The impact of lockdown and isolation and adherence to precautionary measures

Many people have registered grave concerns about the impact of lockdown, quarantine and isolation on families and adherence of the public to the regulations whether or not people had contracted COVID-19. We include in the series three papers based on differing methodologies to draw out some potentially positive and challenging aspects of the ways in which the pandemic is managed.

First, Stallard et al report a mixed-methods study in the first wave of 385 caregivers, mainly parents, in two European countries, Portugal (n = 185) and the UK (n = 200) of whom 70% were working exclusively from home.^{Reference Stallard, Pereira and Barros5} ‘Almost half reported a reduction in income, [and] most children were home taught’.^{Reference Stallard, Pereira and Barros5} Importantly, although the cross-sectional sampling methods reported in this paper impose limitations, a single qualitative question enabled the authors to identify that nearly 90% of participants reported elements of post-traumatic growth despite their experiencing considerable adversity.

Second, Drury et al explore the literature on the role of group processes in the COVID-19 pandemic in three domains: recognition of threat; adherence by the public to the required public health behaviours; and the factors that increased such adherence during lockdown.^{Reference Drury, Carter, Ntontis and Guven6} They draw upon accumulated research on public behaviour in emergencies and the COVID-19 pandemic to show that social identity processes make better sense of the patterns of evidence than alternatives. They offer recommendations on facilitating public responses to COVID-19 by harnessing group processes.

Third, a short report by Bartrés-Faz et al documents a decrease in loneliness in 2 weeks of confinement in participants in Spain compared with their scores prior to the pandemic.^{Reference Bartrés-Faz, Macià, Cattaneo, Borràs, Tarrero and Solana7} Their research involved a continuing longitudinal subsample of 1604 people. It is noticeable that participants had a mean age of 55.7 years, 75% had completed higher education, and loneliness was higher before and during COVID-19 for people confined alone.

Other research suggests that we have witnessed a broad range of people's reactions to, mainly, the strictures on people's lives and employment that have evidenced a socioeconomic trajectory of greater problems experienced by less advantaged people. The UCL COVID-19 Social Study has followed 70 000 people in the UK since March 2020.⁸ That survey's researchers say, ‘The increased adherence to self-isolation rules among those with a higher household income suggests that many of those not isolating are breaking guidelines due to financial concerns …’.⁸ They have fewer choices, are more likely to have lost their jobs and have greater problems isolating in cramped housing and in coping with home schooling when they cannot fund the technology required. Fancourt et al, report elsewhere on their COVID-19 Social Study in the period March to August 2020.^{Reference Fancourt, Steptoe and Bu9} They include three follow-up measures completed by 36 520 participants and observed a decline in levels of anxiety and depression across the first 20 weeks after the introduction of lockdown in England. This finding emphasises the importance of longitudinal research.

Importantly, Fancourt & Bradbury's social study has found that people's experiences of the pandemic were dependent on their life situations prior to the lockdown.^{Reference Fancourt and Bradbury10} These findings well illustrate one of the defining features of secondary stressors.^{Reference Williams, Ntontis, Alfadhli, Drury and Amlôt11} Ethnic minorities, people who are more deprived, and young people struggled much more with distress and other problems than those who are better off.^{Reference Fancourt and Bradbury10} These findings confirm the importance of the cross-disciplinary research agenda advocated by Fancourt et al, in an article in BJPsych Open, in which she and her colleagues identify the importance of focusing on social, contextual and protective factors, mechanisms of action and public engagement.^{Reference Fancourt, Bhui, Chatterjee, Crawford, Crossick and DeNora12} Marmot et al have called for us all to ‘build back fairer’ in response to the pandemic.^{Reference Marmot, Allen, Goldblatt, Herd and Morrisson13}

