A Multi-Country Assessment in Eurasia: Alignment of Physician Perspectives on Palliative Care Integration in Pediatric Oncology with WHO Guidelines

Abstract

Background

The World Health Organization (WHO) advocates for early integration of palliative care for all children with life-threatening illness. Provider awareness and misperceptions, however, can impede this imperative. In the Eurasian Region, little is known about physician knowledge and perspectives on palliative care.

Methods

The “Assessing Doctors’ Attitudes on Palliative Treatment” (ADAPT) survey was developed as an evidence-based and culturally-relevant assessment of physician perceptions on palliative care integration into childhood cancer care in Eurasia. Iteratively tested by American and Eurasian palliative care experts, the survey was culturally adapted, translated, and piloted in English, Russian, and Mongolian. The survey was distributed to physicians caring for children with cancer. Fifteen statements were scored in accordance with WHO guidelines to evaluate provider knowledge. Statistical analysis was complemented by qualitative analysis of open-ended responses.

Results

We received 424 responses from 11 countries in Eurasia. Mean alignment between provider perspectives and WHO recommendations was 70% (range 7%−100%). Significant independent predictors of higher alignment included country, prior palliative care education, and greater experience with patient death. Respondents primarily described palliative care as end-of-life care and symptom management. Two-thirds of respondents (67%) reported not feeling confident delivering at least one component of palliative care.

Conclusion

This is the first study assessing physician perspectives and knowledge of palliative care in Eurasia and reveals wide variability in alignment with WHO guidelines and limited confidence in providing palliative care. Study findings will inform targeted educational interventions which must be tailored to the local political, economic, and cultural context.

Condensed Abstract

This is the first study assessing physician attitudes on the integration of palliative care into pediatric oncology in 11 countries in Eurasia. This work reveals wide variability in alignment of provider knowledge with WHO guidelines and limited confidence in providing palliative care to children with cancer.

Background

The World Health Organization (WHO) defines palliative care as the prevention and relief of physical, developmental, psychosocial, and spiritual suffering of patients and their families facing life-threatening illness.¹ Early integration of palliative care is identified as “an ethical responsibility of health systems” by the WHO in the management of children with serious illness, regardless of local resource limitations.^1,2

Early integration of palliative care is critical for pediatric oncology patients, shown to significantly improve symptom management, family communication, and quality of life for pediatric cancer patients and caregivers, while decreasing patient suffering and parental psychological distress.^3–6 Further, palliative care integration is associated with fewer invasive procedures, increased hospice involvement, and lower odds of dying in an intensive care unit, representing a cost-effective intervention in resource-limited settings.^7–11 Notably, all literature evaluating the outcomes of pediatric palliative care is performed in high-income countries.

Despite the known benefits of palliative care integration for children with cancer, palliative care services do not exist in over 60% of countries, with decreased access in low- and middle-income countries (LMICs).^12,13 In resource-limited settings, decreased reporting of pain symptom control, lack of emotional and bereavement support, and limited interdisciplinary communication hinder the quality of cancer care.^12,14–17 The disproportionate global burden of childhood cancer in LMICs demands increased availability of high quality, comprehensive palliative care to address patient and family suffering and needs in these settings.¹⁷

While structural barriers, such as lack of access to services and other financial constraints, limit utilization of palliative care services, underlying stigma among physicians treating children with cancer further impede integration of palliative care into disease-directed therapy, even when resources are available.^18,19 Studies analyzing provider perceptions of palliative care in pediatric cancer, predominantly conducted in high-income countries (HICs), find that nearly half of physicians do not initiate palliative care consultation until after termination of disease-directed treatment or at the end of life.²⁰ Furthermore, discrepancies persist between physicians’ and families’ perceptions of adequate pain control in children with cancer at the end of life, impacting the quality and effectiveness of delivered care.³ To date, however, limited studies have analyzed provider perspectives on these issues in LMICs.

