Home-based hospice care alleviates the burden of mental disorders in older patients with advanced cancer

Abstract

BACKGROUND

With global aging accelerating, older patients with advanced cancer face a severe burden of mental disorders, such as anxiety, depression, and sleep disturbances, which is worsened by physical symptoms and psychosocial stress. However, conventional hospital care does not address their holistic needs.

AIM

To explore the effect of home-based hospice care on the burden of mental disorders among older patients with advanced cancer.

METHODS

In this retrospective study, we retrospectively reviewed data from 252 older patients with advanced cancer. The data were extracted from the existing clinical database of a home-based hospice care program (April 2023 to May 2025). Patients were divided into two groups based on the care they actually received: Group A (control) received conventional care; Group B (observation) received home-based hospice care. Pre-existing scores from the following instruments were retrieved from the database and compared between the two groups: Hamilton Anxiety Scale (HAMA), Revised Piper Fatigue Scale (PFS-R), 17-item Hamilton Depression Scale (HAMD-17), Edmonton Symptom Assessment System (ESAS), Pittsburgh Sleep Quality Index (PSQI), and Functional Assessment of Cancer Therapy-General (FACT-G).

RESULTS

Prior to the intervention, HAMA, PFS-R, HAMD-17, ESAS, PSQI, and FACT-G scores showed no statistical differences between the two groups (all P > 0.05). Post-intervention, both groups demonstrated reductions in scores on the first five scales (all P < 0.05), with group B’s scores significantly lower than those of group A (all P < 0.05). For FACT-G, physical scores fell in both groups, while emotional, functional, and social/family well-being scores increased significantly in group B, which also recorded a notably higher total score (all P < 0.05).

CONCLUSION

Home-based hospice care in older patients with advanced cancer results in the alleviation of mental disorder-related psychological burden, improvement in sleep quality, and enhancement of overall quality of life.

Key Words: Home-based hospice care; Older patients; Advanced cancer; Anxiety; Depression; Sleep quality; Quality of life

Core Tip: In older patients with advanced cancer, a high burden of mental disorders and poor sleep quality are common challenges. The present research investigated the clinical utility of home hospice services for this population. Compared with traditional treatment and care approaches, home-based end-of-life care incorporating multidisciplinary nursing, symptom control, psychological support, and caregiver training markedly mitigates anxiety, depressive symptoms, and sleep disturbances, while also contributing to better overall quality of life. This conclusion provides solid evidence and practical guidance for end-of-life care practice in older patients with advanced malignant tumors.

INTRODUCTION

Older patients with stage III or IV advanced cancer not only suffer long-term physical complications such as cancer pain, tumor-related fatigue, and cachexia, but also face multiple psychosocial stressors, including fear of death, impaired body image, reduced social participation, and a heavy financial burden on their families[1]. The combined effects of physical discomfort and psychological distress often lead to persistent anxiety, depressive symptoms, and sleep disorders. If not addressed in a timely manner, these problems may reduce treatment adherence, aggravate physical discomfort, and seriously impair quality of life and dignity at the end of life[2].

Routine hospital care mainly focuses on disease control and physiological symptom management. Owing to limitations in medical resource allocation and environmental factors, it is difficult to comprehensively address the complex and profound psychological, spiritual, and humanistic care needs of terminally ill patients[3].

Home-based hospice care is a continuing care model that integrates medical treatment, nursing, psychological support, and social services. It emphasizes comfort-oriented care for the whole person, the whole family, and the entire care trajectory within the familiar home environment. According to each patients’ individual condition and care needs, counselors, social workers, pharmacists, dietitians, and family caregivers may be flexibly involved. The main goals of this service are to reduce patients’ fear of the end of life and their sense of fragmented care, ensure early management of pain and other symptoms, identify emotional distress in a timely manner, and help patients preserve privacy, family connection, and dignity[4].

Although this concept is clinically rational, data on home hospice care programs in China remain scattered. Most published literature is descriptive, and few studies have used a relatively large sample to systematically track anxiety, depressive symptoms, and sleep status in older adults with advanced cancer. This study reviewed 252 medical records from April 2023 to May 2025 to compare differences in psychological and sleep-related outcomes between patients receiving home hospice care and those receiving conventional care, aiming to provide a reference for daily psycho-oncology practice and particularly to help clinicians identify patients who require more intensive end-of-life support.

MATERIALS AND METHODS

Study design and data source

We performed a retrospective study using the hospital database for its home hospice program. After screening the records against the inclusion and exclusion criteria, propensity score matching was applied to make the two care groups more comparable at baseline. The database contained 252 eligible older patients with advanced cancer treated between April 2023 and May 2025. A logistic model including age, sex, tumor type, and clinical stage generated the probability that a patient would receive home-based hospice care. Patients were then matched 1:1 by nearest neighbor matching with a caliper of 0.05. The final matched cohort included 126 patients in group A [(control) received conventional care] and 126 in group B [(observation) received home-based hospice care]; all standardized differences for matched covariates were less than 0.1. Group A was managed with conventional care, and group B with the home hospice program. The Ethics Committee of Shanxi Province Cancer Hospital reviewed and approved the study.

Inclusion criteria: (1) Patients aged 65-80 years; (2) Diagnosed with stage III or IV malignant tumors based on imaging and histopathological examinations; and (3) Previous receipt of conventional anti-tumor therapy, including radiotherapy or chemotherapy when indicated; and availability of complete medical records for analysis. The required records included the cancer diagnosis and stage, documentation of hospice services, and psychological assessment results.

