Continuous Deep Sedation in Medical Practice: A Descriptive Study

Article Outline

• Abstract
• Introduction
• Methods
  • Questionnaire
  • Sample and Questionnaire Return
  • Response Rate and Response Bias
  • Analysis
• Results
• Discussion
  • Strengths and Weaknesses of the Study
• Acknowledgments
• References
• Copyright

Abstract 

Context

Continuous deep sedation (CDS) until death can form an effective part of palliative care. In The Netherlands and Belgium, CDS is sometimes regarded as an alternative to euthanasia, and the involvement of palliative care specialists is low. Provision of CDS through opioids alone is not recommended.

Objectives

This study investigates the use of CDS in the United Kingdom.

Methods

A survey of 8,857 doctors, of whom 3,733 (42%) replied, with 2,923 reporting on their last patient who died.

Results

In total, 18.7% (17.3–20.1) of the doctors attending a dying patient reported the use of CDS. CDS was more likely when patients were younger or were dying of cancer. Specialists in care of the elderly were least likely to report the use of CDS; doctors in other hospital specialties were most likely to report its use. CDS was associated with a higher rate of requests from patients or relatives for a hastened death and with a greater incidence of other end-of-life decisions containing some intent to end life by the doctor. Doctors supporting legalization of euthanasia or physician-assisted suicide, or who were nonreligious, were more likely to report using CDS. There was palliative care team involvement in half of all CDS cases, and prescription of opioids alone for sedation occurred in one-fifth of the cases but was not reported by specialists in palliative care.

Conclusion

This study provides baseline data for monitoring future trends in the United Kingdom and highlights the need for a fuller understanding of the circumstances in which CDS occurs in particular care settings.

Key Words: Sedation, end-of-life decisions, palliative care, end-of-life care

Back to Article Outline

Introduction 

Sedation to address refractory symptoms in patients nearing death can sometimes take the form of continuous deep sedation (CDS) until death, also known as “terminal” or “palliative” sedation.¹ Establishing the prevalence of CDS, the factors that trigger its use, and the degree to which its implementation conforms to ethical and clinical standards is important in ensuring that it is used appropriately. Yet, in prevalence studies, comparisons are difficult when investigators define and measure CDS in different ways.², ³, ⁴, ⁵ Additionally, many studies of CDS concern specific patient groups, such as those receiving palliative care, rather than representing a full range of medical practice.

A group of studies that use the same approach to measurement, and are, therefore, comparable, have reported on the prevalence of CDS in national populations of deaths in European countries,⁶, ⁷, ⁸, ⁹, ¹⁰, ¹¹ with rates ranging between 2.5% of deaths in Denmark in 2001–2002⁶ and 16.5% in the United Kingdom in 2007–2008.¹¹ CDS is more common in patients who are younger or dying from cancer.⁶, ⁷, ⁸ In The Netherlands, trends over time have been reported, showing a rate rising from 5.6% to 7.1% of all deaths between 2001 and 2005.⁸ This rise was accompanied by a decline in the rate of euthanasia, prompting the suggestion that CDS may sometimes be regarded as an acceptable alternative to euthanasia by Dutch physicians.⁸ In Belgium, too, rates rose from 8.2% of all deaths in 2002 to 14.5% in 2007.⁹

The ethical issues involved in opting for CDS are experienced by some as similar to those involved in opting for euthanasia or physician-assisted suicide (PAS).¹², ¹³, ¹⁴ Some commentators encourage this perception,¹⁵, ¹⁶ which is more common in The Netherlands and Belgium, where euthanasia is legal, than in the United Kingdom, where it is not.¹³ Between 6% and 14.5% of CDS cases in The Netherlands are associated with an explicit request from the patient for euthanasia or assisted suicide.⁸, ¹⁷ A survey asking a national sample of Dutch physicians to report on their last case of CDS¹⁸ found that hastening death was an explicit intention in 17% of cases and partly the intention in a further 47%. Withholding artificial nutrition and hydration alongside CDS occurs in 39% of CDS cases in Belgium⁶, ¹⁰ and 56%–80% of CDS cases in The Netherlands.⁶, ⁸, ¹⁷ In The Netherlands, CDS is believed to have shortened life by more than a week in 27% of cases.¹⁹

