ORIGINAL RESEARCH The impact of COVID-19 on oncology professionalsdone year on: lessons learned from the ESMO Resilience Task Force survey series K. H. J. Lim1,2,3y, K. Murali4y, E. Thorne5, K. Punie6, K. Kamposioras1, C. Oing7, M. O’Connor8, E. Élez9, T. Amaral10, P. Garrido11, M. Lambertini12,13, B. Devnani14, C. B. Westphalen15, G. Morgan16, J. B. A. G. Haanen17, C. Hardy5 & S. Banerjee18,19* 1 Department of Medical Oncology, The Christie NHS Foundation Trust, Manchester; 2Department of Immunology and Inflammation, Imperial College London, London; 3 Immunobiology Laboratory, The Francis Crick Institute, London, UK; 4Victorian Clinical Genetics Services & Murdoch Children’s Research Institute, The Royal Children’s Hospital, Melbourne, Australia; 5Division of Health Research, Faculty of Health and Medicine, Lancaster University, Lancaster, UK; 6Department of General Medical Oncology and Multidisciplinary Breast Center, Leuven Cancer Institute, University Hospitals Leuven, Leuven, Belgium; 7Department of Oncology, Hematology and Bone Marrow Transplantation with Division of Pneumology, Mildred Scheel Cancer Career Centre HaTriCs4, University Medical Center Hamburg-Eppendorf, Hamburg, Germany; 8University Hospital Waterford, Waterford, Ireland; 9Vall d’Hebron Barcelona Hospital Campus, Vall d’Hebron Institute of Oncology (VHIO), Universitat Autònoma de Barcelona, Barcelona, Spain; 10Department of Dermatology, University Hospital Tübingen, Tübingen, Germany; 11Medical Oncology, Hospital Ramón y Cajal, IRYCIS, Alcalá University, Madrid, Spain; 12Department of Medical Oncology, U.O. Clinica di Oncologia Medica, IRCCS Ospedale Policlinico San Martino, Genoa; 13 Department of Internal Medicine and Medical Sciences (DiMI), School of Medicine, University of Genoa, Genoa, Italy; 14Department of Radiotherapy and Oncology, All India Institute of Medical Sciences, Jodhpur, India; 15Department of Internal Medicine III, University Hospital, LMU Munich and Comprehensive Cancer Center, Munich, Germany; 16Department of Medical and Radiation Oncology, Skåne University Hospital, Lund, Sweden; 17Division of Medical Oncology, Netherlands Cancer Institute, Amsterdam, The Netherlands; 18Gynaecology Unit, The Royal Marsden NHS Foundation Trust, London; 19The Institute of Cancer Research, London, UK Available online 17 December 2021 Background: COVID-19 has had a significant impact on the well-being and job performance of oncology professionals globally. The European Society for Medical Oncology (ESMO) Resilience Task Force collaboration set out to investigate and monitor well-being since COVID-19 in relation to work, lifestyle and support factors in oncology professionals 1 year on since the start of the pandemic. Methods: An online, anonymous survey was conducted in February/March 2021 (Survey III). Key outcome variables included risk of poor well-being or distress (expanded Well-Being Index), feeling burnout (single item from expanded Well-Being Index), and job performance since COVID-19. Longitudinal analysis of responses to the series of three surveys since COVID-19 was carried out, and responses to job demands and resources questions were interrogated. SPSS V.26.0/V.27.0 and GraphPad Prism V9.0 were used for statistical analyses. Results: Responses from 1269 participants from 104 countries were analysed in Survey III: 55% (n ¼ 699/1269) female, 54% (n ¼ 686/1269) >40 years, and 69% (n ¼ 852/1230) of white ethnicity. There continues to be an increased risk of poor well-being or distress (n ¼ 464/1169, 40%) and feeling burnout (n ¼ 660/1169, 57%) compared with Survey I (25% and 38% respectively, P < 0.0001), despite improved job performance. Compared with the initial period of the pandemic, more participants report feeling overwhelmed with workload (45% versus 29%, P < 0.0001). There remain concerns about the negative impact of the pandemic on career development/training (43%), job security (37%). and international fellowship opportunities (76%). Alarmingly, 25% (n ¼ 266/1086) are considering changing their future career with 38% (n ¼ 100/266) contemplating leaving the profession. Conclusion: Oncology professionals continue to face increased job demands. There is now significant concern regarding potential attrition in the oncology workforce. National and international stakeholders must act immediately and work closely with oncology professionals to draw up future-proof recovery plans. Key words: well-being, burnout, job performance, oncology professionals, resilience, COVID-19 @devnani_bharti, @BenWestphalen, @weoncologists, @OncoAlert, @Haa- *Correspondence to: Dr Susana Banerjee, The Royal Marsden NHS Foundation nenJohn, @BanerjeeSusana. Trust and Institute of Cancer Research, 203 Fulham Road, London SW3 6JJ, UK. Tel: þ44-(0)208-661-3563 yEqual contribution. E-mail: susana.banerjee@rmh.nhs.uk (S. Banerjee). 2059-7029/© 2021 Published by Elsevier Ltd on behalf of European Society Twitter handles (for co-authors with Twitter accounts): @kevinpunie, for Medical Oncology. This is an open access article under the CC BY-NC-ND @OingC1, @elena_elez, @TeresaSAmaral, @Piuchagarrido, @matteolambe, license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Volume 7 - Issue 1 - 2022 https://doi.org/10.1016/j.esmoop.2021.100374 1 ESMO Open K. H. J. Lim et al. INTRODUCTION 19 (JP-CV; score 3.5, favourable) have been previously The discovery of the novel coronavirus SARS-CoV-2 in described in detail.4,5 In Survey III, we also added relevant Wuhan, China in late December 2019 and the official questions regarding participants’ perception of the COVID- declaration of the COVID-19 pandemic in March 2020, has 19 situation in their respective country of work including triggered unprecedented changes