'Reluctant pioneer': A qualitative study of doctors' experiences as patients with long COVID

Abstract Background The coronavirus disease (COVID‐19) pandemic has had far‐reaching effects upon lives, healthcare systems and society. Some who had an apparently 'mild' COVID‐19 infection continue to suffer from persistent symptoms, including chest pain, breathlessness, fatigue, cognitive impairment, paraesthesia, muscle and joint pains. This has been labelled 'long COVID'. This paper reports the experiences of doctors with long COVID. Methods A qualitative study; interviews with doctors experiencing persistent symptoms were conducted by telephone or video call. Interviews were transcribed and analysis conducted using an inductive and thematic approach. Results Thirteen doctors participated. The following themes are reported: making sense of symptoms, feeling let down, using medical knowledge and connections, wanting to help and be helped, combining patient and professional identity. Experiencing long COVID can be transformative: many expressed hope that good would come of their experiences. Distress related to feelings of being ‘let down’ and the hard work of trying to access care. Participants highlighted that they felt better able to care for, and empathize with, patients with chronic conditions, particularly where symptoms are unexplained. Conclusions The study adds to the literature on the experiences of doctors as patients, in particular where evidence is emerging and the patient has to take the lead in finding solutions to their problems and accessing their own care. Patient and Public contribution The study was developed with experts by experience (including co‐authors HA and TAB) who contributed to the protocol and ethics application, and commented on analysis and implications. All participants were given the opportunity to comment on findings.


| BACKG ROU N D
The novel coronavirus disease  pandemic declared by the World Health Organization in March 2020 has had far-reaching effects upon people's lives, healthcare systems and wider society.
Acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes dramatic variation in clinical outcome, from asymptomatic infection through to multi-organ failure and death. 1 In the United Kingdom (UK), over 190 000 people have been hospitalized with COVID-19 as of November 2020, with 17% requiring intensive care and 26% of these people dying. 2,3 In June 2020, NHS England published 'Aftercare needs of inpatients recovering from COVID- 19', estimating that up to that date, more than 95 000 patients had been admitted to hospitals across England with COVID-19 and it assumed 45% would need ongoing support. 4 Indeed, some estimates suggest that up to 50% of people hospitalized will need formal rehabilitation services. 5 Some people who had an apparently 'mild' COVID-19 infection (whether confirmed or suspected) continue to suffer from persistent symptoms, including fatigue, cognitive impairment ('brain fog'), neuropathy and paraesthesia, chest pain and palpitations, muscle and joint aches and shortness of breath. [6][7][8][9][10][11][12][13] This has been termed 'long COVID' by people with the persisting symptoms. 12,14,15 The NICE scoping document defines this as post-COVID-19 syndrome. 16 Evidence for the extent of long COVID is growing: for example, a team from Italy, one of the earliest affected countries in Europe, reported that 87% of people discharged from a Rome hospital were still experiencing at least one symptom 60 days after the onset of COVID-19 and 55% had three or more symptoms including fatigue (53%), difficulty in breathing (43%), joint pain (27%) and chest pain (22%), with 40% reporting that it had reduced their quality of life. 9 Public Health England published guidance on 7th September 2020 stating that around 10% of 'mild' COVID-19 cases who were not admitted to hospital have reported symptoms lasting more than four weeks and a number of hospitalized cases reported continuing symptoms for eight or more weeks following discharge. 17 Similarly, in a UK study (preprint) of a young, low-risk cohort of 201 participants affected by COVID-19, 70% were found to have organ damage with impairment in one or more organs four months after symptom onset; of note, 31% of participants were healthcare workers. 18 From early in the pandemic, doctors highlighted the vulnerability of healthcare workers in terms of both physical health and mental health. 19 Indeed, in the first wave, a sixth of admitted COVID-19 cases were healthcare workers and their households; for patient facing healthcare workers, there was a threefold increased risk of admission with COVID-19 compared with the general population. 20 What is notable about long COVID is that several doctors, including those who have become patients with long COVID, have reflected publicly on their experiences. 7,21 Previous literature on doctors who become patients describes the need to be seen as a person, and treated as a 'patient', but at the same time, doctors experience a loss of professional identity when they become ill. 22 They have also frequently reflected on the standards of care they have experienced and how they wanted to change clinical practice as a result. 22 Doctors may be less likely to seek help than the general public due to structural barriers such as difficulties accessing care during normal working hours, or inconsistent primary care provision due to frequent moves. 23 However, there may also be psychological barriers to help-seeking with the perception that doctors cannot become ill, and the patient is 'the one with the disease'. 24 Additionally, doctors have been described as having certain personality traits including perfectionism and denigration of vulnerability, which may also prevent, or at least reduce, help-seeking as they develop a 'medical identity' in which their professional responsibilities permeate all aspects of their lives. 23,25,26 Those doctors who do have time out of work for health reasons often describe isolation, frequently experiencing stigma from colleagues and family as well as expressing self-stigmatizing views of themselves as 'failures'. 27,28 The limited literature suggests that it is hard to be a doctor and a patient. This may be particularly problematic when the condition you are experiencing is novel and poorly understood. This qualitative study reports the experiences of doctors with long COVID. The analysis of interviews reported here is part of a wider study, 29 the aim of which was to explore the experiences of people with persisting symptoms following suspected or confirmed COVID-19 infection in March or April 2020, which became known as 'long COVID' during the course of their illness.