However, different interpretations may be applied in jurisdictions that prohibit euthanasia or PAS. A survey of American physicians²⁰ showed that most distinguished between “terminal sedation” and PAS. They supported sedation but opposed assisted suicide. A study of patients in a hospice setting indicated no support for the view that life is shortened for patients receiving sedation when medication is carefully titrated against distress.⁵ After reviewing several other studies of this sort, De Graeff and Dean²¹ argue that “palliative sedation therapy” is distinct from euthanasia, because its main intention is symptom relief rather than bringing about death, and it is an intervention applied proportionately to the severity of distress. These authors outline the now widely accepted view that midazolam is the first-line choice for providing palliative sedation, that use of opioids alone for this purpose is to be discouraged, and that a decision to provide sedation should be accompanied by advice from palliative care specialists. Evidence from studies of Dutch practice shows low levels of consultation with palliative care specialists⁸ and a proportion of CDS cases where opioids alone were used,²², ²³ although this proportion is lower when there is palliative care involvement.²² The introduction of national guidelines for palliative sedation has been associated with changes toward the use of medications recommended in the guidelines and more frequent involvement of patients in the decision-making process.¹⁷

In the United Kingdom, end-of-life care is provided in hospitals, freestanding hospices and palliative care units within hospitals, in care homes for elderly people, and in private homes. Specialist palliative care for people with cancer, and increasingly other conditions, is delivered by a combination of domiciliary services, hospital support teams, and specialist inpatient provision. This supports general practitioners attending people who are approaching death in their own homes or in care homes and hospital specialists attending inpatients. Measured in terms of the ratio of services to one million of the population, the United Kingdom has the highest level of development of palliative care in Europe.²⁴

This study reports a survey of U.K. doctors, which obtained information about the care of their last patient who died, focusing on the characteristics of doctors and patients when CDS was used and comparing this with deaths that did not involve CDS. This allowed an assessment of the degree to which recommendations for good practice in implementing CDS are being followed.

Back to Article Outline

Methods 

Questionnaire 

A structured questionnaire designed for doctors included questions about the respondent's age, gender, medical specialty, and the number of deaths attended in a normal year. Respondents also were asked questions about their attitudes toward the legalization of euthanasia and PAS, and the strength of any religious beliefs, worded as for the British Social Attitudes survey.²⁵, ²⁶ If doctors reported having attended a patient who died in the past year, they were asked a series of other questions about the age, gender, place of death, and cause of death of the most recently deceased patient. They were asked whether they withheld or withdrew a treatment with the explicit intention of hastening the end of life and whether they used any drugs “partly intending to end life” or with the “explicit intention of hastening the end of life.”²⁷ Respondents also were asked whether “an explicit request to hasten the end of the patient's life” was made by relatives of the patient, nursing or other care staff, or someone else, and whether the patients had themselves expressed “a wish for the end of his or her life to be hastened.” A further question asked was whether a “palliative care team” had been “involved in the care for the patient during the last month before death.”

Doctors were then asked, “Was the patient continuously and deeply sedated or kept in a coma before death?” Those who answered “yes” were then asked, “Which medication was given for sedation?” and could check any or all of the four options: 1) midazolam, 2) other benzodiazepine, 3) morphine or another opioid, and 4) other type of medication. Further questions asked were what time before death CDS was started (expressed in either hours, days, or weeks) and what were “the most important reasons for this sedation,” with response options as shown in Table 1. All of the questions about sedation were taken from a direct translation of the survey instrument used in studies in The Netherlands.⁷, ⁸, ²⁸

Table 1. Reasons Given for Continuous Deep Sedation and Time Before Death That Sedation Was Started in 519 Cases

	%	95% CI
Reasons (more than one can apply to each case)a
Patient had intractable pain	39.1	34.8–43.4
Patient had intractable psychological distress	24.6	20.8–28.4
Patient had other intractable symptoms	42.2	37.8–46.6
Request or wish of the patient	5.4	3.4–7.4
Request or wish of relatives	10.4	7.7–13.1
Other	32.8	28.6–37.0
Time before death when sedation beganb
<24 hours	30.6	26.5–34.7
1–7 days	61.1	56.7–65.5
More than 1 week	8.3	5.8–10.8

CI = confidence interval.

