Non-systematic review of the literature on palliative sedation. The aim was to identify commonalities in the current literature that could be turned to items used to evaluate the definition.

Our search yielded a total of 11 national and international guidelines and one chapter in a reference book. Since the paediatric literature on the subject mainly consisted of case series,4,5,7–10,14 and due to the lack of specific guidelines beyond some references to paediatric palliative care,6,15,16 we selected the general guidelines for adult palliative care as the framework for the definition, adapting those aspects affected by the particularities of the paediatric population (drug dosage, informed consent…). Specifically, we used as reference the guidelines of the European Association for Palliative Care (EAPC)2 and of the European Society of Medical Oncology (ESMO)1 at the international level and the guidelines of the Sociedad Española de Cuidados Paliativos (Spanish Society of Palliative Care, SECPAL) and Council General of Official Boards of Physicians of Spain17 and of the Spanish Ministry of Health18 at the national level.

We identified 5 common points in these documents: 1) Presence of intractable suffering caused by refractory symptoms. 2) Presence of “terminal” illness. 3) Need to have complex discussions with patients and informed consent process. 4) Need to consult with experts in symptom relief. 5) Delivery of sedative drug(s) to reduce the level of consciousness and prevent suffering, monitoring the response of the patient. We created the definition of palliative sedation taking these points into account. In the subsequent analysis, we excluded items 2 and 4, as all patients managed at the PPCU-CM had life-limiting or life-threatening conditions and this unit is the referral paediatric palliative care unit in the region, the scope of which includes the management of refractory symptoms.

Thus, we addressed mainly 3 items in our analysis: 1. Indication 2. Consent 3. Delivery (Table 1). For item 1a, we used the guidelines for end of life care for infants, children and young people with life-limiting conditions of the National Institute for Health and Care Excellence (NICE) of the United Kingdom.19 We selected this guidance based on methodological considerations, although other sources may be consulted to obtain more detailed information on symptom management.20,21 For item 2a, we adapted the pertinent aspects of the law on patient autonomy of Spain,22 which considers any individual aged more than 16 years free of cognitive or legal limitations, legally emancipated or deemed as such by the physician in charge competent for the purpose of decision-making. Table 2 summarises the information on the drugs and dosage corresponding to items 3a and 3b.6,23

We defined adherence to an item (1, 2 or 3) as fulfilment of all of its subitems (a, b and c).

We conducted a retrospective review of the health records of deceased patients managed at the PPCU-CM in a 5-year period (January 2013-December 2017) to identify those who received palliative sedation and analyse the instances of palliative sedation delivered at the PPCU-CM.

We obtained data on deceased patients from the internal records of the PPCU-CM. We excluded patients whose records did not include a death summary (a report that summarises the care received by the patient at the PPCU-CM). We identified patients that received palliative selection applying the following criteria:

• •

  Mention of the term “palliative sedation”

• •

  Intentional lowering of awareness used for symptom management.

• •

  Use of a sedative drug to manage a symptom described as “refractory” or “difficult to control”.

• •

  Delivery of a sedative drug through intravenous infusion at an appropriate dose.

The definition of palliative sedation excluded the use of a sedative drug primarily for symptom management (benzodiazepines for management of anxiety, epileptic seizures or dyspnoea) in symptoms not considered “refractory” and in the absence of a specific mention of “palliative sedation” in the records.

To assess the need of palliative sedation in our patients, we also classified those that did not receive palliative sedation into the following groups:

• a)

  Patients who did not need sedation: the health records showed adequate symptom control.

• b)

  Patients in whom the need of palliative sedation could not be completely ruled out: adequate symptom control was not clearly documented in the records.

We collected data on the following variables for every patient: sex, age at time of death, classification as oncological and non-oncological patient, primary diagnosis, duration of management in the PPCU-CM and place of death. We compared the characteristics of the patients that received palliative sedation and the patients that did not.

In patients that received sedation at the PPCU-CM, we also collected data for the items used in the definition and the following variables: symptom(s) based on which sedation was considered indicated, continuous or intermittent sedation, and mild or deep sedation. If the intervention was monitored, we documented whether monitoring focused on the assessment of sedation itself and/or of the symptoms, and whether scales were used for the purpose. Lastly, we collected data on the drugs used, the maximum dose, adverse events, the reasons to discontinue sedation and whether sedation achieved comfort.