to health systems lockdown restrictions and national vaccination pro- worldwide, with cancer services being no exception.1,2 grammes. Further questions about participants’ personal Globally, oncology services have experienced significant experience of COVID-19, and their personal and profes- disruption due to staff redeployment, deterioration in sional future outlook were also included. In total, there working conditions, reduction in oncology clinical trials, and were 38 closed response questions with 1 open text research activities.3-5 Additional temporising measures such response question at the end of the survey. as suspension of ‘non-essential’ palliative chemotherapy, cancer screening services, and favouring of less intensive treatment regimens have also impacted the nature of our Statistical analysis work with potential for longer-term consequences.6-9 Key outcome variables: eWBI, burnout, and JP-CV were Since the COVID-19 pandemic, clinicians have been longitudinally compared with results from Surveys I4 and particularly challenged in their ability to provide cancer care II.5 Chi-square analysis was used to compare categorical to aspired standards with a concomitant diminution in variables and paired or unpaired t-tests were used to meaningful professional activities.4,5 All of this whilst also analyse continuous variables. We also carried out the chi- contending with the risk to personal safety at work from square test for trend when comparing proportions across COVID-19 infection and associated morbidity.5 The Euro- time (Survey I versus Survey II versus Survey III, where pean Society for Medical Oncology (ESMO) Resilience Task data available). P values were two-tailed and were Force launched a unique longitudinal series of global sur- considered significant if <0.05. A series of cross-sectional veys since April 2020 to provide contemporary insights into between-subject analyses were also carried out in a sub- the daily practice and well-being of oncology professionals group of participants who completed all three surveys to during the COVID-19 pandemic. Our findings from Survey I examine relationships over time. Descriptive data were conducted in April/May 2020 and Survey II in July/August presented as median (interquartile range) or mean  2020 indicated that COVID-19 has had a detrimental impact standard deviation, and proportions were expressed as a on the lives of oncology professionals with rising rates of percentage (%). All statistical analyses were carried out distress and burnout, and uncovered significant concerns using SPSS V.26.0 or V.27.0 (IBM Corp., Armonk, NY) and regarding job security and their future outlook.4,5 Here, we data represented using GraphPad Prism V9.0 (GraphPad provide a further update 1 year on reporting on the key Software, San Diego, CA). findings from Survey III conducted in February/March 2021. METHODS RESULTS Survey design Participant demographics Survey III followed on from the series of online global sur- A total of 1432 participants from 104 countries responded veys designed by the ESMO Resilience Task Force, in to the Survey III invitation. In the final analysis, we included collaboration with the ESMO Young Oncologists Committee, data from all participants who consented and responded to ESMO Women for Oncology Committee, ESMO Leaders key questions placed at the beginning of the survey on their Generation Programme Alumni members, and the OncoA- perception of the current status of COVID-19 (trend, lock- lert Network, launched at specific timepoints during the down, and vaccination) and provided basic demographic course of the COVID-19 pandemic. Survey III was hosted on details including their country of work [n ¼ 1269 (88.6%)]. the Qualtrics platform (open from 9 February 2021 to 3 The majority (n ¼ 1158/1269, 91.3%) completed the survey March 2021) and was available on the ESMO website, to the end. The personal and professional demographic ESMO membership emails, and was promoted through so- characteristics of Survey III participants are outlined in cial media. Participation was voluntary and anonymous. Table 1 and Supplementary Table S1, available at https:// Participants who consented to longitudinal evaluation of doi.org/10.1016/j.esmoop.2021.100374. Overall, 55.1% their responses at different timepoints of the survey series (n ¼ 699/1269) were female, 54.1% (n ¼ 686/1269) >40 were assigned a trackable unique identifier code. The years of age, and 69.3% (n ¼ 852/1230) of white ethnicity. project was approved by the ESMO Executive Board. Collectively, participants have a mean of 15 years of expe- rience working in oncology, with medical oncology most represented (n ¼ 905/1245, 72.7%). Trainees contributed to Survey measures 22.1% (n ¼ 281/1269) of responses. The majority of re- Key outcome variables used throughout the survey series spondents were ESMO members (n ¼ 1073/1238, 86.7%) including risk of poor well-being or distress (expanded Well- and were from Europe (n ¼ 854/1269, 67.3%) Being Index, eWBI; score 4, at risk),10,11 feeling burnout (Supplementary Table S1, available at https://doi.org/10. (single item from eWBI), and job performance since COVID- 1016/j.esmoop.2021.100374). 