Sample and Questionnaire Return 

Binley's database (www.binleys.com) of U.K. medical practitioners was used to send questionnaires to 8,857 randomly sampled, working U.K. medical practitioners, comprising 2,829 general practitioners, 443 neurologists, 836 specialists in care of the elderly, 462 specialists in palliative medicine, and 4,287 in other hospital specialties (excluding specialties, such as public health, where doctors do not normally treat people who die). The sampling strategy oversampled certain specialties (e.g., palliative medicine) in proportion to the population; hence, weighting of results (described later) adjusts for this. An initial mailing and two follow-up reminders were sent between November 2007 and April 2008. The sensitive nature of the subject matter was addressed by ensuring (as in earlier surveys using this method) that respondents knew their replies could not be traced back to them. No identifying marks were placed on the questionnaire, and a card was returned by respondents separately to indicate that a response had been made and no further reminders should be sent.

Response Rate and Response Bias 

The overall response rate was 42.1%, that is, 3,733 doctors responded, of whom 2,923 (78.3% of respondents) had attended a patient who had died in the previous 12 months. Specialists in palliative medicine produced the highest response rate (67.3%), followed by specialists in care of the elderly (48.1%), neurologists (42.9%), other hospital specialists (40.1%), and general practitioners (39.3%). This compares with a response rate of 35.9% in a recent Dutch study of CDS based on a similar retrospective questionnaire.¹

An investigation of response bias is reported elsewhere.²⁶ Comparisons of responding doctors with national medical workforce statistics and a survey of nonresponders (as in Fischer et al.²⁹) were done. As in Fischer et al., nonresponders were not significantly different from responders in their degree of support for euthanasia or PAS. Nonresponders tended to be younger, to have inadequate time to complete the questionnaire, and to believe it was only relevant to reply if they normally attended to dying patients or were involved in terminal care. The patients reported on by responders were more likely to have died from cancer and less likely to have died from cardiovascular disease than those in national mortality statistics.

Analysis 

Neurologists were combined with “other hospital” specialists for the analysis reported in this article. The four questions about attitudes toward the legalization of euthanasia and PAS were combined to produce a 4-point scale ranging from 4, meaning high support for assisted dying, to 1, meaning high opposition to assisted dying (reported in Table 2). If doctors answered “yes” to any of the questions about having withheld or withdrawn a treatment or having used a drug with either the partial or explicit intention of ending life, they were recorded as having taken an end-of-life decision with some intent to end life (reported in Table 3).

Table 2. Continuous Deep Sedation Reported by 2,782 Doctors With Regard to Doctor Characteristics

Characteristics	% Sedated	95% CI	n (100%)	P-value
Doctor's age, years
<35	16.7	13.1–20.3	420
36–55	19.5	17.7–21.3	1,928	0.28
56–65	17.3	13.7–20.9	420
Total			2,768
No data on age			14
Doctor's gender
Female	16.6	14.3–18.9	990
Male	19.9	18.0–21.8	1,757	0.03
Total			2,747
No data on gender			35
Doctor's specialty
General practice	17.5	14.7–20.3	732
Palliative medicine	15.0	10.5–19.5	246
Elder care	10.6	7.1–14.1	302	<0.001
Other hospital	21.4	19.3–23.5	1,502
Total			2,782
Level of religious faith
High	16.4	12.6–20.2	359
Medium	18.3	16.5–20.1	1,683	0.03
Low	22.7	19.2–26.2	541
Total			2,583
No data on faith			199
Attitude toward assisted dying
Very opposed	14.3	12.0–16.6	900
Moderately opposed	20.2	17.6–22.8	902
Moderately in favor	20.9	18.0–23.8	737	<0.001
Very much in favor	25.6	19.2–32.0	176
Total			2,715
No data on attitudes			67

CI = confidence interval.

Table 3. Continuous Deep Sedation Reported by 2,782 Doctors With Regard to Intention of Caregiver, Wish of Patient, and Involvement of Palliative Care Team

	% Sedated	95% CI	n (100%)	P-value
Patient expressed wish for a hastened death
Wished for	24.1	19.0–29.2	274
Not wished for	18.0	16.5–19.5	2,490	0.01
Total			2,764
No data on wishes			18
Request to hasten end from someone elsea
Requested	39.1	30.1–47.6	128
Not requested	17.7	16.2–19.2	2,590	<0.001
Total			2,718
No data on requests			64
ELD with some intent to end lifeb
ELD present	42.8	36.1–49.5	208
ELD not present	16.4	15.0–17.8	2,529	<0.001
Total			2,737
No data on ELD			45
Palliative care team involved
Involved	19.9	17.7–22.1	1,255
Not involved	17.3	15.3–19.3	1,438	0.08
Total			2,693
No data on team care			89

CI = confidence interval; ELD = end-of-life decision.

