Sedation in Palliative Care

Abstract

Background

The appropriate provision of sedation as a last resort for the relief of suffering in palliative care is dealt with variably in actual practice. This article is intended as an overview of practically relevant information found in treatment recommendations and guidelines.

Methods

A systematic literature search was conducted in the PubMed, Scopus, and Google Scholar databases, and a manual search was carried out online. Recommendations that were not available in either German or English, or that were specific to pediatric practice, were excluded. Publication quality was assessed with the AGREE II instrument (Appraisal of Guidelines for Research & Evaluation II). The recommendations in the documents were qualitatively evaluated.

Results

29 publications (11 journal articles, 18 other) of varying quality according to AGREE II were included. All recommendations and guidelines were essentially based on expert consensus. The common indications for sedation are otherwise intractable delirium, dyspnea, and pain, in patients with a life expectancy of no more than two weeks. Existential distress is a controversial indication. The drug of first choice is midazolam. As the sedating effect of opioids is hard to predict, they should not be used as sedatives. The risks of sedation include respiratory and circulatory depression, as well as the loss of communicative ability, control, and autonomy. It is generally recommended that the patient’s symptom burden and depth of sedation should be monitored; clinical and technically supported monitoring are recommended in some publications as well, depending on the situation.

Conclusion

There is a broad consensus in favor of sedation to relieve suffering in the last days and hours of life. Recommendations vary for patients with a longer life expectancy and for those with existential distress, and with respect to monitoring.

The use of drugs with a sedative effect is an established element of palliative care. Apart from using these medications to relieve symptoms such as fear or restlessness, targeted use is also made of sedation in palliative care (SiP) to reduce patient consciousness in order to avoid suffering from otherwise intractable, or refractory, symptoms (box 1).

BOX 1. Definition of terms.

Refractory to treatment is defined as a situation where all other procedures have failed or cannot be utilized to relieve the symptom within an appropriate period of time (e21).

Sedation in palliative care (SiP): Various terms are used for discussion (1). Examples include “palliative sedation”, “terminal sedation” as well as “intentional sedation”. The last-named term is used in conjunction with an analytical definition and explanation of terms provided in the German recommendations (2). Since we refer in the present review article to different terminologies used in the various guidelines, we have chosen the generic term “SiP” as an overall expression.

Studies reveal that the prevalence of SiP varies considerably (compare, for example, the 0–80% prevalence in the study by Schur et al. [3]). The distinction between SiP (in the sense of targeted sedation for symptom relief) and symptom relief with a sedative effect (3– 5) is often vague. In any case, a distinction is made between the dimensions “depth of sedation” and “duration of sedation” (continuous until death versus temporary). The competent provision of sedation not only has medical, but also ethical and legal implications, for example, when it comes to taking steps that may involve intentionally shortening life (6, 7).

International guidelines and recommendations, including the framework developed by the European Association for Palliative Care (EAPC) in 2009, have been drawn up to raise the level of safety when using this method of treatment (8– 11). Recommendations for sedation in specialist palliative care settings were published in 2021 for Germany under the collaboration of the present team of authors, together with the German Society for Palliative Medicine (DGP) (2).

A number of guideline overviews have been released in recent years (1, 9, 12). However, there have been no studies to date which focus on similarities and differences between the recommendations with regard to clinically relevant and practical aspects. The aim of the present review article is to fill this gap and to extract and discuss the converging and diverging recommendations for everyday clinical practice.

Methods

This study originated from a project called SedPall (Sedation at the End of Life in Specialist Palliative Care) (registered under DRKS00015047). It was not registered as a systematic review. A systematic literature search was conducted in the databases PubMed, Scopus and Google Scholar for the period from January 2000 until May 2022. It included clinical guidelines and recommendations on sedation in palliative care which are at least regionally valid and were created by staff of an institution or by a national or international group of experts. Recommendations were excluded which were not available in English or German, as were those exclusively covering pediatric palliative care. The search strategies are presented in the eBox. Furthermore, a search was also conducted using internet search engines for relevant sites with recommendations from professional societies or organizations which have not been published as journal articles.

