Initially, clinical academics (professionals holding both higher education and healthcare appointments) nationwide were strongly encouraged and often required by national funding bodies to redeploy from research roles to provide direct healthcare. This prompted tension and anxiety. We felt we were balancing prioritised clinical service with both sensed external expectations to support pandemic-related research and internal expectations to meet existing research milestones. Supporting the pandemic response through short-term increased clinical work was necessary but did not fit with our various academic aims around improving the long-term quality, effectiveness, and person-centeredness of geriatric emergency medicine. Many redeployed healthcare professionals experienced anxiety and stress, and although this has not been separately studied in academic clinicians, the sense of duty to continue academic work seemed to create an additional source of stress.³

While planning our approach to restarting research, we felt the responsibility to prevent virus transmission and carefully deliberated the risks to potential patient participants. We were also aware of the health risks to the wider research team undertaking research in an emergency care environment.

One of our studies had been investigating healthcare system dynamics and was using ethnography with prolonged emergency care field observation by non-clinicians. We considered the observers to be at too great a risk of exposure. We also felt that workplace observations might add to the pressures under which healthcare professionals were working. We considered restarting to be unfeasible, and therefore kept this study suspended.

Our other studies had been using interviews to investigate older people’s emergency care experiences and outcome goals. We had previously used visits to older people’s homes to support in-depth freedom of expression through minimized professional deference. This approach became impossible due to the risk of researchers transmitting coronavirus to older people, who were already known to be particularly vulnerable.

However, with the support of the university, the hospital, and our lay research partners, we were able to restart our interview study with minimized researcher-participant contact using a modified schedule. Direct contact during initial hospital attendance was still required in order to identify potential study participants and to obtain their consent. We felt that these short periods of additional patient contact for research recruitment, using the recommended PPE, were acceptable as they would be unlikely to increase the virus exposure risk. We obtained ethics permissions for a hybrid approach where this in-person recruitment was followed by fully remote or short face-to-face followed by remote interviews, depending on patient preference.

Fewer older people consented to participate when approached during the early (October 2020 to March 2021) phase of restarting research (30% vs 75% prior to the pandemic). Reasons for declining included a reluctance to join interactions in general and feeling ‘fed-up’ of discussing health issues. Some people described feeling distracted by fear of catching COVID-19 while in hospital. Many older people lived in social isolation before and during pandemic restrictions, and when acutely unwell seemed to need time to adjust to the sudden social interactions of hospital and research.⁴ We also felt that as there were fewer non-COVID admissions at this time due to fear of the virus and hospital avoidance, the older people who were attending hospital, and hence who could be approached for research, were more unwell and therefore less likely to participate in an interview study.

Our interview studies allowed consultee consent by relatives or close friends if the person was unable to consent for themselves. As visitors were not allowed into hospitals this restricted our ability to recruit; therefore, our sampling may have become biased against people with barriers to independent participation, including cognitive impairment. Debate is needed with researchers, regulators, and ethics committees about the use of remote methods to process consultee consent for unaccompanied participants.

Recruitment of professionals for interviews was also difficult, as staff working under pressure may have lacked motivation, time, or capacity to contribute to extra activities. We found this especially among more junior staff, for whom contracts often rotate between departments and hospitals, who seemed particularly strained.

The use of PPE facemasks may have been a barrier to recruitment. Communication was made difficult by voices being muffled, particularly where participants had hearing difficulties. Masks could be perceived as intimidating, particularly by people who had cognitive impairment. Altered researcher appearance may have impacted data collection in interviews. Despite introductions as university researchers and seeking negative as well as positive views, the uniformity of facemasks, visors, and aprons in the healthcare setting may have blurred participants’ distinction between academic and clinical professionals, causing them to feel unable to criticize clinical staff or processes. Poorer recording quality also made transcription of interviews more challenging.

We initially offered remote interviews conducted by telephone. Few participants agreed to telephone interviews, citing difficulties hearing or using equipment, or the impersonal nature of calls. We had mistakenly assumed that most older people would be comfortable using a telephone, whereas researchers recently were unable to contact 40% individuals by that method.⁵ During telephone interviews, non-verbal cues including use of silence were hard to judge, and as reported previously, we felt questions were often misheard or misunderstood.⁶

While acceptance of telephone interviews was generally poor, some people we approached requested discussions by video call, leading us to challenge our preconception that older people would have preferred telephone over internet communication. Many older people use the internet regularly, and feasibility has previously been demonstrated for delivering health screening and information electronically.⁷ Some participants used mobile phone or tablet apps. During the consent process, we discussed participants’ preferences towards a shorter in-person, longer remote, or hybrid combination interview.

We found that participants seemed comfortable and secure sharing in-depth reflections from their homes. This enabled deeper discussion around healthcare and existential topics which we had found difficult to explore with participants who were in hospital. Participants had considered their experiences and outcomes from health and healthcare, and more often discussed negative events or unfulfilled goals when participating from home.

Prior to pandemic restrictions, the in-person meetings with our lay collaborators had been attended by around five lay representatives. However, the virtual sessions during lockdown were regularly attended by over ten members allowing a broader range of opinions to be heard. Between meetings we corresponded by email and by telephone. Continuing a virtual option may be a method to increase lay representative collaboration in research in future studies.