The Eurasian region, representing eastern Europe and Central Asia, is a geographically and culturally distinct region with a shared history and similar medical system. Across this region, countries experience a wide variability in outcomes and age-standardized DALY rates for childhood cancer.^14,17 Facilitated by St. Jude Global and the WHO Global Initiative for Childhood Cancer,²¹ 11 initial countries formed the Eurasian Alliance of Pediatric Oncologists (EurADO) to improve childhood cancer care in the region. Participating countries are primarily low and middle-income and, with the exception of Mongolia, are all Russian-speaking. To date, there has been no comprehensive evaluation of physician perspectives on palliative care in the region. The objective of this study was to assess perspectives and knowledge of physicians caring for children with cancer on palliative care in 11 countries in Eurasia.

Methods

Instrument Design

The “Assessing Doctors’ Attitudes on Palliative Treatment” (ADAPT) survey was designed for physicians of all specialties who care for pediatric oncology patients in the Eurasian region. This study population was chosen as physicians are typically those making decisions around palliative care consultation and integration in this region. Initial survey items were developed based on WHO guidelines and a review of published literature on provider perspectives on palliative care, physician responsibility, and other ethical considerations. ^1,22–25 Two panels of five American and three Eurasian palliative care experts serially tested survey items in English until final consensus was achieved. Bilingual collaborators with expertise in pediatric palliative care and pediatric hematology-oncology further revised the survey for content validity and cultural relevance. The survey was pilot tested by 12 English-speaking physicians, representative of provider specialties in the target survey population, including pediatric oncologists/hematologists, palliative care specialists, intensivists, and surgeons. Feedback then was used to optimize item content and linguistic face validity.

Subsequently, survey questions were translated into Russian and Mongolian by bilingual native speakers. The translations underwent iterative rounds of review by regional collaborators to optimize colloquial syntax, comprehension, and construct, as well as cultural and linguistic face validity. The survey was pilot tested by 13 Russian- and five Mongolian-speaking physicians, representative of the diversity of specialties in the target population. Adjustments were made based on feedback, and surveys were back-translated and reviewed to ensure item construct consistency.

The final survey questions comprise the following five domains: timing and role of palliative care integration, patient suffering and harm, common misperceptions, interdisciplinary communication, and family communication and perspectives (see Supplemental Figure 1 for the final survey). ADAPT relies on quantitative items utilizing Likert scale and multiple-choice with open-ended questions to enrich data interpretation. The survey was distributed anonymously via two methods: an electronic platform Qualtrics²⁶ or a paper-based option in settings with limited technology access.

Target Sample and Distribution Strategy

The ADAPT survey was distributed in English, Russian, and Mongolian to physicians responsible for managing pediatric oncology patients across 11 participating countries: Armenia, Azerbaijan, Belarus, Kazakhstan, Kyrgyzstan, Moldova, Mongolia, Russia, Tajikistan, Ukraine, and Uzbekistan. The distribution strategy was individualized to each country’s respective medical structure, workforce, and specific needs to maximally circulate the survey to all eligible participants. Within each country, potential survey participants were identified by local study leaders who have comprehensive knowledge of the relevant professional societies, medical institutions, and workforce. Study leaders were instructed to distribute the survey to physicians of all specialties who treat children with cancer in their country. Distribution strategies ranged from pre-determined contact lists within hospital centers to official national pediatric oncology and hematology listservs, depending on their availability in each country (see Table 1 for details of each country’s distribution strategy).

Table 1.

