Forum.
Literature Review
Without exception, all papers dealing with palliative sedation specified the existence of intractable/refractory physical and/or mental symptoms as a prerequisite for palliative sedation. Obviously, this implies that optimal palliative treatment has been provided. However, what is considered ‘optimal’ is highly dependent on personal judgement, experience and expertise (Morita 2002-1). Very few of the papers dealing with sedation provided a framework within which to assess as optimal the treatment already given.
a) Type of intractable/refractory symptoms
We found 19 case series that described the symptoms necessitating sedation as their treatment (Table 1). Some of these papers reported the effectiveness of sedation in alleviating these symptoms. Patients receiving sedation for other purposes (e.g. invasive procedures, prophylaxis against seizures) were excluded from our analysis. Patients treated with haloperidol solely for delirium/terminal restlessness or with opioids solely for pain or dyspnoea were also excluded since the primary goal of these treatments was specific symptom control that was often achieved without the side effect of sedation.
In these studies, delirium and/or terminal restlessness were the most frequent reasons for palliative sedation (57%), followed by dyspnoea (23%), pain (16%) and nausea/vomiting (4%). Other symptoms (malaise, insomnia, anxiety, distress, mental anguish, existential distress) are rarely mentioned. A problem in interpreting these studies is that the terminology is not consistent; some symptoms or signs, such as insomnia, agitation, restlessness, (mental) anguish and distress might be related to delirium, to psychological distress or to other causes.
Two literature reviews found the following symptom frequencies for the use of sedation:
Cowan 2001 Sykes 2003-1
Delirium/agitation/restlessness/confusion 40% 65%
Breathlessness/respiratory distress 12% 26%
Pain 21% 14%
Chater surveyed 61 selected palliative care experts from eight countries to assess the reasons for the use of sedation (Chater 1998). These were reported as follows: pain (20%), anguish (14%), respiratory distress (12%), agitation/delirium/confusion/hallucinations (12%), restlessness (10%), fear/panic/anxiety/terror (10%) and emotional/psychological/spiritual distress (10%). This survey again illustrates the variability of the terminology. This inexact terminology may be one of the reasons for the variability of the symptom frequencies for the use of sedation found in our own literature review, as well as the reviews by Cowan and Sykes, and by Chater.
Reference |
No. |
Sedative |
Delirium/ |
Dys- |
Pain1 |
Nau- |
Other |
Effectiveness |
|
Bottomley ‘90 |
23 |
Midazolam |
23 |
|
|
|
|
22/23 good relief |
|
Ventafridda ‘90 |
63 |
Diazepam, chlorpromazine |
11 |
33 |
31 |
5 |
|
NR2 |
|
Burke ‘91 |
86 |
Midazolam |
86 |
|
|
|
|
85/86 improved |
|
FaiNRinger ‘91 |
16 |
Midazolam, lorazepam, |
10 |
|
6 |
|
|
NR |
|
Greene ‘91 |
17 |
Barbiturates |
2 |
|
15 |
4 |
|
17/17 improved |
|
McNamara ‘91 |
61 |
Midazolam |
61 |
|
|
|
|
NR |
|
Stiefel ‘92 |
10 |
Midazolam (9) |
10 |
|
|
|
|
10/10 control |
|
McIver ‘94 |
20 |
Chlorpromazine |
10 |
10 |
|
|
|
18 complete and |
|
Morita ‘96 |
69 |
Midazolam, diazepam, |
23 |
34 |
27 |
7 |
Malaise: 26 |
NR |
|
Stone ‘97 |
30 |
Midazolam (24) |
18 |
6 |
6 |
|
Mental anguish: 8 |
NR |
|
Fainsinger ’98-1 |
23 |
Midazolam cont. (14) |
21 |
3 |
|
|
|
NR |
|
Fainsinger ’98-2 |
2 |
Midazolam |
2 |
|
|
|
|
2 complete relief |
|
Morita ‘99 |
71 |
Midazolam, diazepam, |
30 |
29 |
9 |
1 |
Myoclonus: 1 |
NR |
|
Chiu ‘01 |
70 |
Midazolam |
40 |
7 |
16 |
|
Insomnia: 5 |
NR |
|
Morita ‘00 |
20 |
Midazolam |
8 |
1 |
10 |
|
Psychol. distress: 1 |
NR |
|
Fainsinger ’00-1 |
10 |
Midazolam (5), |
9 |
1 |
|
|
|
NR |
|
Fainsinger ’00-23 |
97 |
Midazolam (80%) |
59 |
25 |
7 |
9 |
Existential distress 7 |
NR |
|
Soares ‘02 |
3 |
Dexmedetomidine |
2 |
|
2 |
|
Anxiety/distress : 1 |
2/3 good effect |
Muller-Busch ‘03 |
80 |
Midazolam (mostly) |
11 |
28 |
2 |
6(?) |
Anxiety/distress: 32 |
NR |
Total |
771 |
Midazolam (majority), |
436 |
177 (23%)1 |
131 |
32 |
|
|
Table 1. Sedation and symptom control: case series in the literature.