2 https://doi.org/10.1016/j.esmoop.2021.100374 Volume 7 - Issue 1 - 2022 K. H. J. Lim et al. ESMO Open Table 1. Participant demographics for Survey III (n [ 1269) Table 2. Personal experience of COVID-19 as reported by participants in Survey III (n [ 1269), reflecting the period from February to March 2021 Number, n (%) Number, n (%) Age (years), n ¼ 1269 40 583 (45.9) Perception of COVID-19 death rate in region of work >40 686 (54.1) Increasing 536 (42.2) Gender, n ¼ 1269 No change 176 (13.9) Female 699 (55.1) Decreasing 533 (42.0) Male 566 (44.6) Free of COVID-19 24 (1.9) Prefer not to say 4 (0.3) Current restrictions in region of work Ethnicity, n ¼ 1230 Full lockdown 280 (22.1) White 852 (69.3) Partial lockdown 718 (56.6) Asian (East/Southeast) 156 (12.7) End of lockdown 157 (12.4) Asian (South) 63 (5.1) No lockdown imposed so far 114 (9.0) Hispanic 59 (4.8) Current status of COVID-19 vaccination programme in Arab 32 (2.6) country of work Mixed 17 (1.4) National programme has started 1112 (87.6) Black 16 (1.3) National programme planned but has not started yet 138 (10.9) Other 17 (1.4) No plans for a national programme so far 19 (1.5) Prefer not to say 18 (1.5) Personally received vaccination against COVID-19 Lives alone, n ¼ 1232 Yes, 2 doses 554 (43.7) Yes 187 (15.2) Yes, 1 dose 295 (23.2) No 1045 (84.8) No 408 (32.2) Have children/dependents, n ¼ 1232 Prefer not to say 12 (0.9) Yes 738 (59.9) Regular asymptomatic testing for COVID-19, n ¼ 1082 No 494 (40.1) Yes 423 (39.1) Regiona, n ¼ 1269 No 659 (60.9) Europeb Have had to undergo isolation/take sick leave due to Southwestern Europe 264 (20.8) COVID-19 symptoms, n ¼ 1080 Central Europe 220 (17.3) Yes 263 (24.4) Northern Europe and British Isles 175 (13.8) No 817 (75.6) Southeastern Europe 87 (6.9) Tested positive for COVID-19, n ¼ 1081 Western Europe 78 (6.1) Yes 160 (14.8) Eastern Europe 30 (2.4) No 921 (85.2) Asia 223 (17.6) Required hospitalisation for COVID-19, n ¼ 160 North America 78 (6.1) Yes 10 (6.3) South America 56 (4.4) No 150 (93.8) Africa 43 (3.4) Feel given appropriate time to recover (if had symptomatic Oceania 15 (1.2) COVID-19), n ¼ 160 Primary place of work, n ¼ 1238 Yes 117 (73.1) General hospital 628 (50.7) No 43 (26.9) Cancer centre 460 (37.2) Feel completely recovered upon return to work, n ¼ 160 Private outpatient clinic 21 (1.7) Yes 100 (62.5) Pharmaceutical/technology company 44 (3.6) No 60 (37.5) Health care organisation 21 (1.7) Had colleague who has died from COVID-19, n ¼ 1079 Other 64 (5.2) Yes 215 (19.9) Specialtyc, n ¼ 1245 No 841 (77.9) Medical oncology 905 (72.7) Prefer not to say 23 (2.1) Radiation/clinical oncology 177 (14.2) Haemato-oncology 130 (10.4) Palliative care 75 (6.0) Surgical oncology 44 (3.5) Personal experience with COVID-19 Laboratory-based researcher/scientist 38 (3.1) Nursing 11 (0.9) As the COVID-19 pandemic continues, the majority of par- Other 81 (6.5) ticipants have encountered some form of lockdown re- Trainee, n ¼ 1269 Yes 281 (22.1) strictions in their region of work (Table 2). Almost all No 988 (77.9) participants (n ¼ 1250/1269, 98.5%) reported that a na- ESMO member, n ¼ 1238 tional vaccination programme against COVID-19 was un- Yes 1073 (86.7) derway, if not planned, and two-thirds (n ¼ 849/1269, No 165 (13.3) See also Supplementary Table S1, available at https://doi.org/10.1016/j.esmoop. 66.9%) have personally had at least one dose of the vaccine 2021.100374 for further details of countries/regions represented. when asked in February/March 2021 (Table 2). ESMO, European Society for Medical Oncology. a Countries most represented in Survey III were UK (n ¼ 112), Germany (n ¼ 99), Spain (n ¼ 98), Italy (n ¼ 85), Portugal (n ¼ 81), and India (n ¼ 78). See Supplementary Table S1, available at https://doi.org/10.1016/j.esmoop.2021. 100374 for complete list of countries/regions, and the corresponding number of Serbia, Turkey; Western EuropedBelgium, France, Luxembourg, The Netherlands; participants per country. and Eastern EuropedBelarus, Estonia, Georgia, Latvia, Lithuania, Moldova, b Russian Federation, Ukraine. Southwestern EuropedItaly, Portugal, Spain; Central EuropedAustria, Czech c Republic, Germany, Hungary, Poland, Romania, Slovakia, Slovenia, Switzerland; Note that some participants have selected two or more specialties within their job Northern Europe and the British IslesdDenmark, Finland, Norway, Republic of role (to encompass differences in the scope of practice between countries/regions), Ireland, Sweden, UK; Southeastern EuropedAlbania, Bosnia and Herzegovina, and proportion of representation is summarised as such. Overall, participants have Bulgaria, Croatia, Cyprus, Greece, Israel, Kosovo, Montenegro, North Macedonia, reported a mean of 15.1  10.5 years of experience in the field of oncology. Volume 7 - Issue 1 - 2022 https://doi.org/10.1016/j.esmoop.2021.100374 3 ESMO Open K. H. J. Lim et al. A Risk of poor wellbeing/distress B Feeling burnout C Job performance since COVID-19 (eWBI 4) (single-item) (JP-CV 3.5) 100 Survey I (n = 386/1520) 100 100 Survey I (n = 572/1520) Survey I (n = 523/1520) Survey II (n = 310/942) Survey II (n = 444/873) Survey II (n = 460/942) Survey III (n = 464/1169) Survey III (n = 545/1108) Survey III (n = 660/1169) 80 80 80 **** **** **** Proportion of participants (%) Proportion of participants (%) Proportion of participants (%) ** *** ns **** **** 60 **** 60 56% 60 51% 49% 49% 40% 40 40 40 38% 33% 34% 25% 20 20 20 0 0 0 Apr/May Jul/Aug Feb/Mar Apr/May Jul/Aug Feb/Mar 2020 2020 2021 Apr/May Jul/Aug Feb/Mar 2020 2020 2021 2020 2020 2021 Odds ratio Odds ratio Odds ratio S I versus S III: 1.934 (95% CI 1.640-2.275) S I versus S III: 2.149 (95% CI 1.840-2.509) S I versus S III: 1.845 (95% CI 1.575-2.159) S I versus S II: 1.441 (95% CI 1.206-1.721) S I versus S II: 1.582 (95% CI 1.340-1.867) S I versus S II: 1.973 (95% CI 1.667-2.337) S II versus S III: 1.342 (95% CI 1.122-1.603) S II versus S III: 1.359 (95% CI 1.142-1.611) S II versus S III: 0.935 (95% CI 0.784-1.115) Chi-square test for trend Chi-square test for trend Chi-square test for trend S I versus S II versus S III – P < 0.0001 S I versus S II versus S III – P < 0.0001 S I versus S II versus S III – P < 0.0001 Figure 1. Comparison of key outcome variables across three survey timepoints since the outbreak of the COVID-19 pandemic (Survey I: April-May 2020, Survey II: July-August 2020, and Survey III: February-March 2021). (A) Risk of poor well-being/distress. (B) Burnout. (C) Job performance since COVID-19 (JP-CV). Groups were compared using chi-square analyses. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. CI, confidence interval; eWBI, expanded Well-Being Index; ns, not significant; S I, Survey I; S II, Survey II; S III, Survey III. See also Supplementary Figures S1 and S2, available at https://doi.org/10.1016/j.esmoop.2021.100374. More than a third of participants reported having regular increased use of remote or virtual meetings and consulta- asymptomatic testing for COVID-19 (n ¼ 423/1082, 39.1%) tions (>80%). Of concern, more than half of participants (Table 2). A total of 160 participants (14.8%) had tested reported a decrease in clinical trial activity (n ¼ 392/648, positive for COVID-19, of whom 10 had required hospital- 60.5%) and general research activity (n ¼ 366/664, 55.1%). isation. Of those who had tested positive for COVID-19, About two-thirds of participants (n ¼ 708/1108, 63.9%) were around a third (n ¼ 43/160, 26.9%) neither felt they had worried that COVID-19 would have a negative impact on the been given appropriate time to recover nor felt completely quality of cancer research at their institution. recovered upon their return to work (n ¼ 60/160, 37.5%). By February/March 2021, one in five participants disclosed that they had a work colleague who died of COVID-19 (n ¼ Well-being, burnout, and resilience throughout the 215/1079, 19.9%). pandemic Compared with Surveys I4 and II,5 there were now signifi- cantly more participants who were at risk of poor well-being Ongoing changes in professional role and delivery of care or distress (n ¼ 464/1169, 39.7%) and feeling burnout (n ¼ Work routine has yet to return to the pre-COVID-19 situation 660/1169, 56.5%) (chi-square test for trend, P < 0.0001) for more than half of participants (n ¼ 578/1269, 52.3%). In (Figure 1A and B). These trends were confirmed in a longi- Survey III, around a fifth of participants (n ¼ 238/1107, tudinal subgroup of n ¼ 127 participants from 47 countries 21.5%) were either partially or fully redeployed. Participants (50.0% female, 39.4% 40 years old, and 69.3% of white continued to report several changes to their professional ethnicity) who had completed all three surveys, in whom roles and duties as detailed in Supplementary Table S2, there was a progressive increase in the proportion of those at available at https://doi.org/10.1016/j.esmoop.2021.100374, risk of poor well-being (20.5% versus 28.9% versus 31.6%, with nearly half (n ¼ 354/756, 46.8%) reporting an increase P ¼ 0.0516) and feeling burnout (34.4% versus 47.9% versus in overall hours of work. As expected, there remained an 56.4%, P ¼ 0.0006) (Supplementary Figure S1A and B, 4 https://doi.org/10.1016/j.esmoop.2021.100374 Volume 7 - Issue 1 - 2022 K. H. J. Lim et al. ESMO Open A Gender B Age 100 Female (n = 644) 100 40 years (n = 533) Male (n = 523) >40 years (n = 636) **** ** 80 80 Proportion of participants (%) Proportion of participants (%) **** 63% **** 61% 60 60 52% 48% 49% 45% 40 40 33% 32% 20 20 0 0 eWBI 4 Feeling burnout eWBI 4 Feeling burnout OR 1.643 OR 1.859 OR 1.987 OR 1.444 (95% CI 1.298-2.091) (95% CI 1.469-2.342) (95% CI 1.562-2.519) (95% CI 1.144-1.825) Figure 2. Subgroup comparisons of key outcome variables (risk of poor well-being/distress and burnout) analysed by (A) age and (B) gender, respectively, of participants in Survey III. Groups were compared using chi-square analyses. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. CI, confidence interval; eWBI, expanded Well-Being Index; ns, not significant; OR, odds ratio. available at https://doi.org/10.1016/j.esmoop.2021.100374). Alarmingly, we observed that in general there has been a These observations are despite the perception of improved progressive increase in job demands, with considerable in- job performance, where JP-CV plateaued at w50% since the creases in the proportion of participants who have reported Survey II timepoint (Figure 1C and Supplementary Figure S1C, an increase in the perception of feeling overwhelmed with available at https://doi.org/10.1016/j.esmoop.2021.100374). workload (29.1% versus 35.6% versus 45.2%, P < 0.0001) Of note, there appear to be sustained levels of psychological and overall working hours (16.7% versus 38.4% versus resilience amongst participants throughout the pandemic 46.8%, P < 0.0001) (Figure 3A). Moreover, more partici- (Supplementary Figure S2, available at https://doi.org/10. pants were now also burdened with increased COVID-19 1016/j.esmoop.2021.100374). As gender and age were inpatient work (13.6% versus 40.3% versus 58.1%, P < found to be significant predictors of risk of poor well-being in 0.0001) and COVID-19-related research (15.6% versus 58.7% Survey I,4 we assessed whether this disparity persisted over versus 64.5%, P < 0.0001). Whereas the majority have the course of the COVID-19 pandemic. Female colleagues continued to work in pleasant physical conditions, their continued to be at higher risk of poor well-being {odds ratio concern for personal safety at work due to COVID-19 has (OR) 1.643 [95% confidence interval (CI) 1.298-2.091], P < remained persistently high (78.3% versus 72.5% versus 0.0001} and feeling burnout [OR 1.859 (95% CI 1.469-2.342), 63.0%, P < 0.0001) (Figure 3A). P < 0.0001] (Figure 2). Similarly, younger (40 years old) Almost half (n ¼ 527/1169, 45.1%) were reporting having colleagues were also at continued risk of poor well-being [OR inadequate time for personal and/or family life compared 1.987 (95% CI 1.562-2.519), P < 