Country-Specific Distribution Strategies

Country	Distribution Method	Total Distributed Surveys	Total Answered Surveys	Response rate
Armenia	Electronic distribution via formalized contact list; paper version for those with limited technological access	27	16	59%
Azerbaijan	Electronic distribution via formalized contact list	22	17	77%
Belarus	Electronic distribution via formalized contact list	55	23	42%
Kazakhstan	Electronic distribution via formalized contact list	71	53	75%
Kyrgyzstan	Electronic distribution via formalized contact list	19	16	84%
Moldova	Electronic distribution via formalized contact lists by specialty	35	21	60%
Mongolia	Paper distribution in 3 main oncology centers in Ulaanbaatar	29	25	86%
Russia	Electronic distribution via official list-serve of National Society of Pediatric Hematologists Oncologists (NSPHO); partial follow-up via overlapping contact list	1275	186	15%
Tajikistan	Electronic distribution via contact list by specialty, formalized by the Ministry of Health	17	16	94%
Ukraine	Electronic distribution via formalized contact list in all regions and centers	140	31	22%
Uzbekistan	Electronic distribution in urban hospitals via phone	40	20	50%

The ADAPT survey was primarily distributed electronically using the Qualtrics platform. Countries with limited technology or internet access were identified prior to distribution as requiring paper surveys, which were completed by respondents, delivered to a designated drop-off location, scanned by local study leaders, and manually entered into the electronic system for analysis. Both electronic and paper surveys were completed anonymously, and participation was optional at all centers. Only completed surveys were recorded by the electronic platform, and there was no missing data in either paper or electronic surveys used in analysis. Local study leaders recorded how many providers received the survey to calculate response rates. The survey was closed after three months.

WHO Alignment

In accordance with WHO guidelines for pediatric palliative care, fifteen statements in the survey were categorized to be factually accurate or inaccurate (Supplemental Table 1).¹ Each participant received a point for scoring correctly on a survey item if they indicated moderate or strong alignment with the correct response, and received an overall score for percent alignment with WHO recommendations based on the number of correct responses of fifteen statements.

Statistical Analysis

Cumulative regional and country-specific demographic data was reported using descriptive statistics. The Shapiro-Wilk test tested the assumption of normality, based on which either the two-sample t-test or Wilcoxon rank-sum was employed to test differences among continuous variables between groups. The Pearson chi-square assessed differences among categorical variables when all cell counts were greater than 5; Fisher’s exact test was used in other cases to test associations when any cell count was smaller than or equal to 5. Both univariate and multiple regression models were used to determine the predictors of WHO alignment scoring. In the multiple regression analysis, variables with p<0.2 in the univariate analysis were included, and the stepwise variable selection was used to choose the variables with p<0.05 based on the Wald test. Russia was used as the reference group for calculating the country effect as the median scoring country and largest responder group. For secondary analyses, the 5-point Likert scales were assigned numeric value to conduct two-sample paired t-tests. The 5-point Likert scale was also collapsed to three categories (often/always, sometimes, never/rarely) to compare the effect of demographic variables via Fisher’s exact test or Pearson chi-square. Two-sided significance of p<0.05 was employed for the statistical methods. All analyses were performed using SAS software version 9.4, and the heat map was generated in R software version 3.6.2.

Qualitative Analysis

Written survey responses to free-text questions were translated to English by bilingual native speakers. A preliminary codebook was created from a combination of an inductive and deductive approach, encompassing identified common themes and recurrent phrases used in the text.²⁷ Each response served as one segment as a unit of analysis. Test coding on set groups of segments was conducted by two coders to refine codes and definitions, and the codebook was finalized through a reconciliation process to reach consensus within the analytic team. Fluent members of the analytic team were able to refer to Russian responses to provide further clarification as needed. Responses were coded by two coders, with each response serving as the unit of analysis. After double-coding, a kappa of 0.90 ensured inter-rater reliability, and a third reviewer assisted with resolution of code disagreements. All data analysis was performed using MAXQDA software.

