Introduction
At the end of life, certain symptoms commonly appear as the disease progresses, and although different treatment options are available to control these symptoms, in some cases they lose effectiveness (in terms of benefit achieved or duration) or produce results that are more negative than positive (de Graeff and Dean Reference de Graeff and Dean2007). When the symptoms are very severe, becoming intolerable for the patient, and can be considered refractory, palliative sedation (PS) may be considered as a treatment option (Arantzamendi et al. Reference Arantzamendi, Belar and Payne2021).
PS in the context of palliative medicine is the monitored use of medications intended to induce a state of decreased or absent awareness (unconsciousness) in order to relieve the burden of otherwise intractable suffering in a manner that is ethically acceptable to the patient, the family, and health-care providers (Cherny and Radbruch Reference Cherny and Radbruch2009). PS may be continuous or intermittent and its depth is graduated to provide the minimum level of sedation that achieves symptomatic relief (guia_sedaccion_paliativa.pdf 2021). The intent of PS, therefore, is to relieve the burden of otherwise intolerable suffering for terminally ill patients (Cherny and Portenoy Reference Cherny and Portenoy1994).
This measure is indicated for adult and pediatric palliative care patients with advanced, incurable illness (Palliative sedation – UpToDate 2023), under the condition that its introduction does not directly provoke a detrimental effect on survival (Maeda et al. Reference Maeda, Morita and Yamaguchi2016; Maltoni et al. Reference Maltoni, Scarpi and Rosati2012).
Several studies of this question, referring to palliative care for adults, have been published, mainly focusing on the development of clinical guidelines (Cherny Reference Cherny2014; Cherny and Radbruch Reference Cherny and Radbruch2009; guia_sedaccion_paliativa.pdf 2021), and usually based on expert opinion and retrospective reviews. In contrast, hardly any case series have been published with respect to the pediatric population (Chen et al. Reference Chen, Jiang and Peng2022; de Noriega et al. Reference de Noriega, Rigal Andrés and Martino Alba2021; Kiman et al. Reference Kiman, Wuiloud and Requena2011), although this lack of data does not mean that the practice is uncommon (Korzeniewska-Eksterowicz et al. Reference Korzeniewska-Eksterowicz, Przysło and Fendler2014). More knowledge is needed of the habitual approaches of health-care professionals regarding pediatric PS, and of the clinical and contextual circumstances in which it is carried out, in order to identify areas for improvement, and to provide indications for future research, thus strengthening the evidence supporting protocols and guidelines for clinical practice.
The aim of the present study is to describe the incidence of PS in patients treated by pediatric palliative care teams (PPCT), to determine the indications considered, the drugs used and, in a more innovative way, the relationship between PS, the place of death, and the characteristics of the care teams.
Methods
This ambispective, analytical, multicenter study was conducted from 1 January to 31 December 2019. Data collection was performed by 14 PPCT at 219 public hospitals in Spain with pediatric services (Consulta Interactiva del SNS 2021). Of this total, only 26 hospitals had a PPCT when the study period began.
All patients treated by the PPCT were followed up during the study period, and data were also compiled on those who died. All data collection was performed through the analysis of medical records.
The following study variables were recorded for each patient: sex, age, underlying disease, medication, doses provided during the week prior to death, reason for death, place of death, expression of preference by the family regarding the place of death, and the performance of and reason for sedation. In addition, the year of creation of each PPCT was recorded, to determine its length of experience.
The patient’s underlying disease was described according to the classification system for pediatric complex chronic diseases (v.2), developed by Feudtner et al. (Reference Feudtner, Feinstein and Zhong2014).
