Palliative care among cancer patients in Georgia: System-level gaps identified through health-related quality of life discrepancies

Abstract

Background

Palliative care aims to improve health-related quality of life (HRQoL) through holistic, patient-centered approaches. In low- and middle-income countries, including Georgia, care remains largely medically focused, with limited psychological, social, and spiritual support. Discrepancies in HRQoL assessments between patients, families, and healthcare providers may obscure unmet needs and hinder comprehensive care delivery.

Design and Methods

We conducted a multicenter, cross-sectional study across nine palliative care clinics in Georgia from November 2024 to June 2025. HRQoL was assessed using the validated EORTC QLQ-C30 questionnaire, independently completed by 298 cancer patients, their family members (n=298), and 80 healthcare staff.

Results

The QLQ-C30 demonstrated acceptable to excellent internal consistency across all respondent groups. Family members tended to report higher symptom burden and greater functional impairment compared with patient self-reports. Healthcare personnel reported higher functioning levels in emotional, social, and role domains, suggesting potential underestimation of impairments in these areas, while certain symptom domains (pain) were overestimated. Agreement was highest for observable physical symptoms and lowest for psychosocial domains. Despite widespread recognition of the importance of multidisciplinary care, such services were largely absent. Most participants expressed strong support for including psychologists and spiritual care providers, highlighting substantial unmet non-medical needs.

Conclusions

Integrating perspectives from patients, family members, and healthcare personnel reveals critical gaps in Georgian palliative care. The findings underscore the importance of multi-source HRQoL assessment and culturally adapted, multidisciplinary care models. This evidence can guide public health policy, workforce training, and service design, ultimately improving patient-centered care in Georgia and similar settings.

Keywords

Significance for public health

This study exposes critical systemic gaps in palliative care in Georgia, revealing significant discrepancies in health-related quality of life (HRQoL) assessments among patients, family members, and healthcare personnel. The near-absence of structured psychological, social, and spiritual support, as well as multidisciplinary services, undermines patient-centered care and exacerbates caregiver burden. These deficiencies have broad public health implications: they hinder effective symptom management, increase psychosocial stress, and compromise health outcomes at the population level. By demonstrating the urgent need for culturally adapted, integrated, multidisciplinary palliative care models, this research informs policy priorities, guiding targeted interventions and resource allocation. Implementation of such models has the potential to improve population health, optimize quality of life for patients with life-limiting illnesses, and reduce systemic inequities in care delivery, thereby addressing a critical gap in the public health infrastructure of transitional healthcare systems.

Introduction

Modern healthcare has shifted its focus from disease-oriented to patient-centered care. Palliative care has become a core component of healthcare systems, aiming to improve the health-related quality of life (HRQoL) of patients with life-limiting illnesses by addressing physical, psychological, social, and spiritual distress. According to the WHO and the European Association for Palliative Care (EAPC), palliative care should adopt a holistic, multidisciplinary, patient-centered approach that integrates medical interventions with structured psychological, social, and spiritual support.¹ Despite these recommendations, palliative care services remain incomplete worldwide, particularly in low- and middle-income countries, where psychosocial and spiritual care are underdeveloped. Structured, multidimensional evaluation of HRQoL is therefore essential to support patient-centered care and inform public health policy and resource allocation.² Evidence indicates that multidisciplinary care significantly improves HRQoL and supports systematic assessment of patient outcomes.³

Assessment of patients’ health condition requires a multidimensional approach, as perceptions differ among patients, family members, and healthcare personnel. Prior studies indicate that these groups often evaluate HRQoL differently.⁴ Patient self-report is generally considered the most accurate, while assessments by family members and healthcare personnel may be influenced by bias; however, these perspectives can complement patient reports and support a more comprehensive evaluation.^5,6 These discrepancies highlight the limitations of relying on a single perspective and underscore the importance of multi-source assessment in providing a more comprehensive understanding of patient status. Importantly, such discrepancies may reflect not only differences in perception but also underlying system-level gaps in palliative care delivery, including limitations in communication, coordination, and patient-centered assessment practices.

Despite this, studies simultaneously comparing HRQoL assessments across patients, family members, and healthcare personnel remain scarce globally, particularly in low- and middle-income countries, where systemic constraints, limited resources, and insufficient workforce training hinder the implementation of patient-centered and multidisciplinary care models. In these settings, non-medical needs, as well as the quality and alignment of services with contemporary standards, remain insufficiently addressed and poorly documented.

Georgia exemplifies a context in which the burden of cancer and other progressive chronic diseases is rising, yet palliative care remains predominantly medically focused. This context provides an important opportunity to examine how system-level constraints are reflected in differences in HRQoL assessments across stakeholders. Multidisciplinary care approaches are virtually absent, and HRQoL assessments rely mainly on subjective clinical observations rather than structured, patient-centered evaluations. Palliative care was formally established as an independent medical specialty in Georgia in 2004, through the combined efforts of national medical and non-governmental sectors, supported financially and intellectually by international foundations. Initially, the system aligned with global best practices, including standards for multidisciplinary care. Over time, palliative care was integrated into national legislation, the number of service providers expanded, and a regulatory framework was established.

Despite progressive policy development and recognition of palliative care as a continuous, coordinated medical service, 25 years later, state funding remains insufficient to meet actual needs. Government resources cover physicians’ and nurses’ work, but psychological, social, and spiritual services are virtually unfunded. Service accessibility remains a critical challenge, both quantitatively and geographically, with the majority of services concentrated in the capital.⁷ Regulatory barriers also impede access to essential medications, including opioids, limiting adequate management of pain and other symptoms.⁸

Currently, palliative care services in Georgia do not fully meet core objectives, as comprehensive biopsychosocial and spiritual care remains insufficient, limiting patients’ and families’ HRQoL. Patient-centered evaluation of service quality is also underdeveloped; healthcare personnel often rely on family assessments, while policy decisions are primarily based on data from healthcare providers. In the present study, a multidimensional approach is applied to assess HRQoL from the perspectives of patients, family members, and healthcare personnel. We argue that a multidimensional, multi-perspective approach is essential not only for understanding patient suffering but also for identifying system-level gaps in the delivery of high-quality, patient-centered palliative care.

The study aimed to identify system-level gaps in palliative care in Georgia by examining discrepancies in HRQoL assessments among patients, family members, and healthcare personnel.

Design and methods

Study design

This cross-sectional study was conducted in nine palliative care clinics across Georgia between November 2024 and June 2025. Clinics were selected using a pragmatic convenience sampling approach based on accessibility, while ensuring geographical representation across the capital and both eastern and western regions of Georgia. Due to the unequal geographical distribution and functional availability of inpatient palliative care services, random or proportional sampling was not feasible.

According to national data, 18 inpatient palliative care facilities are formally registered in Georgia. Of these, 2 are located in western Georgia, with only 1 functionally active and included; in eastern regions outside the capital, 8 are registered, of which 2 were operational and both included. In Tbilisi, where services are concentrated (8 facilities), 6 high-volume, actively functioning clinics were selected. All participating clinics operate under the same national state-funded palliative care program, ensuring a relatively uniform organizational, regulatory, and financing framework, which supports comparability across sites. Sampling was based on actual service activity and patient flow rather than nominal facility distribution, enabling inclusion of all active regional services and prioritization of high-volume centers. This approach aligns with the study’s aim to assess HRQoL discrepancies in real-world palliative care settings.

However, this strategy may introduce selection bias, with potential overrepresentation of more active and better-resourced facilities. Participating clinics may therefore differ systematically from non-participating or non-operational services, limiting generalizability to less active settings.

Participants were recruited using consecutive sampling, whereby all individuals meeting predefined inclusion and exclusion criteria during the study period were invited to participate until the target sample size was reached. Corresponding family members and healthcare personnel were included following patient consent.