Results

Sample Demographics

The ADAPT survey was completed by 424 physicians from 11 countries in the Eurasian region, with a median country response rate of 60% (range: 15–100%, Table 1). Most surveys were completed electronically (91%), with the remainder completing paper surveys (n=38, 9%). The majority of respondents were pediatric hematologists/oncologists (67%), female (73%), below the age of 50 (78%), and had low experience with patient death (five or fewer patients die under their care in the last year, 85%). A minority (19%) reported receiving some form of palliative care education or self-identified as palliative care specialists (5.2%, see Table 2 for complete demographics). Approximately half (54%) of respondents reported access to a pediatric palliative care consultation in their practice setting. Of those who reported availability of palliative care consultation, providers typically indicated access to a physician (92%), with more limited access to interdisciplinary specialists such as psychologists (27%), social workers (16%), or nurses (13%).

Table 2.

Respondent Demographic Characteristics

Demographics	Overall Sample n=424 n (%)
Country
Armenia	16 (4 %)
Azerbaijan	17 (4 %)
Belarus	23 (5 %)
Kazakhstan	53 (13 %)
Kyrgyzstan	16 (4 %)
Moldova	21 (5 %)
Mongolia	25 (6 %)
Russia	186 (44 %)
Tajikistan	16 (4 %)
Ukraine	31 (7 %)
Uzbekistan	20 (5 %)
Age
<35	152 (36 %)
35+	272 (64 %)
Sex
Female	308 (73 %)
Male	116 (27 %)
Primary medical specialty
Adult/Pediatric Palliative Care	22 (5 %)
Pediatric Hematology/Oncology	285 (67 %)
All others	102 (24 %)
Years of experience
0–10 years	168 (40 %)
11+ years	256 (60 %)
Training in Palliative Care
Yes	81 (19 %)
No	343 (81 %)
Access to Consultation
Yes	230 (54 %)
No	194 (46 %)
Number of patients died in care
0–5 patients	361 (85 %)
6+ patients	63 (15 %)

Role of Palliative Care in Childhood Cancer Care

The majority of respondents described the role of palliative care in the care of children with cancer to include pain and symptom management (97%) and psychological support (92%). Significantly fewer respondents, however, identified other established components of palliative care, such as help communicating bad news to patients and families (60%), aid in communication between the patient, family, and medical teams (56%), aid in family decision-making around treatment options (51%), and help clarifying the goals of care of the patient and family (39%). Few respondents indicated that palliative care has no role in cancer care for children (3.8%) (Figure 1a).

Figure 1. Components of Palliative Care: Quantitative and Qualitative Analysis.

1a. Quantitative analysis of responses to question “The role of palliative care in the care of children with cancer” (n=424). Respondents could choose all components that they identified as appropriate. Components are listed from most to least commonly indicated by % of respondents.

1b. Qualitative analysis of responses to question “What does palliative care meant to you?” (n=379). The code map depicts the relatively frequency of all codes identified in qualitative analysis of question responses. The codebook was developed through a combination of an inductive and deductive approach (see Supplemental Table 3 for code definitions and Supplemental Table 4 for code frequencies).

These findings were supported by qualitative analysis of free-response survey questions, answered by 89% (n=379) of respondents (see Figure 1b for code frequency map and Supplemental Table 2 for code definitions). In response to the question “What does palliative care mean to you?”, those mentioning components of palliative care (n=346) most often described palliative care as addressing patient quality of life (72%), symptom management (37%), psychological support (30%), end-of-life care (24%), and medical care (23%), while religious and spiritual support (8.4%) and interdisciplinary communication (2.0%) were mentioned less frequently (Supplemental Table 3). Of respondents who described the timing of palliative care, the majority indicated that palliative care is administered when no other curative options are available (57%), as well as at the end of life (36%). Many responses focused on “improving the quality of life of incurable patients” as well as mitigating patient “pain and suffering.” A minority of respondents described palliative care more positively (7.7%) as a celebration of life (“decorate last days”, “giving life”), as a form of compassion (“without loneliness”, “caring and loving”), or as a mechanism by which to promote life and death with dignity (“dignified passing into another world”).

