Objectives Development of evidence-based good practice recommendations for clinicians considering the use of antibiotics in patients towards the end of life.
Design A multiprofessional group of experts in end-of-life care and antimicrobial stewardship was convened. Findings from a scoping review of the literature and a consultation of clinicians were triangulated. Expert discussion was used to generate consensus on how to approach decision-making.
Setting Representatives from hospital and a range of community health and care settings.
Participants Medical, pharmacy and nursing professionals.
Main outcome measures Good practice recommendations based on published evidence and the experience of prescribers in Scotland.
Results The findings of 88 uncontrolled, observational studies of variable quality were considered alongside a survey of over 200 prescribers. No national or international guidelines were identified. Antibiotic use towards the end of life was common but practice was highly variable. The potential harms associated with giving antibiotics tended to be less well considered than the potential benefits. Antibiotics often extended the length of time to death but this was sometimes at the cost of higher symptom burden. There was strong consensus around the importance of effective communication with patients and their families and making treatment decisions aligned to a patient’s goals and priorities.
Conclusions Good practice recommendations were agreed with focus on three areas: making shared decisions about future care; agreeing clear goals and limits of therapy; reviewing all antibiotic prescribing decisions regularly. These will be disseminated widely to support optimal care for patients towards the end of life. A patient version of the recommendations has also been produced to support implementation.
- symptoms and symptom management
- end of life care
- clinical decisions
Statistics from Altmetric.com
Antibiotics are used widely for patients in all care settings in the months, weeks, days and even hours before death.1–3 Towards the end of life, patients can be more susceptible to infection and more likely to develop symptoms which may be mistaken for infection. The decision to initiate, continue or discontinue antibiotic therapy in this context requires careful consideration. Although giving an antibiotic may extend life in certain circumstances,4 it may not necessarily improve quality of life5 and may lead to care being overly medicalised at end of life, detracting from what matters most to the individual. There is also the potential for harm, both for the individual (through side-effects or complications, including Clostridioides difficile)6 and for the wider community (through promotion of antimicrobial resistance).7 8 Decisions can become more complex if the patient’s prognosis is uncertain or if their wishes regarding future treatment of infection are not known.9–11 To date, there are no UK national guidelines that focus on the use of antibiotics towards the end of life, despite an acknowledgement from clinicians that these would be beneficial.12 13
The Scottish Antimicrobial Prescribing Group (SAPG)14 works with regional National Health Service (NHS) boards and other national stakeholders to ensure the judicious use of antibiotics and to achieve best outcomes for patients in all care settings. In 2019, SAPG convened a short-life multiprofessional steering group to develop consensus on the use of antibiotics in patients towards the end of life in all care settings. Membership of the group included general practitioners, physicians in acute medicine, geriatric medicine, palliative medicine, infectious disease, microbiology and public health, old age psychiatrists, specialist nurses and clinical pharmacists.
The aim of the group was to optimise antibiotic prescribing practices towards the end of life and to promote antimicrobial stewardship. There was also recognition that practice should be aligned more closely with realistic medicine, a philosophy first championed by the chief medical officer for Scotland in 2016,15 with core principles of building a personalised approach to care through shared decision-making and reducing harm, waste and unwarranted variation in practice. The objective of this work was to review the available evidence on antibiotic use towards end of life and to develop good practice recommendations to support clinicians in their approach to decision-making.
Two pieces of work were undertaken to inform the good practice recommendations. First, a scoping review was conducted to provide an overview of the available research evidence on the broad topic area of use of antibiotics towards the end of life. Studies across all care settings and all life-limiting health conditions were included.15 Second, information was gathered via a consultation process using an electronic survey. The survey was developed by steering group members using an iterative process and tested prior to dissemination. The link to the survey was communicated via a standard email invitation sent by steering group members to their networks and other contacts who may be involved in prescribing antibiotics for patients towards the end of life. Within the survey, clinicians from across Scotland were invited to discuss and reflect on their approach for using antibiotics towards the end of life (online supplementary information 1).
