Four Essential Drugs Needed for Quality Care of the Dying: A Delphi-Study Based International Expert Consensus Opinion

Abstract

Purpose:

The majority of dying patients do not have access to necessary drugs to alleviate their most common symptoms, despite evidence of drug efficacy. Our aim was to explore the degree of consensus about appropriate pharmacological treatment for common symptoms in the last days of life for patients with cancer, among physicians working in specialist palliative care.

Material and methods:

Within OPCARE9, a European Union seventh framework project aiming to optimize end-of-life cancer care, we conducted a Delphi survey among 135 palliative care clinicians in nine countries. Physicians were initially asked about first and second choice of drugs to alleviate anxiety, dyspnea, nausea and vomiting, pain, respiratory tract secretions (RTS), as well as terminal restlessness.

Results:

Based on a list of 35 drugs mentioned at least twice in the first round (n=93), a second Delphi round was performed to determine ≤5 essential drugs for symptom alleviation in the last 48 hours of life that should be available even outside specialist palliative care. There was ≥80% consensus among the participants (n=90) regarding morphine, midazolam, and haloperidol as essential drugs. For RTS, there was consensus about use of an antimuscarinic drug, with 9%–27% of the physicians each choosing one of four different drugs.

Conclusion:

Based on this consensus opinion and other literature, we suggest four drugs that should be made available in all settings caring for dying patients with cancer, to decrease the gap between knowledge and practice: morphine (i.e., an opioid), midazolam (a benzodiazepine), haloperidol (a neuroleptic), and an antimuscarinic.

Background and Objective

It is well recognized that high symptom prevalence and distress at the end of life are major problems for patients with cancer and those caring for them.^1,2 Symptoms tend to increase during the last days and weeks of life³ and pharmacological interventions are essential for adequate alleviation of common symptoms like pain, dyspnea, nausea, anxiety, and terminal restlessness.^3–5 A large component of palliative care (PC) focuses on efforts to achieve appropriate pharmacological alleviation.^6–8 Despite a robust formal and informal knowledge base, there are strong indications that a majority of dying patients even in well-developed European countries do not receive appropriate pharmacological relief for even the most common symptoms,^9,10 a problem that is particularly severe in nonspecialist settings caring for dying people. Gomes and colleagues¹¹ point out numerous challenges in providing adequate end-of-life care as the population ages, and settings for dying increasingly include a variety of hospital, home, and residential care facility environments with staff often without specific expertise in symptom alleviation.

Within OPCARE9, a European Union seventh framework project,¹² we aimed to better address such deficits in end-of-life care in a work package led by the Swedish OPCARE9 team, focusing on both pharmacological and nonpharmacological alleviation of suffering during the last days of life.^13,14 The participating partner countries were Argentina, Germany, Italy, New Zealand, Slovenia, Sweden, Switzerland, the Netherlands, and the United Kingdom. Each country had a representative assigned to this work package, which sought to systematize existing knowledge and identify knowledge gaps,¹² in part through Delphi investigations. In a first exploration within this work package to examine the variety of drugs used in the last three days of life, 29 physicians from 26 PC services in eight countries reported a total of 103 different drugs administered to 105 patients with cancer. This heterogeneity in drugs prescribed even in the very last days of life in specialist PC was the catalyst for the present study. The aim of the Delphi study presented here was to explore the degree of consensus about appropriate pharmacological symptom alleviation for pain, dyspnea, nausea, anxiety, respiratory tract secretions (RTS), and terminal restlessness in the last days of life for patients with cancer, among physicians working in specialist PC facilities in the nine OPCARE9 countries.

Design

The Delphi survey technique employed here is a systematic research method using an iterative multistage process in which a series, known as ‘rounds,’ of questionnaires are used to anonymously develop consensus based on the views of a carefully selected panel of experts.^15,16

All participants were recruited through each OPCARE9 country representative. Country representatives were asked to contact expert, English-speaking physicians for the panel. In this situation, ‘expert’ was defined as physicians working in specialist PC units, with at least three years practice as PC physician, and who were presently working in PC. Each country representative was asked to recruit at least 10 experts. The selected physicians received an e-mail invitation to participate in each of the two Delphi rounds, with a link to the web survey in each invitation.