We used the software STATA® version 16.1. We compared variables in patients that received and did not receive sedation by means of parametric tests ( χ2 and Student t test to compare means) or nonparametric tests (Wilcoxon rank test and Mann-Whitney U test). In the descriptive analysis, we summarised the data using the median and interquartile range. Statistical significance was defined as a P value of less than 0.05.

In the period under study, 202 patients managed by the PPCU-CM died (Fig. 1), of whom we excluded 38 due to the absence of a death summary; 20 patients (12.2%) received palliative sedation, in 17 cases through the PPCU-CM and in 3 in other units (1 through the department of haematology, 1 through the department of oncology and 1 through the department of paediatrics). In the group of 144 patients that did not receive palliative sedation, the records indicated adequate symptom control in 93.1%, while the need for sedation could not be ruled out in 6.9%.

Figure 1.

Distribution of patients under study.

PS, palliative sedation.

(0.14MB).

Table 3 compares the characteristics of patients that did and did not receive sedation. Of the 20 patients that received it, 70% were male, and the median age in this group was 11.9 years (IQR, 8.0-16.8); 12 patients (60%) had cancer (8 had extracranial solid tumours, 2 tumours of the central nervous system and 2 blood tumours), 7 (35%) had neurologic diseases (6 had static encephalopathies and 1 spinal muscular atrophy) and 1 had a polymalformative syndrome in which the major feature was bone marrow aplasia. The median duration of follow-up was 2.7 meses (IQR, 0.4-14.5). Patients who received sedation were significantly older and died at home more frequently compared to those who did not receive it. The frequency of palliative sedation was not higher in patients with cancer compared to patients with other diagnoses.

Of the 17 patients that received sedation through the PPCU-CM, 9 had more than 1 refractory symptom at the time sedation was initiated (Table 4). The most frequent intractable symptoms were pain (64.7%) and dyspnoea (58.8%). Four patients had agitation, in every case associated with some other symptom (pain in 3, dyspnoea in 1). One patient had convulsions and dyspnoea. Sedation was administered at hospital in 52.9% of the patients and at home in the remaining 47.1%.

Of all patients that received sedation, 76.4% received continuous sedation and 11.8% intermittent sedation (missing data in 2 cases). The depth of sedation was recorded in 10 patients: deep in 8 and mild in 2. Midazolam was the most frequently used drug (88.2%), administered alone (64.7%) or combined with haloperidol (17.6%) or propofol (5.9%). One patient received propofol as monotherapy. The maximum dose for each drug was 0.35 mg/kg/h for midazolam, 0.8 mg/kg/h for haloperidol and 7.3 mg/kg/h for propofol.

In 16 cases, sedation was discontinued due to the death of the patient; the remaining patient developed septic shock that resulted in coma, a state in which sedation was no longer needed. Only one patient experienced an adverse event (mild bronchorrhoea that did not require treatment). The median duration of sedation was 1.5 days (IQR, 1.5-3.5).

Fig. 2 presents the main results in this section:

• 1

  Indication: all patients had previously received appropriate treatment (1a = 100%). In 11 patients, there was an identified refractory symptom. Adherence to item 1: 64.7%.

• 2

  Informed consent: 16 patients were not considered competent to consent due to age (< 16 years) or cognitive impairment, placing the onus of consent on the legal guardians. Patient 17 (age 14 years) was deemed competent by the physician in charge and participated in the decision-making process. Patient 14 was older than 16 years and did not exhibit cognitive impairment, but there was no documentation of the patient’s participation in the consent process. For this reason, item 2a was only fulfilled in 94.1% of the patients. The informed consent process (item 2b) was only documented in 17.6% patients. Adherence to item 2: 17.6%.

• 3

  Delivery of sedation: In 16 patients, the selected drug was appropriate, and in the rest it had not been documented (item 3a = 94.1%). In 11 patients, the initial dose was adequate, in 2 lower than recommended and in 4 it was not documented (item 3b = 11/17). Monitoring of symptom control (17/17) or sedation (5/17; item 3c = 17/17) was performed in every patient, but scales were not used in any case, and it was always based on subjective measures. Adherence to item 3: 64.7%.

Figure 2.

Fulfilment of the definition. Items and subitems completed appropriately are marked with an X. The overall assessment row shows the number of items that were completely fulfilled in each patient.

(0.29MB).

In the overall assessment of the standard, we found that all three items had been fulfilled in 17.6% of the patients, two in 23.5%, one in 47.1% and none in 11.8%.