All procedures were conducted in accordance with national legislation and the 1964 Helsinki Declaration. The study was approved by the Institutional Review Board (Study No. 11-13083; 03/07/2024), and written informed consent was obtained from all participants prior to enrollment.

Inclusion criteria

• Adult patients (>18 years) with cancer receiving palliative care;

• Provision of written informed consent to participate in the study;

• Provision of informed consent by the patient for the survey of family members and healthcare personnel.

Exclusion criteria

• Patients deemed psychologically or clinically inadequate, as determined by consultation and clinical data;

• Patients and family members under 18 years old;

• Refusal to participate in the study.

Sample size

The study employed a pragmatic sampling approach based on consecutive recruitment of eligible participants across participating centers during the study period.

Each patient constituted the primary analytical unit and was assessed in parallel by a corresponding family member and a healthcare personnel, forming matched triads that enabled comparative and correlational analyses across respondent groups.

A formal a priori sample size calculation based on expected effect sizes for paired comparisons or correlation coefficients was not performed.

Study subjects

Based on the inclusion and exclusion criteria and the achieved sample, 298 patients with cancer of various sites and stages (Group 1, Table 1), 298 family members (Group 2), and 80 healthcare personnel (physicians and nurses; Group 3) were included in the study.

Table 1.

Distribution of patients by cancer localization according to ICD codes (N = 298).

ICD code*	N	%
C01	3	1.0%
C02	15	5.0%
C07	3	1.0%
C16	11	3.7%
C18	33	11.1%
C19	3	1.0%
C20	7	2.3%
C22	15	5.0%
C25	18	6.0%
C31	3	1.0%
C32	3	1.0%
C34	57	19.1%
C41	3	1.0%
C43	3	1.0%
C50	19	6.4%
C51	3	1.0%
C56	15	5.0%
C61	11	3.7%
C62	3	1.0%
C64	3	1.0%
C67	11	3.7%
C71	11	3.7%
C72	3	1.0%
C77	3	1.0%
C79	3	1.0%
C80	3	1.0%
C81	11	3.7%
C82	3	1.0%
C83	3	1.0%
C85	3	1.0%
C90	3	1.0%
C91	3	1.0%
C96	7	2.3%
Total	298	100.0%

^*ICD – International Classification of Diseases.

The healthcare personnel group comprised 36 physicians and 44 nurses. The majority of both physicians and nurses had more than 3 years of clinical experience.

Questionnaires

Patients enrolled in the study were provided with a set of questionnaires between the 7th and 14th day following their initial contact with the department. The questionnaires, including a structured socio-demographic form and the EORTC QLQ-C30, were completed on the same day or the following day.

All questionnaires were administered in a supervised clinical setting. After completion, questionnaires were briefly checked for completeness, and participants were politely asked to review unintentionally missed items when applicable. Interviewers did not complete responses on behalf of participants, and all responses were recorded directly by the respondents themselves. No missing item-level data were identified across patient, family member, or healthcare personnel questionnaires; therefore, no imputation procedures were required.

To be included in the analyses, assessments from the patient, a corresponding family member, and a healthcare professional were required to be completed within a maximum of 1 day of each other. This restriction minimized the likelihood that differences in ratings were due to changes in the patient’s condition over time.

QLQ-C30 questionnaire

The study utilized the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-C30, version 3.0).⁹ This widely used cancer-specific instrument consists of 30 items, including nine multi-item scales (six functional and three symptom scales) and six single-item symptom measures. Scores are linearly transformed to a 0–100 scale; higher scores indicate better functioning on functional scales and greater symptom burden on symptom scales.

Paired comparisons were conducted between patients and their corresponding family members, and between patients and the healthcare professional directly responsible for their care. For each patient, a single healthcare professional (physician or nurse) primarily involved in their management completed the assessment, ensuring patient-level matching across all three respondent groups.

However, as individual healthcare professionals evaluated multiple patients, the assumption of independence between observations is partially violated, introducing clustering at the clinician level. To account for this hierarchical data structure, analyses involving healthcare personnel were conducted using linear mixed-effects models with a random intercept for clinician, thereby accounting for clustering of patients within clinicians. Findings were consistent with the original analyses; however, results involving healthcare personnel should be interpreted with caution.

The Georgian version of the EORTC QLQ-C30 used in this study was obtained directly from the EORTC Quality of Life Group following formal authorization and developed in accordance with the standardized EORTC Quality of Life Group (QLG) Translation Procedure,¹⁰ ensuring appropriate linguistic and cultural adaptation. To our knowledge, no separate published psychometric validation study specific to the Georgian population is currently available. However, the EORTC QLQ-C30 has been extensively validated across diverse international settings, supporting its use in different cultural contexts.

For proxy respondents (family members and healthcare personnel), the questionnaire was administered using a minimally adapted third-person format referring to the patient. Specifically, item wording was converted from first-person to third-person phrasing, while no modifications were made to item content, response scales, or scoring procedures. This approach preserves the structural integrity of the original instrument and is consistent with prior studies employing proxy-reported QLQ-C30 assessments in oncology and palliative care settings.

Socio-demographic questionnaire

All participants completed a structured socio-demographic questionnaire, which included items on age, age group, gender, education, and employment. Additional questions addressed:

• Beneficiary status within the state program;

• Availability of non-medical specialists in the clinic;

• Preferences regarding involvement of additional specialists;

• Satisfaction with medical services;

• Satisfaction with non-medical services (psychological, social, or spiritual).

Overall satisfaction with care was assessed using a study-specific 7-point Likert-type scale ranging from 1 (“very bad”) to 7 (“very good”). Although this measure has not undergone formal psychometric validation, its structure aligns with commonly used Likert-type scales in healthcare research and was considered appropriate for exploratory assessment.

Healthcare personnel were additionally asked:

• Whether the team includes specialists beyond medical staff;

• Whether a multidisciplinary approach is practiced;

• Whether such an approach is considered necessary.

Responses from all three groups, along with socio-demographic data, are presented in Table 2.

Table 2.

The responses of study participants on socio-demographic and other questions.

Question	Group 1 (n=298)		Group 2 (n=298)		Group 3 (n=80)
Question	Mean or N	SD or %	Mean or N	SD or %	Mean or N	SD or %
Age	71.4	11.4	47.9	12.9
Median	72		45
Min	36		20
Max	90		88
Age group
<25	0	0.0%	2	0.7%
25-29	0	0.0%	5	1.7%
30-34	0	0.0%	27	9.1%
35-39	2	0.7%	42	14.1%
40-44	1	0.3%	69	23.2%
45-49	11	3.7%	60	20.1%
50-54	14	4.7%	18	6.0%
55-59	14	4.7%	13	4.4%
60-64	28	9.4%	21	7.0%
65-69	53	17.8%	13	4.4%
70-74	49	16.4%	18	6.0%
75-79	46	15.4%	6	2.0%
80-84	44	14.8%	2	0.7%
85+	36	12.1%	2	0.7%
Education
Middle School	251	84.2%	139	46.6%	44	55.0%
High School	47	15.8%	159	53.4%	36	45.0%
Employment
Student	0	0.0%	1	0.3%
Permanent Job	8	2.7%	135	45.3%
Temporary Job	1	0.3%	24	8.1%
Pensioner	239	80.2%	50	16.8%
Unemployed	43	14.4%	40	13.4%
Self-employed	7	2.3%	48	16.1%
Beneficiary of the state program
100% co-financing	14	4.7%	16	5.4%
90% co-financing	240	80.5%	243	81.5%
80% co-financing	6	2.0%	4	1.3%
70% co-financing	38	12.8%	35	11.7%

Group 1 – Patients; Group 2 – Family Members; Group 3 – Healthcare Personnel. Data for employment status and beneficiary of the state program were not collected for Group 3 (Healthcare Personnel), as these variables were not applicable to this group.

