GPs’ views on emergency care treatment plans: an online survey

Background A holistic approach to emergency care treatment planning is needed to ensure that patients’ preferences are considered should their clinical condition deteriorate. To address this, emergency care and treatment plans (ECTPs) have been introduced. Little is known about their use in general practice. Aim To find out GPs’ experiences of, and views on, using ECTPs. Design & setting Online survey of GPs practising in England. Method A total of 841 GPs were surveyed using the monthly online survey provided by medeConnect, a market research company. Results Forty-one per cent of responders' practices used Recommended Summary Plan for Emergency Care and Treatment (ReSPECT) plans for ECTP, 8% used other ECTPs, and 51% used Do Not Attempt Cardiopulmonary Resuscitation (DNACPR) forms. GPs were the predominant professional group completing ECTPs in the community. There was broad support for a wider range of community-based health and social care professionals being able to complete ECTPs. There was no system for reviewing ECTPs in 20% of responders’ practices. When compared with using a DNACPR form, GPs using a ReSPECT form for ECTP were more comfortable having conversations about emergency care treatment with patients (odds ratio [OR] = 1.72, 95% confidence interval [CI] = 1.1 to 2.69) and family members (OR =1.85, 95% CI = 1.19 to 2.87). Conclusion The potential benefits and challenges of widening the pool of health and social care professionals initiating and/or completing the ECTP process needs consideration. ReSPECT plans appear to make GPs more comfortable with ECTP discussions, supporting their implementation. Practice-based systems for reviewing ECTP decisions should be strengthened.


Introduction
3][4] In response, there has been a move to a more holistic approach to recording recommendations about future treatment decisions with the development of emergency care treatment plans (ECTPs).These plans encompass broader clinical decision making, while still describing DNACPR recommendations.6][7][8] In 2016, the Resuscitation Council UK (RCUK) developed a model of ECTP that was intended to be used nationally across primary and secondary care. 9,10By 2023, this model, the Recommended Summary Plan for Emergency Care and Treatment (ReSPECT), had been adopted to some extent in 65% of integrated care systems in England (personal communication, RCUK).
An evaluation of ReSPECT in early adopting acute NHS trusts identified challenges, with clinicians suggesting that the conversation, and completion of a ReSPECT plan, would be better in primary care with conversation(s) taking place over a period of time, with a GP with whom the patient has an ongoing relationship. 11Focus groups with GPs identified challenges to using ReSPECT in primary care.GPs with experience of completing ReSPECT forms conceptualised them as end-of-life planning documents, limiting the population for whom a plan might be initiated.Recommendations on GP-initiated plans differed from those completed in hospitals, reflecting the context in which they were expected to be used. 12he COVID-19 pandemic increased focus on the role of ECTP.Regulatory authorities identified the importance of individualised conversations with patients about future treatment decisions, carried out by healthcare professionals with the requisite skills, knowledge, and confidence. 13,14ittle is known about how GPs view and make use of ECTPs with their patients.We report a national survey measuring GPs' use of ECTPs, their views on using ECTPs in primary care; their readiness to complete plans with their patients,

Method
The survey is part of a larger mixed-methods evaluation of the use of ECTPs in primary care. 15nformed by our qualitative work in GP practices, and with involvement from our patient and public advisory group, we developed a questionnaire survey to measure the views of GPs working in England regarding the use of ECTPs in primary care (Table 1).For the survey, we included DNACPR forms as a type of ECTP, albeit one limited to a single emergency treatment decision.Key questions of interest were to identify which factors might predict how comfortable GPs were in having ECTP conversations with a patient or family member, assessed using 5-point Likert scales.After developing our initial questions, we refined these using think-aloud interviews with six GPs.
We outsourced data collection to medeConnect, a market research company providing a monthly online survey of 1000, regionally representative UK GPs (Appendix 1). 16There were no restrictions on multiple GPs from the same practice completing the survey.The final questions, formatted in an online survey, were tested by the company and the research team.These are presented in Tables 2-4.