This situation, and the comparisons made by politicians and others early in the pandemic of its effects on populations with those of war, reminds us of similar concerns in the Second World War. The Blitz in the UK, from 1940, was forecast well before it started. Psychiatric casualties because of people's exposure to the bombing were anticipated at a ratio of three for every physical casualty,^{Reference Gardiner14} and large numbers of mental illness hospital beds were cleared to create space to which people in the UK could be admitted. In fact, few of those beds were used in this way although reviews of the history reveal the great suffering as well as the stoicism of the public and an inverse socioeconomic gradient of the impact. The way in which the UK Government's Ministry of Information used daily reports of the impact of the bombing to manage worries about public compliance contributed a partially mythical element to what is still referred to as the Blitz Spirit. In 2021, Jones contrasted the impact on the UK public of wartime aerial bombing with those of COVID-19 and the public health responses to it indicating that, by March 2021, the mortality and serious morbidity statistics were similar.^{Reference Jones15} Whether this comparison remains appropriate is a matter for future scrutiny. The number of deaths in the UK continued to rise in the third wave and is rising further in the fourth, and the matter of persisting serious ‘injury’ related to Long COVID or long-haul disorders is continuing to grow. Thus, as in the Second World War, the comparison turns on which issues are included.

A truth that was instrumental in how people coped in the Blitz was identified in a report on its impact on East London by Calder who observed that the thing that sustains people is ‘not letting the other fellow down and forgetting self in the common danger’.^{Reference Calder16} Similarly, this is a lesson we are learning again in this pandemic about sustaining the public and healthcare staff and the importance of cohesion of teams has been, and is being, emphasised recurrently.

The impact of COVID-19 on the mental health of the public

As before the Blitz, there were claims early in the pandemic, that there would be a tidal wave of mental disorders.^{Reference Lamb, Greenberg, Stevelink and Wessely1} Many journal articles have reported that the COVID-19 pandemic has been associated with an increased frequency of a broad array of symptoms of mental disorders in the general population, mainly symptoms of anxiety, depression and post-traumatic stress. But it seems to us that the focus on these findings reflects the contents of the screening tools used rather more than an exhaustive survey of the actual array of the impact on mental health. These findings have, to the best of our knowledge, yet to be supported by substantial clinical research, but studies comprised of longitudinal surveys, structured clinical examinations and reports of people's own accounts through qualitative enquires are reaching publication or are underway.^{Reference Van der Goot, Duvivier, Van Yperen, de Carvalho-Filho, Noot and Ikink4} We introduce one such paper later in this review. Other studies that used cross-sectional methods but asked different questions offer a wider array of sources of anxiety.^{Reference Siddiqui, Aurelia, Gupta, Blythe and Khanji17} Now, 22 months after the onset of the pandemic in the UK, is still relatively early in mental illness and research terms.

It is important to identify several longitudinal studies. First, is a paper that reports research on participants in the Trøndelag Health Study in Trondheim, Norway. Participants (n = 2154) were recruited by repeated probability sampling to estimate prevalence of mental disorders and suicidal ideation across 2020. The authors carried out repeated cross-sectional analyses. Knudsen et al's paper supports caution about a tidal wave of impact on the mental health of the population.^{Reference Knudsen, Stene-Larsen, Gustavson, Hotopf, Kessler and Krokstad18} The authors showed that levels of mental disorders, suicidal ideation and suicide deaths were stable in the first 6 months of the pandemic and comparable with findings before the pandemic. Appleby also expressed caution based on reports from seven countries of national or state-level suicide data pointing out the suicide rates had not risen by spring 2021.^{Reference Appleby19}

In this context, a short report in our series, by Cobo et al, documents the authors’ prospective assessment of 36 adult patients in treatment in Spain because of their high risk of suicide.^{Reference Cobo, Porras-Segovia, Pérez-Rodriguez, Artés-Rodriquez, Barrigón and Courtet20} They administered a questionnaire using a smartphone app to attendees of a suicide-prevention out-patient clinic using an ingenious process of asking two random questions each day at random times to reduce the burden on users. They also used the Columbia Suicide Severity Rating Scale at baseline and follow-up during lockdown. The authors direct attention to two other studies that showed findings consistent with their own, that self-reported suicide risk decreased during the lockdown. It is possible, of course, that the decrease might have been temporary or an early artefact of the research and COVID-19 circumstances. We should follow these three sources to see how rates progress over time.