Statistical analysis

All subscale and symptom scores of the QLQ-C30-GEO were calculated based on responses from patients, family members, and healthcare personnel. Statistical analyses were performed using SPSS version 22.0 (IBM Corp., Chicago, IL, USA). Continuous variables are presented as mean ± standard deviation (SD), median, and range (minimum–maximum).

Comparisons between groups were conducted using paired t-tests. Multivariate regression models were applied to adjust for potential confounders, including age, gender, cancer type, disease stage, and relevant socio-demographic variables (education level and employment status). Clustering and healthcare personnel experience were also taken into account in the models.

Effect sizes for group comparisons were calculated using Cohen’s d and interpreted according to conventional thresholds (small < 0.2, moderate = 0.5, large ≥ 0.8). Clinical relevance of observed differences was interpreted with reference to published minimally important difference (MID) estimates for the EORTC QLQ-C30 across multiple cancer populations.¹¹

Associations between respondent groups were assessed using Pearson’s correlation analysis. To address multiple comparisons across QLQ-C30 scales and symptom domains, p-values were adjusted using the Benjamini–Hochberg false discovery rate (FDR) method. Adjusted p-values are reported alongside unadjusted values, and statistical significance was determined based on adjusted results (p < 0.05).

Internal consistency of the questionnaire was evaluated using Cronbach’s alpha (CA) with 95% confidence intervals. Values >0.7 were considered acceptable, >0.8 good, and >0.9 excellent.

Results

A total of 298 patient–family–healthcare personnel triads completed the full set of questionnaires, with no missing item-level data. To account for clustering of patients within healthcare personnel, linear mixed-effects models with a random intercept for clinician were applied. These analyses yielded results consistent in both direction and statistical significance with the original paired comparisons, indicating that the main findings are robust to adjustment for clinician-level clustering.

Internal consistency of the QLQ-C30 scales and symptom subscales was assessed using Cronbach’s alpha (α) with 95% confidence intervals across all three groups: patients, family members, and healthcare personnel (Table 3). Cronbach’s alpha values ranged from acceptable to excellent in all groups, indicating that the QLQ-C30 reliably captured HRQoL domains across diverse respondent populations.

Table 3.

Internal consistency of scales and symptoms of the QLQ-30 questionnaire in all groups.

Domain	Cronbach’s alpha	95%CI		Internal consistency
Domain	Cronbach’s alpha	Lower	Upper	Internal consistency
Global Health Status
Group 1 (patients)	0.795	0.747	0.833	Acceptable
Group 2 (family members)	0.773	0.721	0.815	Acceptable
Group 3 (healthcare personnel)	0.880	0.814	0.923	Good
Functional Scales Physical Functioning
Group 1 (patients)	0.747	0.702	0.785	Acceptable
Group 2 (family members)	0.799	0.764	0.830	Acceptable
Group 3 (healthcare personnel)	0.914	0.879	0.939	Excellent
Role Functioning
Group 1 (patients)	0.909	0.888	0.926	Excellent
Group 2 (family members)	0.821	0.779	0.854	Good
Group 3 (healthcare personnel)	0.817	0.716	0.882	Good
Emotional Functioning
Group 1 (patients)	0.761	0.717	0.799	Acceptable
Group 2 (family members)	0.764	0.721	0.801	Acceptable
Group 3 (healthcare personnel)	0.892	0.846	0.925	Good
Cognitive Functioning
Group 1 (patients)	0.766	0.712	0.810	Acceptable
Group 2 (family members)	0.784	0.735	0.824	Acceptable
Group 3 (healthcare personnel)	0.750	0.613	0.839	Acceptable
Social Functioning
Group 1 (patients)	0.955	0.944	0.963	Excellent
Group 2 (family members)	0.881	0.853	0.903	Good
Group 3 (healthcare personnel)	0.982	0.972	0.988	Excellent
Symptom Scales Fatigue
Group 1 (patients)	0.778	0.734	0.814	Acceptable
Group 2 (family members)	0.838	0.806	0.864	Good
Group 3 (healthcare personnel)	0.792	0.696	0.858	Acceptable
Nausea and Vomiting
Group 1 (patients)	0.782	0.732	0.823	Acceptable
Group 2 (family members)	0.768	0.715	0.811	Acceptable
Group 3 (healthcare personnel)	0.818	0.717	0.882	Good
Pain
Group 1 (patients)	0.931	0.915	0.944	Excellent
Group 2 (family members)	0.943	0.930	0.953	Excellent
Group 3 (healthcare personnel)	0.832	0.739	0.892	Good

Adjusted mean scores with corresponding 95% confidence intervals for all QLQ-C30 scales and subscales are presented in Table 4 following adjustment for age, gender, cancer type, disease stage, education, employment status, and clustering at the clinician level. Supplementary Table S1 presents detailed adjusted results with standard errors, while Supplementary Table S2 presents the corresponding unadjusted descriptive scores. Pearson’s coefficients of correlation between these scores and all three groups are presented in Table 5.

Table 4.

Adjusted mean scores of QLQ-C30 scales and subscales (95% CI) across respondent groups.

Scale	Group 1 – Patients			Group 2 – Family members			Group 3 – Healthcare personnel
Scale	Adjusted mean	95%CI		Adjusted mean	95%CI		Adjusted mean	95%CI
GHS	14.1 ^ab	(13.5	14.7)	6.3 ^ac	(5.7	6.9)	58.4 ^bc	(57.1	59.7)
PF	21.4 ^a	(20.5	22.3)	8.2 ^ac	(7.4	9.0)	18.2 ^c	(17.4	19.0)
RF	8.5 ^ab	(7.6	9.4)	2.9 ^ac	(2.3	3.5)	18.0 ^bc	(16.7	19.3)
EF	10.0 ^ab	(9.3	10.7)	4.9 ^ac	(4.0	5.8)	27.0 ^bc	(25.1	28.9)
CF	23.5 ^ab	(22.0	25.0)	10.2 ^ac	(8.8	11.6)	55.5 ^bc	(53.6	57.4)
SF	13.7 ^ab	(12.5	14.9)	5.9 ^ac	(5.1	6.7)	69.2 ^bc	(67.0	71.4)
FA	88.1 ^ab	(87.3	88.9)	96.7 ^ac	(96.0	97.4)	62.1 ^bc	(60.4	63.8)
NV	32.9 ^ab	(31.1	34.7)	43.0 ^ac	(40.9	45.1)	19.7 ^bc	(17.9	21.5)
PA	46.1 ^b	(43.6	48.6)	50.4 ^c	(47.7	53.1)	62.9 ^bc	(61.1	64.7)
DY	75.5 ^ab	(73.5	77.5)	85.0 ^ac	(83.0	87.0)	67.2 ^bc	(65.3	69.1)
SL	84.1 ^ab	(82.7	85.5)	91.6 ^ac	(90.3	92.9)	60.0 ^bc	(57.9	62.1)
AL	71.5 ^ab	(69.4	73.6)	88.6 ^ac	(87.0	90.2)	45.7 ^bc	(43.3	48.1)
CO	79.4 ^b	(77.2	81.6)	82.7 ^c	(80.5	84.9)	37.8 ^bc	(35.2	40.4)
DI	9.4 ^b	(8.0	10.8)	7.7	(6.4	9.0)	5.9 ^b	(4.8	7.0)
FI	71.3 ^ab	(69.4	73.2)	32.2 ^ac	(30.1	34.3)	10.7 ^bc	(9.2	12.2)

^aGroup 1 vs. Group 2 – FDR-adjusted p < 0.05.

^bGroup 1 vs. Group 3 – FDR-adjusted p < 0.05.

^cGroup 2 vs. Group 3 – FDR-adjusted p < 0.05.