For the analyses reported in this article, all data for results reported after this point are weighted by the doctor's specialty to mirror proportions of specialties in the population and by the number of deaths normally attended by each doctor in one year, so that doctors attending fewer deaths are not overrepresented. Data were analyzed using SPSS 16.0 (SPSS Inc., Chicago, IL). Chi-squared tests and 95% confidence intervals were used to identify statistically significant differences between groups in bivariate analysis. For multivariate analysis, logistic regression was used to determine the independent association of selected variables whether CDS was reported or not. Variables for the multivariate analysis were selected on the basis of being significantly associated with the dependent variable at the bivariate stage, excluding variables showing multicollinearity with included variables. Odds ratios were considered statistically significant at the P < 0.05 levels.

Back to Article Outline

Results 

Five hundred and nineteen (18.7%; 17.3–20.1) of 2,782 doctors attending a patient who died stated that CDS had occurred (141 doctors who said they attended a dying patient did not know the answer to the question about CDS or failed to provide an answer). This figure differs from the 16.5% of deaths receiving CDS reported elsewhere from this survey²⁷ because of different approaches to weighting, which in the earlier report, took account of the cause of death to generalize to the population of deaths. In contrast, this report generalizes to the population of doctors who have attended people who died in the previous year.

Table 2 shows the incidence of CDS by selected characteristics of doctors, and Table 4 shows this by patients' characteristics. Although the doctor's age was not significantly related to sedation, male doctors and those in “other hospital” specialties were more likely to report a case involving sedation. Doctors specializing in the care of elderly patients were least likely to report sedation, a finding consistent with the observation that sedation is more common in younger patients. It is also consistent with the data for place of death, which indicates sedation is common in hospitals and least common in care homes. Dying from cancer is particularly associated with sedation but not with the patient's gender. Doctors who were nonreligious and, in particular, those who supported the legalization of assisted dying were more likely to report a case of sedation. When the mean number of annual deaths attended by doctors reporting sedation cases (mean [standard deviation (SD)]: 36.5 [50.2]) was compared with the mean number of annual deaths attended by those who did not report a sedation case (mean [SD]: 37.9 [54.5]), no significant difference was found (t-test P = 0.595).

Table 4. Continuous Deep Sedation Reported by 2,782 Doctors With Regard to Patient Characteristics

Characteristics	% Sedated	95% CI	n (100%)	P-value
Gender of patient
Female	17.7	15.6–19.8	1,296
Male	19.7	17.5–21.9	1,433	0.16
Total			2,729
No data on gender			53
Age of patient, years
0–59	26.4	22.7–30.1	560
60–79	19.7	17.5–21.9	1,220	<0.001
80+	12.9	10.8–15.0	989
Total			2,769
No data on age			13
Cause of death
Cancer	20.9	18.6–23.2	1,207
Cardiovascular	11.5	9.0–17.0	520
Respiratory	17.4	13.1–21.7	298	<0.001
Nervous system	11.3	6.2–16.4	150
Other	23.5	19.6–27.4	452
Total			2,627
No data on cause			155
Place of death
Hospital	19.9	18.1–21.7	1,835
Hospice	13.0	9.0–17.0	270
Care home	9.8	5.5–14.1	183	0.001
Private home	19.6	16.0–23.2	474
Total			2,762
No data or “other” place			20

CI = confidence interval.

A request for a hastened death from the patient occurred in 12.8% (66 of 514) of CDS cases and a request from someone else—for example, relatives or nurses—in 9.8% (50 of 458). Table 3 shows that both types of requests are associated with a higher likelihood that CDS was reported. In 17.6% (89 of 504) of CDS cases, an end-of-life decision with some intent to end life (withdrawing or withholding treatment or providing a drug with the partial or explicit intention of ending life) was reported; Table 3 shows that this was associated with a higher likelihood that CDS was provided.

