Introduction
The estimated global total number of children in need of palliative care in 2017 was almost 4 million. Children and adolescents aged 0–19 years make up 7% of the total global palliative care demand. The health conditions that generate the greatest need for palliative care among children are Human immunodeficiency virus/Acquired Immune Deficiency Syndrome, premature birth and birth trauma, congenital anomalies, injuries, malignant neoplasms, and other rare diseases (Connor Reference Connor2020). The treatment of children with chronic disease is complex and extends over a long period of time (Levine et al. Reference Levine, Cuviello and Nelson2021) and, during this period, the child’s well-being may fluctuate depending on their state of health, as well as the support received and resources available (Donnelly et al. Reference Donnelly, Downing and Cloen2018; Presedo and Wenk Reference Presedo and Wenk2007).
The literature generally shows that the way children and adolescents experience severe illness is related to adaptive psychological reactions and not to specific clinical or psychopathological disorders (Hernández et al. Reference Hernández, López and Durá2009). This is because patients gradually become more familiar with their situation and, as a result, their maladaptive reactions and symptoms decrease. Suffering is related to emotions, such as fear, anger, sadness, helplessness, and despair (Krikorian and Limonero Reference Krikorian and Limonero2012). Most young people with life-threatening illnesses often experience symptoms of anxiety and depression, which are especially common in adolescents (Weaver et al. Reference Weaver, Heinze and Bell2016), as well as nervousness or irritability (Collins et al. Reference Collins, Devine and Dick2002), while the most prevalent mental health problems are adaptive and behavioral disorders, and post-traumatic stress, together with depression and anxiety (DeJong and Fombonne Reference DeJong and Fombonne2006; Krikorian and Limonero Reference Krikorian and Limonero2012). It is particularly difficult for children to differentiate psychological symptoms from their actual physical symptoms (Woodgate et al. Reference Woodgate, Degner and Yanofsky2003).
In clinical practice, treatment decisions are often, at least partly, based upon quality-of-life considerations. Therefore, it is important to be aware of differences in the perception of emotional well-being expressed by children, physicians, and parents (Janse et al. Reference Janse, Sinnema and Uiterwaal2008). In the pediatric setting, to assess the well-being of children and adolescents, the perspective of different informants is crucial. The children’s own perception of their well-being should be noted whenever possible, but when they cannot communicate, information provided by parents, who know their children better than anyone, can help toward making an assessment. In addition, health-care professionals who work closely with patients (Barry et al. Reference Barry, Frick, Kamphaus and Geisinger2013; Mash and Hunsley Reference Mash and Hunsley2010; Whitcomb and Merrell Reference Whitcomb and Merrell2013) are also key informants. In palliative pediatric care, children’s developmental stages (biological maturation) or their capacity for verbal communication, resulting from their cognitive development or neurological deterioration caused by the disease itself (Bogetz and Lemmon Reference Bogetz and Lemmon2021; Wood et al. Reference Wood, Simpson and Barnes2010), may make interpretation of a child’s emotional sphere even more complicated than it already is in adults.
However, clinically relevant discrepancies are known to exist between the reports of parents and children, and those of the practitioners (Janse et al. Reference Janse, Sinnema and Uiterwaal2008). Some studies compared perceptions of quality of life among children, parents, and physicians and showed the objective attributes had higher percentage agreement between children/parents and physicians than the subjective attributes, like emotion and well-being (Janse et al. Reference Janse, Gemke and Uiterwaal2004). In the case of a pediatric cancer patient, parents are known to be more optimistic about their child’s prognosis and that the physician’s assessment tends to be less so (Levine et al. Reference Levine, Cuviello and Nelson2021). Other studies indicate that successful symptom treatment may not be the only crucial variable influencing main outcome measures, such as satisfaction with care, quality of life of children and parents, or peacefulness of the dying phase. Concordant symptom perception between parents and professionals appears to be a pivotal factor in predicting both satisfaction with the care team and parental quality of life (Vollenbroich et al. Reference Vollenbroich, Borasio and Duroux2016).
Our research aims to compare the perception that children in the pediatric palliative care setting have of their emotional well-being, or that expressed by the parents of those children who cannot communicate, with the perception held by the professionals involved in their care. Our hypothesis is that there will be differences between the perception expressed by the children and their parents, compared to that expressed by health-care professionals.
Methods
Study design
This is a cross-sectional study of the perception of the emotional well-being of children nearing the end of their life.
Settings
In this study, we used a convenience sample of children being treated at the Pediatric Palliative Care Unit (PPCU) of the Sant Joan de Déu Hospital (Barcelona, Spain). We evaluated the perception of well-being of all the children through assessments, attempting to gauge the opinions of the children (themselves if they had verbal communication or, if not, of their parents) and health-care professionals.