Table 5.

Correlation of scores of QLQ-30 scales and symptoms between the groups.

Scale	Group 1 vs. Group 2 Patients vs. Family members		Group 1 vs. Group 3 Patients vs. Healthcare Personnel		Group 2 vs. Group 3 Family members vs. Healthcare Personnel
Scale	R	p-value (adjusted p-value)	R	p-value (adjusted p-value)	R	p-value (adjusted p-value)
Global Health Status – GHS	0.158	0.006 (0.021) *	0.129	0.026 (0.077)	-0.018	0.757 (0.831)
Physical Functioning – PF	0.216	<0.001 (0.001) *	0.594	<0.001 (<0.001) *	0.086	0.140 (0.300)
Role Functioning – RF	-0.018	0.757 (0.852)	0.349	<0.001 (0.001) *	-0.169	0.003 (0.013) *
Emotional Functioning – EF	0.072	0.215 (0.393)	-0.059	0.309 (0.535)	-0.015	0.790 (0.846)
Cognitive Functioning – CF	0.216	<0.001 (0.002) *	-0.061	0.296 (0.533)	0.035	0.543 (0.718)
Social Functioning – SF	0.052	0.371 (0.596)	-0.009	0.878 (0.878)	0.026	0.654 (0.795)
Fatigue – FA	0.049	0.399 (0.602)	0.033	0.570 (0.733)	-0.014	0.814 (0.852)
Nausea and Vomiting – NV	0.430	<0.001 (0.001) *	0.050	0.388 (0.602)	0.074	0.201 (0.393)
Pain – PA	0.760	<0.001 (<0.001) *	0.090	0.120 (0.271)	0.080	0.167 (0.341)
Dyspnea – DY	0.367	<0.001 (0.001) *	-0.050	0.390 (0.585)	-0.011	0.849 (0.868)
Insomnia – SL	0.327	<0.001 (0.001) *	-0.114	0.048 (0.136)	-0.039	0.498 (0.679)
Appetite Loss – AL	0.182	<0.001 (0.007) *	-0.030	0.611 (0.764)	-0.020	0.732 (0.867)
Constipation – CO	0.219	<0.001 (0.001) *	-0.091	0.118 (0.280)	-0.055	0.347 (0.578)
Diarrhea – DI	0.287	<0.001 (0.001) *	0.100	0.086 (0.228)	-0.042	0.469 (0.660)
Financial Difficulties – FI	0.152	<0.001 (0.028) *	-0.094	0.106 (0.265)	-0.019	0.747 (0.861)

^*Correlation was significant. Pearson correlation coefficients (R) are presented for pairwise comparisons between respondent groups: Group 1 vs. Group 2 (Patients vs. Family members), Group 1 vs. Group 3 (Patients vs. Healthcare Personnel), and Group 2 vs. Group 3 (Family members vs. Healthcare Personnel).

Healthcare personnel reported significantly higher Global Health Status (GHS) scores than patients and family members (Cohen’s d = 2.52), exceeding clinical relevance thresholds, while family members reported lower scores than patients (Cohen’s d = 0.77). A significant correlation was observed only between patients and family members (adjusted p = 0.021), with no significant correlations between patients and healthcare personnel (adjusted p = 0.071) or between family members and healthcare personnel (adjusted p = 0.877).

Family members reported significantly lower Physical Functioning (PF) scores than both patients (Cohen’s d = 0.77) and healthcare personnel (Cohen’s d = 0.71), while no meaningful difference was observed between patients and healthcare personnel (Cohen’s d = 0.07). Correlations were strongest between patients and healthcare personnel (adjusted p < 0.001), followed by patients and family members (adjusted p = 0.001), with no significant correlation between family members and healthcare personnel (adjusted p = 0.391).

Family members reported significantly lower Role Functioning (RF) scores than both patients (Cohen’s d = 0.43) and healthcare personnel (Cohen’s d = 0.88), while patient scores were lower than those of healthcare personnel (Cohen’s d = 0.51). Significant correlations were observed between patients and healthcare personnel (adjusted p < 0.001) and between family members and healthcare personnel (adjusted p = 0.008), with no significant correlation between patients and family members (adjusted p = 0.852).

Family members reported significantly lower Emotional Functioning (EF) scores than both patients (Cohen’s d = 0.38) and healthcare personnel (Cohen’s d = 0.87), while patient scores were lower than those of healthcare personnel (Cohen’s d = 0.69). No significant correlations were observed between patients and family members (adjusted p = 0.421), patients and healthcare personnel (adjusted p = 0.537), or family members and healthcare personnel (adjusted p = 0.889).

Family members reported significantly lower Cognitive Functioning (CF) scores than both patients (Cohen’s d = 0.53) and healthcare personnel (Cohen’s d = 1.57), while patient scores were lower than those of healthcare personnel (Cohen’s d = 1.06). A significant correlation was observed only between patients and family members (adjusted p = 0.001), with no significant correlations between patients and healthcare personnel (adjusted p = 0.544) or between family members and healthcare personnel (adjusted p = 0.709).

Family members reported significantly lower Social Functioning (SF) scores than both patients (Cohen’s d = 0.43) and healthcare personnel (Cohen’s d = 2.23), while patient scores were lower than those of healthcare personnel (Cohen’s d = 1.83). No significant correlations were observed between patients and family members (adjusted p = 0.596), patients and healthcare personnel (adjusted p = 0.897), or family members and healthcare personnel (adjusted p = 0.797).

Analysis of symptom subscales between groups showed that the healthcare personnel reported significantly lower Fatigue (FA) scores than both patients (Cohen’s d = 1.14) and family members (Cohen’s d = 1.53), while patient scores were lower than those of family members (Cohen’s d = 0.67). No significant correlations were observed between patients and family members (adjusted p = 0.619), patients and healthcare personnel (adjusted p = 0.733), or family members and healthcare personnel (adjusted p = 0.868).

Healthcare personnel reported significantly lower Nausea/Vomiting (NV) scores than both patients (Cohen’s d = 0.43) and family members (Cohen’s d = 0.71), while family members reported higher scores than patients (Cohen’s d = 0.30). A significant correlation was observed only between patients and family members (adjusted p = 0.001), with no significant correlations between patients and healthcare personnel (adjusted p = 0.614) or between family members and healthcare personnel (adjusted p = 0.416).

Healthcare personnel reported significantly higher Pain (PA) scores than both patients (Cohen’s d = 0.45) and family members (Cohen’s d = 0.32), while no meaningful difference was observed between patients and family members (Cohen’s d = 0.10). A significant correlation was observed only between patients and family members (adjusted p < 0.001), with no significant correlations between patients and healthcare personnel (adjusted p = 0.272) or between family members and healthcare personnel (adjusted p = 0.349).

Healthcare personnel reported significantly lower Dyspnea (DY) scores than both patients (Cohen’s d = 0.25) and family members (Cohen’s d = 0.54), while family members reported higher scores than patients (Cohen’s d = 0.28). A significant correlation was observed only between patients and family members (adjusted p = 0.001), with no significant correlations between patients and healthcare personnel (adjusted p = 0.233) or between family members and healthcare personnel (adjusted p = 0.627).

Healthcare personnel reported significantly lower Insomnia (SL) scores than both patients (Cohen’s d = 0.77) and family members (Cohen’s d = 1.04), while family members reported higher scores than patients (Cohen’s d = 0.31). A significant correlation was observed only between patients and family members (adjusted p = 0.001), with no significant correlations between patients and healthcare personnel (adjusted p = 0.123) or between family members and healthcare personnel (adjusted p = 0.647).