Ettman and colleagues report a survey conducted in March and April 2020 of 1441 respondents in the USA and compared the findings with returns from 5065 respondents from before the pandemic.^{Reference Ettman, Abdall, Cohen, Sampson, Vivier and Galea21} They found a three-fold higher prevalence of symptoms of depression. The researchers conclude, ‘These findings suggest that there is a high burden of depression symptoms in the US associated with the COVID-19 pandemic and that this burden falls disproportionately on individuals who are already at increased risk’ in the form of having lower incomes, fewer savings and exposure to more stressors. We conclude from the works of Fancourt et al, Marmot et al and Ettman et al that there is a clear finding emerging that secondary stressors within people's life circumstances are having profound effects on the risks run by populations of their developing symptoms of mental disorders and that these stressors may be exacerbated by events associated with the pandemic. This is a good illustration of how secondary stressors operate.^{Reference Williams, Ntontis, Alfadhli, Drury and Amlôt11} Murray et al also make this point and we return to that paper later.^{Reference Murray, Kaufman and Williams22}

Another paper considers the longitudinal trajectories of the public's mental health responses to the COVID pandemic in the UK.^{Reference Pierce, McManus, Hope, Hotopf, Ford and Hatch23} Pierce et al, report a secondary analysis of the large UK Household Longitudinal Study that has been collecting data since 2009. The assessment covered the period April to October 2020 and used the 12-item General Health Questionnaire (GHQ-12) to assess mental health in 19 763 adults. This research showed that most people remained resilient, or their mental health returned to pre-pandemic levels after a deterioration in the population mean in April 2020. ‘One smaller group of people had sustained elevated scores (4.1% of the cohort) and another group (7.0% of the cohort) had little deterioration …’ but showed a steady decline in mental health over time. This paper also addresses predictors of deterioration of this nature and coming from an ethnic minority background, pre-existing mental illness, financial adversity and SARS-CoV-2 infection emerged as risk factors, as they have in other research. Pierce et al remind us of research on trajectories of psychosocial impact after major incidents that has documented a number of patterns of change over time of people's experiences or symptoms of stress and distress.^{Reference Norris, Tracy and Galea24} The two most common patterns or longitudinal ‘trajectories’ are: (a) mild–stable levels of distress; and (b) decreasing symptoms from moderate to mild over months and years. Smaller proportions of people affected by major incidents may show: (c) deteriorating patterns of worsening stress; or (d) high-stable levels of stress. These last two groups appear more at risk of psychopathology.

We include in this thematic series two papers that report public prevalence data for possible mental health disorders.^{Reference Ben-Ezra, Hou and Goodwin25,Reference Chamberlain, Grant, Trender, Hellyer and Hampshire28} Both these papers are based on self-reported symptoms measured using screening questionnaires given to cross-sectional convenience samples. The response rates were either not stated or low.

The first, by Ben-Ezra et al, reports two studies of 1293 and 1073 UK participants in the early weeks of the pandemic.^{Reference Ben-Ezra, Hou and Goodwin25} In the first, the authors used random stratified sampling to construct an internet panel and the K6 to screen for ‘serious mental illness’. The response rate for study 1 was 37% and it found 16.6% of the participants who reported COVID-19-related social life or occupationally stressful events to have elevated levels of symptoms of serious mental illness. The second study was intended to replicate the findings from the first. The authors report an elevated risk of serious mental illness and probable ICD-11 adjustment disorder among people who say they have had stressful events related to COVID-19. Both studies showed stratified increased risk based on number of stressful events. This is an interesting finding because it resonates with the ‘cumulative risk model’. This model is the dominant thesis regarding children stating that ‘… long-term adverse outcomes are better predicted by the total number, rather than the specific nature of environmental risk exposures’ to ‘… different adverse experiences and events’ early in childhood.^{Reference Williams, Kemp, Williams, Kemp, Haslam, Haslam, Bhui and Bailey26,Reference Zeanah and Sonuga-Barke27}

In the second paper, Chamberlain et al report a study that examined, in May 2020, post-traumatic stress disorder (PTSD) symptoms in 13 049 people in the UK aged 16 or older who self-reported suspected or confirmed COVID-19, predominantly from the UK general population.^{Reference Chamberlain, Grant, Trender, Hellyer and Hampshire28} Participants were recruited through television webpages. This project identified escalating effect sizes with increased severity of COVID-19 and degree of treatment.