Participants
The inclusion criteria were as follows: (a) children with a life-threatening or life-limiting illness, (b) the child’s first multidisciplinary evaluation had been carried out by the PPCU, (c) the father and/or mother were the child’s legal tutors, and (d) both had signed the consent form.
Variables and data collection
Senior psychologists from the palliative care service collected data via interviews with the participating children or parents and health-care professionals (doctors and nurses from the PPCU). To obtain the evaluation of the well-being of children, the researchers created 2 groups: children or parents and professionals. The group of children or parents was made up of the children themselves when they could answer according to their evolutionary stage or the parents of the children if the child was under 8 years of age or had severe neurological impairment that did not allow him/her to answer. According to the literature (Feeny et al. Reference Feeny, Juniper, Ferrie and Drotar1998), children of 7 years and older can answer questions regarding their own health. The group of professionals was made up of pediatricians and/or nurses from the PPCU who were referents for that child and family.
Sociodemographic information on the participating children was obtained, such as age, gender, whether they had any siblings, national origin, family cohabitation, and school attendance.
Illness data collected included diagnosis, clinical situation, and length of time they had been visiting the PPCU.
Perception of emotional well-being
To collect data on the perception of emotional well-being of the children, we used an adaptation to Distress Thermometer (DT) (Graham-Wisener et al. Reference Graham-Wisener, Dempster and Sadler2021; Mitchell et al. Reference Mitchell, Hussain and Grainger2011) which is a one-item instrument indicating a patient’s general distress level on a 0–10 visual analog scale. The DT (0 for no distress; 10 for extreme distress), developed by the National Comprehensive Cancer Network panel, was used within their guidelines for the management of emotional distress. In our research, children were asked: “Over the last few days, in general, how have you been feeling? Ranging from 0 to 10, where 0 corresponds to very bad and 10 to very good.” For parents and professionals, the question was adapted as follows: “Over the last few days, how would you assess the well-being of your/this child?” The score ranges from 0 (very bad) to 10 (very good). The lower the scores, the worse the emotional state. This question is adapted from the “Detection of Emotional Distress (DED) scale” (Limonero et al. Reference Limonero, Mateo and Maté-Méndez2012, 148; Maté et al. Reference Maté, Mateo and Bayés2009, 517) which measures the emotional distress of patients and end-of-life situations. Once the child’s well-being was initially assessed, the health workers were asked for their evaluations within a 3-day period to ensure that the time difference did not result in a confounding variable, thus avoiding potential sources of bias.
Statistical methods
Data analysis was carried out using the software program R Core Team v.4.2.1. Descriptive means and standard deviations (SDs) were calculated. The t-test and F-statistic for the variance of the group means were carried out, along with one-sided tests for paired samples and nonparametric Wilcoxon–Mann–Whitney tests.
Results
Figure 1 is a flow chart representing the participants in the study. Of the 40 families who were approached to participate in the study, 3 (7.5%) refused to participate and 7 children (17.5%) could not be assessed in the evaluation window. Of the 30 children in the final sample, 14 children were themselves able to express their level of well-being, while that of the other 16 was expressed by parents since those children could not communicate. All children were assessed by a health professional.
Fig. 1. Flow chart of participants in the evaluation of well-being in the palliative pediatric unit.
The mean age of children was 10.8 years (SD = 6.1; range: 0 < age < 21), 11 were females (37%), and 19 were males (63%). In total, 12 children (40%) attended school (Table 1). Oncohematological diseases were present in 53.3% of the cases and neurological disease in 36.7%. The clinical situation was stable for 20 of the children (66.7%) and in progression for 10 children (33.4%). On average, children had been treated in the palliative care unit for 7.8 months (SD = 12.6) at the time of the interview.
Table 1. Sociodemographic and clinical characteristics of the children (N = 30)
A two-sided t-test for mean differences is used to compare the differences between the “children or parents” group and the “health professionals” group when they rate children’s emotional well-being (Table 2). The results reveal that there is no statistically significant difference in the mean scores of males vs. females, nor in the children’s or the parents’ evaluations (p = .979) or in the health professionals’ assessments (p = .373). No difference is found by groups of children or parents in any of the comparisons by informants, by speciality (oncology or neurology), age group, schooling, siblings and cohabitation condition, origin, or time visiting the PPCU.
Table 2. Emotional well-being evaluation of children (N = 30) by informants
* p-value: two-sided t-test of the mean difference between the 2 groups evaluated by the same informant.