0.0001] and feeling burnout with 34.6% (n ¼ 526/1520) and 43.0% (n ¼ 405/942) in [OR 1.444 (95% CI 1.144-1.825), P ¼ 0.002] (Figure 2). Surveys I4 and II,5 respectively (Figure 3A). Moreover, the majority of participants in Survey III reported that they had not been able to take time off for annual leave or holidays Factors contributing to perception of increased job (n ¼ 650/1084, 60.0%), and/or study leave (n ¼ 863/1084, demands 79.6%) (Figure 3A). We analysed responses to questions associated with job demands over the different timepoints studied in this sur- vey series to identify reasons which may have potentially Personal and professional job resources currently available contributed to worsening distress and burnout (Figure 3A). In addition, we also interrogated several domains in terms A chi-square test for trend was carried out to compare re- of personal and professional resources and coping strate- sponses to Survey I versus Survey II versus Survey III. gies available to participants which could potentially Volume 7 - Issue 1 - 2022 https://doi.org/10.1016/j.esmoop.2021.100374 5 ESMO Open K. H. J. Lim et al. A JOB DEMANDS Survey I Survey II Survey III n = 1520 n = 942 n = 1269 (Apr/May 2020) (Jul/Aug 2020) (Feb/Mar 2021) P value Perception of overwhelming workload 29 36 45 **** Increased COVID-19 inpatient work 14 40 58 **** Workload Increased cross-covering of non-oncology specialties 11 32 42 **** Increased COVID-19 research 16 59 65 **** Increased out-of-hours in-hospital work 16 36 41 **** Shift work Increased weekend shifts 12 19 26 **** Increased overnight shifts 8 15 18 **** Increased overall working hours 17 38 47 **** Inadequate time for personal/family life 35 43 45 **** Time Unable to take adequate rest periods/breaks during work 45 n/a pressure Unable to take all allocated annual leave/holiday 59 60 ns Unable to take all allocated study leave 80 n/a Physical Unpleasant physical working conditions 21 19 24 ns environment Increased concern for personal safety at work due to COVID-19 78 73 63 **** Proportion of participants 20% 40% 60% 80% B JOB RESOURCES Survey I Survey II Survey III n = 1520 n = 942 n = 1269 (Apr/May 2020) (Jul/Aug 2020) (Feb/Mar 2021) P value Adequate resources to do job 62 67 59 ns General Communication Adequate communication to do job 68 66 54 **** Participation Control over changes to patient care 83 81 ns Feeling valued by work organisation 60 54 51 **** Feedback Feeling valued by the public 75 64 57 **** Rewards Work is meaningful 86 84 84 ns Adequate control over most aspects of job 56 62 56 ns Job control Job performance since COVID-19 (JP-CV) (favourable 3.5) 34 51 49 **** Career development/training 38 37 34 ns Job security Job security 43 43 ns International fellowship opportunities 11 12 ns Felt well supported by management at workplace 57 54 45 **** Support at work Felt well supported by colleagues at workplace 83 78 73 **** Access to well-being support services 51 45 44 **** Other support Felt well supported by global/national societies/groups 57 51 42 **** Felt well supported by the government 39 34 28 **** Proportion of participants 80% 60% 40% 20% Figure 3. Heatmaps of factors which may be contributing to worsening distress and feeling burnout since COVID-19. 6 https://doi.org/10.1016/j.esmoop.2021.100374 Volume 7 - Issue 1 - 2022 K. H. J. Lim et al. ESMO Open alleviate job demands (Figure 3B). Here, the main areas of the significant threat of workforce attrition. A year on concern were regarding job security and participants’ sup- since the COVID-19 pandemic, there is marked increase in port systems (Figure 3B). In particular, participants were risk of poor wellbeing/distress (40% versus 25%) and concerned that the pandemic would have a negative impact burnout (57% versus 38%) compared with Survey I (April/ on their personal job security (n ¼ 386/1057, 36.5%). Just May 2020).4 Job demands have continued to increase, more than a third of participants (n ¼ 346/1023, 33.8%) meanwhile there appears to be a perceived decline in were not concerned that the COVID-19 pandemic would clinical trial activity and research, career development have a negative impact on their personal career develop- opportunities, and available resources and support. ment/training, and very few were not concerned about Almost half of survey participants now feel overwhelmed international fellowship opportunities (n ¼ 116/973, 11.9%) with their workload. (Figure 3B). Female and younger (40 years) colleagues continue to Participants were questioned about their support sys- be at higher risk of poor well-being and burnout, two de- tems, access to well-being support services, and the coping mographic groups noted to be particularly vulnerable in our strategies they had been using to help themselves during previous surveys4,5 and in the published literature.12,13 the pandemic (details summarised in Figure 3B, and There is evidence that the pandemic has had an unequal Supplementary Table S3, available at https://doi.org/10. impact on gender, with female clinicians shouldering the 1016/j.esmoop.2021.100374). There has been a gradual greater burden of domestic responsibilities which may be decline in the proportion of participants who felt well exacerbating the pressure of growing professional de- supported by the management at their workplace since mands.14,15 In Survey II, we previously highlighted concerns Survey I, with now less than half (n ¼ 476/1058, 45.0%) raised regarding career development, training, and job se- feeling well supported (Figure 3B). Similarly, their percep- curity disproportionately impacting trainee and early-career tion of support from global and/or national societies/groups oncologists.5 Survey III broadens these concerns to that of a and governments had declined over time (56.8% versus more widespread impending crisis in workforce retention. 