Alignment with WHO Guidelines on Palliative Care Integration

Using the 15 questions aligning with WHO guidelines on palliative care integration with pediatrics (Supplemental Table 1), the overall mean WHO-alignment of respondents was 70% (range 7%−100%). In univariate analysis, palliative care education, country of practice, specialty (adult or pediatric palliative care vs. others), more than 10 years of work experience, access to palliative care consultation, and having greater experience with patient death (six or more patients die in the last year) were significant predictors of participant alignment with WHO guidelines, with the difference between groups ranging from 3 to 12%, and medical specialty having the largest difference between groups (12%) (Table 3). Mean WHO alignment significantly varied across countries, with a range between 62% and 81%, (Figure 2 and Supplemental Figure 2). Multiple regression analysis demonstrated that country, palliative care education, and having six or more patients die in the last year were independent predictors of WHO alignment (Table 3).

Table 3.

Analysis of Demographic Factors and Perspectives Significantly Impacting WHO Alignment

Demographic	Category	Mean WHO Alignment (%)	Univariate model (p-value)	Multivariate Model (p-value)	Percent Difference (% Alignment)
Age			0.067	Not significant	3%
	<35 years	68%
	>35 years	71%
Sex			0.066	Not significant	4%
	Female	71%
	Male	67%
Primary medical specialty			0.001	Not significant	12%
	Adult/pediatric palliative care specialists	81%
	All others	69%
Years of experience			0.034	Not significant	4%
	0–10	67%
	11+	71%
Palliative care education			<.001	0.002	7%
	Yes	75%
	No	68%
Access to consultation			.043	Not significant	3%
	Yes	71%
	No	68%
Patients died in care in last 12 months			0.038	0.047	5%
	0–5	69%
	6+	74%
Country	Individual countries		0.002	0.001	See Figure 2

Figure 2. Eurasia Regional Country Map by Percent WHO Alignment.

Based on the 15 factual statements supported by the WHO guidelines on palliative care (Supplemental Table 1), each respondent received an overall score for percent alignment with WHO recommendations based on the number of correct responses of 15 statements (overall mean 70%, n=424). The mean country WHO alignment was calculated based on respondents’ performance from each country, ranging from 62% to 81%. The map displays the distribution of WHO alignment across 11 countries in the Eurasian region from highest (green) to lowest (red) alignment.

Among the questions used to measure WHO alignment, the most common regional misconceptions were: 1) early consultation with palliative care causes increased parental burden and anxiety (23% disagree), 2) it is difficult to know when a patient with cancer would most benefit from meeting the palliative care team (34% disagree), and 3) palliative care is synonymous with “end-of-life’ care: (44% disagree) (see Table 4 for complete ranked list).

Table 4.

Misperceptions of WHO Alignment Statements: Most to Least Common

Question	% Correct
Early consultation with palliative care causes increased parental burden and anxiety.	23
It is difficult to know when a patient with cancer would most benefit from meeting the palliative care team.	34
Palliative care is synonymous with “end-of-life” care.	44
Palliative care for children with cancer can be delivered by health care workers of all disciplines, not only by palliative care specialists.	60
Involving palliative care suggests the oncologist has failed in the mission to cure the patient.	68
Involving the palliative care team early has negative effects on the relationship between the oncologist and the patient and family.	71
Palliative care can be integrated with disease-directed therapy.	74
Palliative care is incompatible with curative care.	76
Involvement of palliative care undermines the role of the pediatric oncologist as the physician in charge of patient care.	81
Children with cancer who receive palliative care die earlier than those who do not.	81
Palliative care is appropriate at any stage of treatment in a child with high-risk cancer.	84
Early integration of palliative care for all children diagnosed with cancer would decrease patient suffering.	86
Children with advanced and incurable cancer often suffer at the end-of-life	87
Early integration of pediatric palliative care with cancer care would improve interdisciplinary communication.	88
Involvement of palliative care during cancer therapy gives greater attention to quality of life and symptom management (e.g. pain, constipation, dyspnea, fatigue).	89

Impact of Palliative Care Education

Prior palliative care education not only impacted physician knowledge, but also influenced perceptions and experiences around palliative care and the end of life (Table 5). Respondents with palliative care education were significantly less likely to believe that children receiving palliative care die earlier than those who do not (90% disagree vs 79%, p=0.014) and that administering opioids to patients in pain hastens death due to respiratory depression (94% disagree vs 74%, p<0.001).