The study-specific questions utilized in the surveys were jointly developed by the country representatives in OPCARE9, based on existing literature and clinical experience. The survey questions were repeatedly trialled and discussed in the Swedish multidisciplinary OPCARE9 team, and pre-tested with reference persons from OPCARE9 countries. The design of the second Delphi round built further on the findings from the first round. The demographic data collected about participants consisted of country, gender, age, number of years working as PC physician, and place of work. In Delphi round 1 (DR1), three reminders were sent out through the web survey software (Textalk Websurvey^®); in Delphi round 2 (DR2), if the participant had not answered two weeks after the second reminder, we contacted the country representative, who sent the third and final reminder.

Procedures

Delphi Round 1

DR1 was conducted in June 2009. The aim of this first round was to identify drugs seen as essential in the last days of life. The participants were asked to list their first and second choice of generic drugs to alleviate the following common and potentially treatable symptoms: anxiety, dyspnea/breathlessness, nausea and vomiting, pain, RTS, and terminal restlessness/agitation in the last days of life. The responses were in free text, with additional space provided for comments. We requested generic drug names, but indicated that trade names would also be accepted. An open question, formulated in the same manner, allowed each respondent to detail pharmacological treatment for three additional symptoms viewed as important by the participant.

Delphi Round 2

Based on the results from DR1, DR2 was conducted in January 2010, with the same palliative care physicians as in DR1 invited to participate. The aim of this round was to reach consensus around essential drugs for symptom alleviation in the last 48 hours of life in patients with cancer, which could be made available even outside specialist PC settings. Hospital-based nonspecialist PC settings were defined as geriatric wards, oncology wards, and other acute hospital wards. Community-based nonspecialist PC settings were defined as nursing homes, residential care facilities, and the patient's own home. The instructions stated that the chosen drugs should be available for general symptom alleviation irrespective of symptoms experienced by individual patients, and that other drugs could still be prescribed when appropriate for specific patients for symptom alleviation.

Physicians were asked to choose a maximum of five essential drugs from a checklist based on the suggestions elicited from DR1. The list consisted of the 35 generic drugs mentioned at least twice in DR1 presented in alphabetical order. It was possible for the respondent to add up to five additional generic drugs if they found the list insufficient. The instructions also stated that the essential drugs could but need not necessarily be for treatment of the same symptoms as listed above in DR1.

Analysis

Descriptive statistics were used to analyze responses including demographic variables. In the Delphi study we considered consensus to be achieved by ≥75% agreement.¹⁶

Ethics

Ethical principles for research have been followed in accordance with norms in each of the nine involved countries. All physicians contributing data were aware of the purpose of the study and agreed to contribute. Participation by the expert panel was voluntary, and all data collected remained confidential. The identity of the participants was not known to other expert panel members, but researchers had access to name, country, and e-mail addresses.^15,16

Results

Delphi round 1

Of the 135 PC physicians invited to participate, 93 (69%) with ≥3 years experience in PC responded (see Table 1). Another four physicians responded, but were excluded as they had <3 years of experience in PC (three from Slovenia and one from Switzerland). The characteristics of the participants are shown in Table 2.

Table 1.

Participation in Delphi Rounds by Country

		Participants in round 1 (n=93)		Participants in round 2 (n=90)
Country	Participants invited	No.	Response rate (%)	No.	Response rate (%)
Argentina	15	11	73	11	73
Germany	10	9	90	8	80
Italy	11	8	73	9	82
Netherlands	23	15	65	17	74
New Zealand	12	10	83	9	75
Slovenia	14	3	21	2	14
Sweden	29	22	76	23	79
Switzerland	10	5	50	5	50
UK	11	10	91	6	55
Total	135	93	69	90	67

Table 2.

Description of Participants

	Round 1 (n=93)		Round 2 (n=90)
	n	%	n	%
Gender
Female	43	46	36	40
Male	50	54	54	60
Total	93	100	90	100
Age	Years	-	Years	-
Median	49	-	52	-
Range	34–64	-	34–65	-
Years working as PC physician		-		-
Median	11	-	12
Range	3–29	-	3–30	-
Place of work ^a	n		n
Hospice/PC unit in hospital	58	62	61	68
Palliative homecare	40	43	36	40
Hospital	32	34	33	37
Nursing home	7	8	9	10
Total	137	-	139	-

PC, palliative care

It was possible to give multiple answers in this question.