** p-value: two-sided F-test of the mean difference between several groups evaluated by the same informant.
Health professionals rated well-being when the disease status was in progression (mean = 5.1, SD = 1.1) significantly (p = .037) lower than when the disease was stable (mean = 6.2, SD = 1.4).
The scores provided by children or parents (mean = 7.1, SD = 1.6) and their corresponding paired health professionals (mean = 5.8, SD = 1.4) indicate that children or parents provide higher scores (means and SDs are shown in Table 2). In order to see if the difference made in the evaluation of the children well-being between “children or parents” and “professionals” is maintained considering the different sociodemographic and clinical variables, Table 3 shows the difference between the mean score given by children or parents minus the mean score given by health professionals. The one-sided t-test for the difference in means for paired samples reveals a mean assessment of “children or parents” higher than that of professionals (difference = 1.3, t-test = 4.6, p < 0.001).
Table 3. Differences in well-being mean scores of children or parents minus health professional rating
** p-value: paired sample one-sided test of the mean difference between children’s or parents’ and professional’s evaluation.
Similarly, a significantly higher difference in the average score is obtained for professionals compared to children or parents independently of who is assessing the child. When children evaluate their own status (mean = 7.3, SD = 1.5), their rate is higher than that of their corresponding health professional (mean = 5.9, SD = 1.6). The one-sided t-test for paired samples indicates that the difference is significant (difference = 1.4, t-test = 3.1, p = .004). When parents evaluate their child (mean = 7.0, SD = 1.7), their rate is also higher than that of the health professionals evaluating their child (mean = 5.7, SD = 1.2), and the one-sided test again indicates that the score is higher (difference = 1.3, t-test = 3.3, p = .003).
Higher average scores of children or parents compared to health professionals are found independently of the disease status (p < .001 when the disease status is stable, and p = .012 when the disease status is in progression), for children aged 10 years or under (p = .001) or more than 10 years (p = .008), for those attending school (p = .012) and for those that do not attend school (p < .001), and by medical speciality (p = .009 for oncology and p = .002 for neurology). No difference is found in the mean score of well-being provided by children or parents and professionals for medical specialities other than oncology and neurology, although there are only 3 children in this group. Similarly, no evidence is found for children of origins other than Spanish (p = .085) and children with no siblings (p = .086), but again these are small groups with fewer than 10 children.
When assessing children by sex, the mean score provided by male children (mean = 7.1, SD = 1.4) and their corresponding health professionals (mean = 5.6, SD = 1.2) is higher (t-test = 4.7, p < .001). A significantly higher score cannot be detected (t-test = 1.8, p = .052) when comparing female children (mean = 7.1, SD = 1.9) and their corresponding health professionals (mean = 6.1, SD = 1.7).
Figure 2 presents the individual paired evaluations provided by children or parents (left) and the health professionals (right) together with a box-plot. In 20 cases, the connecting dashed line is seen to fall, indicating that the evaluation made by children or parents is higher than that made by health professionals (and in 6 cases, the difference is by more than 2 points). There is no difference in the score provided by children or parents and their health professional in 7 specific cases; only in 3 cases, did children or parents feel worse than the health professionals’ assessment.
Fig. 2. Paired scores in the well-being assessment (N=30).
Figure 3 presents the individual differences of the emotional well-being scores provided by children or parents and their corresponding health professional, ordered from highest to lowest. Where the child has provided a self-assessment, the bar is shaded in a dark color, and where the parents have provided the emotional assessment, the bar is shown in a lighter color. The figure does not show a systematically higher difference between children and professionals compared to the difference between parents and professionals.
Fig. 3. Individual differences of the emotional well-being scores.
Discussion
Our study highlights that there is a different assessment of emotional well-being between the children (children themselves or their parents) and the health professionals who treat them, usually with a more positive assessment being made by children or parents.