All identified publications were independently assessed by two researchers (JS and CK) who looked at title and abstract to see whether they contained reference to a relevant guideline corresponding to the inclusion criteria. If this was not certain, then the full text was reviewed. All relevant recommendations were entered into a software program for qualitative text analysis (MaxQDA, Verbi GmbH Berlin). Two independent teams then assessed quality by applying the AGREE II criteria (AGREE, Appraisal of Guidelines for Research & Evaluation) in the domains “Scope and purpose”, “Stakeholder involvement”, “Rigor of guideline development”, “Clarity of presentation”, “Applicability”, and “Editorial independence” (13, 14), and their appraisals were then combined to form an overall evaluation. A mean comparison test was conducted to assess for any potential quality improvement achieved in more recent recommendations (Mann-Whitney U test, SPSS 28).

A framework for a category of systems was created for the content-related evaluation in the form of a structuring content analysis (15) based on the EAPC framework (10), which was inductively extended. The recommendations published in the documents were assigned to these categories by at least two scientists (RV, CK, and others). The analysis focused on clinically relevant recommendations on the aspects “indications”, “clinical requirements“, “medication”, “risks”, and “monitoring”.

Results

The systematic database search identified a total of 781 publications (figure). The search using internet search engines for guidelines and recommendations from professional associations or organizations produced 39 hits, with 621 publications remaining after exclusion of duplicates. Based on a review of title and abstract, 562 articles were considered irrelevant and therefore excluded, a further 30 after full-text analysis. A total of 29 recommendations published between 2004 and 2021 were included from Germany, France, Spain, the United Kingdom, Austria, Switzerland, Norway, the Netherlands, the USA, Mexico, Canada, Australia, Japan, together with those with a European relevance (e1– e29). Of these, 11 were available as articles in scientific journals and 18 as (online) publications. All recommendations refer to deep sedation until death, with some also including temporary or light sedation. Only seven recommendations explicitly provide the underlying level of evidence, with evidence consistently reported as low (based on expert consensus or non-analytical studies). Contrary to this, two publications rate “C” for medication choice (one exploratory study with a low risk of bias or concordant studies with high risk of bias or investigative or non-analytical studies) and “3” (non-analytical studies), respectively (table 1).

Figure.

Presentation of the search process (Ger., German; int., international)

Table 1. Evidence statements, forms of sedation, clinical preconditions and mental/existential suffering as indication.

Ref.	Publishing institution	Year	Evidence statements/literature search statements	Forms of sedation				Life-limiting disease/ DNR as precondition for sedation	Precondition: Prognosis Death expected within				Mental/existential suffering as an indication
				Sedation until death	Deep sedation	Temporary sedation	Mild sedation		hours to days	days to weeks	a few days	2 weeks
e1	British Columbia Center for Palliative Care, Canada	2019	n.s.	X	X		X	X / X			X		+SC
e2	Winnipeg Regional Health Authority, Canada	2017	general Level IV/n.s.	X	X		X	- / -				X	+
e3	Haute Autorité de Santé, France	2018	n.s.	X	X			X / -	X				+
e4	International Panel of Experts, Netherlands/Canada	2007	overall: D, except for medication selection: C/ syst. review	X	X	X	X	X / -	X				+SC
e5	Alberta Health Service, Canada	2015	n.s./review (search strategy stated)	X	X			X / X	X				NR
e6	Hospice & Palliative Care Federation of Massachusetts, USA	2004	n.s.	X	X	X	X	- / X	X				n.s.
e7	Fraser Health Authority, Canada	2011	n.s./review (search strategy stated)	X	X		X	X / X	X				+SC
e8	Champlain Hospice Palliative Care Program, Canada	2018	review	X	X		X	X / X			X		+SC
e9	Norwegian Medical Association, Norway	2014	n.s.	X	X	X	X	- / -			X		+SC
e10	Swiss Society for Palliative Medicine, Switzerland	2005	n.s.	X	X	X	X	X / -		X			n.a.
e11	Waterloo Wellington Interdisciplinary Hospice and Palliative Care Education Committee, Canada	2013	n.s.	X	X			X / X		X			+SC
e12	The Mississauga Halton Palliative Sedation Therapy Development Committee, Canada	2012	n.s.	X	X		X	X / X	X				NR
e13	Palliative Care working group of the Comprehensive Cancer Center, Germany, funded by the Germany Cancer Aid	2017	n.s.	X	X	X	X	- / -					+SC
e14	National Health Service Scotland, Scotland	2020	n.s.	X	X	X	X	- / -	X				+
e15	Ministry of Health, Mexico	2015	n.s./review	X	X	X	X	X / -					+SC
e16	Collège des médecins du Québec, Canada	2016	n.s.	X	X	X	X	X / -		X			+SC
e17	National Comprehensive Cancer Network, USA	2019	2A/review (search strategy stated)	X	X			- / X	X				n.a.
e18	National Hospice and Palliative Care Organization, USA	2010	n.s.	X	X		X	- / X				X	+SC
e19	Royal Dutch Medical Association (KNMG), Netherlands	2009	Level 4/n.s.	X	X	X	X	- / -		X			+SC
e20	Spanish National Health System, Spain	2008	overall D, except for medication selection 3/review	X	X	X	X	X / -					+
e21	Canadian Society of Palliative Care Physicians, Canada	2012	insufficient evidence to provide information/review	X	X		X	- / -				X	+SC
e22	European Association for Palliative Care	2009	n.s./review (search strategy stated)	X	X	X	X	X / -	X				SC
e23	European Society for Medical Oncology	2014	overall Level V, expert opinion/n.s.	X	X	X	X	X / -					+SC
e24	Academy for Ethics in Medicine, Germany	2010	n.s.	X	X	X	X	- / -					+
e25	Sedation Guidline Taskforce, Japan	2020	Suggestion of overall low level of evidence/syst. review	X	X	X	X	X / X	X				+SC
e26	National Group of Experts, Canada	2002	n.s./review	X	X	X	X	- / -					n.a.
e27	Austrian Palliative Society (OPG), Austria	2017	n.s./review of 8 international guidelines	X	X	X	X	- / -	X				+SC
e28	State of Victoria, Australia, Safer Care Victoria, Australia	2020	n.s./review	X	X		X	X / -				X	+SC
e29	Government of Western Australia, Department of Health, Australia	2021	n.s.	X	X	X	X	X / -	X				+SC