A total of 35 drugs were identified. The most common first-choice drugs per symptom were:

• for anxiety: midazolam (34% of the participating physicians) and lorazepam (26%)

• for dyspnea/breathlessness, morphine (86%)

• for nausea and vomiting: metoclopramide (51%) and haloperidol (36%)

• for pain: morphine (71%)

• for RTS: hyoscine (reported as hyoscine, hyoscine hydrobromide, or hyoscine butylbromide: total 67%) and glycopyrronium (25%)

• for terminal restlessness and agitation: midazolam (40%) and haloperidol (36%)

As seen in Table 3, the additional symptom most commonly reported was bleeding (by 10 participants), with midazolam combined with tranexamic acid most often suggested as treatment.

Table 3.

Other Important Symptoms in the Last Days of Life Reported in Delphi Round 1

Symptom	n	Drugs suggested by ≥2 participants
Bleeding	10	midazolam (n=9), tranexamic acid (n=6), aminocaproic acid (n=2)
Delirium	9	haloperidol (n=9), levomepromazine (n=5), midazolam (n=2)
Seizures	7	midazolam (n=7), phenobarbital (n=3)
Hiccups	6	baclofen (n=2), haloperidol (n=2), levomepromazine (n=2), methoclopramide (n=2)
Constipation	4	sodium picosulfate (n=2)
Enteral obstruction	4	hyoscine butylbromide (n=4), octreotide (n=2)
Itching	4	-
Fever	3	-
Hallucinations	3	haloperidol (n=2)
Confusion	2	haloperidol (n=2)
Fatigue	2	-

Includes only symptoms reported by ≥2 physicians. Physicians may suggest two drugs for each symptom (n=number of physicians).

Delphi round 2

Ninety physicians with ≥3 years experience in PC (67%) answered the DR2 web survey (see Tables 1 and 2), with 79 having previously responded to DR1. The responses of an additional five physicians were excluded as they had <3 years of experience in PC (four from Slovenia and one from the UK).

The expert physicians selected 29 of the 35 listed drugs as those which should be readily available in all settings (see Table 4). Four additional drugs—ketamine, ketorolac, omeprazole and propofol—were suggested by one participant each.

Table 4.

Delphi Round 2: Degree of Consensus about Essential Drugs for Pharmacological Symptom Alleviation in the Last 48 Hours of Life (n=90 Physicians)

Drug	n	%
Morphine	85	94
Midazolam	76	84
Haloperidol	72	80
Levomepromazine	26	29
Glycopyrronium	24	27
Hyoscine butylbromide	21	23
Hyoscine hydrobromide	20	22
Methoclopramide	17	19
Furosemid	13	14
Lorazepam	13	14
Hydromorphone	10	11
Corticosteroids	9	10
Diazepam	8	9
Morphine–Scopolamine	8	9
Chlorpromazine	7	8
Fentanyl	7	8
Cyclizine	5	6
Octreotide	4	4
Methadone	3	3
Alfentanil/Kebemidone	2	2
Clonazepam	2	2
Dimenhydrinate	2	2
Domperidone	2	2
Oxycodone	2	2
Atropine	1	1
Flunitrazepam	1	1
Ondansetrone	1	1
Oxazepam	1	1
Thiazinamium	1	1

There was a high degree of consensus among the PC physicians regarding morphine (94% of the physicians), midazolam (84%), and haloperidol (80%) as essential drugs in the last 48 hours of life (see Table 4). There was less consensus about drugs for RTS, with the physicians choosing five different generic drugs: glycopyrronium (27%), hyoscine butylbromide (23%), hyoscine hydrobromide (22%), morphine-scopolamine (9%), and atropine (1%).

Discussion

We found a high level of consensus among the international clinical PC experts that morphine, midazolam, haloperidol, and an antimuscarinic drug should be available in all settings in which patients are cared for in the last days of life. The experts represent three different continents, different health care systems, structures, and cultures making this study valuable in an international context.