For the group of children or parents, no differences were found in the evaluation of emotional well-being in relation to sociodemographic and clinical variables. In other studies that compare the prognostic vision of patients and professionals, a significant difference was found among patients in different diagnostic groups (Levine et al. Reference Levine, Cuviello and Nelson2021). In the study of Wolfe et al. (Reference Wolfe, Orellana and Ullrich2015), which evaluated symptoms in the last 12 weeks of life of 25 children and adolescents (with ages ranging from 7 to 18 years old), symptoms were reported along with a high degree of emotional well-being, with physical symptoms being particularly frequent in the final weeks of life. This is not observed in the evaluations of our children or parents, which prompts us to question how much of their discomfort is based on the clinical situation of the disease and how much is due more to psychosocial problems or anticipation of future scenarios, as other studies point out (Baggott et al. Reference Baggott, Cooper and Marina2014). In our research, we have not compared the perception of well-being that children express with that expressed by their own parents, but it is important to highlight that the evidence points to significant differences. Parents reported a higher prevalence of physical and psychological symptoms than children for every symptom except shortness of breath. The overestimation of symptom prevalence was most significant for the physical symptoms of fatigue, nausea, and lack of appetite and for the psychological symptoms of feeling nervous and sadness (Baggott et al. Reference Baggott, Cooper and Marina2014; Montgomery et al. Reference Montgomery, Vos and Raybin2020). Although including the child’s perspective is an essential component of understanding the individual experience of suffering, and thus of improving overall care during advanced stages of an illness, it is a perspective that is often overlooked (Montgomery et al. Reference Montgomery, Sawin and Hendricks-Ferguson2016; Wolfe et al. Reference Wolfe, Orellana and Ullrich2015). The group of professionals tended to evaluate the well-being of the child differently, depending on whether the child’s clinical situation was progressive or stable, something that did not occur in the children or parents group. Health professionals when assessing a clinical situation of an illness in progression tended to perceive the children as having a worse emotional state. Clinicians may be more influenced by, and in turn more focused on, the course of the disease when assessing a child’s well-being.
In line with the main objective of the study, significant differences were found between the evaluation of children’s emotional well-being made by children or parents and that of health professionals. Children or parents always value children’s emotional well-being more positively than do the professionals. Even if differentiating between the group of children or parents of responding children and that of responding parents, the differences remain. Our findings are consistent with the study of Wolfe et al. (Reference Wolfe, Grier and Klar2000) who described a staggering difference between the parents’ and the health-care providers’ perception of symptoms during the dying phase, and that this might contribute to the child’s suffering. In our study, parents really appreciated appropriate communication in advance about possible distressing symptoms that might arise during the last days of their child’s life (Vollenbroich et al. Reference Vollenbroich, Borasio and Duroux2016). Others studies found that parent–provider agreement regarding prognosis and goals of therapy was generally poor. Overall, parents tended to be more optimistic than providers (Levine et al. Reference Levine, Cuviello and Nelson2021; Rosenberg et al. Reference Rosenberg, Orellana and Kang2014). One possible explanation is that parents give a higher score to emotional well-being and socialization, while health professionals focus on evaluating objective components (the attributes of vision, hearing, speech, ambulation, or dexterity) related to the quality of life (Janse et al. Reference Janse, Sinnema and Uiterwaal2008; Morrow et al. Reference Morrow, Quine and Loughlin2008). Family members are often managing the conflicting emotions of not wanting their child to suffer but not wanting their child to die and have to adapt rapidly and repeatedly to new situations. Coping strategies include trying to maintain hope and “staying positive” (Mitchell et al. Reference Mitchell, Slowther and Coad2022).
The findings of this study may have important clinical implications. First, it points to the need for health professionals to ensure there is effective communication with children and their parents. Clinicians should remember that even moderate levels of discord could potentially lead to disagreements, misunderstandings, and loss of trust if children or parents feel their concerns are not being addressed by the clinical team (Hill et al. Reference Hill, Nathanson and Fenderson2017). Some studies suggest that optimism may be conveyed directly from the physician through overly optimistic statements or that optimism may be implied through the avoidance of difficult discussions (Levine et al. Reference Levine, Cuviello and Nelson2021). Although some clinicians may be reticent in sharing realistic information for fear of taking away hope, honest disclosure of a poor prognosis may in fact foster hope (Mack et al. Reference Mack, Wolfe and Cook2007) by allowing families to make the best decision possible for their child’s care. In contrast, excessive optimism from the physician or lack of sharing of information can threaten hope and harm the parent–physician relationship, since parents often depend on their physician for direct information and guidance in decision-making (Nyborn et al. Reference Nyborn, Olcese and Nickerson2016).
Second, we must remember that optimism is a coping strategy that can be temporarily adaptive for children and their parents. From the results obtained in our study and those already presented by other studies which also compare the perception of children, parents, and professionals, clinicians must bear in mind that these differences in perception do not have to lead to a discrepancy in the emotional well-being of the child, but rather to the use of scale or words. It could be that children and parents, armed with optimism as a coping strategy, find it difficult to negatively evaluate their well-being and that it is easier for professionals to give low scores. There is also a possibility that participants may have expressed prognostic optimism as a function of performative language, maintaining a positive mindset to influence a preferred outcome (Levine et al. Reference Levine, Cuviello and Nelson2021). This also leads us to consider the importance of using this type of scale as a comparative measure for the same subject at different times, and not to compare between subjects. Recognizing this tendency toward optimism in patients and families may enable clinicians to guide communication and aid in decision-making, supporting hope but also clearly communicating facts and realistic probabilities. Hope is essential for people who are coping with serious and prolonged psychological stress, and it is not a perpetually self-renewing resource; it has peaks and troughs and, at times, can be absent altogether (Folkman Reference Folkman2010). Without a proper understanding of the variation in the parental sense of duty, professionals may be inadequately prepared to understand parents’ decisions and help them prioritize their duties based on an accurate prognosis (Feudtner et al. Reference Feudtner, Walter and Faerber2015).