(e2) Level IV-expert committee reports or opinions and/or clinical experience of respected authorities; (e4) C. One Level III study or consistent Level IV studies; D. Level V evidence or inconsistent or inconclusive studies of any level; III. Exploratory studies with a low risk of bias; IV. Any type of study with a high risk of bias, or investigative studies, or nonanalytic studies; V. Experts’ opinion;

(e17) 2A. Based upon lower-level evidence, there is uniform NCCN consensus that the intervention is appropriate.; (e19) Level 4. Opinion of respected authorities; (e20) D. Evidence of Llevel 3 or 4, or extrapolated evidence from studies rated as 2+;

3. Non-analytic studies, e.g. case reports, case series; 4. Expert opinion

+: stated as an indication without restriction; +SC: stated as an indication only under special circumstances; DNR: Do Not Resuscitate; n.s.: not stated/not mentioned;

CO: subject of controversy, no conclusive assessment; NR: not recommended; Ref.: reference; syst.: systematic

Quality assessment using the AGREE II tool

Good results were often noted for the domains “Scope and purpose“ (78%; 39–100%; mean value in each case; minimum–maximum) and “Clarity of presentation” (71%; 31–92%). Moderate assessments resulted for the domains “Applicability” (54%; 21–85%) and “Stakeholder involvement” (51%; 11–94%). Quality was rated worst with regard to the domains “Rigor of development” (42%; 0–81%) and “Editorial independence” (38%; 0–100%) (etable 1).

eTable 1. Presentation of the AGREE II assessment of the analyzed guidelines and recommendations.

