Developing Kano’s model of customer satisfaction-driven intensive care unit end-of-life care strategies for gastrointestinal cancer patients

Abstract

BACKGROUND

In the intensive care unit (ICU) setting, patients with end-stage gastrointestinal cancer (GIC) frequently bear a heavier symptom burden, see faster functional regression, and struggle with eating impairments. These further increase the demands and complexities of nursing care.

AIM

To discuss the development of Kano model-driven ICU end-of-life care (ICU-EOLC) strategies for patients with GIC and its implications for quality of death (QoD).

METHODS

This study enrolled 115 patients with end-stage GIC admitted to the ICU from June 2021 to June 2024. A Kano model-driven ICU-EOLC protocol was applied to the observation group (n = 65), contrasting with routine care in the control group (n = 50). Pre-intervention and post-intervention comparisons were made in terms of psychological well-being [Hospital Anxiety and Depression Scale (HADS)], quality of life (QoL) [European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core 15 Palliative Care Version (EORTC QLQ-C15-PAL)], QoD (Chinese version of the Quality of Dying and Death in the ICU Scale), death-related attitudes [Revised Death Attitude Profile (DAP-R)], and nursing care satisfaction.

RESULTS

Compared with controls, patients in the observation group scored markedly lower on HADS (all domains), EORTC QLQ-C15-PAL (symptom domain), and DAP-R (escape acceptance, fear of death, and death avoidance), while achieving statistically higher scores on EORTC QLQ-C15-PAL (physical function and overall QoL dimensions), C-QDD-ICU, DAP-R (natural and approaching acceptance), and nursing care satisfaction measures.

CONCLUSION

Kano model-driven ICU-EOLC interventions significantly reduced anxiety and depression in patients with end-stage GIC, and patients’ quality of life, QoD, death-related attitudes, and nursing satisfaction were improved.

Key Words: Kano model; End-of-life care in intensive care unit; Gastrointestinal cancer; Construction of nursing strategies; Quality of death

Core Tip: Focused on end-stage gastrointestinal cancer patients in intensive care unit (ICU), this research centers on constructing a Kano model-driven ICU end-of-life care model and examining how it influences patients’ quality of death. Involving 115 patients, our multi-faceted analysis, which incorporated metrics on psychological wellbeing, life and death quality, death attitudes, and care satisfaction, demonstrated that the Kano model-driven nursing intervention is clinically more advantageous than standard nursing methods for this patient population.

INTRODUCTION

Modern medical advances have enabled various life-sustaining measures, leading many patients with terminal conditions to depend on intensive care unit (ICU) support. However, the inevitability of death persists[1,2]. As indicated by relevant statistics, elderly patients in the ICU have high mortality rates, with approximately 51.0% dying during their ICU stays, 76.0% during hospitalization, 58% at 6 months, and 72% within 1 year[3]. Such an increased death risk among patients in the ICU is significantly affected by three or more organ failures, sepsis upon admission, and other critical factors[4]. Research has revealed the phenomenon of end-of-life isolation among patients in the ICU due to restricted visitation policies, leading to solitary deaths and compromised end-of-life experiences[5]. Additionally, 30% of global cancer diagnoses involve digestive tract tumors, a category associated with 40% of cancer fatalities[6]. Affected individuals frequently experience severe symptomatology, rapid physical deterioration, and feeding difficulties, causing substantially increased care requirements and clinical complexities[7,8]. The clinical analysis in this study was centered on patients in the ICU suffering from terminal gastrointestinal cancer (GIC) to develop tailored nursing protocols for optimizing their end-of-life care (EOLC).

Family members still perceived the quality of death (QoD) among patients in the ICU as inadequate. Further care approach exploration and enhancement are imperative, considering the significant role of nursing interventions in quality of dying and death (QODD)[9]. Currently, research on QODD among patients in the ICU in China is rather limited. Broader awareness and deeper investigation are warranted despite some studies being conducted. Relevant research that investigates dying experiences from diverse viewpoints is scarce, and mixed-methods approaches remain underused[10]. Furthermore, holistic approaches to improve death care quality remain underdeveloped globally. Routine nursing interventions struggle to adequately support patients’ end-of-life experiences, requiring the development of specialized and more effective intervention frameworks[11]. The Kano model-driven nursing approach improves care delivery through a three-phase process: (1) Identifying patient requirements; (2) Categorizing these needs; and (3) Establishing priority levels. This methodology enables precisely customized treatments to fully accommodate individual care needs[12]. Yao et al[13] revealed that 75.1% of patients with chronic obstructive pulmonary disease in Zhejiang Province, China, were willing to adopt Kano model-driven lung rehabilitation. Notably, service pricing and insurance reimbursement ratios were the primary determinants affecting their acceptance of this novel care model, each influencing 96.2% of the patients’ decisions.