Participants’ exclusion criteria: (1) Presence of severe dementia, delirium, or other conditions during the observation period that hindered effective psychological assessment or communication; (2) Voluntary engagement in professional psychotherapy or systematic use of psychiatric medications during the observation period; (3) Comorbid severe dysfunction of major organs or hematological diseases; (4) History of drug dependence or diagnosis of alcohol use disorder according to criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition; (5) Lack of awareness of the medical condition; and (6) Expected survival of less than 6 months.

Methods

Group A: Conventional care plan. Patients in group A underwent standardized assessment and outpatient follow-up. During hospitalization or before discharge, nurses assessed pain, sleep, emotional status, nutrition, bowel and urinary function, mobility, exercise tolerance, and fall risk, while clinicians reviewed antitumor treatment and symptom-control medications. On the basis of these findings, the team prepared health education materials, a follow-up schedule, and instructions for clinical deterioration.

Pain management followed the three-step analgesic ladder, with an individualized regimen developed for each patient. Nurses checked patients’ adherence to analgesic medications, monitored symptoms such as constipation, nausea, excessive sedation, and respiratory depression, and instructed caregivers on pain assessment and recording rescue doses. Educational content also included management of oral mucositis, skin protection, fatigue management, activity restriction, indwelling catheter care, prevention of falls and aspiration, dietary management, sleep habits, home safety measures, and division of caregiving responsibilities. If patients developed persistent psychological distress, insomnia, abnormal behaviors, or if caregivers showed signs of burnout, psychiatric or psychological assessment was arranged immediately. When relevant clinical problems exceeded the scope of routine follow-up care, specialist consultations in pain management, nutritional support, and rehabilitation were arranged.

After discharge, patients received regular outpatient and telephone follow-up. We asked in detail about recent changes in symptoms, medication use, adverse reactions, dietary intake, body weight, sleep status, and caregiver burden. Key information was entered into the electronic medical record system. Family members recorded patients’ symptoms and medication use on a daily basis. Any medical warning signs were required to be reported promptly.

Group B: Home hospice service. Module for multidisciplinary collaboration and dynamic monitoring. Patients in group B were cared for by a home-based hospice care team. Oncology and palliative care staff were responsible for core assessments and the development of individualized care plans, pharmacists reviewed medication regimens, and dietitians assessed nutritional intake and changes in body weight. When patients experienced mood fluctuations, family conflict, financial stress, or increasing caregiving burden, counselors or social workers provided timely support. During the first home visit, nurses explained the scope of services to patients and their families, assessed symptoms and care needs, identified primary and backup caregivers, confirmed emergency contact information, and established a home care record for each patient. Care plans were developed after team discussion and adjusted dynamically according to changes in the patient’s condition and the family’s caregiving capacity.

Patients in group B received home visits once weekly and telephone follow-up once every two weeks. When symptom fluctuations or increased care risks occurred, follow-up frequency was increased and urgent home visits were arranged. During follow-up, Hamilton Anxiety Scale (HAMA), 17-item Hamilton Depression Scale (HAMD-17), Pittsburgh Sleep Quality Index (PSQI), Edmonton Symptom Assessment System (ESAS), Revised Piper Fatigue Scale (PFS-R), and Functional Assessment of Cancer Therapy-General (FACT-G) scores were recorded. We also reviewed symptom diaries and medication records daily to identify the exact timing and potential triggers of worsening anxiety, depressed mood, insomnia, pain, or fatigue. Based on the assessment results, individualized care plans were continuously optimized. All adjustments and workflow steps were formally documented, forming a traceable process covering assessment, targeted intervention, follow-up, and reassessment.

We established standardized procedures covering early warning, triage assessment, and referral coordination, including risk checklists and graded response measures. Particular attention was given to critical conditions such as uncontrolled pain, persistent vomiting, progressive dyspnea, altered consciousness, bleeding, fever, chills, and severe dehydration. Once abnormalities were identified, telephone-based assessment was first used to provide timely guidance. Severe complications prompted urgent home visits or referral to hospital care. Standardized handover was implemented before and after referral, with updates on medications, symptom control goals, and key nursing priorities to maintain continuity of information. All referred patients received follow-up within 24 hours. For patients who repeatedly triggered emergency alerts, multidisciplinary consultations were conducted to refine care strategies, and monitoring frequency and nighttime care measures were adjusted according to caregiving capacity and individual risk level.

Physical symptom and sleep rhythm management module. In holistic physical symptom management, the team comprehensively assessed and documented pain location, characteristics, triggers, and relieving factors, adjusted analgesic regimens, and developed emergency response plans. Nurses monitored medication adherence and adverse reactions such as constipation, nausea, drowsiness, and respiratory distress, while instructing caregivers to record the use and effectiveness of rescue medications.

Dyspnea was managed through position adjustment, breathing training, oxygen therapy, and sputum care guidance. Nausea, anorexia, and fatigue were relieved through small frequent meals, oral care, dietary modification, and appropriate medication. For patients at high risk of pressure injuries, turning schedules, skin assessment, and pressure-relief measures were implemented to improve comfort and reduce complications.

Sleep management began with bedside assessment, with priority given to major sleep-disrupting factors such as pain, dyspnea, cough, increased nocturia, and adverse drug reactions. Healthcare professionals instructed caregivers to maintain stable daytime and bedtime routines, reduce prolonged daytime naps, arrange safe daytime activities, and improve environmental factors such as bright light, noise, poor ventilation, and uncomfortable room temperature. Patients were advised to avoid stimulating foods and reduce exposure to electronic screens before bedtime. When sedative medication was considered, physicians assessed its potential benefits against the risks of falls and next-day drowsiness, while nurses simultaneously reinforced safety precautions. Subsequent follow-up frequency and nursing intensity were dynamically adjusted according to PSQI scores and bedside observations by family members.

