Practitioners and patients were recruited through GP practices in Bristol and the surrounding areas. The West of England Clinical Research Network (CRN) informed these practices about the study and passed details of interested practices to the research team. The research team selected practices to support study recruitment based on their varying deprivation deciles and sociodemographic characteristics of their patient populations.

Practitioners were informed about the study by practice managers. Eligible practitioners were those working in to take part in the study and providing mental healthcare to patients with anxiety or depression. Practitioners were excluded if they did not meet both of the inclusion criteria. Practitioners willing to be interviewed emailed the research team their contact details, clinical role and demographic details (eg, age and gender). This information, alongside knowledge of their practice, was used to purposively sample practitioners of varying roles, age and gender. Initially, the research team aimed to recruit GPs and practice nurses, as they both have a role in supporting primary care patients with mental health problems. However, as data collection progressed, it became apparent that many practices had other allied health professionals (AHPs) who also provided mental healthcare. Therefore, the study’s inclusion criteria were expanded to include AHPs, including practice pharmacists and well-being coaches.12

Patients were invited for interview via their GP practices. Practices identified potential participants through searches of their practice electronic databases. Eligible patients were aged ≥18 years, with a current diagnosis of anxiety and/or depression, who had consulted their practice for their CMD either remotely or in-person, in the past 6 months. Patients were excluded if they had bipolar, schizophrenia, personality disorder, dementia or substance (alcohol/drug) misuse in the past year. Practices posted an invitation letter and information sheet to potential participants. Those interested in participating returned response forms (contact details, sociodemographic (age, gender, ethnicity), CMD (diagnosis, duration, past episodes and medication) and mode of mental healthcare access (in-person, telephone, video and/or e-consult)) using reply paid envelopes. This information, alongside their practice’s deprivation decile, was used to purposively sample individuals varying in these attributes.

Interviews were conducted by CA (an experienced qualitative researcher) by telephone or videocall. Verbal consent to participate was audio-recorded immediately prior to interview. Two topic guides (one for practitioners; one for patients) were used to ensure consistency across the interviews (online supplemental file 1, 2). The guides were developed in parallel and were based on the aims of the study, relevant literature and discussions with the research team and patient and public involvement (PPI) contributors.

Practitioners and patients were asked about their views and experiences of remote consultations for CMDs, the benefits, challenges and facilitators of these consultations and when they thought in-person appointments were preferable to remote modes of consultation. After each interview, patients and practitioners were asked to complete a brief questionnaire (employment status, education, marital status), the information from which was used to describe those interviewed and considered during data analysis. All interviewees were financially reimbursed to thank them for their time.

Interviews were audio-recorded using an encrypted voice recorder, professionally transcribed, fully anonymised and checked for accuracy. Data collection and analysis proceeded in parallel so that analytical insights from early data collection could shape later data collection and so that data collection would end when data saturation had been reached, that is, no new themes were identified in later interviews. Thematic analysis was undertaken following the steps defined by Braun and Clarke,13 to allow comparisons to be made within and across patient and practitioner interviews.

Each data set was analysed separately. Two investigators (CA and KT) read and re-read a sample of practitioner and patient transcripts to identify possible codes and then met to compare and discuss their coding and interpretation of the data. Two preliminary coding frameworks (one for practitioners; one for patients) were developed in parallel, to ensure codes common to both were included in each. Each coding frame was revised as new codes were identified in later transcripts, and previously coded transcripts were re-coded where necessary. All transcripts were electronically coded in NVivo (V.12) so that data relating to each code could be easily extracted. To aid interpretation of the data, extracted data were summarised in a table where rows represented interviewees and columns specific codes. The tables were then read and re-read to identify key themes and deviant cases. Once each data set was fully analysed, the views and experiences of practitioners and patients were compared, allowing similarities and differences to be identified.

A PPI co-investigator (LT) provided input at every stage of the research. LT and two PPI contributors, all of whom had lived experience of anxiety or depression, met with CA and commented on initial ideas for the study and advised on patient-facing materials and topic guides. Seventeen months later, CA, LT and six PPI contributors discussed the study findings via an online meeting and written feedback. All felt that the results were important and relevant and highlighted key messages and avenues for dissemination.

Table 1

Characteristics of practitioners interviewed and practice deprivation deciles

Table 2

Characteristics of patients interviewed and practice deprivation deciles

Practitioners and patients view remote consultations as improving access to care for patients with CMDs, but appropriateness depends on patients’ preferences and circumstances, previous contact and symptom severity. Remote consultations can be easier for patients who are feeling very low or anxious and can be more convenient, particularly for follow-up appointments. However, practitioners say that risk is harder to assess remotely due to the loss of information, and some are more likely to prescribe medication if they were not confident in their ability to assess patient risk when consulting remotely. Patients state that they may withhold information, find it harder to disclose sensitive information or have concerns about not being believed or their distress recognised. For these reasons, they note that in-person appointments are important for initial consultations and to provide a safe space to facilitate disclosure.

Remote consultations may provide patients with quicker access to care, and hence patients may accept this offer as they are aware of the pressures on general practice. However, patients find waiting for a telephone call from the GP difficult because of the wide calling window and are often uncertain how quickly they will receive a response to an e-consult. While remote consultations may improve access to care, practitioners report that remote consultations for mental health problems are not necessarily more time-efficient, with patient contacts via e-consult often needing a follow-up telephone call.