Sample size and statistical analysis
For a binary outcome (very comfortable or fairly comfortable versus all other responses) a sample size of 1000 (the size of the medeConnect monthly survey) would: if 50% were 'comfortable', provide precision of 6.2%; or, if 80% were comfortable, provide a precision of 5%.
In addition to descriptive statistics for each question, we present logistic regression analyses investigating the variables associated with how comfortable GPs are having ECTP discussions with patients or their families.We initially did unadjusted logistic regression analyses with gender, GP role, NHS region, type of area (that is, major conurbation, large town), years since completion of GP training, and use of ReSPECT form versus DNACPR or another ECTP as explanatory variables.We then constructed a fully adjusted logistic regression model including all the explanatory variables.As a sensitivity analysis, we repeated this using a backward elimination approach.

Results
The survey ran in November 2022.Only the 841 (of 1000) GPs surveyed who practised in England were invited to complete it (Appendix 2).We did not achieve our original sample because of the need to gain additional ethics approval to include the devolved nations.Responders' demographic characteristics were broadly representative of GPs in England, although males and GP partners or principals were overrepresented (Table 1).Just over half (51%) of responders reported their practice used standalone DNACPR forms.ReSPECT forms were used by 41%, and 8% used other ECTP forms (Table 2).There were substantial regional differences in the forms used ranging from just over threequarters of GPs in East (79%) and West Midlands (80%) using ReSPECT forms to DNACPR forms being predominantly used by GPs in London (76%), the North East (75%), and North West (77%) (Supplementary Table S1).
Overwhelmingly (93%), responders reported that GPs completed ECTPs within their practices, and they felt that GPs should be able to complete them (Table 2).However, there were substantial disparities between who was reported as completing ECTPs and who they thought could complete  them.Consistently, responders suggested that a wider range of healthcare professionals should be able to complete the forms.For example, no responders reported GP trainees (registrars) completing ECTPs in their practice, whereas 62% felt they should be able to do this task (Table 2).Just over threequarters (77%) of responders thought that advanced nurse practitioners should be able to complete ECTPs but only about one-quarter (28%) reported that this currently happened (Table 2).Similarly, there was broad support for a wide range of community-based health and social care professionals being able to complete ECTPs; for example, more than 80% supported senior nurses completing them (82%-95%), half supported less senior nurses completing the forms (51%-56%), and one-quarter (25%) supported senior care home staff to do this task (Table 2).When GPs would consider completing an ECTP was primarily influenced by the patient's health state; 97% would consider completing a form if they felt the patient had a life expectancy of less than 1 year, 86% when a patient has been diagnosed with a life-threatening condition, and 71% when a patient entered a care home.Just under one in four (24%) GPs would consider completing a plan based on the patient's age alone.An ECTP was considered by fewer responders for people who were severely disabled (59%) or living with a long-term condition (61%) (Table 2).
A mixed pattern was seen for when GPs might consider reviewing ECTPs.Strikingly, one in five (20%) responders reported that their practices had no system for reviewing forms.Only a minority had routine systems in place for reviewing these; annually (37%), 6-monthly (12%), or at the annual health check for patients aged ≥75 (28%).Even when there was a patient request (57%), or a change in health state (71%), it was far from standard practice to review forms.Only 46% would consider reviewing an ECTP following a hospital admission (Table 2).
Overall, ECTP was viewed positively; 89% agreed that having a plan ensures treating clinicians know the patient's wishes and 82% agreed it can avoid patients' families making difficult decisions.Nevertheless, half (51%) agreed that a patient's current health condition may not be reflected in the plan when implemented and there is a serious risk it could be out of date (Table 3).
Considering the last time they had completed an ECTP, a small minority (9%) reported that a family member was not involved.Most commonly this was because the patient had capacity 54/72 (75%), although 18 (25%) reported that the family was not available and 11 (15%) that the patient didn't want the family involved.One responder reported that the family did not want to be involved (Supplementary Table S2).
GPs reported being at least fairly comfortable having ECTP conversations with both the patient (81%) and the patient's family (79%) (Table 4).
In our adjusted logistic regression analyses for conversations with patients, locum and salaried GPs were substantially (around 48%) less likely to be comfortable having ECTP conversations compared with partner or principal GPs; odds ratio (OR) = 0.51 (95% confidence interval [CI] = 0.31 to 0.82) and 0.53 (95% CI = 0.34 to 0.82), respectively (Table 5, Supplementary Table S3).For conversations with family members the difference was only statistically significant for salaried GPs; OR 0.58 (95% CI = 0.38 to 0.88; Supplementary Table S4).
When compared with London, GPs in the South West and the North East were substantially more likely (odds around 4.2 times greater) to be comfortable with ECTP conversations; OR = 4.30 (95% CI = 1.45 to 12.7) and 4.10 (95% CI = 1.22 to 13.8), respectively.For conversations with family members, GPs from the South East, East Midlands, or Yorkshire and Humber were also more comfortable than GPs from London (Table 5, Supplementary Table S4).