The involvement of a palliative care team in the last month of life was reported in 50.1% (250 of 499) cases of CDS, but Table 3 shows that this was not associated with variation in reporting of CDS. Further analysis of this shows that CDS accompanied by an end-of-life decision that contained some intent to end life was rarer when there was palliative care team involvement in the last month of life (10.7% [26 of 244] of cases with team involvement vs. 25.6% [61 of 238] of cases without such involvement [P < 0.001 based on Chi-squared test]). Palliative care team involvement was more common in CDS cases involving a patient reported as having made a request for a hastened death (18.2% [45 of 247]) vs. when a request was not reported (8.5% [21 of 247]) (P = 0.002 based on Chi-squared test).

Table 5 reports the results of a logistic regression, showing significantly higher odds of CDS being reported in deaths from cancer, younger patients, and by doctors working in “other hospital” specialties. Being in favor of assisted dying also increases the odds of reporting a case of CDS. The gender of the patient remained insignificant.

Table 5. Variables Associated With Reports of Continuous Deep Sedation: Logistic Regression (n = 2,525)^a

OR = odds ratio; CI = confidence interval.

The reasons given for sedation are shown in Table 1, which shows that doctors more often regard the decision to sedate as a response to physical, psychological, and other symptoms, than a response to requests for sedation. Table 1 also reports the time before death at which sedation was started.

The drugs that the doctors reported having used for CDS are shown in Table 6; 20.7% (100 of 484) of CDS cases involved morphine or another opioid alone. Table 6 shows that this was never done by palliative medicine specialists, all of whom used midazolam. Doctors in “other hospital” specialties were the least likely to use midazolam and the most likely to use opioids alone. Table 6 also shows that doctors working in “other hospital” specialties reporting cases of CDS were the least likely to report the involvement of a palliative care team in the last month of life. Overall, 50.1% (250 of 499) of doctors reporting CDS cases reported such involvement.

Table 6. Drugs Used for Continuous Deep Sedation and Involvement of Palliative Care Team, With Breakdown by Specialty, in 519 Cases

	%	95% CI	n (100%)	P-value
a. Drugs used (more than one can apply to each case)
Midazolam	59.7	55.3–64.1
Other benzodiazepine	5.3	3.3–7.3
Opioid (e.g., morphine)	67.1	62.9–71.3
Other	21.0	17.4–24.6
n (100%)	484
No data on drugs used	35
b. Drugs used by specialty
Only morphine/other opioid
General practice	17.2	10.5–23.9	122
Palliative medicine	0.0	–	37
Elder care	23.3	8.2–38.4	30	0.004
Other hospital	24.3	19.4–29.2	296
Total			484
No data			35
Used midazolam
General practice	80.5	73.5–87.5	123
Palliative medicine	100.0	–	37
Elder care	66.7	49.8–83.6	30	<0.001
Other hospital	45.4	39.7–51.1	295
Total			484
No data			35
c. Palliative care team involved in last month, by specialty
General practice	72.7	65.0–80.4	128
Palliative medicine	89.2	79.2–99.2	37
Elder care	64.5	47.7–81.3	31	<0.005
Other hospital	34.3	29.0–39.6	303
Total			499
No data			20

CI = confidence interval.

Doctors working in “other hospital” specialties were asked to provide the name of their specialty. CDS was particularly common in deaths reported by specialists in intensive and critical care, anesthetics and intensive care unit (ICU) anesthetics, and cardiac surgery (63.3% [88 of 139] in these specialties vs. 17.2% [233 of 1,349] in other hospital specialties; P < 0.0005). Because doctors in these specialties were also particularly likely to report an “other” reason for sedation (83.8% [67 of 80] vs. 31.5% [68 of 216]; P < 0.0005), the qualitative comments about these cases (made by 56 of the doctors in these specialties who reported an “other” reason for a CDS case) were examined. Seventy-one percent (40 of 56) of these indicated that sedation was provided to enable patients to tolerate ventilation, and a further 23% (13 of 56) wrote that sedation was a routine in the ICU to enable treatment or was needed pre- or postoperatively.