Psychological, emotional, and spiritual care module. At each follow-up visit, patients’ emotional status and stress levels were routinely assessed to provide psychological comfort, emotional support, and spiritual care. The focus was on identifying potential psychological distress, including helplessness, fear of death, family stress, loneliness, and sleep disorders. Supportive communication and reflective counseling were used to help patients express fear, sadness, anger, and unresolved psychological distress. On the basis of symptom control, we worked with patients to set short-term achievable goals to enhance their sense of control over their illness. When communication barriers existed among family members, family meetings were arranged to further clarify care plans, communication methods, emergency response procedures, and priorities for end-of-life care.

When patients entered the terminal stage, the care plan was further simplified and shifted toward comfort care. Unnecessary routine procedures were reduced, with emphasis placed on bedside care, including oral care, position adjustment, skin protection, secretion management, analgesia, dyspnea relief, and agitation control. Caregivers were taught what changes might occur during the dying process and were given clear instructions on when to call the team or request hospital referral. Patients presenting with severe depressive symptoms or safety concerns would receive psychological or psychiatric intervention, along with practical advice for family members regarding nighttime care and emergency response measures.

After the patient’s death, the medical team continued to provide bereavement follow-up care. Family members’ grief responses, sleep status, appetite, and physical discomfort were assessed. For those who had difficulty returning to daily life, appropriate guidance was provided, along with professional psychological support when necessary. Meanwhile, medical records were reviewed to identify deficiencies in communication, symptom control, and caregiver support, thereby continuously optimizing subsequent home-based hospice care services.

Module for home environment optimization and caregiver empowerment. Caregiver training was based on procedures that could be practically performed at home. The training covered symptom observation, nursing record documentation, medication administration, use of rescue analgesics, operation of oxygen and other equipment, nutritional management, oral care, skin care, and emergency response. Nurses first demonstrated each procedure, then observed caregivers performing the tasks and provided guidance. A concise checklist of key steps was placed at the bedside to help family members verify the main procedures during daily care.

Caregiver stress and financial burden were assessed. When needed, families were helped to access community resources, volunteer support, social work services, or charitable programs, and care arrangements were adjusted to prevent prolonged overburdening of a single caregiver. If family members disagreed on treatment decisions or daily care, we facilitated communication and provided external support resources.

A home safety assessment was completed within 72 hours after discharge. The assessment covered the sickbed, side rails, anti-slip measures, lighting, ventilation, temperature and humidity control, noise level, oxygen therapy and nebulization equipment, medication storage, and risks of pressure injury, falls, aspiration, and infection. Nurses guided caregivers in practicing position adjustment, pressure point inspection, oral and gastrointestinal care, and safe transfer procedures. Before the assessment was completed, the team organized commonly used supplies and emergency equipment, placed contact information in a visible location, documented home safety hazards, and proposed corrective measures.

Observed indicators and scoring criteria

HAMA: The HAMA scale[5] consists of 14 items, with possible total scores ranging from 0 to 56. A score higher than 7 suggests possible anxiety; above 14 confirms anxiety; exceeding 21 indicates significant anxiety; and a score over 29 reflects severe anxiety.

HAMD-17: The HAMD-17 scale[6] includes 17 items, with total scores ranging from 0 to 54. A score below 7 is considered normal, while 7-17 suggests possible depression, 17-24 confirms depression, and scores above 24 indicate severe depression.

PFS-R: The PFS-R scale[7] comprises 22 items divided into four dimensions: Cognitive, sensory, emotional, and behavioral fatigue. Each item is rated on a 0-10 Visual Analog Scale (VAS), with higher scores indicating more severe fatigue.

ESAS: The ESAS[8] assesses 10 core symptoms: Drowsiness, weakness, nausea, pain, shortness of breath, anxiety, depression, insomnia, poor appetite, and poor overall well-being, each rated on a 0-10 VAS, with higher scores reflecting greater symptom severity.

PSQI: The PSQI scale[9] comprises seven components: Sleep quality, sleep latency, sleep duration, sleep efficiency, sleep disturbances, use of sleep medication, and daytime dysfunction, each scored on a scale ranging from 0 to 3, with the cumulative total ranging from 0 to 21. Higher scores indicate poorer sleep quality.

FACT-G: The FACT-G scale[10] comprises four dimensions: Physical well-being (7 items), social/family well-being (7 items), emotional well-being (6 items), and functional well-being (7 items). Each item is scored on a scale of 0-4, with higher scores reflecting better quality of life for the patient.

Statistical analysis

Data organized in Excel were analyzed using SPSS 25.0. Normally distributed measurement data are presented as mean ± SD. Independent and paired samples t-tests were used for between-group and within-group comparisons, respectively. Categorical data were described as n (%) and comparisons were made using the χ² test. Multiple linear regression analysis was employed to identify the factors contributing to improvements in anxiety, depression, and sleep disturbances through home-based palliative care. The reduction scores (pre-intervention score minus post-intervention score) of HAMA, HAMD-17, and PSQI in the observation group were used as dependent variables, and demographic characteristics, disease features, and baseline symptom burden indicators were included. A stepwise regression method was applied to select variables and assess the predictors of improvement. A significance level of α = 0.05 was set as the threshold.