Drug usage and availability across the participating countries varies, which may be seen as a strength of this study. On the other hand, there is also a potential limitation in the geographical spread, since a majority of the participating experts work in countries that are among the most developed in terms of PC.¹⁷

The consensus opinion from this study is not meant to substitute the need for individual symptom assessment in the dying patient and personalized choice of treatment. It is instead proposed as guidance to support improved quality of care on both an organizational and individual level. While other lists of drugs essential for PC have been formulated, they have addressed broader time frames in PC, rather than, as in this case, being specific to the last days of life.^9,10,18,19 For example, the essential drug recommendations from the International Association for Hospice & Palliative Care (IAHPC) also includes morphine along with three other opioids, as well as midazolam, diazepam, and lorazepam, among the 33 different pharmacological substances suggested for a wider variety of common symptoms.^9,10 The drugs suggested in this Delphi study are included in most other expert recommendations, and they may be particularly useful, as the same medication can alleviate more than one symptom. Other studies with expert ranking of important drugs support our suggestions;^20–22 for example, the list from Australia includes both midazolam and diazepam.²¹ The present consensus is also supported by the recommendations for drugs on demand in the Liverpool Care Pathway for the dying patient.²³

Despite relative consensus in opinions and recommendations from different expert groups, some problems remain in implementation in practice. This may particularly be the case for opioids, as these drugs are often not readily accessible due to legal restraints. A lack of competency and experience among physicians and nurses may also inhibit their administration and use, with fear of opioids still playing a role in many countries and settings.

Other factors may hinder use of appropriate drugs; for example, fear of the negative effects on breathing in combination with limited knowledge and experience may lead to reluctance to use opioids to alleviate dyspnea. Haloperidol is often used in specialist PC for the alleviation of both delirium and nausea, but in geriatric and residential care facilities, the use of neuroleptic drugs has been criticized for reasons not related to care of the dying.²⁴ While we found consensus on using an antimuscarinic drug among the participating PC experts, several different drugs were proposed. This may reflect the lack of strong scientific evidence showing clear benefit for one drug over the other in treating RTS.^25–27 This is also reflected in other similar drug lists, which mention different glycopyrronium/hyoscine formulations.^9,10,20–23 The geographical location of experts and availability of drugs in each setting influenced the choice of drugs we report here. For example, there is no diamorphine in the list, despite this substance's wide use in the UK, especially in community-based care.

It should however be remembered that the participating experts in our study were all experienced palliative medicine physicians. This is in contrast to other expert panels, which were composed of participants chosen by other selection principles^9,10 or who had less experience in specialist palliative medicine.²¹ The high response rate of close to 70% in our study also adds to its impact; other similar lists have been troubled by notably lower participation rates, e.g., the Australian list is based on a 29% response rate.²¹

While carrying out the Delphi study, differences in available knowledge and expertise across countries even in specialist PC care became apparent. One example is that it was more difficult to find PC physicians meeting our criteria for ‘expert’ in Slovenia than in the other participating countries, and Slovenia also had a notably lower participation rate than other countries. This was in part due to the exclusion of practicing PC physicians with <3 years' experience. This is likely to be a structural feature, indicating that specialist PC has not been available in Slovenia as long as in some other countries, that it is a smaller country, and that English is less commonly used. This points to the importance and potential benefits of international collaborative projects, such as OPCARE9.

Summary and Conclusions

As previously noted, there is a lack of research-based literature on pharmacological treatment in the last days of life. Based on this consensus, other existing recommendations, and with consideration given both variety in local treatment traditions and drug availability, we suggest that four classes of essential drugs should be available for all patients in the last days of life. These classes are: (1) an opioid—when possible, morphine, as it is well known and relatively inexpensive; (2) a benzodiazepine—preferably midazolam, since this is a short acting, well-tolerated substance, which can be used both for subcutaneous injections as well as in combination with other drugs in syringe drivers; (3) an antipsychotic—preferably haloperidol, which is also useful for alleviation of nausea and vomiting, delirium, and hallucinations; and (4) a glycopyrronium/hyoscine formulation for RTS.

