health 23.7% (Salas-Nicás et al., during the state of alert (dropping from 37.8% to 36.7%) and an increase 2018). This means that workers with very frequent sleep problems have of 0.9% of employees worried about finding another job if they were to almost quadrupled compared to 2016 and those at risk of poor mental lose the present one (from 75.6% to 76.5%). Estimates related to health health have more than doubled. This worsening in poor mental health in would vary by at most 0.5%, specifically for those who report a wors­ the context of the COVID-19 pandemic is due to multiple causes. Among ened health status, the estimate would fall from 36.7% to 36.2%. Of the work-related causes will surely be the increase in exposure to high strain total of 32 estimates made, 27 would vary by under 0.5%. (Harvey et al., 2017), uncertainty related to job insecurity and financial concern (Wilson et al., 2020), fear of becoming infected, etc., but also 4. Discussion the increase in social isolation and loneliness (Holmes et al., 2020). In any case, it is evident that if this situation continues over time, the The effects of the COVID-19 pandemic on all spheres of life have been damage to the health of workers may be very significant. enormous, and that of occupation has been no exception. Our study Although for the set of indicators studied the level of harmfulness is estimates that approximately a third of Spanish employees continued to generally high, there are also considerable inequalities: manual workers, attend their company or institution regularly, some lost their jobs, others women, young people and those with lower wages are undoubtedly were affected by temporary lay-offs, others had to telework from home, worse off. As in the 2008 crisis (Utzet et al., 2015), manual workers, etc. It is obvious that this abrupt break with the occupational routine can occupations less prepared for teleworking and many “front-line” have important consequences on working conditions and the health of workers face the high demands deriving from staff cuts, from short­ workers. comings in technologies and processes aggravated by the pandemic, or The unexpected emergence of the pandemic surprised those are unable to influence their daily work situation due to the authori­ responsible for Public Administrations, Social Agents and even most tarian or neo-Taylorist forms of work-organisation that characterise researchers, something which, together with initial doubts regarding the Spanish labour management practices in this segment of workers (Llo­ severity of the disease and how it is transmitted within workplace set­ rens et al., 2019). tings or to the usefulness of personal protective equipment (Semple and Young people, already with the highest levels of precariousness and Cherrie, 2020), could explain to a large extent why almost three out of suffering the attendant job insecurity and job destruction, are probably four workers who continued going to their workplace consider that they the main victims of the pandemic from an occupational point of view, had done so without adequate protection measures. since they run the risk of their entire working life being affected by the Furthermore, it appears that working conditions worsened notably. current situation, becoming “the lockdown generation” (ILO, 2020). In Thus, based on the ERP-2016 we may estimate, using exactly the same addition to the usual unfavorable situation of women, the pandemic has methodology, certain figures which permit a comparison with obser­ a high feminization of many of the so-called ”front-line“ occupations vations made during the pandemic. For example, in 2016 the percentage (Wenham et al., 2020), which have been more exposed and have suf­ of workers in a situation of high strain (according to the con­ fered more intensely from the effect of working in very stressful condi­ ceptualisation of the “demand-control” model) (Karasek et al., 1981) tions, with high quantitative demands and with a higher probability of was 22.3%, 58.1% were concerned about losing their jobs, 68.4% about being infected. Salary is also revealed as a powerful axis of inequality, not finding another job if the present one were to be lost, or 63% about being strongly associated with most of the items studied, for example, their salary being reduced (unpublished data). Therefore, during the going to work with symptoms compatible with COVID-19, which surely pandemic, the number of workers in conditions of high strain (high indicates that for many people staying home is not an option, for eco­ demands but low control) has doubled, while in the case of insecurity nomic reasons. Moreover, other studies have shown the vulnerability of there are also increases, although not so pronounced, of around 6–7% in other occupational groups, for example migrant workers, many of them terms of fear about not finding another job or about a decrease in wages, employed in low-skilled occupations and with poor housing conditions which were already very high in 2016, the result of a context of high (Koh, 2020). insecurity that can almost be considered a characteristic