1 Patients may have received sedation for more than one symptom
2 NR = Not reported
3 Four different centres in three countries (South Africa, Spain and Israel)
The controversial issue of psychological and existential distress as an indication for sedation is seldom mentioned (Shaiova 1998, Morita 2000, Cherny 2003). In a recent publication anxiety/psychological distress was an indication for sedation in 32/80 cases (40%); in this study over a period of 7 years there was a (non-significant) trend for an increased use of sedation for this indication between 1995 and 2002 (Muller-Busch 2003). This issue has implications for the definition of refractory physical symptoms since in the concept of ‘total’ pain (or any other physical symptom) psychological or existential distress may amplify a physical symptom (Bruera 1994, Robinson 1995, Zaza 2002). A recent study in 81 centres in Japan found varying prevalences of continuous deep sedation for psycho-existential suffering (0%, 0,5-5% and >10% in 64%, 32% and 4% of institutions respectively)(Morita 2004). Most palliative care specialists would agree that before a symptom is considered to be refractory, the patient must have received skilled multidimensional management directed at the physical, psychological and existential dimensions of the symptom.
There are differences between centres with regard to the reported frequency of and indications for sedation. In seven relatively large single-centre studies and two multicentre studies the percentage of patients requiring sedation for refractory symptoms varied from <10% to 52% (Table 2). There were considerable differences between these centres with regard to the percentages of patients sedated for delirium (varying from 14 to 91%), dyspnoea (0-52%) and pain (3-49%). These differences may be due to differences in terminology, selection of patients (in particular with regard to diagnosis and life expectancy), to differences in multidimensional management and/or to regional, international or cultural differences.
|
Reference |
Percentage of patients |
|
Ventafridda 1990 |
52 |
|
Fainsinger 1991 |
16 |
|
Morita 1996 |
43 |
|
Stone 1997 |
26 |
|
Fainsinger 1998-1 |
30 |
|
Fainsinger 2000-1 |
15 |
|
Sykes 2003-2 |
48 |
|
Muller-Busch 2003 |
15 |
|
Morita 2004 (81 centers in |
<10->50% |
Table 2. Percentage of patients requiring sedation in different centres.
b) Definitions of optimal symptom management
Only one of the papers reviewed provided a framework of optimal palliative care standards required before considering palliative sedation. In this report a refractory terminal cancer symptom was defined as one incapable of adequate relief, that caused severe suffering, and was unlikely to lessen within a short time (Chiu 2001), but there was no definition of what is adequate relief, severe suffering, or a short time. A review paper referred to symptoms not having responded to conventional management (Cowan 2001) without defining the meaning of conventional management.
The Edmonton group developed a consistent assessment system and treatment standards of important symptoms and syndromes, such as pain, delirium and vomiting. They reported quite low percentages (4% - 10%) of patients requiring palliative sedation (Fainsinger 1991). This paper suggests that optimal assessment and treatment reduces the need for sedation
c) Level and type of sedation
In most studies the level of sedation was titrated against the distress response (proportional sedation) and unconsciousness was not an invariable outcome (Sykes and Thorns 2003-1). As relief of distress and discomfort is the primary aim of palliative sedation, this is appropriate. Intermittent sedation should also be considered (Morita 1996). A survey in the Japanese general population indicated a preference for intermittent instead of continuous sedation (Morita 2002-2).
Comments