50.7% versus 41.7%, P < 0.0001) (Figure 3B). Overall, in Alarmingly, one in four participants are considering chang- Survey III, 59.5% (n ¼ 635/1068) reported having adequate ing their future career direction, of whom 38% are resources to do their jobs (Figure 3B). There has also been a contemplating leaving the oncology profession altogether decrease in the proportion of participants who felt valued and 28% deliberating moving to a role within industry. This by their work organisation (59.7% versus 53.7% versus is on a background of perceived declining support from 50.8%, P < 0.0001) and the public (75.1% versus 63.9% employers and national/global bodies. versus 57.0%, P < 0.0001) (Supplementary Table S3, avail- Although our findings are compelling, this study has its able at https://doi.org/10.1016/j.esmoop.2021.100374). limitations. Survey III had a substantial number of partici- Whereas access to well-being support services was only pants but this still only constitutes approximately 5% of the perceived as available for less than half of participants (n ¼ ESMO membership base. Participants were mainly medical 472/1077, 43.8%), this represented a reassuring increase oncologists, more established in their careers with the compared with the pre-pandemic level (n ¼ 210/1076, majority based in Europe. Thus, these findings may not 19.5%) (P < 0.0001) (Supplementary Table S3, available at necessarily be representative of the needs of the global https://doi.org/10.1016/j.esmoop.2021.100374). oncology workforce, particularly those in more resource- constrained health care systems. Findings from the 2003 Potential risk of workforce attrition SARS experience suggests that those with the most direct contact with patients, including nurses, administrative staff, Finally, in Survey III, concerns with regards to attrition in the and ancillary workers, had the highest level of stress.16 oncology workforce were flagged. A quarter of participants These occupational groups are not well represented in (n ¼ 266/1086, 24.5%) disclosed that they had considered our survey series. Attrition in the oncology nursing work- changing their future career, with 37.6% (n ¼ 100/266) force in particular has already impacted significantly on thinking of leaving the oncology profession and 27.8% (n ¼ cancer treatment delivery.2 Nurses are also predominantly 74/266) considering moving to work in industry. female and are the largest health care professional group providing frontline care, both of which are risk factors for DISCUSSION burnout.2,12 Survey III draws on the lived experiences of 1269 oncology Our study methodology involved optional online sur- professionals from 104 countries to shed light on the veys which are associated with the risk of participant self- ongoing deterioration in well-being in oncology and raise selection bias. On balance, however, this remains the (A) various factors related to job demands. (B) personal and professional resources available to participants. Proportions (percentage, %) are displayed as colours ranging from blue to red [with red denoting cause for concern i.e. increased job demands for (A) and decreased job resources in (B)], as shown in the key. Groups were compared using Chi-square test for trend (Survey I versus Survey II versus Survey III, where available). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. ns, not significant; n/a, not applicable. See also Supplementary Table S3, available at https://doi.org/10.1016/j.esmoop.2021.100374. Volume 7 - Issue 1 - 2022 https://doi.org/10.1016/j.esmoop.2021.100374 7 ESMO Open K. H. J. Lim et al. optimal methodology for accessing a wider audience A focus group study of American oncologists during the within the constraints and work pressures of oncology pandemic identified that many participants were strongly professionals working on the frontline during the COVID- considering working part-time or taking early retirement.23 19 pandemic. Our survey provides indicators of distress Although this may seem concerning from a workforce in participants but does not provide a definitive under- planning viewpoint similar to our study, these career de- standing of their mental health. As the WBI has been cisions were triggered by COVID-19-related work changes cross-validated with an increased risk of suicide, depres- providing opportunity for self-care and reprioritisation of sion, and anxiety there may be even graver psychological work-life balance. This examination of personal and pro- consequences associated with our study findings.10,17 fessional values should be encouraged as essential for Importantly, the impact of additional factors such as healthy careers and workplaces. It also highlights the prolonged separation from loved ones should also not be importance of workplace arrangements that promote flex- underestimated. ibility and work-life balance. Despite Melbourne being the Although the findings of the ESMO Resilience Taskforce city worst affected by the COVID-19 pandemic in Austral- Survey series are cause for concern, there are also reasons asia, modifications to practice in response to the pandemic for optimism. Survey III highlights a significant increase were well received by radiation oncologists, with the vast in available well-being support services compared with majority of staff (the majority of whom were female) pre-pandemic levels. Our data have allowed us to infer reporting satisfaction with their work arrangements, longitudinal trends, but they do not provide us with an particularly those with children.24 The ability to work from indication of future trends particularly