continued on next page
GPs using a ReSPECT form were more comfortable having these conversations with patients (72% more odds) and family members (85% more odds) when compared with GPs using a DNACPR form; OR = 1.72 (95% CI = 1.10 to 2.69) and OR = 1.85 (95% CI = 1.19 to 2.87), respectively.Results from our sensitivity analysis using a backwards elimination model were not materially different (Supplementary Tables S5 and S6).

Discussion Summary
This study shows that ECTPs have become a standard part of general practice with 100% of responders reporting using some form of ECTP.Nevertheless, the finding that just over half (51%) of our responders are still using standalone DNACPR forms is potentially a cause for concern, when the limitations of DNACPR for making holistic patient-centred decisions have been recognised since at least 2016. 17Ps who used ReSPECT when compared with DNACPR were more likely to feel comfortable in having ECTP conversations with patients and their relatives.The main trigger for initiating an ECTP conversation is diagnosis of a life-limiting or life-threatening condition.While completion of an ECTP in primary care is currently carried out predominantly by GPs, responders suggested that this could be carried out by a much broader range of health and social care professionals.
Responders were very supportive of a wider spectrum of health and social care professionals being able to complete ECTPs.Support for specialist nurse practitioners for palliative care completing these forms is not surprising.The finding that a substantial minority (25%) of GPs support senior care home staff completing ECTP forms is perhaps more surprising, as these are not designed for completion by non-clinicians.Possibly our responders had in mind senior care home staff having the initial conversations with their residents rather than formal completion of the form without clinician input.Indeed, an interview study of GPs and care home staff found that GPs value the input of care home staff in ReSPECT conversations. 18While care home staff were generally positive about being involved, they had concerns about taking responsibility for the form's content. 18ost responders reported the patient's clinical condition as the stimulus for initiating ECTP conversations, predominantly in the context of life-limiting diagnosis or terminal prognosis.This conceptualises ECTPs as being associated with end-of-life care.This contrasts with how its developers envisaged ReSPECT but is consistent with the previous studies of the ReSPECT process. 12,19,20It is unclear whether time constrains the GP staff to focus on patients who are perceived to be likely to have an acute need for emergency care in the foreseeable future, or they are conflating ECTPs with advance care planning.How GPs conceptualise ECTPs may affect their views on who can complete them and how often they need reviewing.We are exploring this question in our related qualitative study. 15he finding that one in five (20%) practices have no system for reviewing ECTPs with only 58% having any routine system for review is of concern.This is concerning particularly since 41% of our responders agreed that there might be a serious risk of the plan being out of date and not reflecting the patient's views, and half (50%) felt the patient's current health condition might not be reflected in the plan when implemented.
Caution is needed when interpreting the apparent regional differences observed in how comfortable GPs felt in having ECTP conversations because of the large number of comparisons and small numbers in some groups.Nevertheless, there appears to be differences between London, and both the South West and the North East.This might reflect the impact of the presence or absence

Table 1
Responder characteristics, totals, and percentages Underwood M et al.BJGP Open 2024; DOI: 10.3399/BJGPO.2023.0192patients'families, or someone important to the patient (henceforth included within the term 'families'); and the factors that might influence this process.

Table 1 ContinuedTable 2
Totals and percentages for the emergency care and treatment planning form completion (N = 841) What form of emergency care and treatment plans does your practice use?

Table 4
How comfortable or uncomfortable do you feel having conversations about an emergency care and treatment plans?(N = 841)How comfortable or uncomfortable do you feel having conversations about an emergency care and treatment plan with patients?

Table 5
Odds ratios (95% confidence intervals[CIs]) and P values of the predictors for being comfortable having emergency care and treatment planning conversations with patients (n = 841)