In contrast to this group, a further subset of 200 CDS cases were selected in which there was a high likelihood of CDS being a palliative measure in response to intractable symptoms. This included only deaths from cancer, excluding those reported by the hospital specialists listed in the earlier paragraph and excluding those where “other” was given as a reason for sedation. Use of midazolam for CDS was higher in this group (75.5% [151 of 200]) than for all CDS cases (Table 6), but the rate at which morphine or another opioid alone was used was also as high in this group as for all CDS cases (18.5% [37 of 200]). Use of midazolam for CDS among hospital specialists (including neurologists) reporting for this group of patients became more common than in general CDS cases (63.5% [47 of 74]), as did the involvement of a palliative care team in the last month (78.4% [58 of 74]). But 24.7% (18 of 73) of these hospital specialists still reported using morphine or another opioid alone for these CDS patients (and 18.4% [16 of 87] of general practitioners in this subsample reported this).

Back to Article Outline

Discussion 

As in other studies in national populations,⁶, ⁷, ⁸ CDS in the United Kingdom is more common in patients who are younger or dying of cancer. Doctors reporting on older patients are often those specializing in the care of the elderly or reporting on people who die in care homes, and in all of these categories, CDS is less often reported. Doctors working in hospital specialties (apart from care of the elderly) are most likely to report cases of CDS. Unlike Belgium, where CDS is much less common in deaths at home than in hospital,⁶ and The Netherlands, where general practitioners are less likely than hospital specialists to report CDS,⁸ in the United Kingdom, CDS rates for general practitioners are not significantly different from hospital specialists or in private homes compared with hospitals. The high frequency of CDS in the United Kingdom²⁷ may in part be because of a greater willingness to use this in deaths occurring at home as well as encouragement from national end-of-life care guidelines.³⁰

Although the involvement of a palliative care team is not associated with variation in the rate of CDS, such involvement (50% of all U.K. cases) is considerably higher than that in The Netherlands, where it is present for just 9% of CDS cases.⁸ Palliative care team involvement is associated with a reduced rate of sedation accompanied by an end-of-life decision with intent to hasten death; such involvement is more common in cases of sedation where a patient has made a request for a hastened death.

The fact that cases of CDS were more likely to co-occur with other end-of-life decisions that involved intent to end life does not necessarily mean that this intention was being applied to the decision to give CDS nor is the fact that CDS cases were associated with a higher rate of requests for a hastened death in itself evidence that CDS was a response to such requests. Doctors indicate more often that CDS was a response to intractable pain, psychological distress, or other symptoms than a response to a request for it from either patients or relatives. It is arguable that such intractable problems cause a variety of things quite independently, including requests for a hastened death, decisions to provide CDS, and decisions to take other actions partly or wholly intended to hasten the end of life.

However, CDS is more likely to be reported by doctors who support the legalization of euthanasia or PAS and who are nonreligious. This suggests that the decision to provide CDS may be influenced by having attitudes that permit medical actions that shorten life.

The incidence of opioid-only prescriptions for CDS occurs at a similar rate in the U.K. as it did in The Netherlands in 2003–2005, where it was reported to be 22% of the CDS cases;²² the rate in The Netherlands declined to just 9% in a 2007 follow-up survey of Dutch doctors after guidelines had been introduced.¹⁷ In the present study, all palliative care specialists conform to the recommendation²¹ that midazolam is used for sedation and that opioids alone should not be used for this purpose. Doctors working in “other hospital” specialties were the least likely to conform to these recommendations.

However, further investigation of CDS in hospital settings, such as the ICU, showed that this survey method includes cases of sedation that are rather different in purpose and in choice of medication than those encountered in palliative and terminal care settings. Applying recommendations for good practice developed for palliative care²¹ to ICU settings is probably inappropriate. Guidelines for sedation in critical care³¹ make it clear, for example, that a wider range of medications are considered appropriate than in palliative care.

This is the first national survey of CDS in U.K. medical practice, providing baseline data for monitoring future trends. Certain aspects of the findings indicate a need for better understanding of the circumstances in which CDS is provided. For example, there is a need to know what leads to the use of opioids alone for sedation in spite of generally high levels of palliative care team involvement. The association between providing CDS and doctors' attitudes toward assisted dying and religious beliefs as well as with the incidence of requests for an assisted death requires a better understanding of the ethical reasoning of doctors who decide to provide CDS in particular cases. The degree to which the decision to provide CDS is shared with patients or families, or is associated with withdrawal of artificial nutrition and hydration, awaits further research in the U.K. context.