RESULTS

Comparison of baseline data between the two groups

Following 1:1 propensity score matching, the control group (group A) and observation group (group B) each included 126 older patients with advanced cancer. Baseline characteristics included demographic data, disease-specific factors, living arrangements, educational background, and concurrent medical conditions. Furthermore, demographic data included age and sex distribution; disease-specific factors involved clinical staging (stages III and IV) and tumor types (lung, gastric, liver, colorectal, breast, esophageal cancer, etc.); comorbidities included hyperlipidemia, hypertension, type 2 diabetes, and coronary heart disease. Baseline indicators showed no significant between-group differences across all comparisons (all P > 0.05), indicating satisfactory comparability and supporting the validity of follow-up outcome evaluations (Table 1). The median follow-up time for the entire study population was 7.2 months (interquartile range 4.5-10.8 months), and the two groups did not differ significantly in this regard (P > 0.05).

Table 1 Comparison of baseline data between the two groups, mean ± SD/n (%).

Baseline data	Group A (n = 126)	Group B (n = 126)	t/χ2 value	P value
Male/female	67/59	65/61	0.064	0.801
Age (years)	71.14 ± 5.36	70.48 ± 5.89	0.930	0.353
Clinical stage			0.398	0.528
Stage III	57 (45.24)	62 (49.20)
Stage IV	69 (54.76)	64 (50.79)
Cancer type			0.886	0.971
Lung cancer	32 (25.40)	29 (23.02)
Gastric cancer	26 (20.63)	30 (23.81)
Liver cancer	23 (18.25)	21 (16.67)
Colorectal cancer	31 (24.60)	34 (26.98)
Breast cancer	10 (7.94)	8 (6.35)
Esophageal cancer	4 (3.17)	4 (3.17)
Living situation			0.941	0.816
Living alone	10 (7.94)	8 (6.35)
Living with spouse only	51 (40.48)	58 (46.03)
Living with spouse and children	53 (42.06)	50 (39.68)
Other	12 (9.52)	10 (7.94)
Education level			0.862	0.835
Illiterate	12 (9.52)	15 (11.90)
Primary school	44 (34.92)	43 (34.13)
Junior high school	36 (28.57)	39 (30.95)
High school or above	34 (26.98)	29 (23.02)
Comorbidities
Hyperlipidemia	57 (45.24)	52 (41.27)	0.404	0.525
Hypertension	46 (36.51)	53 (42.06)	0.815	0.367
Type 2 diabetes	31 (24.60)	36 (28.57)	0.508	0.476
Coronary heart disease	19 (15.08)	22 (17.46)	0.262	0.609

Group A (control) received conventional care; group B (observation) received home-based hospice care.

Comparison of HAMA and HAMD-17 scores between the two groups

Prior to the intervention, the two groups did not differ significantly in HAMA or HAMD-17 scores, indicating comparable baseline anxiety and depression status. After the implementation of distinct care strategies, both groups showed a statistically significant improvement, i.e., lower scores on both rating scales relative to their own pre-intervention levels, confirming the beneficial role of either care model in alleviating anxious and depressive symptoms. Moreover, group B achieved significantly better outcomes than group A at the post-intervention time point, supporting the conclusion that home-based hospice care is more effective in reducing the burden of anxiety and depression in older adults with advanced cancer (Table 2).

Table 2 Comparison of Hamilton Anxiety scale and Hamilton Depression scale scores between the two groups, mean ± SD.

Group	n	HAMA		HAMD
		Before intervention	After intervention	Before intervention	After intervention
Group A	126	18.85 ± 2.96	14.24 ± 2.12a	17.11 ± 3.42	12.74 ± 2.69a
Group B	126	19.01 ± 2.87	11.78 ± 1.63a	17.05 ± 3.56	10.15 ± 2.21a
t value		-0.436	10.326	0.136	8.351
P value		0.663	0.000	0.892	0.000

^aP < 0.05.

P compared with before the intervention. Group A (control) received conventional care; group B (observation) received home-based hospice care. HAMA: Hamilton Anxiety Scale; HAMD: Hamilton Depression Scale.

Comparison of PFS-R scale scores between the two groups

Prior to the intervention, the two groups did not differ significantly in the cognitive, sensory, emotional, or behavioral fatigue scores of the PFS-R, indicating comparable baseline fatigue levels. After the implementation of different care regimens, both groups showed significant within-group reductions in all four subscale scores relative to pre-intervention measures, suggesting that either approach can effectively combat fatigue. Nevertheless, group B had consistently larger decreases than group A across every fatigue dimension, with the differences achieving statistical significance. Thus, home-based hospice care proves more effective than conventional care in improving multidimensional fatigue (cognitive, sensory, emotional, and behavioral) in older patients with advanced cancer (Table 3).

Table 3 Comparison of Revised Piper Fatigue Scale scores between the two groups, mean ± SD.

Group	n	Cognition		Sense		Emotion		Behavior
		Before intervention	After intervention	Before intervention	After intervention	Before intervention	After intervention	Before intervention	After intervention
Group A	126	6.36 ± 0.74	5.14 ± 0.52a	5.93 ± 0.78	4.86 ± 0.64a	6.47 ± 0.69	5.42 ± 0.59a	5.67 ± 0.65	4.02 ± 0.42a
Group B	126	6.29 ± 0.81	4.55 ± 0.68a	6.02 ± 0.74	4.47 ± 0.60a	6.52 ± 0.71	4.60 ± 0.61a	5.59 ± 0.71	3.54 ± 0.50a
t value		0.716	7.737	-0.940	4.990	-0.567	10.846	0.933	8.251
P value		0.475	0.000	0.348	0.000	0.571	0.000	0.352	0.000

^aP < 0.05.

P compared with before the intervention. Group A (control) received conventional care; group B (observation) received home-based hospice care.