Reference	Publishing institution	Domain 1	Domain 2	Domain 3	Domain 4	Domain 5	Domain 6	Mean %	Journals Publication	Year of publication
		Scope and purpose %	Stakeholder involvement %	Rigor of GL development %	Clarity of presentation %	Applicability %	Editorial independence %
e1	British Columbia Center for Palliative Care, Canada	83	92	53	83	71	8	65	no	2019
e2	Winnipeg Regional Health Authority, Canada	89	58	28	67	75	4	54	no	2017
e3	Haute Autorité de Santé, France	100	81	74	86	85	4	72	no	2018
e4	International Panel of Experts, Netherlands/ Canada	72	47	55	83	21	0	46	yes	2007
e5	Alberta Health Service, Canada	83	53	61	69	48	100	69	no	2015
e6	Hospice & Palliative Care Federation of Massachusetts, USA	42	11	7	61	52	0	29	no	2004
e7	Fraser Health Authority, Canada	100	31	58	75	67	100	72	no	2011
e8	Champlain Hospice Palliative Care Program, Canada	89	42	55	78	73	100	73	no	2018
e9	Norwegian Medical Association, Norway	64	19	0	44	27	0	26	no	2014
e10	Swiss Society for Palliative Medicine, Switzerland	67	58	8	50	44	0	38	no	2005
e11	Waterloo Wellington Interdisciplinary Hospice and Palliative Care Education Committee, Canada	89	50	38	89	81	0	58	no	2013
e12	The Mississauga Halton Palliative Sedation Therapy Development Committee, Canada	89	53	34	72	54	4	51	no	2012
e13	Palliative Care working group of the Comprehensive Cancer Center, Germany funded by the German Cancer Aid	39	44	10	86	46	25	42	yes	2017
e14	National Health Service Scotland, Scotland	50	50	26	78	31	0	39	no	2020
e15	Ministry of Health, Mexico	56	50	30	64	48	100	58	yes	2015
e16	Collège des médecins du Québec, Canada	86	58	23	89	63	29	58	no	2016
e17	National Comprehensive Cancer Network, USA	47	83	66	36	33	79	57	no	2019
e18	National Hospice and Palliative Care Organization, USA	89	39	31	31	44	58	49	yes	2010
e19	Royal Dutch Medical Association (KNMG), Netherlands	86	53	52	92	54	0	56	no	2009
e20	Spanish National Health System, Spain	83	83	81	75	50	83	76	no	2008
e21	Canadian Society of Palliative Care Physicians, Canada	89	28	54	69	67	0	51	yes	2012
e22	European Association for Palliative Care	89	47	66	92	71	4	61	yes	2009
e23	European Society for Medical Oncology	69	11	39	61	33	0	36	yes	2014
e24	Academy for Ethics in Medicine, Germany	58	42	10	56	58	50	46	yes	2010
e25	Sedation Guideline Task Force, Japan	86	53	68	67	44	71	65	yes	2020
e26	National Group of Experts, Canada	83	44	55	67	46	83	63	yes	2002
e27	Austrian Palliative Society (OPG), Austria	100	94	81	89	75	92	92	yes	2017
e28	State of Victoria, Australia, Safer Care Victoria, Australia	100	78	44	83	50	92	74	no	2020
e29	Government of Western Australia, Department of Health, Australia	94	22	14	56	65	8	43	no	2021
Domain means (min. – max.)		78(39–100)	51(11–94)	42(0–81)	71(31–92)	54(21–85)	38(0–100)	56(26–92)

Mean of all guidelines published in professional journals: 55%; mean of all other guidelines: 56%; GL, guideline; max., maximum; min., minimum

The mean values of the recommendations published as journal articles (55 %; 36–92 %) and those of the other recommendations (56%; 26–76%) did not substantially differ. A comparison of the mean values did not show any improvement of the five youngest in comparison with the five oldest recommendations (57 versus 50%; p = 0.548).

Indications

All recommendations act on the assumption of a refractory symptom burden as an indication for SiP; 16 recommendations detail possible symptoms: common: shortness of breath (n = 15), pain (n = 15), delirium (n = 13); less common: nausea/vomiting (n = 10), epileptic seizures/myoclonia (n = 8), agitation/restlessness (n = 7), massive hemorrhage (n = 5), asphyxia (n = 3); single mentions only: bronchial secretion (n = 1), airway obstruction (n = 1), fatigue (n = 1) (etable 2).

eTable 2. Indictions for sedation in palliative care.