Geographical location and specific setting will have impact on which individual drug is selected. A recommendation based on a broad consensus—not only on the opinions resulting from this international Delphi study, but supported by other existing documentation—has great implications for change of policy, education, and research, and in some countries may exert influence on legislation and drug availability. With the help of this consensus opinion in regard to appropriate pharmacological care for common symptoms in the last days of life, there is a potential to develop recommendations for standard care of patients with cancer approaching death. We suggest four key elements for optimizing pharmacological treatment of symptoms in the last days of life:

• Awareness of which drugs are essential

• Knowledge about their appropriate use

• Accessibility to the above mentioned essential drugs for symptom alleviation

• Individualized prescriptions for the dying patient

However, the lack of accessibility of different drugs in some countries and settings is an important factor leading to inequitable implementation and inhibiting immediate use of many recommendations. The consensus opinion found here about essential drugs for symptom treatment in the last days of life can provide one means of exerting pressure toward constructive change in national, regional, and local drug policies, and can provide a basis for minimum standards of care.

To decrease the gap between knowledge and practice, we recommend that these four classes of essential drugs should be available for every dying patient, not only those with cancer. As people approach death, the less disease specific symptoms tend to be. The lack of good scientific evidence for pharmacological treatment of symptoms in the dying patient calls for efforts to design and conduct appropriate interventions and studies in this vulnerable group to minimize avoidable distress in proximity to death.

Footnotes

Acknowledgments

The authors would like to thank the international experts who participated in this study. We would also like to thank the country representatives in the OPCARE9 countries for their assistance in different phases of the project: Maria L. Daud (Argentina); Maren Galushko (Germany); Guido Miccinesi (Italy); Jean Clark and Simon Allan (New Zealand); Franzisca Domeisen Benedetti (Switzerland); and Lia van Zuylen, Natasja Raijmakers (the Netherlands).

OPCARE9 is funded by the European Commission's Seventh Framework Programme (contract number HEALTH-F2-2008-202112) with the aim of improving care in the last days of life by systematically identifying existing knowledge as well as knowledge gaps. The project aims to do this collaboratively across Europe and beyond to integrate knowledge from a range of health care environments and cultures to avoid duplication of resources and effort. Project outputs include the dissemination of key findings and recommendations for further research and development for care in the last days of life. Further information on OPCARE9, its members, work packages, outputs, and contact details, can be found online at .

Author Disclosure Statement

OPCARE9 is a coordination and support action project funded by Framework Programme 7, European Union. No conflicting financial interests exist.

References

Teunissen

, Wesker

, Kruitwagen

, de Haes

, Voest

, de Graeff

. Symptom prevalence in patients with incurable cancer: A systematic review. J Pain Symptom Manage, 2007; 34,1:94–104.

Seow

, Barbera

, Sutradhar

, Howell

, Dudgeon

, Atzema

et al. Trajectory of performance status and symptom scores for patients with cancer during the last six months of life. J Clin Oncol, 2011; 29,9:1151–1158.

Lichter

, Hunt

. The last 48 hours of life. J Palliat Care, 1990; 6,4:7–15.

Klinkenberg

, Willems

, van der Wal

, Deeg

. Symptom burden in the last week of life. J Pain Symptom Manage, 2004; 27,1:5–13.

Lokker

, van Zuylen

, Veerbeek

, van der Rijt

, van der Heide

. Awareness of dying: It needs words. Support Care Cancer, 2012; 20,6:1227–1233.

Caraceni

, Hanks

, Kaasa

, Bennett

, Brunelli

, Cherny

et al. Use of opioid analgesics in the treatment of cancer pain: Evidence-based recommendations from the EAPC. Lancet Oncol, 2012; 13,2:58–68.

Glare

, Pereira

, Kristjanson

, Stockler

, Tattersall

. Systematic review of the efficacy of antiemetics in the treatment of nausea in patients with far-advanced cancer. Support Care Cancer, 2004; 12,6:432–440.

Jackson

, Lipman

. Drug therapy for anxiety in palliative care. Cochrane Database Syst Rev, 2004; 1:CD004596.