Strengths and Weaknesses of the Study 

The study benefits from using questions worded in ways that permit comparability with other national studies and from a disproportionate stratified sample that allows for comparison among specialties. However, the overall response rate, though it compares favorably with a Dutch study using a similar retrospective method,¹ raises concerns about representativeness. Because the investigation of nonresponse showed a bias toward cases involving terminal illness and cancer deaths, it is likely that the figure of 18.7% given for the overall rate of CDS among doctors reporting on a death overestimates the prevalence of CDS. However, the comparisons between groups of doctors and patients are not affected by this bias if it is presumed that the direction of the bias would apply equally to all categories involved in any one comparison.

More detailed, qualitative, observational research of particular care settings is needed to overcome the limitations inherent in any questionnaire relying on retrospective recall and fixed-choice questions. This is needed to provide a better understanding of the issues raised earlier and to understand the range of circumstances involved in providing CDS in hospital practice, which this study shows to be in need of better evidence reflecting the context in which these decisions are taken.

Back to Article Outline

Acknowledgments 

Buket Okucu, Niraj Sachania, and Lavenhams Press assisted with data collection and data entry. The study was supported and advised by a group including representatives of the funding bodies (The National Council for Palliative Care, Age Concern, the Motor Neurone Disease Association, the Multiple Sclerosis Society, Help the Hospices, Macmillan Cancer Support, and Sue Ryder Care). Dr. Nigel Sykes and Professor Jane Seymour also participated in this committee and commented on drafts of this article. The author is grateful to them and to the doctors who found time to fill in the questionnaire.