Table 5.

Provider Perspectives Significantly Impacted by Palliative Care Education

	Palliative Care Education (n=81)	No Palliative Care Education (n=343)	P-value
Involving palliative care suggests the oncologist has failed in the mission to cure the patient.			0.029 a
Disagree	62 (76.5%)	225 (65.6%)	.
Neutral	1 (1.2%)	29 (8.5%)	.
Agree	18 (22.2%)	89 (25.9%)	.
Palliative care for children with cancer can be delivered by health care workers of all disciplines, not only by palliative care specialists.			0.013 b
Disagree	15 (18.5%)	119 (34.7%)	.
Neutral	6 (7.4%)	29 (8.5%)	.
Agree	60 (74.1%)	195 (56.9%)	.
Children with cancer who receive palliative care die earlier than those who do not.			0.005 a
Disagree	73 (90.1%)	272 (79.3%)	.
Neutral	1 (1.2%)	41 (12.0%)	.
Agree	7 (8.6%)	30 (8.7%)	.
Involving the palliative care team early has negative effects on the relationship between the oncologist and the patient and family.			0.054 a
Disagree	66 (81.5%)	233 (67.9%)	.
Neutral	5 (6.2%)	45 (13.1%)	.
Agree	10 (12.3%)	65 (19.0%)	.
Administering opioids to patients in pain hastens death due to respiratory depression.			<.001 a
Disagree	76 (93.8%)	254 (74.1%)	.
Neutral	1 (1.2%)	32 (9.3%)	.
Agree	4 (4.9%)	57 (16.6%)	.
In my setting of practice, palliative care consultation is available when I feel it is needed for a child with cancer.			<.001 b
Never/Rarely	13 (16.0%)	126 (36.7%)	.
Sometimes	14 (17.3%)	63 (18.4%)	.
Often/Always	54 (66.7%)	154 (44.9%)	.
I have felt burdened by my inability to control the suffering of children at the end-of-life.			0.021 b
Never/Rarely	18 (22.2%)	37 (10.8%)	.
Sometimes	21 (25.9%)	94 (27.4%)	.
Often/Always	42 (51.9%)	212 (61.8%)	.

^aFisher’s exact test

^bPearson chi-square test

Overall, 60% of respondents reported they always or often felt burdened by their inability to control the suffering of children at the end-of-life; feelings of burden were significantly less prevalent among those with palliative care education (52% vs. 62%, p=0.021) and among palliative care specialists as opposed to other specialties (36% vs. 61%, p=0.048).

Ultimately, the majority of respondents (95%) reported they would like to receive more palliative care education. Those providers desiring more education scored 19% higher in WHO alignment than those that showed less interest in learning more about palliative care (71% vs. 52%, p<0.001).

Attitudes Toward Palliative Care

Almost half of providers (45%) indicated that quality of life is often overlooked in the face of cancer-directed treatment. Majority believed that children suffer at the end-of-life (87%) and that early integration of palliative care would decrease patient suffering (86%). However, 57% agreed that physicians typically continue to recommend cancer-directed treatment for children with advanced and incurable disease (Supplemental Table 4).

Respondents reported generally being confident managing the physical (57%) and emotional (63%) needs of their patients, while less felt comfortable addressing grief and bereavement (45%), with 67% of respondents reporting a lack of confidence managing at least one of these components of palliative care. Palliative care education did not significantly influence confidence delivering any component of palliative care (Supplemental Table 5). Respondents’ assessment of their comfort was higher than their assessment of the comfort of other physicians in their setting in providing all three types of care: physical needs (p=0.011), emotional needs (p<0.001), and grief/bereavement support (p=0.018) (Supplemental Table 6). Further results of relevant physician perspectives are summarized in Supplemental Table 4.