Analysis
Descriptive statistics were obtained by exploratory data analysis, with measures of central tendency, dispersion, and frequency. The normality of the distributions was determined by the Kolmogorov–Smirnov test, together with an evaluation of asymmetry and kurtosis. Variables with a normal distribution are described by the mean and standard deviation; otherwise, by the median and interquartile range (IQR). Bivariate analysis was performed using the chi-square and Fisher’s exact tests where necessary, as well as analysis of variance for the comparative analysis by Feudtner subgroups, with measures of central robustness in the event of non-homoscedasticity (verified with the Levene test) using the Brown–Forsythe test, and determining the difference in means for independent groups using the Mann–Whitney U test. Subsequently, a multivariate logistic regression model was constructed with the presence of sedation as the dependent variable and the introduction of predictors from the variables that presented a significant association in the bivariate analysis, or which had significant clinical plausibility. All analyses were performed using SSPS 26 (IBM Corp. Released 2019. IBM SPSS Statistics for Windows, Version 26.0. Armonk, NY: IBM Corp) and JAMOVI (The jamovi project (2021). jamovi (Version 1.6) [Computer Software]. Retrieved from https://www.jamovi.org 2021) statistical software.
Results
The study sample consisted of 164 patients, of whom 84 (51.2%) were female and 80 (48.8%), male. Table 1 describes the underlying diseases identified, most of which correspond to oncological, neurological, and neuromuscular processes. The median age at death was 6.9 years (IQR 11.2). The reasons for death were the progression of the underlying disease (n = 101 patients; 61.6%), comorbidity (n = 51 patients, 31.1%), or unexpected cause (n = 12 patients, 7.3%).
Baseline disease
In total, 52 patients (31.7%) received sedation in the last 7 days of life, while 83 (50.6%) received PS in the last 24 hours. Among the refractory symptoms, we found terminal suffering in 40 patients (48.2%), dyspnea (n = 9, 10.8%), pain (n = 8, 9.6%), and convulsive state (n = 7, 8.4%). The most frequently used drug for PS was midazolam in continuous infusion (n = 74 patients, 89.2%), which was supplied intravenously to 45 patients (54.2%), with a median dose of 0.1 mg/kg/h (IQR 0.2).
A total of 95 patients (57.9%) died in the hospital and 67 (40.9%), at home. Sedation was more likely to be given in the last 24 hours if the patient died in the hospital (62.9% vs. 33.3%, χ2: 13.7, p < 0.01).
One of the study variables considered was whether the PPCT had previously discussed with the family their preference regarding the place of death. In this regard, 125 families (76.2%) had expressed a preference, with 70 (56%) stating the home should be the place of death. The children whose parents had stated this preference were less likely to be sedated during the last 24 hours of life (45.2% vs. 69.2%, χ2: 6.9, p = 0.009, OR: 0.37 95% CI: 0.17 to 0.79). When no preference was expressed, sedation was supplied up to 3 times more frequently when the child was in a situation of terminal suffering (72.7% vs. 46.2%; OR: 3.1 95% CI: 1.05 to 9.21; χ2: 4.4, p = 0.036).
Focusing on the teams’ experience, those with less than 5 years’ practice sedated 66.7% of their patients during the last 24 hours of life, compared to 45.5% of the units with greater experience (χ2: 5.6, p = 0.018, OR: 0.21, 95% CI: 0.04 to 0.38). The indications for sedation were similar in both groups. Moreover, higher infusion doses were given by the teams with less experience, at both 1 week before and at 24 hours before death (0.31 vs. 0.16 mg/kg/h, z = 2.60; p = 0.009 and 0.27 vs. 0.15 mg/kg/h, z = 2.13, p = 0.033, respectively).
Multivariate analysis, adjusted for age and sex, showed that the likelihood of receiving sedation in the last 24 hours of life was greater in patients who were treated by a PPCT with less than 5 years’ experience (OR: 2.84, 95% CI: 1.09 to 7.37) and that this value decreased when the family’s preference regarding the place of death had been discussed and the child died at home (OR: 2.43, 95% CI: 1.14 to 5.18). In this respect, there was no influence by age or sex (Table 2).
Factors associated with the use of sedation during the 24 hours before death
McFadden’s R 2: 0.07, variance inflation factor range: 1–1.03, tolerance range: 0.97–0.99.
Discussion
This study was conducted to determine the incidence of PS in patients treated by a PPCT, and the association between this value and the clinical care and contextual circumstances in which the sedation was given. To our knowledge, this is the first multicenter investigation of its type to be performed in Spain.