Back to Article Outline

References 

1. Hasselaar JG, Reuzel RP, van den Muijsenbergh ME, et al. Dealing with delicate issues in continuous deep sedation. Varying practices among Dutch medical specialists, general practitioners, and nursing home physicians. Arch Intern Med. 2008;168:537–543
   • View In Article
   • CrossRef
2. Claessens P, Menten J, Schotsmans P, Broeckaert B. Palliative sedation: a review of the research literature. J Pain Symptom Manage. 2008;36:310–333
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (301 KB)
   • CrossRef
3. Murray SA, Boyd K, Byock I. Continuous deep sedation in patients nearing death: imprecise taxonomy makes interpreting trends difficult. BMJ. 2008;336:781–782
   • View In Article
   • CrossRef
4. Bonito V, Caraceni A, Borghi L, et al. The clinical and ethical appropriateness of sedation in palliative neurological treatments. Neurol Sci. 2005;26:370–385
   • View In Article
   • MEDLINE
   • CrossRef
5. Sykes N, Thorns A. The use of opioids and sedatives at the end of life. Lancet Oncol. 2003;4:312–318
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (608 KB)
   • CrossRef
6. Miccinesi G, Rietjens JA, Deliens L, et al. Continuous deep sedation: physicians' experiences in six European countries. J Pain Symptom Manage. 2006;31:122–129
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (95 KB)
   • CrossRef
7. van der Heide A, Onwuteaka-Philipsen BD, Rurup ML, et al. End-of-life practices in The Netherlands under the Euthanasia Act. N Engl J Med. 2007;356:1957–1965
   • View In Article
   • CrossRef
8. Rietjens J, van Delden J, Onwuteaka-Philipsen B, et al. Continuous deep sedation for patients nearing death in The Netherlands: descriptive study. BMJ. 2008;336(7648):810–813
   • View In Article
   • CrossRef
9. Bilsen J, Cohen J, Chambaere K, et al. Medical end-of-life practices under euthanasia law in Belgium. N Engl J Med. 2009;361:1119–1121
   • View In Article
   • CrossRef
10. Bilsen J, van der Stichele R, Broeckaert B, Mortier F, Deliens L. Changes in medical end-of-life practices during the legalization process of euthanasia in Belgium. Soc Sci Med. 2007;65:803–808
    • View In Article
    • CrossRef
11. Seale C. National survey of end-of-life decisions made by UK medical practitioners. Palliat Med. 2006;20:3–10
    • View In Article
    • MEDLINE
    • CrossRef
12. Morita T, Miyashita M, Kimura R, Adachi I, Shima Y. Emotional burden of nurses in palliative sedation therapy. Palliat Med. 2004;18:550–557
    • View In Article
    • MEDLINE
    • CrossRef
13. Seymour JE, Janssens R, Broeckaert B. Relieving suffering at the end of life: practitioners' perspectives on palliative sedation from three European countries. Soc Sci Med. 2007;64:1679–1691
    • View In Article
    • MEDLINE
    • CrossRef
14. Rietjens JAC, Buiting HM, Pasman HRW, et al. Deciding about continuous deep sedation: physicians' perspectives. Palliat Med. 2009;23:410–417
    • View In Article
    • CrossRef
15. Billings JA, Block SD. Slow euthanasia. J Palliat Care. 1996;12:21–30
    • View In Article
    • MEDLINE
16. Quill TE, Lo B, Brock DW. Palliative options of last resort: a comparison of voluntarily stopping eating and drinking, terminal sedation, physician-assisted suicide, and voluntary active euthanasia. JAMA. 1997;278:2099–2104
    • View In Article
    • MEDLINE
17. Hasselaar JGJ, Stans CA, Verhagen SC, et al. Changed patterns in Dutch palliative sedation practices after the introduction of a national guideline. Arch Intern Med. 2009;169:430–437
    • View In Article
    • CrossRef
18. Rietjens JA, van der Heide A, Vrakking AM, et al. Physician reports of terminal sedation without hydration or nutrition for patients nearing death in The Netherlands. Ann Intern Med. 2004;141:178–185
    • View In Article
19. Rietjens JA, van Delden JJ, van der Heide A, et al. Terminal sedation and euthanasia: a comparison of clinical practices. Arch Intern Med. 2006;166:749–753
    • View In Article
    • MEDLINE
    • CrossRef
20. Kaldjian LC, Jekel JF, Bernene JL, et al. Internists' attitudes towards terminal sedation in end of life care. J Med Ethics. 2004;30:499–503
    • View In Article
    • MEDLINE
    • CrossRef
21. de Graeff A, Dean M. Palliative sedation therapy in the last weeks of life: a literature review and recommendations for standards. J Palliat Med. 2007;10:67–85
    • View In Article
    • MEDLINE
    • CrossRef
22. Reuzel RP, Hasselaar GJ, Vissers KC, et al. Inappropriateness of using opioids for end-stage palliative sedation: a Dutch study. Palliat Med. 2008;22:641–646
    • View In Article
    • CrossRef
23. Hasselaar JG, Reuzel RP, Verhagen SC, et al. Improving prescription in palliative sedation: compliance with Dutch guidelines. Arch Intern Med. 2007;167:1166–1171
    • View In Article
    • MEDLINE
    • CrossRef
24. Centeno C, Clark D, Rocafort J, et al. Facts and indicators on palliative care development in 52 countries of the WHO European region: results of an EAPC Task Force. Palliat Med. 2007;21:463–471
    • View In Article
    • CrossRef
25. Clery E, McLean S, Phillips M. Quickening death: the euthanasia debate. In:  Park A,  Curtice J,  Thomson K,  Phillips M,  Johnson M editor. British social attitudes: Perspectives on a changing society. 23rd Annual Report. London, UK: Sage; 2007;p. 35–54
    • View In Article
26. Seale C. Legalisation of euthanasia or physician-assisted suicide: survey of doctors' attitudes. Palliat Med. 2009;23:205–212
    • View In Article
    • CrossRef
27. Seale C. End-of-life decisions in the UK involving medical practitioners. Palliat Med. 2009;23:198–204
    • View In Article
    • CrossRef
28. Rurup ML, Borgsteede SD, ven der Heide A, van der Maas PJ, Onwuteaka-Philipsen BD. Trends in the use of opioids at the end of life and expected effects on hastening death. J Pain Symptom Manage. 2009;37:144–152
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (174 KB)
    • CrossRef
29. Fischer S, Miccinesi G, Hornung R, et al. Responders and non-responders in a study on medical end-of life decisions in Denmark, The Netherlands, Sweden and Switzerland. Soz Praventivmed. 2006;51:24–33
    • View In Article
    • MEDLINE
    • CrossRef
30. Treloar AJ. Dutch research reflects problems with the Liverpool care pathway. BMJ. 2008;336:905
    • View In Article
    • CrossRef
31. Society of Critical Care Medicine and American Society of Health-System Pharmacists. Clinical practice guidelines for the sustained use of sedatives and analgesics in the critically ill adult. Am J Health Syst Pharm. 2002;59:150–178
    • View In Article
    • MEDLINE