| ME THODS
A qualitative study was conducted, using semi-structured interviews with people with long COVID recruited through social media, to generate data. 29 This study was developed in collaboration with experts by experience (people with persisting symptoms, or long COVID).
Authors HA and TAB both have long COVID and thus are experts by experience; HA is a health services researcher and is married to a clinician, and TAB is a clinical academic. One of the authors (CCG) discussed the planned study with members of a post-COVID support group, facilitated by a clinical commissioning group.
A subset of interviews were conducted with doctors experiencing persistent symptoms following COVID-19 or suspected COVID-19 infection. Semi-structured interviews were chosen to allow participants to talk without limitation about the areas they felt were most important, while ensuring that all topics felt relevant by the research team were covered and perspectives were explored in detail. 30,31 The protocol and documents for University ethics application were developed in collaboration with experts by experience.

| Recruitment
Doctors with self-reported experiences of long COVID were recruited using social media posts (Facebook and Twitter). They were invited to contact the research team by email to express interest, and potential participants were sent a participant information sheet and consent form. Those doctors who agreed to participate in an interview were invited to pass on details of the study to others they knew who were experiencing similar symptoms, including sharing to peer support groups on social media ('snowball sampling'). 32

| Data collection
Interviews were conducted by telephone or using software such as Microsoft Teams or Zoom, according to the participant's preference, during July and August 2020. Consent was obtained (consent form completed and sent by email to members of the research team) in advance of the interviews, reconfirmed and audio-recorded at the start and end of the interview. Interviews were conducted by TK (social scientist) and CCG (academic GP), both experienced in qualitative research methods, who both kept reflective notes which contributed to analysis.
The topic guide was developed by the researchers in collaboration with experts by experience. The topic guide was modified iteratively throughout data collection and analysis.
Details including when acute COVID-19 infection was experienced, whether infection was confirmed by antigen [Ag] or antibody [Ab] testing, and demographic information (including age, ethnicity, occupation) were from participants, in order to contextualize the data and support description of the sample.
Following the interview, participants were asked if they wished to receive a summary of the findings and/or publications arising from the study. They were also offered a voucher to compensate them for their time and were sent a 'thank you' email. Comments sent back by participants (by email) who reviewed the summary of findings were incorporated into the ongoing analysis.
Digital audio-recordings of the interviews were transcribed by members of the research team, or a professional transcribing company.