Ref.	Publishing institution	Delirium	Shortness of breath	Pain	Nausea/ vomiting	Epileptic seizures/myoclonia	Agitation/ restlessness	Massive hemorrhage	other indications
e1	British Columbia Center for Palliative Care, Canada	X	X	X	X	X
e2	Winnipeg Regional Health Authority, Canada
e3	Haute Autorité de Santé, France	X	X					X	AS
e4	International Panel of Experts, Netherlands/Canada	X	X	X	X		X
e5	Alberta Health Service, Canada	X	X	X
e6	Hospice & Palliative Care Federation of Massachusetts, USA
e7	Fraser Health Authority, Canada	X	X	X	X
e8	Champlain Hospice Palliative Care Program, Canada	X	X	X	X	X
e9	Norwegian Medical Association, Norway
e10	Swiss Society for Palliative Medicine, Switzerland
e11	Waterloo Wellington Interdisciplinary Hospice and Palliative Care Education Committee, Canada	X	X	X	X	X
e12	The Mississauga Halton Palliative Sedation Therapy Development Committee, Canada
e13	Palliative Care working group of the Comprehensive Cancer Center, Germany, funded by the Germany Cancer Aid
e14	National Health Service Scotland, Scotland
e15	Ministry of Health, Mexico	X	X	X	X			X	AO
e16	Collège des médecins du Québec, Canada	X	X	X	X	X	X	X	BS
e17	National Comprehensive Cancer Network, USA
e18	National Hospice and Palliative Care Organization, USA	X	X	X			X
e19	Royal Dutch Medical Association (KNMG), Netherlands	X	X	X	X
e20	Spanish National Health System, Spain
e21	Canadian Society of Palliative Care Physicians, Canada
e22	European Association for Palliative Care		X	X		X	X	X	AS
e23	European Society for Medical Oncology		X	X		X	X
e24	Academy for Ethics in Medicine, Germany	X	X	X	X	X	X
e25	Sedation Guideline Task Force, Japan								FA
e26	National Group of Experts, Canada			X
e27	Austrian Palliative Society (OPG), Austria
e28	State of Victoria, Australia, Safer Care Victoria, Australia
e29	Government of Western Australia, Department of Health, Australia	X	X	X	X	X	X	X	AS

Other indications: AS: asphyxia; BS: bronchial secretion; AO: airways obstruction; FA: fatigue; Ref.: reference

The presence of non-somatic symptom burden alone (psychological symptoms and existential suffering, see Box 2) is reported either as insufficient for the indication or as vague or not possible to determine conclusively. Only five recommendations mention existential suffering as an absolute indication, while five fail to comment on this. Mental or existential suffering presents an indication for SiP under certain conditions or in exceptional cases for the vast majority of the examined publications. At times, consultations with experts and multidisciplinary assessment of the symptoms or the restriction of SiP to particularly severe existential suffering are required (table 1).

Clinical requirements

The presence of a life-limiting, or immediately fatal, underlying condition is a clinical requirement according to most of the recommendations; ten stipulate a do-not-resuscitate order. Twenty-three name a specific time period for the prognosis, while six recommendations do not provide any particular statement on this. Estimated life expectancy is usually stated as “hours to days”. In four publications this time period is “days to weeks” or “two weeks“, in three “a few days” (table 1).

Medication

The majority of the assessed recommendations include specific medication recommendations. Usually, short-acting benzodiazepine is mentioned as medication of first choice (etable 3). Other drugs commonly listed in the recommendations include the neuroleptic levomepromazine/methotrimeprazine, the narcotic agent propofol, and the barbiturate phenobarbital. Occasionally, the benzodiazepine lorazepam and the neuroleptic chlorpromazine find mention. Only sporadically recommended for sedation are haloperidol (usually only in combination), pentobarbital, diazepam, flunitrazepam, clonazepam, clorazepate, scopolamine, sodium oxybate/gamma-hydroxybutyrate, and the alpha-2-receptor blocker dexmedetomidine, which is primarily used in intensive care. The vast majority of recommendations contain dose and administration information, often providing a dose titration regimen. Table 2 presents an overview of the three most frequently recommended substances, together with dose examples.

eTable 3. Recommended substances for sedation, listed risks, and recommended monitoring.