Ethical approval was not required for the survey as it was considered to be service evaluation, participation was voluntary and all collected data were anonymous.
The findings from the two studies were triangulated according to the framework developed by Hopf et al.16 Key points from the review and survey were discussed by the expert group to underpin the development of good practice recommendations. Agreement was reached by consensus during group discussion.
The scoping review is reported in full elsewhere15 but, in brief, included 91 papers from 88 uncontrolled, observational studies of variable quality but with broad international spread. No national or international guidelines were identified. Most studies involved adults in a variety of care settings with multiple long-term health conditions or diagnoses of advanced cancer or advanced dementia.
Antibiotic use towards the end of life was common but there was a lack of consensus around when a patient should be deemed as approaching the end of life. Several factors influenced the decision to prescribe or not prescribe an antibiotic, including the patient’s underlying diagnosis. The most commonly stated indications for giving an antibiotic were to treat either a urinary tract or a respiratory tract infection but there was wide variation in practice regarding the type of antibiotic given and by which route of administration. When reported, antibiotics often extended the length of time to death but this was sometimes at the cost of higher symptom burden. Only a small number of studies provided detailed data on adverse events and antimicrobial resistance. There was extensive commentary in the studies around attitudes to antibiotic use and decision-making at the end of life.
Survey of prescribers
The survey was completed by 260/402 (65%) clinicians who accessed the link. Respondents comprised prescribers in a range of roles across primary and secondary care, with varying levels of experience in caring for patients approaching the end of life.
Respondents were asked to provide a practical definition of ‘end of life’ and answer all subsequent questions according to this specified timeframe. There was general agreement that end of life should be defined as ‘last weeks of life’. Factors influencing prescribing behaviour were identified and are summarised in Box 1.
Summary of survey findings
The potential harms associated with giving antibiotics tend to be less well considered than the potential benefits.
Almost a third of clinicians would prescribe antibiotics if they thought that the infection was treatable (irrespective of a patient’s prognosis). A few would prescribe antibiotics with the intention of increasing hope for the patient or their family.
Clinicians can feel under pressure from patients and especially families to prescribe antibiotics. Some clinicians prescribe antibiotics to make themselves or their patients/families feel as though they are doing something.
Although symptoms of infection are often treated with medications other than antibiotics, most said they would occasionally prescribe antibiotics to manage symptoms.
More than half of respondents would not escalate antibiotic treatment from an oral to intravenous route of administration. Almost half would not use second-line or third-line antibiotics unless guided by sensitivities from microbiology. There was acknowledgement that escalation of therapy (particularly to intravenous antibiotics) could impact on the patient’s preferred place of care.
There was agreement that healthcare professionals should discuss the decision to give antibiotics with their patients. There was understanding that patients have the right to decline treatment.
There was strong consensus around the importance of: effective communication with patients and their families; making treatment decisions aligned to a patient’s goals and priorities; documenting a patient’s preferences for their future care; keeping patients and families informed; avoiding giving false hope.
There was agreement that clinical decision-making should be tailored to the individual patient and their circumstances but that this is more difficult when clinicians are less familiar with the person (eg, in the acute hospital or in general practitioner out of hours) or when there is uncertainty around prognosis or the potential risks and benefits of antibiotics. There was acknowledgement that peer support is beneficial in such cases.
Antibiotics to improve the symptoms of urinary tract infections, respiratory tract infections and cellulitis are often thought to be beneficial.
There was a recurring theme to ‘avoid creating policy’ and that, while guidance could be helpful, it should support but not replace a personalised approach to care and decision-making.
Table 1 shows the triangulation of key findings from the scoping review and the survey. This type of analysis allows systematic comparison of findings across different studies and categorises the extent to which findings converge: full or partial convergence (agreement); divergence (disagreement); absence of a connection (silence).
The steering group members discussed the findings of the scoping review and the survey and used these to inform the development of the good practice recommendations. These were summarised into three key areas: (1) making shared decisions about future care; (2) agreeing clear goals and limits of therapy; and (3) reviewing all antibiotic prescribing decisions regularly. These are presented in Box 2.