Healthcare personnel reported significantly lower Appetite Loss (AL) scores than both patients (Cohen’s d = 0.67) and family members (Cohen’s d = 1.21), while family members reported higher scores than patients (Cohen’s d = 0.53). A significant correlation was observed only between patients and family members (adjusted p = 0.007), with no significant correlations between patients and healthcare personnel (adjusted p = 0.758) or between family members and healthcare personnel (adjusted p = 0.866).

Healthcare personnel reported significantly lower Constipation (CO) scores than both patients (Cohen’s d = 1.00) and family members (Cohen’s d = 1.09), while no meaningful difference was observed between patients and family members (Cohen’s d = 0.09). A significant correlation was observed only between patients and family members (adjusted p = 0.001), with no significant correlations between patients and healthcare personnel (adjusted p = 0.277) or between family members and healthcare personnel (adjusted p = 0.573).

The Diarrhea (DI) scores did not differ significantly between groups (Cohen’s d: patients vs. family members = 0.07; patients vs. healthcare personnel = 0.16; family members vs. healthcare personnel = 0.08). A significant correlation was observed only between patients and family members (adjusted p = 0.001), with no significant correlations between patients and healthcare personnel (adjusted p = 0.233) or between family members and healthcare personnel (adjusted p = 0.627).

Healthcare personnel reported significantly lower Financial Difficulties (FI) scores than both patients (Cohen’s d = 2.06) and family members (Cohen’s d = 0.68), while family members reported lower scores than patients (Cohen’s d = 1.14). A significant correlation was observed only between patients and family members (adjusted p = 0.028), with no significant correlations between patients and healthcare personnel (adjusted p = 0.263) or between family members and healthcare personnel (adjusted p = 0.858).

Responses to additional questions (Table 6) indicated minimal involvement of non-medical professionals in care across all groups. Interest in integrating psychological support was consistently reported by patients, family members, and healthcare personnel. In contrast, chaplain involvement was less consistently preferred and was more often supported by patients than by family members and healthcare personnel.

Table 6.

The distribution of responses from participants across all three groups to the additional questions for healthcare personnel and to satisfaction with medical and non-medical services.

Question	Group 1 (n=298)		Group 2 (n=298)		Group 3 (n=80)
Question	Mean or N	SD or %	Mean or N	SD or %	Mean or N	SD or %
Is there any specialist other than medical staff involved in the clinic?
No one	297	99.7%	298	100.0%	79	98.8%
Psychologist	0	0.0%	0	0.0%	1	1.2%
Chaplain	1	0.3%	0	0.0%	0	0.0%
Social Worker	0	0.0%	0	0.0%	0	0.0%
Which specialist would you like to include?
Psychologist	229	76.8%	276	92.6%	75	93.8%
Chaplain	69	23.2%	22	7.4%	5	6.3%
Social Worker	0	0.0%	0	0.0%	0	0.0%
Does the team include other specialists besides healthcare personnel?
No one					80	100.0%
Do you practice a multidisciplinary approach in the treatment?
No					80	100.0%
Do you think that a multidisciplinary approach is necessary?
Yes					80	100.0%
Satisfaction with medical services	4.9 ^{* f}	1.2	2.8 ^{* h}	1.1	4.4 ^ft	1.4
1 – very bad	3	1.0%	20	6.7%	0	0.0%
2	4	1.3%	123	41.3%	2	2.5%
3	31	10.4%	89	29.9%	22	27.5%
4	64	21.5%	46	15.4%	21	26.3%
5	97	32.6%	11	3.7%	18	22.5%
6	81	27.2%	7	2.3%	8	10.0%
7 – excellent	18	6.0%	2	0.7%	9	11.3%
Chi2-test, p	Chi2-test=342.0, p<0.001
Satisfaction with non-medical (psychological, social, spiritual) services	1.0 ^f	0.1	1.0 ^h	0.2	2.1 ^f,h	1.7
1 – very bad	297	99.7%	290	97.3%	50	62.5%
2	1	0.3%	7	2.3%	1	1.3%
3	0	0.0%	1	0.3%	12	15.0%
4	0	0.0%	0	0.0%	9	11.3%
5	0	0.0%	0	0.0%	3	3.8%
6	0	0.0%	0	0.0%	3	3.8%
7 – excellent	0	0.0%	0	0.0%	2	2.5%
Chi2-test, p	Chi2-test=221.7, p<0.001

^*Group 1 vs. group 2 – p<0.05.

^fGroup 1 vs. group 3 – p<0.05.

^hGroup 2 vs. group 3 – p<0.05.

Healthcare personnel confirmed the absence of a multidisciplinary approach in the clinics, consistent with reports from patients and family members that no non-medical specialists were involved in care, although such an approach was considered necessary.

Patient-reported satisfaction with medical services was significantly higher than that of both family members (p < 0.001) and healthcare personnel (p = 0.001), while family members reported lower satisfaction than healthcare personnel (p < 0.001).

For non-medical services, only healthcare personnel reported significantly higher satisfaction compared to both patients (p < 0.001) and family members (p < 0.001).

Discussion

Palliative care requires coordinated involvement of patients, family members, and healthcare providers to prevent and relieve suffering and improve patients’ quality of life. Previous studies have shown that these groups may differ in how they perceive, prioritize, and evaluate symptoms, functioning, and overall health status. In this context, the present study assessed HRQoL in patients receiving palliative care in Georgia by integrating perspectives from patients, family members, and healthcare personnel using the QLQ-C30 instrument. This multi-source approach provides a broader understanding of patient well-being.

In palliative care, where physical, emotional, and social dimensions are closely interconnected, early HRQoL assessment may support more accurate evaluation of patient status and more effective, individualized clinical decision-making.¹² The QLQ-C30 demonstrated acceptable internal consistency across patients, family members, and healthcare personnel, indicating reliable assessment of HRQoL domains across respondent groups. However, the interpretation of Cronbach’s alpha differs for proxy respondents. While it reflects the reliability of self-reported experiences in patients, for family members and healthcare personnel, it indicates the coherence of their ratings of the patient’s condition rather than true psychometric reliability. Higher alpha values in proxy groups may be influenced by shared perceptual frameworks or systematic biases and should therefore be interpreted cautiously as descriptive of response consistency.

These findings further support the feasibility of multi-informant HRQoL assessment and are consistent with previous research, reinforcing the potential for integrating patient-reported outcome measures (PROMs) into routine clinical practice.¹³

To strengthen the conceptual grounding of our findings, they can be interpreted through established frameworks in palliative care and public health, including patient-centered care, integrated palliative care models, and implementation science approaches related to person-centered outcome measures (PCOMs). Within this perspective, HRQoL is conceptualized as a multidimensional construct shaped by physical, psychological, and social factors, and delayed integration of palliative care is associated with greater symptom burden and poorer outcomes.¹⁴ The patient-centered care framework emphasizes the importance of incorporating multiple stakeholder perspectives—patients, family members, and healthcare professionals—particularly where proxy assessments are required.¹⁵ Discrepancies observed across these groups may partly reflect inherent limitations of proxy reporting, as agreement is often moderate and varies across domains.^16,17 These differences arise from role-dependent perceptions shaped by clinical focus, emotional involvement, and contextual factors.

Implementation science conceptualizes PCOMs as complex interventions influenced by healthcare professionals’ beliefs, communication practices, and organizational contexts,¹⁸ while integrated care frameworks emphasize coordination, collaboration across clinical disciplines, and communication across care settings to achieve consistent evaluation.¹⁹ From a policy perspective, effective integration of PCOMs requires organizational support and alignment with clinical workflows.^20,21

Within this context, the variability identified in our study reflects both perceptual and system-level factors, enhancing interpretability and situating the findings within the broader public health discourse.

Patient assessment

Patients reported fatigue, along with decreased emotional, role, and social functioning, while cognitive functioning was moderately affected. These findings reflect the multidimensional impact of advanced illness on HRQoL, highlighting domains that may require targeted interventions and aligning with previous findings in palliative care populations.