Ref.	Recommended substances for sedation	Dose recommendation	Listed specific risks	Monitoring: generally recommended			Monitoring: special situations (longer life span, temporary sedation, at times during adjustment phase)
				Sedation depth	Symptom burden	Adverse drug reaction	Breathing, respiratory rate or pattern	Oxygen saturation	possibly, other vital signs
e1	M, Le, Ph, Lo, Pr, (Dex)	yes (M, Le, Ph, Lo)	SP	X	X				X
e2	Le, Bz, Ph, Pr	no	SP, UR	X	X		X
e3	M, Dz, Cz, Cl, Ch, Le, Ph, Pr, Na	yes (M)	RD, AS, AA	X	X	X	X
e4	M, Le, Ph, Pr, Ha	yes	RD, CO, PI, LV, IR	X	X	X
e5	M	yes	SP	X	X
e6	M, Lo, Pe, Ph, Ch, Ha	yes	SP			X	X
e7	Bz, M, Lo, Le, Ph, Pr	yes	SP	X	X
e8	M, Le, Ph, Pr	yes (M, Le, Ph)		X	X			X	X
e9		no		X	X	X	X
e10	M, Lo, Le	yes (M)		X	X			X
e11	M, Le, Ph,	yes	SP, OA, AS	X	X			X
e12	M, Le, Ph,	yes	SP	X	X		X
e13	M, Fl, Pr	yes (M)					X		X
e14	M, Lo, Le, Ph, Pr	yes (M, Ph)	AS, CO
e15	M	no	SP, CO
e16	Lo, M, Le, Ch, Ph, Pr, Sc	yes		X	X	X
e17	M, Pr	no	RD	X	X
e18	M, Le, Ch	no	LV
e19	M, Le, Pr	yes	ULC	X	X	X
e20	M, Le, Pr	yes	CO
e21	M, Ch, Le, Ph,	no	RD, PI	X	X	X
e22	M, Lo, Fl, Le, Ch, Pe, Pr,	yes	AS, LV, IR, AA	X	X	X	X	X	X
e23	M, Le, Ch, Ph, Pr	yes	AS, LV, IR, AA, ULC	X	X	X	X	X	X
e24	Bz	no	CO, LV, AA
e25	M	yes	LV, ULC
e26	Bz	no
e27	M, Lo, Le, Pr,	yes	SP, RD	X	X	X			X
e28	Cl, M, Ha, Le, Ph	yes	SP, RD	X	X	X
e29	M, Cl, Ha, Le, Ph	no	AS, LV, IR, RF	X	X	X	X

Substances Bz, benzodiazepine; Ch, chlorpromazine; Cl, clonazepam; Cz, clorazepate; Dex, dexmedetomidine; Dz, diazepam; Fl, flunitrazepam; Ha, haloperidol; Le, levomepromazine; Lo, lorazepam; M, midazolam; Na, sodium oxybate; Pe, pentobarbital; Ph, phenobarbital; Pr, propofol; Sc, scopolamine

Risks: RD, respiratory depression; AS, aspiration; UR, urinary retention; CO, loss of control (self-help, autonomy, see text); PI, positioning injury; LS, life-shortening; OA, obstruction of upper airways; RF, risk for family members (see text); Ref., reference; SP, specific substance effects; AA, accidental awareness; ULC, unintended (continuing) loss of consciousness; IR, inadequate relief

Risks

Similar to the lack of evidence on which to base recommendations for SiP, clinical data on risks are also absent, so this information depends upon the opinions of experts. Apart from substance-specific risks (for example, respiratory and cardiovascular failure), mention is made of risks from using substances beyond the realms of otherwise clinically accepted standards (for example, propofol on a regular ward without appropriate monitoring). Other adverse side effects include loss of important cognitive functions, such as the ability to communicate and express personal wishes and needs, and inability to eat and drink unassisted. Seven recommendations mention a possible shortening of life. Other risks include inadequate relief of symptoms, unexpected awakening, and failure to regain consciousness under otherwise planned temporary sedation. Another guideline reports the stress for family members from anticipated loss and a possibly longer period of sedation with “feelings of uncertainty, helplessness, and alienation” (e19) (etable 3).

Monitoring

The majority of evaluated recommendations prefer (clinical, non-equipment related) monitoring of depth of sedation and symptom relief or symptom burden (etable 3), often applying assessment scales used in intensive care. For such use, the Richmond Agitation Sedation Scale – Palliative Version (RASS-PAL) is validated in the English language, although so far there has been no validation for its translation into German (16, 17). Side effects of the medications should also be taken into consideration. Only rarely is reference made to the monitoring of respiratory rate, respiratory function, and breathing pattern. Some recommendations include certain measures only for special situations (longer life span, adjustment phase, temporary sedation). At times, monitoring of oxygen saturation and other vital signs are required in these situations. While many recommendations place importance on the interests of family members and the care team, only one points out assessment of the psychological and spiritual stress on family and care team as an element of monitoring (e21).

Discussion

Guidelines and recommendations should support a uniform course of action orientated on generally recognized basic principles. The assessment of the quality of the examined publications using the AGREE II tool shows that this is only partly the case, and in particular for the domains “Rigor of development” and “Editorial independence”. Publication of a guideline in a journal does not necessarily indicate its superior formal quality. With the poor ratings of domain 3 “Rigor of development”, a distinction should be made, from a clinical and scientific perspective, between the acknowledged poor body of evidence (18) and the presentation of the evidence research in the guidelines.