given the dramatic home has been reported as a positive experience associated changes in the global landscape that have ensued since with reduced burnout if adequate information technology Survey III was undertaken, with many regions experiencing and childcare support is available.25 This suggests that greater social freedoms. These are likely to be providing judicious changes to standard workplace practice as a result psychological relief and improvements in personal circum- of COVID-19 restrictions have the potential to improve stances, although the prevailing sense of professional un- workplace flexibility and quality. certainty with the spectre of rising case numbers remains. The ESMO Resilience Taskforce survey series highlights There is already some early evidence, though, of resilience the pivotal importance of ameliorating the distress of can- in health services, with oncology departments experiencing cer professionals with a critical focus on prioritising work- a return back to normal or higher clinical activity than force retention. National and international stakeholders before the pandemic. must act together as we recover from the COVID-19 crisis to The COVID-19 pandemic has also provided an opportu- promote the well-being of oncology professionals. Further nity for positive transformation. There are a number of detailed analyses, including multivariate analyses on factors adaptations to service delivery that should perhaps influencing outcomes of interest and interrogation of continue post-pandemic. For instance, the beneficial impact qualitative data collected from the survey series, are un- on equitable access to clinical trials by facilitating greater derway. The ESMO Resilience Task Force will shortly be access to telehealth encounters has been notable, particu- releasing a position paper with some key recommendations. larly for patients based in non-metropolitan settings who Ultimately, a healthy oncology workforce is a matter of often face greater travel time and financial toxicity to access urgency. It is essential for supporting the well-being of our novel therapies.18 Remote site visits, greater use of elec- patients and their loved ones, many of whom know that tronic signatures, and virtual meetings have also improved their time together is limited. efficiency in clinical trial delivery which has been beneficial for all stakeholders. For example, in Singapore, the phar- ACKNOWLEDGEMENTS macy drug courier service has allowed a substantial number We would like to thank all participants for taking part, and of patient prescriptions to be sent directly to their homes, national societies and organizations who helped distribute improving overall quality of care.19 Virtual outpatient pa- the survey. We would also like to thank Francesca Longo, tient visits conducted by Clinical Pharmacy Specialists in Mariya Lemosle, and Katharine Fumasoli from the ESMO New York City optimised treatment delivery and patient Head Office for providing vital administrative support for safety whilst reducing in-person visits.20 the delivery of this study. Although our survey participants have highlighted ongoing challenges with accessing professional develop- FUNDING ment opportunities, international initiatives such as the introduction of virtual oncology fellowships and dedicated This work was supported by the European Society for Medical Oncology (ESMO): for license to use Qualtrics educational webinars have been examples of positive re- (online survey platform), and publication of figure fee. sponses.21 The increased access to virtual oncology con- ferences and virtual mentorship programmes have also proven popular.22 In the long term, these initiatives may DISCLOSURE increase the ability of more staff to participate in profes- KHJL is currently funded by the Wellcome-Imperial 4i Clin- sional development activities while balancing other com- ical Research Fellowship, and reports speaker honorarium mitments and mitigating financial cost. from Janssen, outside the submitted work. KP’s institution 8 https://doi.org/10.1016/j.esmoop.2021.100374 Volume 7 - Issue 1 - 2022 K. H. J. Lim et al. ESMO Open received speaker fees or honoraria for consultancy/advisory ESMO Resilience Task Force survey II. ESMO Open. 2021;6(4): roles from AstraZeneca, Eli Lilly, Gilead Sciences, Medscape, 100199. 6. Maringe C, Spicer J, Morris M, et al. The impact of the COVID-19 Merck Sharp & Dohme (MSD), Novartis, Pfizer, Pierre Fabre, pandemic on cancer deaths due to delays in diagnosis in England, Hoffmann La Roche, Mundipharma, PharmaMar, Seagen, UK: a national, population-based, modelling study. Lancet Oncol. Teva, and Vifor Pharma; KP’s institution received research 2020;21(8):1023-1034. grants from MSD and Sanofi; KP received travel support 7. Clark JJ, Dwyer D, Pinwill N, Clark P, Johnson P, Hackshaw A. The effect from AstraZeneca, Novartis, Pfizer, PharmaMar, and Roche; of clinical decision making for initiation of systemic anticancer treat- ments in response to the COVID-19 pandemic in England: a retro- all outside the submitted work. CO reports research funding spective analysis. Lancet Oncol. 2021;22(1):66-73. and honoraria from Roche; travel grant and honoraria from 8. Campbell C, Sommerfield T, Clark GRC, et al. COVID-19 and cancer medac Pharma and Ipsen Pharma; and travel grant from screening in Scotland: A national and coordinated approach to mini- PharmaMar; outside the submitted work. EÉ reports mising harm. Prev Med. 2021;151:106606. speaker honoraria, travel support and advisory board: Bayer, 9. Richards M, Anderson M, Carter P, Ebert BL, Mossialos E. The impact of the COVID-19 pandemic on cancer care. Nat Cancer. 