During the study period, slightly more than 30% of the terminal pediatric patients treated by a PPCT received sedation in the last 7 days of life. This value rose to 50% in the last 24 hours. To our knowledge, no previous study has examined such a large sample of pediatric patients treated with PS. In a broader context, the above rates of incidence are lower (Libro-de-ponencias-y-comunicaciones-III-Congreso-PedPal.pdf 2019) or similar to those reported in some cases (Chen et al. Reference Chen, Jiang and Peng2022; Korzeniewska-Eksterowicz et al. Reference Korzeniewska-Eksterowicz, Przysło and Fendler2014; Vallero et al. Reference Vallero, Lijoi and Bertin2014) but higher than has been published in other, recent, publications (Cuviello et al. Reference Cuviello, Pasli and Bhatia2023; de Noriega et al. Reference de Noriega, Rigal Andrés and Martino Alba2021; Maeda et al. Reference Maeda, Kato and Umeda2020). These discrepancies might be explained, at least in part, by the existence of differences between the respective study cohorts and by the lack of a unified definition.
The question of definitions remains controversial, and most have been developed from experience with adult patients. Nevertheless, despite this variability, a common concept in many of the definitions of PS is “the use of sedative medications to alleviate the intolerable suffering of refractory symptoms through a reduction in the patient’s consciousness” (Cherny Reference Cherny2014; Cherny and Radbruch Reference Cherny and Radbruch2009). Given this premise, the presence of refractory symptoms at the end of life is an essential requirement for prescribing PS. Refractory symptoms are usually considered to be those which are physical; in our sample, the most common of these symptoms were dyspnea, pain, and convulsive status. The first 2 of these are corroborated in the literature, where dyspnea, pain, and delirium are most frequently mentioned (Arantzamendi et al. Reference Arantzamendi, Belar and Payne2021; de Noriega et al. Reference de Noriega, Rigal Andrés and Martino Alba2021; Korzeniewska-Eksterowicz et al. Reference Korzeniewska-Eksterowicz, Przysło and Fendler2014; Maltoni et al. Reference Maltoni, Scarpi and Rosati2012). Despite receiving optimal palliative care, some children still experience severe uncontrolled symptoms in their last weeks of life. In these circumstances, no measure to alleviate suffering should be spared and PS must be considered.
However, reducing refractory symptoms to those physically apparent would be too simplistic a view of human suffering (Goldman Reference Goldman, Hain and Liben2012; Rodrigues et al. Reference Rodrigues, Crokaert and Gastmans2018). There is a situation that precedes death and in which there is intense physical deterioration, extreme weakness, high frequency of cognitive disorders and consciousness, difficulty relating, and eating with a prognosis of life limited to hours or days that sometimes causes suffering. The Spanish Society for Palliative Care (SECPAL) coined the expression terminal suffering palliative sedation to refer to the sedation employed when the patient has only hours or a few days to live, and is experiencing intense suffering (guia_sedaccion_paliativa.pdf, 2021). This situation, affecting 48.2% of the patients in our sample, was the most frequent reason for initiating PS. As specific data under this description are compiled only in Spain, the latter result cannot be compared with other research findings.
In this context, there may sometimes be difficulty in treating suffering at the end of life when it is secondary to a symptom, but you do not have enough time. It will also be difficult to distinguish between a refractory symptom and one that is merely difficult to treat, the latter type being one requiring intensive therapeutic intervention: pharmacological, instrumental, and/or psychological (“salud_5af19569afca1_02_definiciones_cpali_sedacion.pdf” 2005). Our analysis shows that when the PPCT in question had sufficient expertise and experience to treat difficult symptoms and did not consider them refractory, rates of PS were lower and, when necessary, they were supplied at lower doses than those given by teams with less experience. Therefore, before classifying a symptom as refractory, professionals should be advised by experienced palliative care specialists (de Graeff and Dean Reference de Graeff and Dean2007). Furthermore, if the physician is unable to relieve a distressing symptom, he/she may feel pressured to use PS, or even to provide disproportionate sedation. It has been suggested that physician fatigue and burnout may be associated with an increased use of PS (Coyle et al. Reference Coyle, Adelhardt and Foley1990; “Practices and attitudes of Japanese oncologists and palliative care physicians concerning terminal sedation: a nationwide survey – PubMed” 2002). In this regard, the development of explicit institutional PS protocols could help reduce uncertainty and variability in this practice (Henderson et al. Reference Henderson, FitzGerald and Hoehn2017).