Apart from agreements, the content analysis of the recommendations also shows differences which are relevant to clinical practice. There is no apparent evidence for a targeted inclusion or exclusion of specific physical symptoms as an indication for SiP. Generally, the symptoms pain, shortness of breath, and delirium are often stated as an indication for SiP, and this is consistent with clinical experience. Existential suffering as a precondition continues to be a controversial issue, being based less on underlying evidence and more on ethical considerations and reflections on treatment options, and on the potential for abuse (e5). An intense interdisciplinary discussion is required about the indications and limitations of the use of SiP for suffering during palliative care (19). There is also a need for tools for the structured method of assessing (non-specific) general symptom burden. The often-recommended RASS-PAL has most recently also been the subject of criticism (20). The Discomfort Scale Dementia of Alzheimer Type (DS-DAT) and the Patient Comfort Score, for example, are currently under discussion as alternatives (21).

The present assessment of the recommendations agrees with the current literature with respect to midazolam as the drug of first choice (18, 22). Lorazepam is occasionally recommended for mild SiP only. There is weak evidence for the use of midazolam in palliation; otherwise, there is no sound data for preferring a particular active substance. Thus, apart from availability and setting, it is the experience of the attending staff, in particular, which is decisive (e4). Most guidelines do not distinguish between inpatient and outpatient settings, although the possibilities in a domestic setting are already limited where labor-intensive clinical monitoring is involved (4). It was these theoretical considerations that gave the authors of the German recommendations reason not to advise propofol for outpatient care.

With regard to the use of second-line substances, clinical studies should examine whether selection should be made based on a specific symptom, for example, selection of a neuroleptic agent for SiP in the presence of refractory delirium (6).

Opioids are still used in actual practice for sedation, contrary to recommendations and against good clinical practice (23, 24). This indicates a need for research into the implementation of recommendations.

Other sedatives are in use in anesthesia and intensive care which are interesting with regard to their risk profile, for example, gamma-hydroxybutyrate and dexmedetomidine, yet still do not receive widespread attention in the official guidelines presented here. Some individual reports on this are available (25, 26).

Depth of sedation and symptom burden are cited as important target parameters for monitoring and controlling SiP. There is no uniform approach put forward in the recommendations on the need for continuous monitoring of respiratory rate and saturation. The decision to dispense with instrumental monitoring of patients with a short life expectancy is often justified in palliative care by its lack of benefit and by its additional burden to patient and relatives. In cases of prolonged life expectancy or temporary sedation, measures are required—in compliance with the German recommendations as well—to minimize the risk of treatment-related complications so that the goal of alleviating suffering without shortening life can be achieved in a transparent manner. Depending on the overall clinical constellation of SiP, however, it must also be ensured that rescue measures are not automatically initiated on deterioration of vital signs, for example, resuscitation performed although no longer indicated. The “Do not resuscitate” order makes allowance for this but is not generally applied.

Apart from a lack of evidence, differences in the recommendations on sedation in palliative care are also due to different individual values and attitudes as well as in basic cultural and legal parameters (9, 27). Therefore, the inclusion of ethics and law in further guideline development would seem appropriate.

Limitations of the present assessment

Despite intense research, it cannot be excluded that some relevant documents were not included in the search. Many of the analyzed guidelines build upon each other or make reference to previous publications. As a result, individual opinions of certain experts may be exaggerated. The comprehensive decision-making processes used to draw up the recommendations lessen this risk.

BOX 2. “Existential suffering”.

The concept of “existential suffering” (sometimes also mental suffering) is used in palliative medicine to describe non-physical suffering. In many cases, it remains vague in the literature whether suffering from a mental health disease is included. Kissane (30) provides a comprehensive presentation of the various forms on existential suffering:

• death anxiety

• the effects of loss and necessary change, mourning

• loss of autonomy or control, dependency, burden for others

• effects on dignity and self-esteem

• fundamental aloneness

• altered quality of relationships (by the disease)

• search for meaning

• mystery about what seems unknowable

The guidelines evaluated in this review use different definitions. The question of whether existential suffering should be considered an indication for sedation is addressed in different ways. The reasons justifying sedation in cases of existential suffering include the fact that specialist expertise in psychology, and possibly psychiatry, are required for such an assessment and that it is less possible to objectify than physical symptoms. Its course often fluctuates spontaneously so that its refractoriness to treatment cannot be clearly determined and sedation is initially recommended as a temporary measure.