Comparison of ESAS scores between the two groups

Prior to intervention, the two groups did not differ significantly in the ESAS scores for drowsiness, weakness, nausea, pain, shortness of breath, anxiety, depression, insomnia, poor appetite, or impaired overall well-being, reflecting comparable baseline symptom status and confirming similar initial symptom profiles. Following the distinct care interventions, both groups demonstrated significant declines in all ten symptom scores from pre-intervention levels, indicating that each care model can alleviate a range of physical and psychological complaints. Notably, group B achieved more pronounced improvements than group A on every symptom item, with the between-group differences being statistically significant. This evidence supports that home-based hospice care is more effective in comprehensively reducing multiple symptoms among elderly patients with advanced cancer (Table 4).

Table 4 Comparison of Edmonton Symptom Assessment System scores between the two groups, mean ± SD.

Dimension	Time	Group A (n = 126)	Group B (n = 126)	t value	P value
Drowsiness	Before intervention	5.92 ± 0.74	6.01 ± 0.74	-0.965	0.335
	After intervention	3.45 ± 0.50a	3.07 ± 0.38a	6.792	0.000
Weakness	Before intervention	5.98 ± 0.78	5.87 ± 0.82	1.091	0.276
	After intervention	4.56 ± 0.64a	3.98 ± 0.27a	9.373	0.000
Nausea	Before intervention	4.54 ± 0.63	4.46 ± 0.67	0.976	0.330
	After intervention	2.91 ± 0.77a	2.44 ± 0.61a	5.371	0.000
Pain	Before intervention	6.15 ± 0.77	6.05 ± 0.70	1.079	0.282
	After intervention	4.41 ± 0.62a	3.95 ± 0.40a	6.998	0.000
Shortness of breath	Before intervention	3.57 ± 0.63	3.59 ± 0.60	-0.258	0.797
	After intervention	2.02 ± 0.31a	1.62 ± 0.49a	7.744	0.000
Anxiety	Before intervention	6.40 ± 0.92	6.33 ± 0.86	0.624	0.533
	After intervention	3.43 ± 0.60a	2.42 ± 0.50a	14.516	0.000
Depression	Before intervention	6.37 ± 0.94	6.44 ± 0.98	-0.579	0.563
	After intervention	3.59 ± 0.62a	2.92 ± 0.40a	10.193	0.000
Insomnia	Before intervention	6.48 ± 0.92	6.56 ± 0.93	-0.686	0.493
	After intervention	4.54 ± 0.62a	4.00 ± 0.36a	8.455	0.000
Poor appetite	Before intervention	5.93 ± 0.65	5.99 ± 0.75	-0.679	0.498
	After intervention	4.52 ± 0.60a	4.03 ± 0.40a	7.627	0.000
Impaired overall well-being	Before intervention	5.54 ± 0.63	5.50 ± 0.58	0.524	0.601
	After intervention	4.49 ± 0.50a	3.96 ± 0.41a	9.201	0.000

^aP < 0.05.

P compared with before the intervention. Group A (control) received conventional care; group B (observation) received home-based hospice care.

Comparison of PSQI scale scores between the two groups

Pre-intervention analysis of the seven components of the PSQI (sleep quality, sleep latency, sleep duration, sleep efficiency, sleep disturbances, use of sleep medication, daytime dysfunction) revealed no statistically significant differences between the two groups, indicating that their initial sleep status was similar. After the intervention, the scores of all seven components in both groups decreased significantly compared with before the intervention, revealing that both care plans were conducive to improving patients’ sleep quality. However, group B had significantly lower scores in each sleep-related component than did group A after the intervention, and the difference was statistically significant, indicating that the care approach in group B mode was more effective in optimizing sleep quality, shortening sleep latency, prolonging sleep duration, improving sleep efficiency, reducing sleep disturbances and the use of sleep medication, and alleviating daytime dysfunction (Table 5).

Table 5 Comparison of Pittsburgh Sleep Quality Index scale scores between the two groups, mean ± SD.

Dimension	Time	Group A (n = 126)	Group B (n = 126)	t value	P value
Sleep quality	Before intervention	2.42 ± 0.50	2.54 ± 0.50	-1.905	0.058
	After intervention	1.51 ± 0.62a	1.00 ± 0.47a	7.358	0.000
Sleep latency	Before intervention	2.01 ± 0.37	1.99 ± 0.37	0.429	0.668
	After intervention	1.05 ± 0.42a	0.57 ± 0.50a	8.251	0.000
Sleep duration	Before intervention	2.49 ± 0.50	2.48 ± 0.50	0.159	0.874
	After intervention	1.42 ± 0.61a	1.06 ± 0.34a	5.786	0.000
Sleep efficiency	Before intervention	2.46 ± 0.50	2.52 ± 0.50	-0.952	0.342
	After intervention	1.95 ± 0.40a	1.53 ± 0.64a	7.734	0.000
Sleep disorders	Before intervention	2.42 ± 0.50	2.54 ± 0.50	-1.905	0.058
	After intervention	2.02 ± 0.46a	1.60 ± 0.59a	6.302	0.000
Hypnotic drugs	Before intervention	2.02 ± 0.39	1.98 ± 0.35	0.857	0.392
	After intervention	1.50 ± 0.50	1.02 ± 0.44	8.090	0.000
Daytime function	Before intervention	2.51 ± 0.50	2.53 ± 0.50	-0.317	0.751
	After intervention	1.97 ± 0.38a	1.51 ± 0.50a	8.222	0.000

^aP < 0.05.

P compared with before the intervention. Group A (control) received conventional care; group B (observation) received home-based hospice care.