Existing literature lacks sufficient exploration of EOLC strategies for patients with GIC in the ICU and how these strategies affect their QoD. This research demonstrates clinical novelty by focusing on this understudied aspect, contributing new knowledge and practical guidance in this field.

MATERIALS AND METHODS

Case selection

This study selected 115 patients with end-stage GIC in the ICU from June 2021 to June 2024. A Kano model-driven ICU-EOLC strategy was employed for the observation group (65 patients), whereas standard care was provided to the control group (50 patients). Eligibility criteria were: (1) Patients with end-stage GIC admitted to the ICU[14]; (2) Age of ≥ 18 years; (3) Severe condition requiring ICU treatment; (4) Estimated survival time under 1 month; and (5) Consciousness and intact verbal communication during the investigation. Exclusion criteria were: (1) Expected duration ≥ 1 month; and (2) Missing or defective clinical data.

Intervention methods

The control group received standard care, primarily involving basic life support (e.g., assistance with repositioning, oral hygiene maintenance), catheter device management, and symptom control measures implementation.

The observation group received the Kano model-driven nursing protocol, implemented as follows. A modified Kano questionnaire was administered to assess patient needs. Attribute-based classification of nursing service items was conducted. Prioritization was subsequently established through quadrant analysis. Essential attributes (top priority) encompassed pain assessment and intervention, in addition to basic life care support. Additional important needs (key areas for improvement), requiring targeted optimization, included psychological counseling services and family-involved caregiving. Desirable but non-essential components comprised personalized wish fulfillment, spiritual care services, and memorial ceremonies (optional enhancements). To ensure effective pain management, a standardized pain management protocol was implemented, with analgesics selected according to dynamic pain assessment. Non-opioids, weak opioids, and strong opioids were administered for scores 1-3 (mild pain), 4-6 (moderate pain), and 7-10 (severe pain), respectively. Basic life care support included tailored oral care adjusted to the patient’s oral pH, 2-hourly repositioning, and pressure-redistribution dressings for bony areas. In parallel, a drainage protocol was developed for incontinence management, comprising indwelling catheters for urinary incontinence and anal drainage for fecal incontinence. Daily 30-minute psychological counseling were provided using cognitive behavioral therapy and mindfulness-based stress reduction techniques to alleviate patients’ distress. Furthermore, regular assessments were conducted to track patients’ emotional fluctuations, thereby facilitating prompt adjustments and targeted psychological support when needed. Regarding family-involved caregiving, basic nursing techniques and first-aid procedures were taught to patients’ relatives. One designated family member was allowed continuous (24/7) accompaniment. Personalized wish fulfillment was facilitated through a “bucket list” initiative, with video recording services made available for documentation. The spiritual care program supported patients in processing key life experiences by applying life retrospective therapy, complemented by individualized soundtracks following their preferences. Regarding the memorial ceremony, a commemorative booklet was created using the patient’s photos and cherished possessions, and a farewell service was arranged according to their wishes.

Data collection and outcome measurement

Psychological state evaluation: The Hospital Anxiety and Depression Scale (HADS) was used to measure psychological well-being. HADS comprised two parallel subscales – anxiety and depression – each containing seven questions scored on a 0-3 scale. Thus, the total possible score for each subscale is 0-21. Scores above seven points indicated clinically relevant symptoms, categorized into mild (8-10 points), moderate (11-14 points), and severe (15-21 points)[15].

Quality of life assessment: The European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core 15 Palliative Care Version (EORTC QLQ-C15-PAL) was used for evaluation. Physical function, symptom (pain, fatigue, nausea/vomiting), and overall quality of life (QoL) domains constitute the primary assessment content. All scoring follows a 0-100 scale. Enhanced QoL is associated with higher scores in physical function and overall QoL dimensions, but with lower scores in symptom domains[16].