Discussion

The WHO advocates for early integration of palliative care as a standard of care for all children with life-threatening illness, and specifically, for children with cancer. Universal implementation is hindered by numerous factors such as provider attitudes, misconceptions, and lack of knowledge on palliative care.¹ Understanding these perspectives is imperative to guide educational interventions designed to align provider practice with WHO guidelines. Provider perspectives likely differ globally, and little is known on this topic in the Eurasian region. To our knowledge, this is the first study of provider perspectives on palliative care in this region.

In this large international survey of physicians who care for children with cancer, we demonstrated suboptimal (70%) alignment with WHO guidelines on this topic, suggesting a knowledge gap that may impact the quality of care provided in this region. There is limited data on physician knowledge of palliative care in other resource-limited settings; however, these results are comparable to previous results where physicians had a mean score of 68%.²⁸ Respondent alignment with WHO guidance on palliative care integration varied greatly in our study, both among individual participants (7% to 100%) and by country (62% to 81%). In multiple regression analysis, independent predictors of higher WHO alignment included previous palliative care education, having more experience with patient death, and country of practice. The effect of these factors on WHO alignment, however, was small, suggesting that additional factors likely influence perspectives.

In line with published literature,²⁹ commonly held misperceptions centered around family communication. This included beliefs that involvement of palliative care has a negative effect on the relationship between the oncologist and patient and family and that early consultation burdens parents and causes increased anxiety. Importantly, recent data demonstrate that, contrary to Eurasian provider beliefs, patients and parents often benefit and value early integration of palliative care in the context of pediatric cancer management.³⁰

Supported by quantitative and qualitative survey data, providers in the Eurasian region perceive palliative care primarily as end-of-life care, pain and symptom management, quality of life, and psychological support, whereas discussion of goals of care and communication of bad news were not readily identified as pillars of palliative care. These findings reveal a narrow understanding of the role of palliative care, potentially explaining the perceived conflict between palliative care and disease-directed treatment. Approximately a third of respondents agreed that involvement of palliative care implies failure on behalf of the oncologist, suggesting that many physicians do not recognize alignment between the goals of cancer care and palliative care.

Knowledge of palliative care, as indicated by WHO alignment, nonetheless, does not indicate proficient clinical skills or comfort in providing components of palliative care to pediatric patients. Less than half of respondents reported that physicians limit cancer-directed therapy in the setting of advanced and incurable disease, signifying that clinical practice, as perceived by physicians, does not align with WHO recommendations. The majority of respondents (67%) reported a lack of confidence in providing at least one of the components of palliative care to their patients, of which respondents were least confident providing grief and bereavement care. This is starkly lower than previously self-reported physician competence managing palliative care needs in high-income settings (92%).²⁹ This lack of confidence likely contributes to 60% of respondents feeling burdened by their inability to manage suffering of their patients at the end of life. While those with palliative education felt less burdened, they did not report higher confidence in providing elements of palliative care, differing from the effects of palliative care education in other low-income settings.²⁸

This study clearly captures low exposure to formal palliative care training among physicians caring for children with cancer in Eurasia (19%), with the majority of respondents (81%) having no prior palliative care education. This represents an urgent need to increase access to palliative care training for physicians treating children with cancer in the region. Our findings further demonstrate the widespread demand for palliative care education among these physicians, with the vast majority (95%) desiring more education. Even among those with prior palliative care education, however, the effect of this training on WHO alignment was relatively small, and it did not improve comfort in providing elements of palliative care, suggesting that available palliative care education in the region may not only be limited, but also not optimally effective or aligned with international standards. These findings should inform the development of palliative care educational curricula for providers, as well as organization of clinical residencies to build regional capacity and expertise. Demonstrated by the response variability across countries, local context significantly influences provider perspectives; thus, education curricula must be tailored to meet the needs of local political, economic, and cultural contexts. This study expands our understanding of the predominant attitudes of physicians regarding the role of palliative care in childhood cancer care in Eurasia and is a prerequisite to the creation of effective interventions aimed at improving the timeliness of palliative care integration and the quality of care provided to pediatric patients with serious illness and their families.