Comparison of FACT-G scale scores between the two groups

Before the intervention, no statistically significant difference was observed in the FACT-G scale scores between the two groups, indicating that their initial quality of life was comparable. After the intervention, the emotional well-being and functional well-being scores of both groups increased significantly compared with before the intervention, while the physical well-being scores of both groups decreased. However, although the social/family well-being scores of group B increased significantly, no significant change was observed in group A. This indicates that although the physical function of patients in both groups revealed a declining trend, the care mode of group B could better improve patients’ emotional state, functional status, and social/family adaptation, thereby comprehensively enhancing the overall quality of life of patients (Table 6).

Table 6 Comparison of Functional Assessment of Cancer Therapy-General scores between the two groups, mean ± SD.

Group	n	Physical		Social/family		Emotion		Function
		Before intervention	After intervention	Before intervention	After intervention	Before intervention	After intervention	Before intervention	After intervention
Group A	126	11.02 ± 1.44	9.98 ± 1.13a	9.56 ± 0.98	9.43 ± 0.93	7.52 ± 0.93	9.99 ± 1.08a	6.52 ± 0.63	8.06 ± 0.76a
Group B	126	11.25 ± 1.41	10.43 ± 0.92a	9.46 ± 0.88	13.74 ± 1.28a	7.43 ± 0.95	13.59 ± 1.40a	6.59 ± 0.54	9.04 ± 0.75a
t value		-1.281	-3.467	0.852	-30.578	0.760	-22.854	-0.947	-10.302
P value		0.201	0.001	0.395	0.000	0.448	0.000	0.345	0.000

^aP < 0.05.

P compared with before the intervention. Group A (control) received conventional care; group B (observation) received home-based hospice care.

Multivariate analysis of factors influencing the improvement of mental disorder burden

Multiple linear regression analysis was conducted for group B, with the reduction in HAMA score (Δ HAMA), reduction in HAMD-17 score (Δ HAMD-17), and reduction in total PSQI score (Δ PSQI) as dependent variables, and demographic characteristics, disease features, baseline burden, and intervention elements as independent variables. Model 1 revealed that higher pre-intervention FACT-G social/family well-being scores and higher caregiver training completion were significantly associated with greater improvement in anxiety, whereas higher pre-intervention ESAS pain scores limited the extent of anxiety improvement.

Model 2 indicated that clinical stage IV and living alone were independent risk factors for reduced improvement in depression, while an education level of high school or above was an independent protective factor for depressive symptom improvement. Model 3 revealed that achieving pain control during the intervention period was an independent protective factor for improved sleep quality, whereas pre-intervention ESAS pain scores and total PFS-R scores were independent risk factors for sleep improvement (Table 7).

Table 7 Results of multiple linear regression analysis on factors influencing the improvement of mental disorder burden (n = 126).

Independent variables	Model 1 Δ HAMA		Model 2 Δ HAMD		Model 3 Δ PSQI
	β (95%CI)	P value	β (95%CI)	P value	β (95%CI)	P value
Age	-0.105 (-0.420 to 0.210)	0.483	-0.084 (-0.389 to 0.221)	0.601	-0.053 (-0.283 to 0.177)	0.661
Sex	0.085 (-0.224 to 0.394)	0.562	0.117 (-0.185 to 0.419)	0.441	0.041 (-0.187 to 0.269)	0.718
Clinical stage	-0.151 (-0.464 to 0.162)	0.339	-0.250 (-0.564 to -0.010)	0.041	-0.125 (-0.357 to 0.107)	0.262
Living status	-0.179 (-0.492 to 0.134)	0.253	-0.310 (-0.618 to -0.046)	0.021	-0.142 (-0.374 to 0.090)	0.230
Education level	0.187 (-0.121 to 0.495)	0.226	0.351 (0.040 to 0.662)	0.027	0.147 (-0.082 to 0.376)	0.203
Pre-intervention ESAS pain score	-0.283 (-0.594 to -0.029)	0.028	-0.162 (-0.473 to 0.149)	0.311	-0.331 (-0.562 to -0.100)	0.005
Pre-intervention FACT-G social/family well-being score	0.316 (0.006 to 0.626)	0.045	0.267 (-0.043 to 0.577)	0.087	0.176 (-0.051 to 0.403)	0.124
Pre-intervention total PFS-R score	-0.214 (-0.525 to 0.097)	0.185	-0.223 (-0.534 to 0.088)	0.157	-0.287 (-0.518 to -0.056)	0.014
Caregiver training completion	0.374 (0.062 to 0.686)	0.020	0.232 (-0.084 to 0.548)	0.142	0.201 (-0.033 to 0.435)	0.092
Pain control target reached	0.243 (-0.071 to 0.557)	0.128	0.203 (-0.113 to 0.519)	0.206	0.408 (0.176 to 0.640)	0.001
Model statistics	R2 = 0.361, after adjustment R2 = 0.294		R2 = 0.332, after adjustment R2 = 0.264		R2 = 0.398, after adjustment R2 = 0.332
	F = 5.115, P < 0.001		F = 4.475, P < 0.001		F = 6.013, P < 0.001

Δ Hamilton Anxiety Scale/Δ Hamilton Depression Scale/Δ Pittsburgh Sleep Quality Index, pre-intervention score minus post-intervention score. β, standardized regression coefficient, the absolute value indicates the relative influence of the independent variable on the dependent variable. Age, actual value. Sex, male = 1, female = 0. Clinical stage, stage Ⅲ = 0, stage IV = 1. Living status, not-living alone = 0, living alone = 1. Education level, below high school = 0, high school or above = 1. Pre-intervention Edmonton Symptom Assessment System pain score, actual value. Pre-intervention Functional Assessment of Cancer Therapy-General social/family well-being score, actual value. Pre-intervention total Revised Piper Fatigue Scale score, actual value. Caregiver training completion, not all training completed = 0, all training completed = 1. Pain control target reached, not reached = 0, reached = 1. HAMA: Hamilton Anxiety Scale; HAMD: Hamilton Depression Scale; PSQI: Pittsburgh Sleep Quality Index; CI: Confidence interval; ESAS: Edmonton Symptom Assessment System; FACT-G: Functional Assessment of Cancer Therapy-General; PFS-R: Revised Piper Fatigue Scale.