Patients and practitioners view continuity of care and training in remote consulting for mental health problems as facilitators. However, they identified training needs around identifying risk in the absence of non-verbal cues, developing skills in active listening, conveying empathy and encouraging patient disclosure.

PPI was embedded in all stages of the study, thereby influencing every stage of the research. We recruited interviewees with a wide range of sociodemographic and clinical characteristics. However, we recruited patients who had anxiety or depression in their recent medical history. This study, therefore, does not capture the views of those who have not yet sought help. Likewise, patients and practitioners were self-selecting in terms of being individuals who responded to the invitation, which stated the research was focused on remote methods (telephone, video, e-consult) for consulting in general practice, and that the interviews were being held by telephone or videocall. It is possible that those interviewed were individuals who had strong preferences for or against remote consultations and were comfortable being interviewed remotely. Patients and practitioners had little experience of receiving or providing care by videocall, thus limiting our understanding of the benefits, challenges and facilitators of this mode of consultation. Future studies could seek to interview primary care patients who have used videocalls for CMDs. That said, most practices do not offer video consultations frequently, so it would be challenging to seek in-depth views on this mode.14

Concerns about the use of telephone and video consulting in primary care highlight missed diagnoses, challenges to therapeutic relationships,15 difficulties in doctor-patient interactions, feeling rushed16 and less engagement with services.17 Our study supports some of these, suggesting risk is more difficult to assess over telephone, potentially taking longer and challenges in establishing patient-practitioner rapport. Practitioners in our study also noted that e-consults may not be that efficient, often leading to additional contacts, and that telephone appointments may take longer than in-person consultations. This is in line with a systematic review investigating the impact of telephone and video primary care consultations—for any indication—finding that they may increase the need for additional GP appointments compared with in-person,18 and with wider literature around mental health appointments taking longer.8 19 20 It is clear from our study and others14 that videocalls are not often used in general practice.

Many of our interviewees said that remote modes increased access to care, and this supports findings by NHS England.5 Unfortunately, increased provision of care by practitioners does not always equate to increased engagement by those accessing the care, with some patients sharing their experience of limiting how much information they disclosed over the telephone, potentially due to the transactional nature and not wanting to take up practitioners’ time. A recent briefing highlights that it is important to consider access to healthcare as more than just provision of appointments; it is also about broader factors, such as health literacy, barriers to reach services and how people identify themselves as candidates for medical attention.21

There is also a contradiction in our data between the importance of continuity of care yet the need to be able to respond rapidly in situations where risk is high. This is challenging to resolve as patients’ ability to see the same practitioner is likely dependent on service delivery pressures and the urgency of the problem, which is often closely linked to risk, further compounding this contradiction.22 Some practitioners in our study suggested they might be more likely to prescribe over the telephone if they were concerned about risk. This is not in line with analyses conducted prior to the pandemic using secondary care electronic health records, which did not find an increase in psychiatric prescribing.23 However, there is some evidence that remote consultations have higher doses of anxiolytics24 and increased antibiotic prescribing rates,25 suggesting mode may be a factor in prescribing decisions.

Most interviewees preferred in-person appointments for initial consultations for a CMD, and a survey of patient experiences during the pandemic—for all indications—also makes this recommendation.26 In our study, in-person consultations at the practice were considered key in providing a safe and private space to discuss concerns. This is consistent with other research exploring the role of place in GP consultations.27 Not all patients are able to access appropriate private spaces for remote consults, and this can negatively impact their experiences of care, thus creating inequities. It was noted that such spaces are particularly important for those discussing sexual or mental health problems.27 Greenhalgh et al28 note that staff training for remote consultations has not kept pace with NHS plans. Although there are training resources covering remote risk assessment and rapport,29–31 there is a lack of training specific to remote primary mental healthcare. Interviewees in our study suggested that mental health consultations may be more complex than appointments for physical health, requiring greater skill to encourage disclosure and assess risk. Other interviews with patients and practitioners have identified training as a facilitator in implementing remote mental health consultations during the pandemic,32 although practitioners in our study said they had not received such training.

This study aims to understand primary care patients’ and practitioners’ views and experiences on the benefits, challenges and facilitators of remote consulting for CMDs. Findings suggest that it is important to offer in-person appointments, especially when patients initially seek help for anxiety or depression, and are unknown to the practitioner. Ideally, practices should provide a narrower window for a telephone call, as many patients find waiting for the telephone call difficult. The challenges in assessing risk remotely and patients’ hesitancy to discuss their mental well-being mean that it is important for practitioners to convey empathy, ask questions that encourage disclosure and demonstrate active listening. Some practitioners may benefit from training in this area. There is a need for future research to review existing training, determine whether it needs to be expanded to cover mental health consultations more specifically and, if so, consider how best to deliver it with busy practitioners with subsequent evaluation.

Not applicable.

This study involved human participants and was approved by the South West - Frenchay Research Ethics Committee (REC reference: 22/SW/0136). Participants gave informed consent to participate in the study before taking part.