Family members assessment

Family members generally rated patients’ symptoms as more severe than the patients themselves. Functioning domains, including emotional and cognitive, were also rated lower by family members. This suggests that caregiver perceptions are influenced by emotional burden and responsibility, providing complementary information that can identify patient needs not fully captured by self-reports. These findings highlight the potential role of supportive professionals, such as social workers and psychologists, in assisting caregivers and addressing unmet patient needs.

Healthcare personnel assessment

Healthcare personnel generally rated HRQoL higher, underestimating patients’ physical, role, and emotional impairments. This highlights discrepancies between clinical perception and patient experience, emphasizing the need for structured PROMs to guide accurate, patient-centered care.

The healthcare personnel group included both physicians and nurses, whose perspectives on patient HRQoL may differ. Due to sample size constraints, subgroup analyses were not performed, potentially obscuring within-group variation.

To further explore these discrepancies, comparative analysis across respondent groups was conducted:

Patients vs. family members

Family members generally reported higher symptom severity and perceived greater functional limitations than patients. While agreement was observed for several symptoms, discrepancies persisted across key domains, consistent with prior studies highlighting the influence of caregiver burden.

Patients vs. healthcare personnel

Healthcare personnel consistently rated patients’ HRQoL higher across functional domains compared with patient self-reports, suggesting underestimation of patients’ subjective experiences. Agreement was limited and domain-specific.

Family members vs. healthcare personnel

Family members tended to report greater functional limitations compared with healthcare personnel, with additional differences in symptom evaluation. This divergence reflects emotional and caregiving influences versus more clinically oriented assessments.

In addition, several domains, particularly Global Health Status, Social Functioning, Emotional Functioning, and Financial Difficulties, showed large differences between respondent groups that appeared to exceed published minimally important difference (MID) estimates for the EORTC QLQ-C30, suggesting potentially clinically meaningful discrepancies in HRQoL perception.

Taken together, the differences observed among patients, family members, and healthcare personnel underscore the importance of multi-source assessment in HRQoL research. These differences likely reflect underlying system-level gaps in palliative care delivery, including fragmented communication, limited coordination across disciplines, and variability in patient-centered assessment practices. Family members’ assessments provide critical insights into aspects of patient well-being that may be overlooked by healthcare personnel, particularly when patients cannot self-report.^4,5 However, they may overestimate symptom severity due to emotional burden or fear of loss,^22,23 requiring cautious interpretation.Consistent with prior studies,^6,24 healthcare personnel frequently underestimated patients’ symptom burden in emotional, social, and cognitive domains, while clinical assessments aligned more closely with observable symptoms such as physical functioning and pain. Longitudinal evidence demonstrates that physician evaluations cannot reliably substitute for patient self-reporting in palliative care settings.²⁵

A three-way comparison provides a more complete illustration of patient experience, mitigating the limitations inherent to any single perspective.

In summary, these findings support a multi-source approach, prioritizing patient self-reports while integrating family members and healthcare personnel as complementary perspectives. These results should be interpreted within a broader sociocultural and health-system context.

In many low- and middle-income settings, family members play a central role in decision-making and caregiving, shaping perceptions of patient quality of life. Evidence indicates that family involvement influences care-related judgments and communication, contributing to discrepancies between patient and proxy assessments.²⁶ These differences may reflect not only perceptual variation but also culturally embedded patterns of family involvement and responsibility.

Structural health system factors further contribute to these discrepancies. In LMICs, palliative care delivery is often fragmented and influenced by broader cultural, social, and system-level determinants.^27,28 This is particularly relevant in Georgia, where services remain limited, geographically concentrated, and insufficiently integrated into primary and community care, with coverage below estimated need and limited rural access.⁷

Financing constraints exacerbate these challenges. High out-of-pocket costs, limited state support, and restricted access to essential services contribute to inequities in care delivery.^27,29 In Georgia, constrained program budgets and weak incentives for rural service provision further affect access and continuity of care, potentially widening gaps between patient-reported and proxy assessments.⁷

These patterns are consistent with other LMIC settings, where limited integration, workforce shortages, and system-level barriers constrain comprehensive palliative care delivery.^27,30

The results reveal a structural gap in palliative care provision, with minimal involvement of non-medical specialists. A multidisciplinary approach was not implemented despite its recognized importance, which may contribute to reduced patient functioning and increased concerns among family members.

Our findings indicate a high demand for multidisciplinary care in Georgia, reflecting a disconnect between current practice and contemporary palliative care standards.

International evidence shows that involvement of these professionals improves patient outcomes, while their absence reflects broader systemic challenges. Similar limitations have been reported in other LMIC settings, where insufficient integration and unclear role definitions constrain care delivery.³¹

Most participants across all three groups expressed a preference for psychological support, while patients specifically indicated a need for spiritual care, suggesting unmet psychosocial needs and the importance of culturally appropriate support.

Notably, the role of social workers was not emphasized by participants, possibly reflecting limited role awareness. Evidence suggests that effective integration of social workers depends on role clarity, system integration, and organizational support.³²

Satisfaction ratings further highlight imbalances in care. Patients reported higher satisfaction with medical services, while family members evaluated care more critically, and healthcare personnel were more aligned with patients. Satisfaction with non-medical services was low across all groups, consistent with findings from other LMIC settings.

Overall, these results indicate that although the multidisciplinary care model is recognized as important, its implementation remains limited, reflecting underlying structural and system-level constraints.

These findings highlight important system-level implications for palliative care delivery, emphasizing the need for integrated, patient-centered approaches to HRQoL assessment. Evidence demonstrates that integrated palliative care improves HRQoL and supports more effective,³³ patient-centered services through coordinated experiences across patients, families, and healthcare professionals.³⁴

In Georgia, palliative care remains fragmented and limited due to resource constraints³⁵ and geographic inequalities,⁷ with insufficient integration of multidisciplinary care and limited availability of non-medical services. Current services are predominantly medically focused, with minimal incorporation of psychosocial components.³⁶ Palliative care is state-funded through the “Palliative Care for Incurable Patients” program (since 2006),³⁷ which is restricted to patients with an expected life expectancy of 3–6 months. Reimbursement tariffs (11 GEL per outpatient visit and 75 GEL per inpatient day), established in 2004 and unchanged despite inflation, further constrain service delivery.Addressing these challenges requires practical and policy-oriented actions. Implementation of multidisciplinary care models should focus on integrating structured, team-based pathways within existing hospital and primary care settings, including formal involvement of psychologists, social workers, and palliative care specialists supported by clear referral mechanisms. Workforce development through targeted training in palliative care, communication, and interdisciplinary collaboration is essential.

From a financing perspective, reforming the state-funded program to include non-medical services is critical, through updated reimbursement tariffs and the introduction of bundled or case-based payment models. Piloting multidisciplinary care models may offer a feasible pathway for gradual scale-up.

At the system level, strengthening coordination across hospital, primary, and community-based services through formal referral pathways and shared care protocols is essential to ensure continuity of care. Embedding PCOMs into routine practice may further improve standardized assessment and alignment across patient, family, and healthcare professional perspectives.

Collectively, these represent concrete and actionable strategies to support the implementation of multidisciplinary, patient-centered palliative care in Georgia.

Limitations

This study has several limitations that should be considered when interpreting the findings. First, although the study was multicentric and included palliative care clinics from different geographic regions of Georgia, the participating institutions were selected using convenience sampling based on accessibility and willingness to participate. However, all included clinics operate under the same national state-funded palliative care program and adhere to a relatively uniform organizational, regulatory, and financing framework. This structural consistency enhances comparability across sites and partially mitigates variability related to institutional differences. Nevertheless, this sampling approach may introduce selection bias, potentially resulting in the overrepresentation of more active, higher-volume, and better-resourced facilities with greater organizational capacity and research engagement. As such, participating clinics may differ systematically from non-participating or non-operational services, limiting generalizability, particularly to less active or resource-constrained settings. Within this context, the findings are most appropriately interpreted as reflective of palliative care delivery in functioning inpatient settings in Georgia.