| Data analysis
Data analysis, following an inductive and interpretive thematic approach 33,34 and applying principles of constant comparison, 35 was conducted by the research team. This form of analysis allows researchers to understand collective experiences of participants. 33 The research team represented a mix of professional backgrounds, allowing for robust analysis of the data from differing perspectives.
AKT and CCG analysed all transcripts; each coauthor analysed a subset. The codes, themes, and any alterations to the topic guide were discussed collectively and agreed upon by the research team.
Sample size was determined by inductive thematic saturation, which was judged to be the point at which no new codes or themes were developed from the analysis. 36,37

| RE SULTS
Thirteen doctors were interviewed. Eleven participants were female and eleven White British. None were working at the time of their interviews due to their persisting symptoms. Four participants were aged 30-39, five were aged 40-49, and three were over the age of 50 (see Table 1 for other participant demographics).
Interviews lasted between 57 and 120 minutes (mean 87 minutes). The following themes will be presented in this paper, with illustrative quotes identified by a pseudonym for each participant: making sense of symptoms, feeling let down, using medical knowledge and connections, wanting to help and wanting to be helped and becoming a more empathic doctor.

| Making sense of symptoms
Participants described both concern and confusion over the cause and nature of their symptoms, and the challenge of attempting to fit the symptoms they were experiencing into their own existing medical knowledge. That's very strange.' Alice They feared that their symptoms represented serious pathology, such as pulmonary embolism or myocarditis, that had not been considered or investigated and which would cause them significant long-term problems.
'If you're getting cardio and respiratory symptoms, pathology needs ruling out before rehab.' Harriet Most participants expressed concern over whether they would ever recover: 'It's difficult because I keep getting new things, which is one of the frustrations of this. The brain stuff seems to be getting better, to the level that I can function. When the brain wasn't working that made me very scared because I need my brain! Not to be blasé, but the chest pain and stuff I can still work because I can work remotely. If I don't have my brain I can't work, I can't plan, I can't string a sentence together… I did get a bit scared when I was ill for so long… at the moment in a way I'm not missing out on a lot of stuff because everything is virtual.' Irena Symptoms persisting after the initial illness were not expected and even contradicted by initial ideas about SARS-CoV-2 from knowledge available in early 2020. Because of this, participants reported seeking advice, support and information from online groups, including specific groups for doctors with persisting symptoms, in order to try to make sense of their own symptoms as well as seeking reassurance that their symptoms were shared by other people.

| Using medical knowledge and connections
Participants described how they were forced to self-advocate because their own GPs were unsure how to investigate and manage their symptoms.

| Wanting to help and wanting to be helped
All participants, while wanting help for their own symptoms, also emphasized how they also wanted to raise awareness of long COVID.
Participants reflected upon the therapeutic value of engaging in the interview process.

| Summary of findings
This is the first qualitative study reporting the perspectives of doctors experiencing persistent symptoms following infection with COVID-19 in early 2020, as new knowledge was emerging. Doctors Doctors had concerns about ongoing significant organ involvement and wanted such pathology to be excluded, having read about sequelae including myocarditis and pulmonary emboli following COVID-19.
These experiences emphasized a moral dilemma of wanting to be treated as a 'patient', yet, professionally, being dissatisfied with their care. All participants described the struggle of adjusting to their new dual identity as both doctor and patient and reflected on their worry that they would be seen as a 'heartsink' patient. 39 Reasons for participating in this study included wanting to help others, to contribute to the growing body of evidence around long COVID, but also to voice their experiences, to be heard. Field notes taken contemporaneously all included reflections from participants that it had been 'so helpful to have a chance to tell my story'. All the participants described how they also wanted help, to enable them and their doctors to understand their symptoms and be referred for investigations to look for underlying pathology. Most participants suggested that such help had not always been forthcoming.
Participants reflected on how the experience of long COVID might lead them to be more empathic when meeting patients with difficult to understand symptoms in the future.

| Strengths and limitations
While there is much discussion on social media and non-peerreviewed publications, [6][7][8]11,12,15 this is the first qualitative study (embedded within a larger study 29 ) that we are aware of that has focussed on the experience of doctors who have had persistent symptoms following COVID-19 infection in early 2020.
Most participants were general practitioners, which reflects recruitment using social media and snowball sampling. Participants were predominantly white, female and in a younger age group which may reflect the likely preponderance of women with long COVID. 40 The apparent homogeneity of the sample has implications for our claims of saturation and may be a limitation. While we feel our claims are justified for the themes identified, we acknowledge that alternative accounts from demographic groups not captured here could add novel insights and should be explored in further research.
Experts by experience (HA and TAB) were key members of the study team. In addition, other people with long COVID contributed to study design, ethics application, the patient information sheet and the interview topic guide. All study participants contributed to ongoing analysis, by providing comments on analysis summaries as analysis progressed.