Research by Iqbal et al aimed to ascertain the psychiatric morbidity associated with SARS-CoV-2 infection by a retrospective study of case notes of 50 consecutive patients for whom psychiatric diagnoses reflected clinical judgement.^{Reference Iqbal, Al Abdulla, Albrahim, Latoo, Kumar and Haddad29} Having COVID-19 was operationally defined as patients having ‘… a positive antigen test for SARS-CoV-2 [real-time polymerase chain reaction test] taken during the patient's current period of hospital admission’. All were in-patients in hospitals in Doha, Qatar who were referred to a consultation-liaison team. The authors report that a psychiatric diagnosis was made in 49 cases and that ‘… a third of our sample had a past psychiatric history, and of these nearly half had a psychotic disorder or bipolar I disorder …’. The authors say that this ‘… suggests that those with prior mental health problems are particularly vulnerable to develop further mental health problems during the pandemic’.

Another retrospective case-notes study, by Williams et al, reports patients’ adherence with measures taken by an acute psychiatric in-patient ward in London to control spread of SARS-CoV-2 among patients.^{Reference Williams, Tweed, Rebolledo, Khalid, Agyeman and Pinto do Costa30} The authors report that 138 of 176 patients did not comply with the measures and that people who had diagnoses of psychotic, personality and substance use disorders were less adherent than those with other disorders. Their adherence improved when they were given direct instructions. We recognise that there are many potential weaknesses in studies that use case notes to glean data retrospectively but this one shows the importance of giving patients clear instructions about protecting themselves and other people.

It has also become clear that many people, including healthcare staff, have developed sustained healthcare problems of a broad nature that stem from embolic and other phenomena that have been caused by the SARS-CoV-2 virus and the immune responses to it. The National Institute for Health Research in England is publishing reviews of ‘long COVID’.³¹ Paterson describes a spectrum of COVID-19 neurology,^{Reference Paterson32} Baig documents deleterious outcomes in ‘long-haulers’ who have a chronic COVID syndrome as a result of SARS-CoV-2,^{Reference Baig33} and Nalbandian et al, provide an account of persistent and prolonged effects after acute COVID-19.^{Reference Nalbandian, Sehgal, Gupta, Madhavan, McGroder and Stevens34} Hypotheses for understanding the origins, and treatments for these persisting effects of COVID-19 are presently emerging. One important observation is that the occurrence of long-haul problems is not necessarily related to the severity of COVID-19 experienced by patients.^{Reference Crook, Raza, Nowell, Young and Edison35}

The impact of COVID-19 on staff working in healthcare

Our thematic series includes a personal account of the opening 6 months of the pandemic as it affected services and their staff in Spain. Pacchiarotti et al, describe how Spain reconfigured its hospitals and redeployed healthcare professionals to manage demand early in the pandemic.^{Reference Pacchiarotti, Anmella, Fico, Verdolini and Vieta37} They say that these actions led to healthcare staff being exposed to extremely stressful situations and the authors projected ‘… a rebound effect on mental health problems … in the medium and long term …’. They called for preventive approaches, use of telepsychiatry and evaluation of altered ways of providing mental healthcare. We have seen each of these proposals in action in the past 18 months.