Second, the cross-sectional design of the study does not allow for causal conclusions regarding the relationships among symptom burden, functional impairment, and HRQoL. Longitudinal studies are needed to explore temporal changes in HRQoL and to assess the impact of multidisciplinary interventions over time.

Third, family members’ assessments may have been influenced by caregiver burden, emotional distress, and the nature of their relationship with the patient. Such factors may introduce systematic bias and partially explain discrepancies between family members’ reported and patient-reported outcomes. In addition, at several participating sites, individual healthcare professionals evaluated HRQoL for multiple patients, introducing clustering at the clinician level. Although this dependency was addressed analytically using linear mixed-effects models with a random intercept for clinician, residual clustering and unmeasured clinician-level factors, as well as the relatively small number of healthcare personnel, may still influence the estimates, potentially affecting the independence and precision of clinician-reported scores. Therefore, findings involving healthcare personnel should be interpreted with caution. The supervised questionnaire completion procedure may have reduced the likelihood of missing responses compared with unsupervised self-administration and could potentially have influenced response behavior to a limited extent. Furthermore, detailed information on clinical specialty and patient load was not systematically collected and therefore could not be examined, which may further limit the interpretation of findings related to healthcare personnel. Future studies with larger clinician samples and fully specified multilevel designs are warranted to further validate these results. The use of proxy respondents represents a methodological limitation. Proxy-reported outcomes are not interchangeable with patient self-reports and may be influenced by subjective perceptions. In this study, the proxy-adapted version of the EORTC QLQ-C30 was not formally psychometrically validated in the Georgian population. Although internal consistency was acceptable, this does not establish construct validity. Therefore, proxy-based assessments should be interpreted with caution and considered complementary to patient-reported outcomes.

Fourth, cultural, religious, and country-specific factors may influence both perceptions and prioritizations of HRQoL domains. Consequently, the findings should be interpreted within the sociocultural context of Georgia and may not be directly transferable to settings with different cultural or healthcare norms.

Another limitation is the restricted study population, which included only patients with cancer. While cancer represents a substantial proportion of palliative care recipients, these findings may not fully capture the experiences of patients with other advanced non-communicable diseases. Future research should extend this approach to broader palliative populations to explore similarities and disease-specific differences in HRQoL assessment.

Finally, the study did not include direct evaluations from non-medical members of the palliative care team, such as psychologists, social workers, or spiritual care providers. As these professionals play a central role in holistic palliative care, their absence may limit the comprehensiveness of the assessment of multidisciplinary perspectives.

Conclusions

This study provides the first comprehensive, multidimensional assessment of HRQoL in Georgian palliative care by systematically integrating the perspectives of patients, family members, and healthcare personnel—a critical advancement in this transitional healthcare context. Our findings directly address the study’s aim, revealing marked discrepancies: family members consistently overestimate symptom burden, while healthcare personnel underestimate functional and emotional impairments. These results highlight the limitations of single-source assessments and demonstrate the value of multi-source evaluation and structured PROMs to ensure accurate, patient-centered care.

Despite universal acknowledgment of the importance of psychological, social, and spiritual support, substantial gaps persist in multidisciplinary care. These findings emphasize the urgent need for culturally tailored, multidisciplinary palliative care models that actively involve non-medical specialists—including psychologists, social workers, and spiritual care providers—to meet complex patient needs.

By capturing these unique, triangulated insights in a Georgian context, this study not only establishes a foundation for longitudinal research and structured interventions but also provides actionable guidance for optimizing palliative care delivery, shaping health policy, and advancing patient-centered practice in transitional healthcare systems, including implementation of multidisciplinary care models and integration of PROMs into clinical practice.

Supplemental material

Supplemental material - Palliative care among cancer patients in Georgia: System-level gaps identified through health-related quality of life discrepancies

Supplemental material for Palliative care among cancer patients in Georgia: System-level gaps identified through health-related quality of life discrepancies by Tamriko Bulia, Ioseb Abesadze, David Tananashvili, Pati Dzotsenidze, and Tina Beruchashvili in Journal of Public Health Research.

Supplemental material

Supplemental material - Palliative care among cancer patients in Georgia: System-level gaps identified through health-related quality of life discrepancies

Footnotes

Acknowledgements

We want to thank all patients who participated in this study, their family members, and the healthcare personnel. We also extend our gratitude to the School of Health Sciences at the University of Georgia for their support.

ORCID iD

Tamriko Bulia

Ethical considerations

The study was conducted in accordance with national legislation and the 1964 Helsinki Declaration of Human Rights. The Research Ethics Committee approved the study protocol for Biomedical Studies at the School of Health Sciences, University of Georgia (Study No. 11-13083, Date: 03/07/2024).

Consent to participate

Written informed consent was obtained from all participants prior to enrollment.

Consent for publication

This manuscript does not contain any personal, identifiable information, images, or videos from participants.

Author contributions

All authors contributed to the conceptualization and design of the study, participant recruitment, data collection, statistical analyses, interpretation of results, manuscript drafting, and critical review. All authors approved the final version and are accountable for all aspects of the work.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.*

Use of AI tools declaration

The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.

Supplemental material

Supplemental material for this article is available online.

References

Best

Leget

Goodhead

, et al. An EAPC white paper on multi-disciplinary education for spiritual care in palliative care. BMC Palliat Care 2020; 19(1): 9. https://doi.org/10.1186/s12904-019-0508-4

Demuro

Bratzu

Lorrai

, et al. Quality of life in palliative care: a systematic meta-review of reviews and meta-analyses. Clin Pract Epidemiol Ment Health 2024; 20(Suppl-1): e17450179183857. https://doi.org/10.2174/0117450179183857240226094258

Gabriel

Creedy

Coyne

. A systematic review of psychosocial interventions to improve quality of life of people with cancer and their family caregivers. Nurs Open 2020; 7(5): 1299–1312. https://doi.org/10.1002/nop2.543

Sprangers

Aaronson

. The role of health care providers and significant others in evaluating the quality of life of patients with chronic disease: a review. J Clin Epidemiol 1992; 45(7): 743–760. https://doi.org/10.1016/0895-4356(92)90052-O

Sneeuw

Aaronson

Sprangers

, et al. Comparison of patient and proxy EORTC QLQ-C30 ratings in assessing the quality of life of cancer patients. J Clin Epidemiol 1998; 51(7): 617–631. https://doi.org/10.1016/s0895-4356(98)00040-7

Petersen

Larsen

Pedersen

, et al. Assessing health-related quality of life in palliative care: comparing patient and physician assessments. Eur J Cancer 2006; 42(8): 1159–1166. https://doi.org/10.1016/j.ejca.2006.01.032

Bulia

Beruchashvili

Abesadze

, et al. Assessing palliative care gaps in Georgia: a public health perspective on service distribution. Eur J Public Health 2025; 35(4): ckaf161–1191. https://doi.org/10.1093/eurpub/ckaf161.1191

Dzotsenidze

Bulia

. Influence of unbalanced opioid policy on pain management and home-based palliative care services in Republic of Georgia. Med Res Arch 2025; 13(5): https://doi.org/10.18103/mra.v13i5.6581

Aaronson

Ahmedzai

Bergman

, et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 1993; 85(5): 365–376, PMID: 8433390. https://doi.org/10.1093/jnci/85.5.365

10.

Kuliś

Bottomley

Velikova

, et al. EORTC QUALITY OF LIFE GROUP TRANSLATION PROCEDURE. Fourth Edition, 2017. [Internet] Available from. https://www.eortc.org/app/uploads/sites/2/2018/02/translation_manual_2017.pdf

11.