Third, detecting extreme cases in these discrepancies, either too high or too low, is crucial. In the case that the difference is negative, that is, children or parents value their well-being more negatively than the professionals, this could be a signal indicating that the child is experiencing greater suffering and requires further attention. On the contrary, extreme cases of optimism may lead to nonadaptive coping styles and could result in the child’s suffering being underplayed. An excessive optimism may impair the ability of families to make informed care decisions throughout the disease trajectory (Levine et al. Reference Levine, Cuviello and Nelson2021).
Finally, for a global evaluation of the well-being or emotional state of a child or adolescent at the end of their life, it is essential to consider both the assessment given by the child, or possibly their parents, and the health professionals (Montgomery et al. Reference Montgomery, Vos and Raybin2020). It should be remembered that the professionals, and to a certain extent the parents, are observers and base their evaluations on both a rational and emotional judgment of what they think may be affecting the child (Limonero et al. Reference Limonero, Maté and Mateo2016; Stein et al. Reference Stein, Dalton and Rapa2019; Vollenbroich et al. Reference Vollenbroich, Borasio and Duroux2016). Thus, we cannot ignore the fact that in pediatric palliative care, health-care professionals are exposed to emotionally demanding clinical experiences with high levels of bonding and involvement when the suffering of a child or family is particularly intense (Kase et al. Reference Kase, Waldman and Weintraub2019), and this could also lead to the overestimation of psychosocial concerns. As we have seen, health-care professionals tend to evaluate emotional well-being more negatively than the parents and the children themselves. Interestingly, a better evaluation given by children or parents compared to that given by their health professional might be a sign of good caring standards.
Conclusions
Despite the difficulties inherent in advanced and end-of-life disease in the pediatric population, children with communication skills can assess their own level of emotional well-being, while the assessment of children without verbal communication relies on evaluations made by their parents. Compared to the assessment made by children or parents, the health professionals provide a significantly lower score. That difference exists even if we consider age groups (children less than 10 years old and children aged 10 years or over), illness status (stable or in progression), type of disease, and school attendance. Differences persist independently of whether the children themselves provided a self-assessment, or whether the assessment was carried out by the parents.
The triad of informants (children, parents, and professionals) must be seen and treated as complementary in order to establish the best possible evaluation of the children’s emotional well-being, especially when the child lacks verbal communication skills.
When dual assessments are made by children or parents and health professionals simultaneously, those children or parents who report an emotional well-being score lower than the assessment made of them by their health professional should be carefully monitored. A lower score on children’s or parents’ side may be a sign of distress that requires further analysis in case additional emotional support is needed.
Limitations and future research
Despite the importance of the findings reported here, this study has several limitations that merit consideration. The first has to do with the use of a convenience sample of children with advanced or end-stage illnesses being treated at a palliative care unit. The second limitation is related to the sample size. Studies are needed that include more children, relatives, and professionals, albeit acknowledging the inherent difficulty of the end-of-life situation in pediatric patients. The third is related to the assessment of emotional well-being for children unable to communicate orally, for which the study relied on the parents’ assessment. Finally, the scale used in the study was designed for adult populations. Specific tools could be developed for the pediatric population, and instruments to assess patients’ well-being could be designed for PPCU health professionals. Likewise, our study had a cross-sectional design. To confirm these results, it would be appropriate to carry out longitudinal studies with the participation of a greater number of health centers to avoid evaluative biases of health professionals.
Acknowledgments
We would like to express our gratitude to all children, parents, and health-care professionals participating in the study who contributed to improving the assessment of children in an end-of-life situation. The authors are also grateful to the psychologist Marta Viñals and the members of the Palliative Care Unit (Sant Joan de Déu Hospital). We also would like to thank the ANJANA Working Group of the Catalan-Balearic Society of Palliative Care and the Catalan Society of Pediatrics for their support and for including this study in their research priorities for the improvement of end-of-life health care.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this paper.
Conflicts of interest
The author(s) declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this paper.
Ethical approval
The study was approved by the Clinical Research Ethics Committee of the hospital with the reference code PIC-38-17.
Open access