Good practice recommendations for use of antibiotics towards the end of life
These recommendations apply to adults with any life-limiting condition. Here, ‘End of life’ is defined as the last few days or weeks of life.
1. Make shared decisions about future care
The evidence suggests this is the most important aspect of care for patients and their families/carers.
Decisions about antibiotic prescribing towards the end of life should be taken jointly between the clinician, or in some settings the multidisciplinary team, through discussion with the patient and, where appropriate, their family/carer. This shared decision-making process not only involves informing the patient of the potential benefits and risks of antibiotics but also taking the time to understand the patient’s priorities for the future.
Current and future antibiotic prescribing decisions should be discussed as part of anticipatory care planning conversations, documented in the clinical notes and included in the patient’s Key Information Summary*. This discussion should include route of antibiotic therapy as intravenous treatment would usually necessitate hospital admission.
2. Agree clear goals and limits of therapy
These should be defined and agreed with the patient/family/carer after considering the following:
The principal purpose of antibiotics at the end of life may be to relieve symptoms or may potentially be to cure infection.
An infection should not necessarily be treated simply because it is treatable. Likewise, a positive microbiology result should not lead to an antibiotic prescription if there are no significant symptoms.
Consider whether hospital admission, if required, for intravenous antibiotics is in keeping with the patient’s preferred place of care towards end of life.
There are risks associated with giving antibiotics (including side effects, Clostridioides difficile and antimicrobial resistance)
Infection may be reversible and clinicians may feel compelled to offer treatment. However, this should be balanced against potential antibiotic-related toxicity.
If an antibiotic is prescribed, follow local guidance on drug choice, dose and duration and ensure a stop date is recorded.
Overall benefit for each individual patient should be the goal of any treatment as per General Medical Council guidance on ‘Treatment and Care Towards End of Life’.19
Where patients lack capacity, guidance from the Adults with Incapacity Act 2000 should be followed, including, for example, involvement of a power of attorney/guardian where appropriate.20 The pre-existing wishes of the patient should be explored and considered in the context of the clinical situation. The benefits and risks of antibiotic therapy should be discussed with any proxy decision maker or family acting in the patient’s overall benefit.
Other medicines including mucolytics, muscle relaxants, analgesics, antipyretics and antitussives should be considered as alternatives to antibiotics for relief of infection-related symptoms.21
Oxygen and non-pharmacological methods such as a hand held fan may be helpful for dyspnoea. See Scottish Palliative Care Guidelines for further advice.21 22
Delirium is very common and often attributed wrongly to infection. It is important to consider other contributing factors (including that the person may be dying and terminally agitated).
Seek advice from palliative care specialists if required.
3. Review all antibiotic prescribing decisions regularly
If, in the context of an acute severe infection, it emerges that the patient is at the end of life, clinical decisions relating to antibiotic prescribing should be reviewed and discussed immediately.
If it emerges that an antibiotic is not helping or is causing side effects, the discontinuation of treatment should be discussed with the patient and/or their carer/family.
If the patient wishes to stop an antibiotic at any time, this decision should be respected and treatment should be discontinued.
Antibiotic therapy should not routinely be escalated in the deteriorating patient at the end of life (this includes use of broad spectrum or intravenous antibiotics).
* An electronic record held in primary care but accessible to most healthcare professionals, including those working out of hours.
The recommendations were disseminated through Scottish Health Board Antimicrobial Management Teams, via local clinical networks and the Care Inspectorate by steering group members. A patient version of the recommendations (online supplementary information 2) was produced to support shared decision-making.
Antibiotic prescribing towards the end of life is common in all care settings but, to our knowledge, no national or international guidelines exist to support decision-making in this area. This work aimed to address some of the gaps by scoping the available evidence base and by exploring clinician attitudes and behaviours via survey. We concluded that the goals of antibiotic therapy towards end of life are often ill-defined and the potential for harm is frequently under-recognised.