An important matter concerns the risks to staff who address the ethical dilemma of choosing who receives services when healthcare systems are overwhelmed. Triage and triage-like decision-making remains with people who make those decisions for a substantial time afterwards.^{Reference Bard and Bursztajn38} Thus, the potential for staff to experience moral distress and moral injury have been widely discussed in the UK though the pandemic. Murray et al introduce both moral distress and moral injury.^{Reference Murray, Kaufman and Williams22} Rimmer summarises claims made by the British Medical Association after its survey of doctors conducted in March and April 2021.^{Reference Rimmer39} She says that around half of responders had heard of moral distress and moral injury and 78% of responders (around 1800 doctors) said that moral distress resonated with their experiences at work whereas 51% said the same for moral injury. Doctors who worked only with patients who had COVID-19 were more likely to give these responses as were doctors from ethnic minority backgrounds compared with doctors who are White.

Two papers in the thematic series reinforce the psychosocial importance of PPE availability with associated workplace training,^{Reference Liu, Zheng, Sha, Liu, Zheng and Su40,Reference Wanigasooriya, Palimar, Naumann, Ismail, Fellows and Logan41} and these concerns continued through the second wave in the UK. In March 2021, the Royal College of Nursing published an independent report that found that the UK's guidelines for ventilating healthcare premises and face protection had not been updated to reflect evidence about airborne transmission of SARS-CoV-2.^{Reference Gould and Purssell42}

Early in the pandemic, Kisely et al, reported a systematic review of the psychological effects of what are now known as high-consequence infectious diseases (HCIDs) on healthcare workers over the past 20 years.^{Reference Kisely, Warren, McMahon, Dalais, Henry and Siskind43} They found that the organisational measures that best decrease the risk of adverse outcomes include: positive feedback to staff; the faith of staff in local infection control procedures; providing protective gear; effective preparation; and training.

Research into the impact on the mental health of healthcare staff

The thematic series includes eight papers that describe surveys, mainly online, of the impact of the pandemic on healthcare staff conducted in a number of countries using self-administered screening tools with convenience samples.^{Reference Liu, Zheng, Sha, Liu, Zheng and Su40,Reference Wanigasooriya, Palimar, Naumann, Ismail, Fellows and Logan41,Reference Hong, Cao, Wei, Duan, Zhao and Jiang44–Reference Chew, Ngiam, Tan, Tham, Tan and Jing49} We identify a mix of primary and secondary stressors emerging from these papers as associated with participants’ experiences of distress and symptoms of mental health disorders. Seven of the papers in the thematic series (dates of sampling in parenthesis) report findings from a single country; two papers are from China (March 2020;^{Reference Liu, Zheng, Sha, Liu, Zheng and Su40} February to March^{Reference Hong, Cao, Wei, Duan, Zhao and Jiang44}); three from the UK (June to July 2020;^{Reference Wanigasooriya, Palimar, Naumann, Ismail, Fellows and Logan41} April to May 2020^{Reference Gilleen, Santaolalla, Valdearenas, Salice and Fusté45} and November 2020 time 1, February 2021 time 2^{Reference Jordan, Shannon, Browne, Carroll, Maguire and Kerrigan46}); one from Ethiopia (March 2020^{Reference Yitayih, Mekonen, Zeynudin, Mengistie and Ambelu47}); and one from Italy (March 2020^{Reference Naldi, Vallelonga, Di Libero, Cavallo, Agnesone and Gonella48}). The eighth paper reports on healthcare staff in India, Indonesia, Malaysia, Singapore and Vietnam (April to June 2020^{Reference Chew, Ngiam, Tan, Tham, Tan and Jing49}). These papers offer differing but helpful snapshots of the mental health of staff in the first wave of the pandemic and all report high levels of psychiatric symptoms.