DISCUSSION

Fear of death, persistent physical suffering, and loss of social roles frequently underlie mental disorder burdens in patients with advanced cancer. Sustained psychological stress in this setting may maintain activation of the physiological stress response, which can in turn exacerbate emotional disturbances. A systematic review by Hwang et al[11] revealed that home-based supportive care providing emotional support and psychological interventions effectively improves emotional functioning in advanced cancer patients. Additionally, a randomized controlled trial by Hosseini et al[12] demonstrated that community-oriented hospice care significantly relieves psychological distress in terminally ill patients. The current study’s results align with these reports, showing that group B had markedly lower post-intervention HAMA and HAMD-17 scores than group A, with a more pronounced reduction. The psychological, emotional, and spiritual care module employed in group B likely assisted patients in regaining perceived control over their lives and, through structured communication and meaning-centered reflection, attenuated their psychological overreaction to fear stimuli. Moreover, the home environment itself has a nonmedical, soothing quality that can mitigate psychological distress. Together with regular professional team visits, this approach reduces patients’ anxiety related to fears of being abandoned by the healthcare system. Furthermore, effective control of physical symptoms, notably pain, in group B also interrupted the pain-anxiety cycle, establishing physiological conditions supportive of emotional improvement. One hypothesis proposes that these psychosocial interventions may operate via downregulation of the hypothalamic-pituitary-adrenal axis, although this awaits confirmation from studies using direct biomarker measurements[13].

Cancer-related fatigue is driven by a range of central and peripheral influences, including inflammation, nutritional status, and diminished physical activity. It also has close associations with depression, anxiety, and sleep quality. Therefore, fatigue relief requires sustained, comprehensive nursing care. Research on nursing-directed home-based hospice care has underscored the pivotal role of nurses in symptom assessment, complex care procedures, and caregiver instruction, which assists in integrating fatigue and functional dependence into a manageable daily care regimen[14]. Concurrently, Valero-Cantero et al[15] demonstrated that fatigue was one of the most severe symptoms in advanced cancer patients and revealed a significant negative correlation with care satisfaction and quality of life. In the current investigation, after the intervention, both groups showed declines in all four PFS-R dimension scores, with a more marked reduction in group B. This aligns with previous work indicating that nurse-led home management is beneficial for reducing symptom cluster burden. In group B, fatigue progression was systematically tracked throughout follow-up, and causative factors were identified via medication reconciliation and symptom diaries, which helped reconceptualize fatigue as an actively manageable phenomenon rather than an inevitable outcome of tumor advancement. Precise pain titration, control of nausea and appetite, and prevention of constipation contributed to decreased energy expenditure and fewer nighttime awakenings, thereby providing a physiological basis for fatigue relief. Simplified exercise routines and postural management within the home environment kept physical activity within tolerable limits, alleviating the emotional distress associated with prolonged bed rest. Incorporating short-term, feasible goals into the care plan helped improve patients’ avoidance behaviors and low motivation in the behavioral dimension, while concurrent symptom improvement further reduced emotional distress[16].

In advanced cancer, symptoms such as pain, dyspnea, and nausea tend to appear in clusters, and through shared neural pathways, they amplify one another, forming difficult-to-manage symptom constellations. A systematic review by Feliciano and Reis-Pina[17] concluded that while home-based hospice care improves overall quality of life, it does not always produce statistically significant differences in specific symptom management outcomes relative to standard care. Goldstein et al[18] demonstrated that home-based care teams primarily composed of nurses and social workers substantially reduced symptom burden. In the current study, group B showed greater improvement than group A on all ten ESAS core symptoms, with especially pronounced benefits for pain relief, anxiety reduction, and alleviation of insomnia. This finding is in line with Goldstein et al[18], but only partially consistent with Feliciano and Reis-Pina[17]. The discrepancy may reflect the implementation of a more aggressive early warning triage and telephone-based assessment system in our study. Conventional outpatient follow-up is prone to delays, whereas the intensive, real-time follow-up approach in group B allowed the medical team to respond instantly to breakthrough pain or acute dyspnea, thereby preventing peripheral nociceptive signals from advancing to central sensitization in a timely manner. Moreover, dietary adjustments and physical interventions performed in the home setting offered complementary support to pharmacological treatment, effectively minimizing additional discomfort caused by drug side effects and enhancing the comprehensive benefits of symptom management[19].

Sleep disturbances in older patients with cancer not only represent a physiological issue but also reflect disrupted circadian rhythms and neurotransmitter imbalances. Prolonged sleep deprivation can impair emotional regulation and amplify pain perception, creating a vicious cycle that worsens psychological distress. These findings corroborate those of Baykal et al[20] and further validate the clinical importance of environmental and behavioral strategies. Hospital noise, lighting, and nighttime nursing procedures often severely compromise sleep microstructure. In contrast, the home-based hospice care protocol includes sleep hygiene and circadian rhythm maintenance strategies, regular daily routines, limited daytime napping, and optimized bedroom lighting, which help realign the body’s natural day-night cycle. Of particular note, targeted management of nocturia, pain, and coughing addresses the underlying causes of sleep fragmentation. When combined with pre-sleep relaxation techniques that lower cortical arousal, these interventions enhance the physiological functioning of restorative sleep while simultaneously reducing dependence on benzodiazepine medications[21].