Breast, Head and Neck, Genito-urinary, Gynecological, Gastro-intestinal, Brain, Lung and Quality of Life Groups Musoro

Coens

Sprangers

MAG

, et al. Minimally important differences for interpreting EORTC QLQ-C30 change scores over time: A synthesis across 21 clinical trials involving nine different cancer types. Eur J Cancer 2023; 188: 171–182. https://doi.org/10.1016/j.ejca.2023.04.027, Epub 2023 May 7. PMID: 37257278.

12.

Vgt

Beyer

Sistermanns

, et al. Symptom burden and palliative care needs of patients with incurable cancer at diagnosis and during the disease course. Oncologist 2021; 26(6): e1058–e1065. https://doi.org/10.1002/onco.13751

13.

Hinz

Wedding

Schulte

, et al. Assessment of quality of life in cancer patients using anchoring vignettes: comparisons between mixed cancer patients, patients receiving palliative care, and the general population. Front Psychol 2025; 16: 1439655. https://doi.org/10.3389/fpsyg.2025.1439655

14.

Gonçalves

Gaudêncio

Cristiano

, et al. Intensity of Symptoms and Perception of Quality of Life on Admission to Palliative Care: Reality of a Portuguese Team. Healthcare 2024; 12(15): 1529–1539. https://doi.org/10.3390/healthcare12151529

15.

Hocaoglu

Oluyase

Smith

, et al. Co-designing a framework to communicate patient-centred outcomes in palliative care: involving patients and the public to reframe understanding. Journal of Patient-Reported Outcomes 2026; 10(1):47.

16.

Scheepens

JCC

Boele

Koekkoek

JAF

. Patient-proxy agreement on health-related quality of life assessment in cancer patients. Current Opinion in Oncology 2025; 37(6): 603–610. https://doi.org/10.1097/cco.0000000000001189

17.

Gandhi

Ang

FJL

Neo

SHS

, et al. Quality of Care for Patients With Advanced Illness Scale: Development, Preference Elicitation, and Evaluation of Measurement Properties. Value in Health 2025; 28(9): 1417–1425. https://doi.org/10.1016/j.jval.2025.05.006

18.

Antunes

Barclay

Kuhn

, et al. Implementing patient-centred outcome measures in palliative care clinical practice. An updated systematic review of facilitators and barriers. BMC Palliative Care 2026; 25: 66.

19.

Chen

Sleeman

Bradshaw

, et al. The Use of Person-Centered Outcome Measures to Support Integrated Palliative Care for Older People: A Systematic Review. Journal of the American Medical Directors Association 2024; 25(8): 105036–105046. https://doi.org/10.1016/j.jamda.2024.105036

20.

Dai

Ding

Chen

, et al. Process evaluation of a person-centred outcome measures-based quality improvement program in a hospital-based palliative care in mainland China. Qual Life Res 2025; 34(9): 2629–2639. https://doi.org/10.1007/s11136-025-03997-w, Epub 2025 May 27. PMID: 40425990; PMCID: PMC12432087.

21.

Bradshaw

Santarelli

Khamis

, et al. Implementing person-centred outcome measures (PCOMs) into routine palliative care: A protocol for a mixed-methods process evaluation of The RESOLVE PCOM Implementation Strategy. BMJ Open 2021; 11(9): e051904. https://doi.org/10.1136/bmjopen-2021-051904

22.

Rigoni

Bruhn

De Cicco

, et al. Quality of life impairment in patients with head and neck cancer and their caregivers: a comparative study. Braz J Otorhinolaryngol 2016; 82(6): 680–686. https://doi.org/10.1016/j.bjorl.2015.12.012

23.

Jones

McPherson

Zimmermann

, et al. Assessing agreement between terminally ill cancer patients' reports of their quality of life and family caregiver and palliative care physician proxy ratings. J Pain Symptom Manage 2011; 42(3): 354–365. https://doi.org/10.1016/j.jpainsymman.2010.11.018

24.

Hladschik-Kermer

Kierner

Heck

, et al. Patients and staff perceptions of cancer patients' quality of life. Eur J Oncol Nurs 2013; 17(1): 70–74. https://doi.org/10.1016/j.ejon.2012.01.005

25.

Petersen

Pedersen

Groenvold

. Does the agreement of patient and physician assessments of health related quality of life in palliative care depend on patient characteristics? Palliat Med 2007; 21(4): 289–294. https://doi.org/10.1177/0269216307077694

26.

Shimul

MMH

Dowel

Roy

, et al. Choices and challenges in end-of-life care and decision-making: a nationwide cross-sectional study in Bangladesh. BMC Palliative Care 2025; 24(1): 283. https://doi.org/10.1186/s12904-025-01926-9

27.

Pushpalata Zanwar

Yalamanchili

, et al. Assessment of needs, challenges, and re-design considerations for culturally sensitive provision and delivery of palliative care supports and services for older adults who prefer to age and die in place in diverse India. Annals of Palliative Medicine 2024; 13(6): 1476-1489. [Internet] Available from. https://apm.amegroups.org/article/view/128520/html

28.

Peeler

Afolabi

Adcock

, et al. Primary palliative care in low- and middle-income countries: A systematic review and thematic synthesis of the evidence for models and outcomes. Palliative medicine 2024; 38(8): 776-789.

29.

Islam

Katapodi

Gulnara

, et al. Improving palliative care outcomes in remote and rural areas of LMICs through family caregivers: lessons from Kazakhstan. Frontiers in Public Health 2023; 11: 1186107.

30.

Bhattacharyya

Hasibul Hossain

Kumar Dutta

, et al. Service coverage and health workforce allocation strategies for geriatric and palliative care in low and middle-income countries: a systematic review. Health Sciences Review 2024; 11: 100166–100176. https://doi.org/10.1016/j.hsr.2024.100166

31.

MacRae

Fazal

O’Donovan

. Community health workers in palliative care provision in low-income and middle-income countries: a systematic scoping review of the literature. BMJ Glob Health 2020; 5(5): e002368. https://doi.org/10.1136/bmjgh-2020-002368

32.

Alanazi

Alotaibi

AMN

Alsulaiman

, et al. The role of social workers in palliative care: a review of prerequisites and practical implementation. J Int Crisis Risk Commun Res 2024; 7(S9): 3081. https://doi.org/10.63278/jicrcr.vi.2139

33.

Assessing the development of palliative care worldwide: a set of actionable indicators . [Internet]. [cited 2026 Jan 24]. Available from. https://iris.who.int/server/api/core/bitstreams/f05157f5-38f4-4069-aec4-fd8cd17a9596/content

34.

Mitchell

Turner

Fryer

, et al. Integration of primary care and palliative care services to improve equality and equity at the end-of-life: Findings from realist stakeholder workshops. Palliative medicine 2024; 38(8): 830–841. https://doi.org/10.1177/02692163241248962

35.

Garralda

Tripodoro

Ling

, et al. EAPC Atlas of Palliative Care in the European Region 2025. [Internet]. Available from. https://palliaweb.nl/getmedia/f7ba7a87-678d-427b-ab6c-5cd5daed6c71/ICS-ATLANTES-Atlas-Europa.pdf

36.

Gogokhia

Dzotsenidze

Kiknadze

, et al. Assessment of palliative care services in Georgia. Open Society Georgia Foundation, 2021. [Internet] [cited 2026 Jan 24].

37.

On Amending the Resolution of the Government of Georgia №474 of 30 December 2024 on the Approval of the 2025 State Health Care Programs. In: სსიპ ”საქართველოს საკანონმდებლო მაცნე”. 2025 [Internet] Available from. https://www.matsne.gov.ge/ka/document/view/6558358?publication=0

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.13 MB

0.29 MB