De Lima

. The international association for hospice and palliative care list of essential medicines for palliative care. Palliat Med, 2006; 20,7:647–651.

10.

De Lima

, Krakauer

, Lorenz

, Praill

, Macdonald

, Doyle

. Ensuring palliative medicine availability: The development of the IAHPC list of essential medicines for palliative care. J Pain Symptom Manage, 2007; 33,5:521–526.

11.

Gomes

, Cohen

, Deliens

, Higginson

. International trends in circumstances of death and dying amongst older people. Gott

. Ingleton

. Living with Ageing and Dying: Palliative and End of Life Care for Older People. Oxford: Oxford University Press, 2011; 3–18.

12.

Mason

, Dowson

, Gambles

, Ellershaw

. OPCARE9: Optimising research for cancer patient care in the last days of life. Eur J Palliat Care, 2012; 19,1:17–19.

13.

Lindqvist

, Lundh Hagelin

, Lundquist

, Sauter

, Lunder

, Rasmussen

et al. OPCARE9 work package 3: Complementary comfort care at the end of life. Eur J Palliat Care, 2012; 19,2:69–72.

14.

Lindqvist

, Tishelman

, Hagelin

, Clark

, Daud

, Dickman

et al. Complexity in non-pharmacological caregiving activities at the end of life: An international qualitative study. PLoS Med, 2012; 9,2:e1001173.

15.

Hasson

, Keeney

, McKenna

. Research guidelines for the Delphi survey technique. J Adv Nurs, 2000; 32,4:1008–1015.

16.

Keeney

, Hasson

, McKenna

. Consulting the oracle: Ten lessons from using the Delphi technique in nursing research. J Adv Nurs, 2006; 53,2:205–212.

17.

Economist Intelligence Unit. The quality of death: Ranking end-of-life care across the world. http://www.eiu.com/site_info.asp?info_name=qualityofdeath_lienfoundation&page=noads

18.

Bishop

, Stephens

, Goodrich

, Byock

. Medication kits for managing symptomatic emergencies in the home: A survey of common hospice practice. J Palliat Med, 2009; 12,1:37–44.

19.

Wowchuk

, Wilson

, Embleton

, Garcia

, Harlos

, Chochinov

. The palliative medication kit: An effective way of extending care in the home for patients nearing death. J Palliat Med, 2009; 12,9:797–803.

20.

Dickerson

. The 20 essential drugs in palliative care. Eur J Palliat Care, 1999; 6,4:130–135.

21.

Good

, Cavenagh

, Currow

, Woods

, Tuffin

, Ravenscroft

. What are the essential medications in pallative care? A survey of Australian palliative care doctors. Aust Fam Physician, 2006; 35,4:261–264.

22.

LeGrand

, Tropiano

, Marx

, Davis

, Walsh

. Dying at home: Emergency medications for terminal symptoms. Am J Hosp Palliat Care, 2001; 18,6:421–423.

23.

Glare

, Dickman

, Goodman

. Symptom control in care of the dying. Ellershaw

. Wilkinson

. Care of the Dying: A Pathway to Excellence, 2nd. Oxford: Oxford University Press, 2011; 33–62.

24.

Stevenson

, Decker

, Dwyer

, Huskamp

, Grabowski

, Metzger

et al. Antipsychotic and benzodiazepine use among nursing home residents: Findings from the 2004 National Nursing Home Survey. Am J Geriatr Psychiatry, 2010; 18,12:1078–1092.

25.

Wee

, Hillier

. Interventions for noisy breathing in patients near to death. Cochrane Database Syst Rev, 2008; 1:CD005177.

26.

Hugel

, Ellershaw

, Gambles

. Respiratory tract secretions in the dying patient: A comparison between glycopyrronium and hyoscine hydrobromide. J Palliat Med, 2006; 9,2:279–284.

27.

Wildiers

, Dhaenekint

, Demeulenaere

, Clement

, Desmet

, Van Nuffelen

et al. Atropine, hyoscine butylbromide, or scopolamine are equally effective for the treatment of death rattle in terminal care. J Pain Symptom Manage, 2009; 38,1:124–133.