In our sample, midazolam was the drug most frequently used for PS, as recommended elsewhere for use with pediatric and adult populations (Cherny Reference Cherny2014; Cherny and Radbruch Reference Cherny and Radbruch2009) and as reported in most previous studies of this question (de Graeff and Dean Reference de Graeff and Dean2007; de Noriega et al. Reference de Noriega, Rigal Andrés and Martino Alba2021; Maeda et al. Reference Maeda, Kato and Umeda2020; Vallero et al. Reference Vallero, Lijoi and Bertin2014), although recent studies have highlighted that pain medicine specialists opted for opioids as the first-line drug for the initiation of PS therapy, followed by benzodiazepines (Cuviello et al. Reference Cuviello, Johnson and Morgan2022).
Although previous research has shown that families prefer the death to take place in the home (Goldman et al. Reference Goldman, Beardsmore and Hunt1990; Woodman et al. Reference Woodman, Mawdsley and Hauser-Cram2015), which is consistent with our findings, in practice more deaths occur in hospital settings. It is here where PS is more likely to be applied in the last 24 hours of life (62.9% vs. 33.3% at home), a proportion similar to that reported for the treatment of adults (Pousset et al. Reference Pousset, Bilsen and Cohen2011). Our study results also show that among the families who expressed a preference in this respect, there was a lower probability of PS. Some authors suggest that PS, in pediatric care, is most appropriately supplied in the home. It is also less costly than the standard care model for hospitalized patients (Korzeniewska-Eksterowicz et al. Reference Korzeniewska-Eksterowicz, Przysło and Fendler2014). However, we must be cautious in this regard; on the one hand, PS in the home might be considered less invasive, but on the other hand, patients with more difficult-to-control symptoms should be hospitalized to ensure optimum monitoring and treatment.
Our study results also show that children whose parents had expressed their preference regarding the place of death were less likely to receive PS in the last 24 hours of life, and that in families where no such preference had been expressed, sedation during terminal suffering was up to 3 times more frequent. We believe these results highlight the importance of advanced care planning, which if carried out by the PPCT can reduce the need for PS (Kiman et al. Reference Kiman, Wuiloud and Requena2011). In this respect, an important consideration arises: any discussion of preferences for the place of death are focused on an impending death, a process in which the patient (when possible and appropriate) and his/her family should be actively involved. In the case of PS, a significant aspect of advanced care planning is the need to clarify the intentions underlying any intervention made (fundamentally, the treatment team should seek to maximize the patient’s well-being and manage the symptoms observed, but not accelerate death), to obtain verbal consent, to provide sufficient time (if circumstances allow it) for the patient and/or family to assimilate the situation, and to provide support to all family members involved.
Among other limitations of this study, it was not possible to record the duration of the PS, nor the level of sedation provided, nor any side effects that may have been caused.
Conclusions
Some children who are terminally ill die peacefully and painlessly, without the need for PS; others, however, present refractory symptoms, for which PS can be considered as an option. In the latter case, the decision to offer sedation to alleviate intolerable suffering during the last weeks of life does not present any ethical problem. Quite the opposite: deciding upon and applying this treatment option represents a continuation of good clinical practice, based on a careful evaluation of the patient and concern for his/her ultimate well-being.
The application of PS requires the professionals involved to have sufficient experience and appropriate communication skills.
Funding
Funding for open access charge: Universidad de Málaga.
Competing interests
None.
Ethical approval
The study was approved by the Research Ethics Committee of Málaga Regional University Hospital (code 1054-N-21).
Open access