In response, we have developed good practice recommendations for clinicians, identifying three important principles for consideration: (1) decisions about whether to prescribe antibiotics should be taken jointly between the clinical team and the patient; (2) the goals and limits of antibiotic therapy should be defined and agreed at the outset; (3) all antibiotic prescribing decisions should be reviewed regularly. Clinicians are encouraged to discuss antibiotic prescribing decisions with their patients when making plans for future care and are discouraged from treating infections simply because they are treatable. Importantly, the recommendations place emphasis not only on the circumstances in which an antibiotic may be started towards end of life but also on those situations in which it may be stopped.
For the purpose of these recommendations, the group defined end of life as the last few days or weeks of life. This timeframe represented the consensus definition reached by the group following extensive discussion and also reflected the definition used by the majority of respondents in our clinician survey. However, we recognise that a shared decision-making approach to antibiotic prescribing is good practice in all clinical contexts, not just in the last weeks of life. Likewise, we acknowledge that the decision to initiate, continue or discontinue an antibiotic in the last weeks of life should not be influenced by prognosis alone.
We believe that the recommendations encapsulate the essence of realistic medicine.17 They endeavour to build a personalised approach to care by supporting clinicians to more fully understand their patient’s priorities before reaching a shared decision on whether to give an antibiotic. They recognise that using antibiotics towards end of life may benefit certain patients if they reduce symptom burden, help achieve comfort or extend life in line with a person’s goals. Likewise, they affirm the potential for antibiotic use to lead to harm through side effects or by undermining what matters most to the person.
Although this work was largely carried out prior to the COVID-19 pandemic (and indeed SAPG have since produced guidance for antibiotic management of respiratory infections in frail older people during COVID-19),18 we believe that these recommendations and, in particular, their emphasis on shared decision-making, have even greater relevance for those practising now.18 Radical enforced changes to way of life have led many in society, both in healthcare and beyond, to reflect more broadly on what matters most to them. This same concept underpins decisions around whether to use antibiotics towards the end of life and, in a postpandemic world, these recommendations provide an ongoing framework to support clinicians.
What was already known?
Clinical decisions around use of antibiotics in patients towards the end of life can be challenging.
No national or international guidance exists on this specific issue in this patient group.
What are the new findings?
Antibiotics are commonly prescribed in the final days or weeks or life despite benefits being unclear and the potential for harm.
Decisions about antibiotic prescribing towards the end of life should be taken jointly between the clinician, or in some settings the multidisciplinary team, through discussion with the patient and, where appropriate, their family/carer.
These good practice recommendations can support clinicians deciding whether to start or not start antibiotics in patients towards the end of life.
What is their significance?
Clinical - our Good Practice recommendations provide clinicians and patients with practical advice to support shared decisions to provide clear person-centred goals for starting and stopping antibiotic treatment towards the end of life
Research - the three recommendations will provide clinicians with a standard of care to audit thier clinical practice against
The authors would like to acknowledge all other members of the SAPG End of life Steering group for their engagement and contribution to this work (Janine Thoulass, Cesar Rodriguez, Frances Downer, Jayne Walden, Dianne Foster, Euan Paterson, Tom Gillespie, David Boggon, Paul Baughan, David GrayGrey, Fiona Baker, Lynsey Fielden, Abigail Mullings, Kate Nelson, David Marshall). The authors also thank Marion Pirie, SAPG Project Officer, for supporting this work and NHS board Antimicrobial Management Teams for their feedback on the recommendations.
Contributors RAS, JS and JF conceived the project. LC and JC designed the studies, collected and analysed data. LC, JS and RAS drafted the manuscript. JF, SP, LF, TD and SG reviewed and commented on manuscript and approved final draft.
Funding LC, JS and RAS contributions to this work were undertaken as part of their roles in the Scottish Antimicrobial Prescribing Group. JF was funded by the Royal College of Physicians and Surgeons of Glasgow for undertaking this work as part of a Scottish Clinical Leadership Fellowship focused on antimicrobial stewardship across several areas of practice including end of life care. All authors contributed to this work as part of their clinical role.
Competing interests None declared.
Patient consent for publication Not required.
Data availability statement No data are available.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.