In the first paper from China, Liu et al report mental health symptoms experienced by obstetricians and midwives who were not infected as they treated hospital in-patients who had COVID-19.^{Reference Liu, Zheng, Sha, Liu, Zheng and Su40} The authors found high rates of symptoms of depression (42%), anxiety (29%) and insomnia (34%) in a total sample of 2126 participants. ‘Regardless of whether or not they had direct contact with patients with COVID-19, obstetricians and midwives were more likely to report mild and moderate depression and anxiety during the COVID-19 pandemic when compared with before the pandemic’. Higher rates of symptoms of depression and insomnia were reported by staff who had direct contact with patients with COVID-19. ‘Those who had sufficient protective equipment or training were less likely to report depression, anxiety and insomnia than those who did not’.^{Reference Liu, Zheng, Sha, Liu, Zheng and Su40}

Hong et al report a cross-sectional study of the psychological impact of the COVID-19 pandemic on the healthcare staff working in a fever clinic of a general hospital in Beijing using qualitative interviewing, and the Sources of Stress questionnaire (developed during the SARS outbreak in Hong Kong) and the Impact of Event Scale-Revised (IES-R) questionnaire.^{Reference Hong, Cao, Wei, Duan, Zhao and Jiang44} There were 102 participants, (39% doctors; 53% nurses and 8% technicians) who stayed in the fever clinic during their rotation, and adjustments were made to their working conditions to help them adapt to the work. Staff received semi-structured interviews from staff of a hotline support service and the records were coded by scenario analysis and some answers were converted to binary variables. The total score on the Sources of Stress correlated moderately with the IES-R score. The top four sources of distress were worry about: the health of one's family or others; the spread of the virus; changes in work; and one's own health.

Wanigasooriya et al report on an electronic survey of hospital healthcare workers in the West Midlands, UK using clinically validated questionnaires.^{Reference Wanigasooriya, Palimar, Naumann, Ismail, Fellows and Logan41} ‘The rates of … clinically significant symptoms of anxiety, depression and PTSD were 34.3%, 31.2% and 24.5%, respectively’. The factors that emerged as negatively correlated with participants reporting symptoms included availability of adequate PPE and well-being support, and lower exposure to moral dilemmas at work.

Gilleen et al, looked at the impact of the pandemic on the well-being and mental health of healthcare workers in the UK.^{Reference Gilleen, Santaolalla, Valdearenas, Salice and Fusté45} They endeavoured to: quantify how the mental health of healthcare workers changed compared with before the pandemic; identify if healthcare workers on the front line, based in London or from ethnic minorities, had more severe symptoms compared with other staff; quantify the prevalence of severe psychiatric symptoms; and identify factors associated with those symptoms. They conducted a Qualtrics web-based survey open to all UK healthcare workers that presented the nine-item Patient Health Questionnaire (PHQ-9), seven-item anxiety scale (GAD-7), IES-R and Perceived Stress Scale to participants. Only healthcare workers ‘… who had experienced a stressful or traumatic event related to COVID-19 were administered the IES-R. Also, the Connor-Davidson Resilience Scale (CD-RISC) was administered’. ‘Nearly a third of HCWs [healthcare workers] reported moderate to severe levels of anxiety and depression; and the number reporting very high symptoms was more than quadruple that pre-COVID-19’ (albeit measured retrospectively on Likert scales). Factors that influenced the results were reported to cluster within the themes of: demographics and roles; workplace readiness; risk management including PPE; experience of stressful events; protective experiences including sharing stress at work. The authors report that ‘Front-line workers were significantly more likely to be more depressed, anxious, have high PTSD symptoms and be more stressed than non-front-line workers…Working in London was associated with lower risk of depression … and anxiety … than outside London (although there was no difference in stress or PTSD). Ethnic minority status (n = 342) was significantly associated with greater risk of high PTSD symptoms (OR [odds ratio] = 1.52), but not high anxiety, stress or depression’. ‘Being able to share stress at work … [was] associated with significantly lower likelihood of being in the high anxiety, stress and depression groups …’.^{Reference Gilleen, Santaolalla, Valdearenas, Salice and Fusté45}