| Comparison with previous literature
There is a developing body of literature on long COVID, and the research team have published an earlier paper reporting analysis of 24 interviews. 29 There are also patients' narratives on social media and articles in journals, including from people with long COVID who are also doctors. 7,12 Physicians have previously described the ways in which their own illness has been ignored by the medical community, how their dignity and identity have been diminished through the experience of illness, poor communication which has left them upset and confused, and the perceived arrogance of the doctors managing them. 24 Other reasons why doctors may be unwilling to seek help for mental health problems are stigma and the culture of invulnerability pervasive across medicine, which can result in asking for help being seen as 'weak' and 'letting down the profession'. 23 Doctors have in the past described themselves as 'wounded healers', borrowing Jung's term thought to derive from the ancient Greek legend of Asclepius who, in recognition of his own wounds, established a sanctuary at Epidaurus where others could be healed. 41 Fox and colleagues interviewed 17 UK GPs who had experienced significant illness, and confirmed that the ethos of invulnerability to illness persists among GPs. 42 Bradley reflects on the challenges that ill doctors face: their prior knowledge of the health system and personnel, possible concerns about the quality of care and difficulty finding the right GP. 43 The narratives of the participants in our study reflect on all these barriers and describe how they have tried to overcome them.  27 Medical training may create a doctor identity or 'medical self', which allows doctors to do their job effectively despite working long and stressful hours in the proximity of sickness and disease. 23 Although these characteristics predominantly enable doctors to work productively, they can distort doctors' ability to seek help when unwell and tolerate the role of a patient, and can make consultations challenging to negotiate due to unfamiliar power dynamics. When doctors are unwell, they may describe a conflicted sense of identity that, for some, manifested as a 'loss of self'. It has previously been reported by doctors that their own identity has been taken over by 'the role of the doctor', and when they were unwell they attempted to reclaim their 'sense of self'. 46 The perception of doctors as 'invulnerable' can make it more challenging for them to return to work due to concern around negative responses and a lack of support from colleagues. 27 The duration, full extent and severity of long COVID is not yet clear. Arnold reported that although patients with COVID-19 may remain symptomatic at 8-12 weeks, clinical abnormalities requiring action are infrequent, especially in those without a supplementary oxygen requirement during the acute phase of illness. It should be noted, however, that the COVERSCAN study suggested organ damage in 70% of patients that may need follow-up or intervention; evolving evidence will clarify this. 18 What is clear from the narratives of our participants is that doctors must play a significant role in validating the experience of those with long COVID, investigating worrying symptoms and offering emotional support while participants recover.
There is an argument that doctors with long COVID may need different or additional support because of the added complexities of navigating contractual issues and the challenge of returning to work (and its associated risk of exposure). 47 The younger age (<45 years) of those who are typically affected by long COVID means that there are longer-term workforce implications of the increase in doctors off sick. This not only may affect staffing levels in the 'second wave' of COVID-19, but staffing levels in the coming months and potentially years following this as the timeline of recovery is as yet uncertain. 47

| Implications for research and practice
It is important to conduct further research to understand the characteristics of doctors who are more or less likely to be affected by long COVID, and cohort studies should explore the process and length of patient recovery. This will have implications for longerterm workforce planning and developing ways to support affected doctors back into work. The experiences of doctors who treat other doctors with long COVID could also be explored.
As these doctors reflected, living with long COVID can be transformative and many expressed hope that some good will come of their experiences and challenges. This study may help to destigmatize illnesses where symptoms are as yet unexplained, such as long COVID, and study participants highlighted that the experience of illness would make them better able to care for and empathize with patients with ongoing and unexplained symptoms.