feature of the This study has certain limitations, mainly related to data collection Spanish labour market (Sora et al., 2014). In contrast, there is a decrease procedures. It is obvious that in the face of such an unexpected situation in concern over job loss of more than 15%. As observed in 2010 in the as the COVID-19 pandemic, many studies have been done opportunis­ context of the financial crisis (Utzet et al., 2014), part of this decrease tically without detailed planning. In our case, the information was ob­ could be due to the fact that part of the most vulnerable group (young tained through an online survey, so that representativeness of the people and temporary workers) would have already been expelled from sample cannot be ensured, possibly suffering a selection bias, with those the labour market when the survey began, a few weeks after the state of most affected by the pandemic being more likely to participate. In this alert was declared, leaving the “survivors”, typically workers with more sense, the authors took certain measures: first, the sampling distribution stable jobs. Another part could be due to the application of temporary of sex, age and occupational groups was weighted according to the lay-off procedures by the administration and the requirement for com­ population structure of the reference survey in Spain (EPA), as well as by panies adopting this measure of not firing anyone during the six months territorial distribution (autonomous community); second, a sensitivity following its termination (requirement that was relaxed for ERTEs analysis was carried out considering that the response rate of health implemented after June 30 and which does not affect our sample). In personnel (presumably more motivated to respond due to their central addition to “classic” job insecurity regarding employment or its funda­ role in the pandemic) was half that actually obtained. mental conditions, a majority (seven out of ten workers), express un­ This analysis detected irrelevant differences showing that, if it really certainty and concern related to the possibility of becoming infected in existed, the selection bias would not act in this sense; third, of the results the workplace or being themselves a transmission vector. obtained here several are consistent with those from other sources: for The worsening of workers’ health is considerable: more than a third example, at the end of April it is estimated that the total number of of the salaried population affirms that their general health status workers in a situation of ERTE stood at 3.4 million, approximately 27% worsened during the pandemic (only 5.2% stated that it improved); the of all employees, very similar to our figure of 26.2%; or those of the first use of tranquilizers and opioid analgesics more than doubled compared published results of a Eurofound survey (Eurofound, 2020) in which it is to the pre-pandemic situation and, in addition, a very significant per­ stated that in Spain 30.7% of workers carried out telework, almost centage of those who were already users increased the dose or switched identical to our figure of 30.1%. 6 S. Salas-Nicás et al. Safety Science 134 (2021) 105064 In contrast, a main strength of this study is the sample size, more than Harvey, S.B., Modini, M., Joyce, S., Milligan-Saville, J.S., Tan, L., Mykletun, A., Bryant, R.A., Christensen, H., Mitchell, P.B., 2017. Can work make you mentally ill? twenty thousand responses with representation of all possible profiles of A systematic meta-review of work-related risk factors for common mental health workers according to gender, age, occupational group or territory. As a problems. Occup. Environ. Med. 74, 301–310. https://doi.org/10.1136/oemed- comparison, the Eurofound survey cited in the previous paragraph 2016-104015. (Eurofound, 2020) produced its first wave report using a total of n = Holmes, E.A., O’Connor, R.C., Perry, V.H., Tracey, I., Wessely, S., Arseneault, L., Ballard, C., Christensen, H., Cohen Silver, R., Everall, I., Ford, T., John, A., Kabir, T., 3064 Spanish participants (Ahrendt et al., 2020). Our survey provides King, K., Madan, I., Michie, S., Przybylski, A.K., Shafran, R., Sweeney, A., unprecedented and highly relevant information regarding the working Worthman, C.M., Yardley, L., Cowan, K., Cope, C., Hotopf, M., Bullmore, E., 2020. conditions and health of workers resident in Spain at an absolutely Multidisciplinary research priorities for the COVID-19 pandemic: a call for action for mental health science. The Lancet Psychiatry 7, 547–560. https://doi.org/10.1016/ exceptional moment, with the use of some previously used and validated S2215-0366(20)30168-1. questionnaires that allow a consistent comparison of the current situa­ ILO, 2020. ILO Monitor: COVID-19 and the world of work, Fourth edition. International tion compared to the recent past. Labour Organization, Geneva. ISTAS, 2020. Manual del método COPSOQ-Istas21 (versión 3) para la evaluación y la prevención de los riesgos psicosociales en el trabajo. Instituto Sindical de Trabajo, 5. Conclusions Ambiente y Salud, Barcelona. ISTAS, 2019. Exposiciones