This study has several limitations. Differences in the size of the physician workforce and availability of physician listservs in each country resulted in different distribution strategies and variable response rates with only two countries falling below 40%. Lower response rates in Russia and Ukraine may be attributable to the use of large listservs of all potentially eligible providers, while countries where the survey was directly distributed to eligible respondents had more robust rates. Listserv-based distribution was chosen in Russia and Ukraine due to a large number of centers providing pediatric cancer care and a larger pediatric hematology-oncology workforce eligible for study participation. In contrast, the remaining countries with a relatively small pediatric hematology-oncology workforce had no available listservs and relied on the country study leaders conducting hospital-based survey distribution. While this approach may have introduced snowball sampling bias, we achieved a high median response rate and are confident we distributed the survey to the majority of eligible respondents through leveraging local knowledge of the provider workforce. While respondents from Russia are overrepresented in our study sample, Russian physicians had the median WHO alignment of all countries surveyed, preventing study results and conclusions from being skewed. We were also able to demonstrate country-specific differences in physician perspectives, allowing for adapted educational interventions tailored to each local context.

While our survey resulted in a smaller sample size in certain demographic categories (such as physicians with palliative care education) that may have limited our statistical power, our study obtained a large number of responses achieving adequate power to detect significant differences between groups. Similarly, we were able to identify small but statistically significant differences, such as the effect of palliative care education on WHO alignment. Smaller sample sizes in some groups, however, did result in very low cell counts for some variables potentially impacting the validity of significant findings, particularly those with p-values close to 0.05. We believe our overall conclusions, however, many of which had much lower p-values, are generally valid.

A minority (9%) completed the paper survey; however, all recorded responses were complete, and paper surveys were entered into the electronic system, unifying data analysis and ensuring accuracy. As this study primarily relied on electronic distribution (91% surveys), our survey under-represented provider perspectives from more remote, lower-resourced regional centers where provider knowledge is likely more limited. Therefore, these findings likely skew toward more well-resourced, specialized centers in the region, with correspondingly higher levels of education and awareness of palliative care. As a result, this study presents the “best-case scenario” with regards to physician perspectives and knowledge. Nonetheless, we believe that these findings are valid to guide targeted development of palliative care educational initiatives in Eurasia, as initial interventions will be targeted to large, specialized pediatric oncology centers managing the majority of childhood cancer in each country.

Finally, our assessment clearly identified a need to increase access to palliative care services in the Eurasian region in the field of pediatric oncology. More than half of respondents did not have access to palliative care consultation and the majority had no access to psychologists, nurses, or social workers, suggesting that other structural factors play a role in the delivery of pediatric palliative care in Eurasia beyond a lack of high-quality education. While assessment of palliative care perceptions and knowledge among pediatric oncology physicians is crucial to inform the development of targeted educational interventions, further analysis of regional and country-specific barriers must be conducted to better understand other challenges to high-quality palliative care integration into childhood cancer care. While this focus lay beyond the scope of the current study, we advocate for investigation of these important questions in future work.

Conclusion

The ADAPT study is the first assessment of provider attitudes on palliative care integration into pediatric cancer care adapted to the language, political, social, and cultural context of the Eurasian region. Our study identified a number of knowledge gaps and misperceptions among physicians treating children with cancer in this region, with large variability in alignment with international standards as described by the WHO. Respondent history of palliative care education impacted their alignment with WHO standards, perspectives, and experiences, and nearly all participants reported a desire for more education on this topic. Our results demonstrate an opportunity to improve palliative care integration into pediatric oncology care in this region through targeted, locally adapted education interventions with the ultimate goal to improve the quality of life for children with cancer.