2020;1:565-567. Roche, Servier, Amgen, Pierre-Fabre, Sanofi Aventis, MSD, 10. Dyrbye LN, Satele D, Sloan J, Shanafelt TD. Utility of a brief screening and Merck Serono; outside the submitted work. TA reports tool to identify physicians in distress. J Gen Intern Med. 2013;28(3): personal fees from Pierre Fabre and CeCaVa; personal fees 421-427. and travel grants from Bristol Myers Squibb (BMS); grants, 11. Dyrbye LN, Satele D, Shanafelt T. Ability of a 9-item Well-Being Index to personal fees, and travel grants from Novartis; and grants identify distress and stratify quality of life in US workers. J Occup En- viron Med. 2016;58(8):810-817. from NeraCare, Sanofi, and SkylineDx; all outside the sub- 12. Murali K, Makker V, Lynch J, Banerjee S. From burnout to resilience: an mitted work. PG reports personal fees from Roche, MSD, update for oncologists. Am Soc Clin Oncol Educ Book. 2018;38:862-872. BMS, Boerhinger-Ingelheim, Pfizer, AbbVie, Novartis, Lilly, 13. Murali K, Banerjee S. Let’s address burnout in oncologists and AstraZeneca, Janssen, Blueprint Medicines, Takeda, Gilead, reimagine the way we work. Nat Rev Clin Oncol. 2019;16(1):1-2. and ROVI, outside the submitted work. ML acted as a 14. Jiwnani S, Ranganathan P, Tiwari V, et al. COVID-19 pandemic and its gendered impact on Indian physicians. JCO Glob Oncol. 2021;7:1093- consultant for Roche, AstraZeneca, Lilly, and Novartis, and 1100. received honoraria from Theramex, Roche, Novartis, 15. Garrido P, Adjei AA, Bajpai J, et al. Has COVID-19 had a greater impact Takeda, Pfizer, Sandoz, and Lilly, outside the submitted on female than male oncologists? Results of the ESMO Women for work. CBW reports speaker honoraria, travel support, and Oncology (W4O) Survey. ESMO Open. 2021;6(3):100131. participation in advisory boards: Bayer, BMS, Celgene, 16. Wu P, Fang Y, Guan Z, et al. The psychological impact of the SARS epidemic on hospital employees in China: exposure, risk perception, GlaxoSmithKline (GSK), Merck, Rafael, RedHil, Roche, Serv- and altruistic acceptance of risk. Can J Psychiatry. 2009;54(5):302-311. ier, Shire/Baxalta, Sirtex, and Taiho, as well as research 17. Dyrbye LN, Satele D, Sloan J, Shanafelt TD. Ability of the physician well- support from Roche, outside the submitted work. JBAGH being index to identify residents in distress. J Grad Med Educ. reports personal fees for advisory role in Neogene Tx; grants 2014;6(1):78-84. and fees paid to institution from BMS, MSD, Novartis, 18. Pennell NA, Dillmon M, Levit LA, et al. American Society of Clinical Oncology road to recovery report: learning from the COVID-19 expe- BioNTech, Amgen; and fees paid to institution from Achilles rience to improve clinical research and cancer care. J Clin Oncol. Tx, GSK, Immunocore, Ipsen, Merck Serono, Molecular 2021;39(2):155-169. Partners, Pfizer, Roche/Genentech, Sanofi, Seattle Genetics, 19. Kanesvaran R, Chia CS, Yap SP, et al. Cancer versus COVID-19: a co- Third Rock Ventures, Vaximm; outside the submitted work. ordinated Disease Outbreak Response System (DORS) to combat CH reports being Director of a private company Hardy COVID-19 at the National Cancer Centre Singapore. Ann Acad Med Singap. 2020;49(10):807-809. People Ltd, outside the submitted work. SB reports research 20. Yerram P, Thackray J, Modelevsky LR, et al. Outpatient clinical phar- grant (institution) from AstraZeneca, Tesaro, and GSK; macy practice in the face of COVID-19 at a cancer center in New York honoraria for advisory boards or lectures from Amgen, City. J Oncol Pharm Pract. 2021;27(2):389-394. AstraZeneca, Genmab, Immunogen, Mersana, MSD, Merck 21. Gillett C, Mason S, Fleming L, Mayer DK, Bryant AL. An academic- Serono, OncXerna, Pfizer, Roche, Shattuck Labs, Clovis, practice partnership during COVID-19 pandemic: Transitioning from a clinical to virtual fellowship. J Clin Nurs. 2021. Takeda outside the submitted work. KM, KK, MOC, ET, BD, 22. Morgan G, Tagliamento M, Lambertini M, et al. Impact of COVID-19 on and GM have declared no conflicts of interest. social media as perceived by the oncology community: results from a survey in collaboration with the European Society for Medical Oncology (ESMO) and the OncoAlert Network. ESMO Open. 2021;6(2): REFERENCES 100104. 1. Mayor S. COVID-19: impact on cancer workforce and delivery of care. 23. Hlubocky FJ, Symington BE, McFarland DC, et al. Impact of the COVID- Lancet Oncol. 2020;21(5):633. 19 pandemic on oncologist burnout, emotional well-being, and moral 2. Paterson C, Gobel B, Gosselin T, et al. Oncology nursing during a distress: considerations for the cancer organization’s response for pandemic: critical reflections in the context of COVID-19. Semin Oncol readiness, mitigation, and resilience. JCO Oncol Pract. 2021;17:365- Nurs. 2020;36(3):151028. 374. 3. Chazan G, Franchini F, Alexander M, et al. Impact of COVID-19 on 24. Rykers K, Tacey M, Bowes J, et al. Victoria (Australia) radiotherapy cancer service delivery: results from an international survey of response to working through the first and second wave of COVID-19: oncology clinicians. ESMO Open. 2020;5(6):e001090. Strategies and staffing. J Med Imaging Radiat Oncol. 2021;65(3):374- 4. Banerjee S, Lim KHJ, Murali K, et al. The impact of COVID-19 on 383. oncology professionals: results of the ESMO Resilience Task Force 25. Hoffman KE, Garner D, Koong AC, Woodward WA. Understanding the survey collaboration. ESMO Open. 2021;6(2):100058. intersection of working from home and burnout to optimize post- 5. Lim KHJ, Murali K, Kamposioras K, et al. The concerns of oncology COVID19 work arrangements in radiation oncology. Int J Radiat professionals during the COVID-19 pandemic: results from the Oncol Biol Phys. 2020;108(2):370-373. Volume 7 - Issue 1 - 2022 https://doi.org/10.1016/j.esmoop.2021.100374 9