Reasons behind the different understanding of, and variety of recommendations for, existential suffering are, on the one hand, the different concepts of suffering (subjective versus objective concepts of suffering) and, on the other hand, different attitudes towards the normative imperative of suffering (does every form of suffering require medical treatment?). Refer to Bozzaro and Schildmann (19) for a detailed analysis.

Table 2. Medication recommendations from international recommendations*.

Substance/ Class	Use	Possible adverse drug reactions (selection)	Advantages	Reported dosages
Midazolam/ short-acting benzodiazepine	Anxiolysis, antiepileptic treatment, sedation; continuous administration usually recommended due to relatively short duration of action; not sufficiently effective in all cases	Respiratory depression, impairment of protective reflexes, paradoxical agitation, risk of accumulation reported (e19), therefore, very conservative approach, at times with intervals of 6–8 hrs recommended before increasing dose, therefore, long titration phase required, other recommendations allow increase after 1 hr (e7) or 30 min (e8).	IV and SC administration possible, short duration of action, well controlled; maximum effect with IV bolus: 2–5 min; 60 min with SC administration (e8)	5–120 mg/24 hrs (e25); max. 50 mg/h (e4), details at times dependent on body weight: 0.025–0.2 mg/kg/hr (e13); bolus 50% (e10) or equal to (e7) hourly dose, 30-min. interval with SC administration (e8)
Levomepromazine/ low-potency neuroleptic	Sedation, for the treatment of restlessness and agitation; often also used as an antiemetic; at times recommended for use for sedation only in combination with midazolam (e19)	Skin irritation, hypotension, ECG changes, early and tardive dyskinesia, neuroleptic malignant syndrome, lowering of seizure threshold, anticholinergic effects	IV and SC administration possible, relatively wide therapeutic range due to its long duration of action	Bolus administration 12.5–25 mg; every 8 hrs and as needed, interval 1 hr; max. 300 mg/24 hrs (e22) also continuously 0.5–8 mg/h IV or SC possible (e7, e27)
Propofol/ short-acting general anesthetic	Sedation, anesthesia; continuous IV administration necessary; infusions must be changed after 12 hrs due to the risk of bacterial contamination; recommended to be used only by staff (e1, e16, e19) or anesthetist (e19) experienced with the substance	Respiratory depression, hypotension, narrow therapeutic index, injection pain, paravenous administration can result in local irritation or even tissue damage, awareness reactions over time possible, close monitoring of the effect, with dose adjustments necessary	Good controllability, high success rate for sedation	Bolus 0.25–0.5 mg/kg; then 0.25 mg/kg/hr up to a maximum of 4 mg/kg/hr (e7); reduction in older patients

IV: intravenous; SC: subcutaneous

* The dose recommendations provided in the examined documents are based either on a few case reports or on expert opinion. Due to their mode of application (SC administration) or the overall conditions (no continuous monitoring), they are to be considered as off-label use (28, 29).

eBOX. Search strategies.

Search strategy for PubMed:

(“2000/01/01”[PDAT]:“2022/05/01”[PDAT]) AND (((“palliative” AND “sedation”) OR (“terminal” AND “sedation”) OR “terminal sedation”) OR ((“sedation” AND “end of life”) OR “palliative sedation”) AND (“guideline”[Title] OR “policy”[Title] OR “framework”[Title] OR “recommendation”[Title] OR “protocol”[Title] OR “guidelines as topic”[MeSH Terms] OR “policy”[MeSH Terms]))

Search strategy for Scopus:

PUBYEAR AFT 1999 AND ((ALL(“palliative”) AND ALL(“sedation”) OR (ALL(“terminal”) AND ALL(“sedation”)) OR ALL(“terminal sedation”)) OR (ALL(“sedation”) AND ALL(“end of life”) OR ALL(“palliative sedation”)) AND (TITLE(“guideline”) OR TITLE(“policy”) OR TITLE(“framework”) OR TITLE(“recommendation”) OR TITLE(“protocol”)))

Search strategy for Google Scholar:

allintitle: recommendation OR guideline OR policy OR framework “palliative sedation”

Inclusion criteria:

• clinical guideline or recommendations (thus stated) on sedation at the end of life

• released/created by an institution or national/international group of experts

• involving patients from the age of 18 years

• of at least regional importance

Exclusion criteria:

• not a guideline or recommendation

• reference exclusively to pediatric patients

• individual opinion, not based on national, international or official consensus

• valid only for one institution/establishment