Quality of life measurement reflects how patients subjectively perceive their overall functional status, incorporating emotional regulation, social role engagement, physical symptom experience, and daily activity capacity. FACT-G data revealed that group B achieved considerable improvements in emotional, social/family, and functional well-being, whereas physical well-being declined in both groups without a statistically significant between-group difference. In a randomized controlled trial of patients with gastrointestinal malignancies, Bojesson et al[22] demonstrated that early integration of specialized hospice care raised total FACT-G scores by 13 points at 24 weeks. A meta-analysis by Rupang et al[23] also concluded that home-based hospice care enhances overall quality of life in cancer patients. The current results partially mirror these earlier findings, with consistent improvements observed in emotional and functional domains. However, no meaningful benefit emerged in the physical domain, which might be attributable to the heterogeneity of cancer types and the predominance of advanced stages in the sample. Furthermore, the average age exceeded 70 years, and conditions like sarcopenia and cancer cachexia are not easily reversible through hospice care alone. Thus, the steady decline in physical function is consistent with the natural evolution of the disease. In contrast, the home-based model substantially strengthens the social/family dimension by enhancing the visibility of family support and through caregiver training, potentially producing a social buffering effect likely by mitigating stress arising from social isolation. The emotional dimension may gain from regular input by psychological professionals and meaning-centered reflection techniques, which help patients re-frame and dampen the intensity of negative emotional experiences. The increased functional dimension probably reflects a greater patient inclination toward self-care once symptom burden diminishes, rather than an actual recovery of physical function. In this study, group A received only outpatient follow-up without mobilization of the family support system, resulting in only limited quality of life improvement among those patients.

Results from the multifactorial regression indicated that higher baseline scores on the FACT-G social/family well-being domain and completion of caregiver training were both significantly and positively linked to greater anxiety reduction, whereas higher pre-intervention pain scores limited the magnitude of anxiety improvement. This pattern suggests that better family support and more competent caregivers may alleviate anxiety by lowering uncertainty and helplessness while enhancing the consistency of symptom management. In contrast, a heavy pain burden may represent a chronic stressor that promotes heightened vigilance and catastrophic appraisals, thereby increasing the difficulty of anxiety interventions. Model 2 further demonstrated that stage IV disease and living alone were independent risk factors for poorer depressive symptom improvement, while having a high school education or above was an independent protective factor. Individuals with stage IV cancer typically endure more intense symptom clusters and a greater sense of end-of-life threat. Living alone reduces daily emotional interactions and caregiving resource access. Both of these conditions can worsen feelings of hopelessness and social isolation, making depression more difficult to treat. Patients with higher educational attainment may have an edge in comprehending disease-related information, selecting appropriate coping strategies, and mobilizing resources, all of which contribute to better depression outcomes.

Model 3 showed that successful pain control during the intervention period was an independent protective factor for improved sleep quality, whereas pre-intervention pain scores and total PFS-R scores were independent risk factors for poorer sleep improvement. This indicates that sufficient pain relief is a key prerequisite for enhancing sleep. Nevertheless, a heavy fatigue burden may reduce responsiveness to sleep interventions by disrupting circadian rhythms, decreasing daytime activity, and altering arousal thresholds. Age and sex were not significant in this model, suggesting that in a study sample with a relatively homogeneous age distribution and advanced disease, psychosocial outcome differences are more likely driven by modifiable factors including symptom burden, social support, and caregiver capacity.

These observations carry significant clinical implications, as they strongly suggest that the treatment effect of home-based hospice care is not uniform across patients. The level of benefit appears to depend heavily on the patient’s social support structure. To illustrate, those who live alone or have stage IV disease seem more at risk and derive less improvement in depressive symptoms from the present model. This finding parallels the improvement observed in the social/family well-being domain, reinforcing that social support acts as a mediating mechanism. Consequently, for patients who lack adequate family support, the intervention needs to be actively intensified. Approaches such as augmenting volunteer visits, leveraging telehealth for daily emotional check-ins, and offering more targeted respite care for exhausted caregivers could specifically address the deficits in these high-risk subgroups, thereby enhancing both the fairness and the efficacy of the care model.

A number of limitations should be kept in mind when drawing conclusions from this study. First, the retrospective approach did not allow us to prospectively measure or adjust for certain known confounders, including a history of mental illness and the quality of pre-existing family relationships, which may have created residual confounding. Second, the lack of random assignment to the intervention gives rise to possible self-selection bias. Individuals who voluntarily opt for and remain in a home-based hospice program may have inherently greater self-efficacy, more active coping strategies, and more robust family support. These unmeasured factors are protective with respect to mental health and could partially enhance the perceived positive effects of the intervention. Third, this was a single-center study with 252 cases, which restricts the generalizability of our findings. Fourth, patients’ limited life expectancy shortened the available observation period, preventing longer-term follow-up. To confirm our results and further investigate the subgroup differences we observed, future work should employ multicenter prospective designs. Additionally, as noted by Lin and Chu[24] and Cruz et al[25], resource constraints, cultural hurdles, and the intricate process of transitioning from hospital to home continue to pose significant challenges, underscoring the urgent need for methodologically sound and context-sensitive future investigations.

CONCLUSION

Home-based hospice care is effective in reducing the psychological impact of mental disorders, promoting sleep quality, and improving the overall quality of life among older patients with advanced cancer.