psicosociales laborales en España: una mirada de clase y de género. Instituto Sindical de Trabajo, Ambiente y Salud, Barcelona. The results presented here paint a very worrying picture of the sit­ Karasek, R., Baker, D., Marxer, F., Ahlbom, A., Theorell, T., 1981. Job decision latitude, uation of working conditions and the state of health of workers residing job demands, and cardiovascular disease: A prospective study of Swedish men. Am. in Spain, during the most intense phase of the pandemic. In general, we J. Public Health 71, 694–705. https://doi.org/10.2105/AJPH.71.7.694. Koh, D., 2020. Migrant workers and COVID-19. Occup. Environ. Med. 77, 634–636. observe harmful working conditions and health indicators with very https://doi.org/10.1136/oemed-2020-106626. important deterioration compared to available references. And if the Llorens, C., Navarro, A., Salas, S., Utzet, M., Moncada, S., 2019. For better or for worse? overall situation seems bad, in certain groups it is even worse. Thus, the Psychosocial work environment and direct participation practices. Saf. Sci. 116, COVID-19 pandemic may have exacerbated inequalities in working 78–85. https://doi.org/10.1016/j.ssci.2019.02.028. Plan Nacional sobre Drogas, 2018. Encuesta sobre alcohol y drogas en España (EDADES). conditions and health in terms of class, gender, or age. In the future, Ministerio de Sanidad, Consumo y Bienestar Social del Gobierno de España, Madrid. once the pandemic has been overcome, the worsening of these condi­ Salas-Nicás, S., Moncada, S., Llorens, C., Navarro, A., 2018. Cognitive and Affective tions should be reversed and the extent to which it returns to the pre­ Insecurity Related to Remaining Employed and Working Conditions: Their Associations with Mental and General Health. J. Occup. Environ. Med. 60, vious situation should be evaluated, since it would be very disturbing if e589–e594. https://doi.org/10.1097/JOM.0000000000001448. both the working conditions and health of workers were to remain Salas-Nicás, S., Sembajwe, G., Navarro, A., Moncada, S., Llorens, C., Buxton, O., 2020. significantly worse than before the pandemic. Job insecurity, economic hardship, and sleep problems in a national sample of salaried workers in Spain. Sleep Heal. 6, 262–269. Semple, S., Cherrie, J.W., 2020. Covid-19: Protecting Worker Health. Ann. Work Expo. Funding Heal. 64, 461–464. https://doi.org/10.1093/annweh/wxaa033. Sim, M.R., 2020. The COVID-19 pandemic: Major risks to healthcare and other workers on the front line. Occup. Environ. Med. 77, 281–282. https://doi.org/10.1136/ This research did not receive any specific grant from funding oemed-2020-106567. agencies in the public, commercial, or not-for-profit sectors. Sora, B., Caballer, A., Peiró, J.M., 2014. La inseguridad laboral y sus consecuencias en un contexto de crisis económica. Papeles del Psicol. 35, 15–21. Utzet, M., Moncada, S., Molinero, E., Llorens, C., Moreno, N., Navarro, A., 2014. The References changing patterns of psychosocial exposures at work in the South of Europe: Spain as a labor market laboratory. Am. J. Ind. Med. 57, 1032–1042. https://doi.org/ Ahrendt, D., Sandor, E., Leončikas, T., Wilkens, M., Mascherini, M., 2020. Living, 10.1002/ajim.22334. working and COVID-19: Methodological note of Wave 1. European Foundation for Utzet, M., Navarro, A., Llorens, C., Moncada, S., 2015. Intensification and isolation: the Improvement of Living and Working Conditions, Dublin. Psychosocial work environment changes in Spain 2005–10. Occup. Med. (Lond) 65, Bambra, C., Riordan, R., Ford, J., Matthews, F., 2020. The COVID-19 pandemic and 405–412. https://doi.org/10.1093/occmed/kqv062. health inequalities. J. Epidemiol. Community Health. 74, 964–968. https://doi.org/ Vilagut, G., Ferrer, M., Rajmil, L., Rebollo, P., Permanyer-Miralda, G., Quintana, J.M., 10.1136/jech-2020-214401. Santed, R., Valderas, J.M., Domingo-Salvany, A., Alonso, J., 2005. El Cuestionario de Burdorf, A., Porru, F., Rugulies, R., 2020. The COVID-19 (Coronavirus) pandemic: Salud SF-36 español: una década de experiencia y nuevos desarrollos. Gac. Sanit. 19, Consequences for occupational health. Scand. J. Work. Environ. Heal. 46, 229–230. 135–150. https://doi.org/10.5271/sjweh.3893. Wenham, C., Smith, J., Morgan, R., 2020. COVID-19: the gendered impacts of the Burr, H., Berthelsen, H., Moncada, S., Nübling, M., Dupret, E., Demiral, Y., Oudyk, J., outbreak. Lancet 395, 846–848. https://doi.org/10.1016/S0140-6736(20)30526-2. Kristensen, T.S., Llorens, C., Navarro, A., Lincke, H.J., Bocéréan, C., Sahan, C., Wilson, J., Lee, J., Fitzgerald, H., Oosterhoff, B., Sevi, B., Shook, N., 2020. Job Insecurity Smith, P., Pohrt, A., 2019. The Third Version of the Copenhagen Psychosocial and Financial Concern during the COVID-19 Pandemic are Associated with Worse Questionnaire. Saf. Health Work 10, 482–503. https://doi.org/10.1016/j. Mental Health. J. Occup. Environ. Med. 62, 686–691. https://doi.org/10.1097/ shaw.2019.10.002. JOM.0000000000001962. Eurofound, 2020. Living, working and COVID-19: First findings. European Foundation for the Improvement of Living and Working Conditions, Dublin. 7