Jordan et al report findings at two of four time points for which analyses are complete in an ongoing, online cross-sectional and longitudinal survey of the prevalence of mental health symptoms affecting health and social care staff in Northern Ireland.^{Reference Jordan, Shannon, Browne, Carroll, Maguire and Kerrigan46} Data collection at time 1 was in November 2020 (3834 staff responded, response rate 4.9%) and at time 2 was in February 2021 (2898 responses, response rate 3.7%). The validated questionaries were the GAD-7, PHQ-9, IES-R and Insomnia Severity Index. As the survey was open to new respondents at time 2, the authors report the initial findings as two cross-sectional studies. A high proportion of staff reported moderate-to-severe symptoms of depression, anxiety, PTSD and insomnia in the cross-sectional samples used at time 1 (26 to 30%) and time 2 (27 to 36%). Significantly more participants reported symptoms suggesting moderate-to-severe depression at time 2 but the other symptoms were not significantly different compared with time 1. These cross-sectional findings, which appeared to Jordan et al higher compared with findings from other studies of general UK and Irish populations, were ‘broadly mirrored’ in their longitudinal analyses comparing findings at times 1 and 2. But the authors also draw attention to another paper that found no significant differences between the prevalence rates for healthcarers and other populations.^{Reference Cénat, Blais-Rochette, Kokou-Kpolou, Noorishad, Mukunzi and McIntee50}

A hospital-based cross-sectional study was conducted in Ethiopia with 249 healthcare professionals by Yitayih et al.^{Reference Yitayih, Mekonen, Zeynudin, Mengistie and Ambelu47} The authors report the ‘prevalence of psychological distress among healthcare professionals was 78.3% … [and prevalence] of insomnia was 50.2%’. Interestingly, we note that higher rates of psychological distress were associated with not having a daily update on COVID-19 and feeling stigmatised or rejected because of working in a hospital.

Naldi et al, report a cross-sectional, survey of 797 healthcare workers in North-West Italy.^{Reference Naldi, Vallelonga, Di Libero, Cavallo, Agnesone and Gonella48} This study used the IES-R to measure distress, the State-Trait Anxiety Inventory (Form Y) to measure anxiety and the Maslach Burnout Inventory. The authors found that one-third of their participants had severe anxiety and distress. They draw attention to higher rates of distress in women and nurses. The authors conclude that detecting factors ‘associated with worst psychological outcome may favour a tailored, preventive, organisational and psychological approach, representing a potential strength in counteracting the effects of future pandemics’.

Chew et al, report on a total of 1146 participants from India (384), Indonesia (250), Malaysia (175), Singapore (277) and Vietnam (60).^{Reference Chew, Ngiam, Tan, Tham, Tan and Jing49} They used the Depression Anxiety Stress Scales (DASS-21) and IES-R to measure symptoms. Their samples included doctors, nurses, administrative staff, pharmacists, cleaners, porters and technicians. The authors conclude that their study shows that ‘the varied prevalence of psychological adversity among healthcare workers is independent of the burden of COVID-19 cases within …’ each of the countries studied. Interestingly, the authors say that Vietnam, which had the lowest volume of cases of COVID-19 per day per million population, also had ‘a higher prevalence of PTSD related to COVID-19 among healthcare workers compared with India’. Participants in Singapore ‘reported the highest number of cases per day/1 million, but had a lower prevalence of depression and anxiety among its healthcare workers, when compared with the Malaysian cohort’. Consistent with the opinions of other researchers, the authors of this paper comment on the potential beneficial effects from early psychological interventions for vulnerable groups of healthcare workers. These findings support our contention that other factors in addition to the risks introduced by COVID-19 were involved, as suggested by Fancourt et al and others.^{8–Reference Fancourt and Bradbury10,Reference Murray, Kaufman and Williams22}

A serosurvey of the staff working in mental health services

One paper in the thematic series reports the first published serosurvey among psychiatric healthcare providers showing a lower seroprevalence compared with the results from other studies in the general population and other healthcare workers in Belgium.^{Reference Coppens, Goossens, Theeten, Van Damme and Berens51} This paper reports 3.2% of psychiatric providers were seropositive,^{Reference Coppens, Goossens, Theeten, Van Damme and Berens51} whereas the proportion of staff in a tertiary medical centre were reported as 6.4%.^{Reference Steensels, Oris and Coninx52} Exposure at home predicted the presence of antibodies, but exposure at work did not. The authors recommend measures to prevent transmission from staff to patients in psychiatric facilities.