QoD determination: The Chinese version of the Quality of Dying and Death in the ICU Scale, a Chinese-modified version of the QODD 3.2 scale, was used to measure QoD. Our team translated and validated this instrument, confirming strong reliability and validity (Cronbach’s α = 0.892, Kaiser-Meyer-Olkin measure = 0.836). The scale retains 21 specific items capturing ICU nurses’ perspectives on patients’ dying and death experiences [5 domains: (1) Personal autonomy; (2) Treatment preferences; (3) Social connections; (4) Perceptions of death and dignity; and (5) Spiritual needs] and four global evaluation items assessing the global assessment of dying experience (e.g., use of life-prolonging interventions, overall dying quality, quality of care provided by healthcare professionals, support from involved caregivers), slightly adjusted for cultural relevance. Nurses rate each patient’s dying process on a 0-10 scale (0: Terrible, 10: Almost perfect), with higher scores indicating superior dying experiences[17].

Death-related attitudes: The Revised Death Attitude Profile (DAP-R), a 32-item instrument, was used to quantify death-related attitudes across five domains: (1) Natural acceptance; (2) Approaching acceptance; (3) Escape acceptance; (4) Fear of death; and (5) Death avoidance. Participants rated statements on a 1-5 scale, with increased scores denoting greater alignment with a given dimension, thereby revealing their genuine psychological perceptions of death and their attitudes toward it[18].

Nursing care satisfaction: A self-developed satisfaction survey was applied to assess nursing satisfaction. The questionnaire adopted a 5-point Likert scale, with responses classified as “very dissatisfied” (1 point), “dissatisfied” (2 points), “fair” (3 points), “satisfied” (4 points), or “very satisfied” (5 points). Total satisfaction was calculated as the proportion of respondents who selected either “satisfied” or “very satisfied” out of the total participants[19].

Statistical analysis

Statistical Package for the Social Sciences version 23.0 was used for data analysis. Categorical variables n (%) were compared using χ² test (two-tailed α = 0.05). Normally distributed continuous data (mean ± SD) were analyzed using paired t-tests (within-group) or repeated-measures analysis of variance (longitudinal), supplemented by Student-Newman-Keuls post hoc tests. P < 0.05 were considered statistically significant (two-tailed).

RESULTS

Patient baseline data

Table 1 shows the baseline characteristics of the observation and control groups. The comparative assessment indicates that the two groups were statistically comparable in terms of age, sex distribution, tumor classification, Acute Physiology and Chronic Health Evaluation II, and Karnofsky Performance Status scores, as well as previous chemotherapy treatments (P > 0.05).

Table 1 Baseline data of participants, n (%).

Indicators	Observation group (n = 65)	Control group (n = 50)	t/χ2	P value
Age (years) (mean ± SD)	56.06 ± 11.26	59.92 ± 10.79	1.856	0.066
Sex			0.070	0.791
Male	38 (58.46)	28 (56.00)
Female	27 (41.54)	22 (44.00)
Tumor classification			0.606	0.895
Gastric cancer	18 (27.69)	15 (30.00)
Rectal cancer	16 (24.62)	13 (26.00)
Colon cancer	14 (21.54)	12 (24.00)
Pancreatic cancer	17 (26.15)	10 (20.00)
Acute Physiology and Chronic Health Evaluation II (points) (mean ± SD)	27.18 ± 6.34	28.18 ± 6.46	0.832	0.407
Karnofsky Performance Status functional status score (points) (mean ± SD)	39.58 ± 11.32	40.40 ± 14.04	0.347	0.729
Previous chemotherapy treatments			0.645	0.422
No	20 (30.77)	12 (24.00)
Yes	45 (69.23)	38 (76.00)

Psychological well-being evaluation

Figure 1 illustrates psychological status quantified with HADS. Baseline anxiety and depression scores did not differ significantly (P > 0.05). Post-intervention, both groups demonstrated statistically significant improvements, with the observation group achieving superior reductions in HADS scores relative to controls (P < 0.05).

Figure 1 Patient psychological outcomes. A: Hospital Anxiety and Depression Scale (HADS)-Anxiety subscale scores pre-intervention and post-intervention; B: HADS-Depression subscale scores pre-intervention and post-intervention. ^aP < 0.05, ^bP < 0.01 vs baseline; ^cP < 0.05 vs control at the same time point. HADS: Hospital Anxiety and Depression Scale.

Analysis of patient QoL

QoL was measured using the EORTC QLQ-C15-PAL scale (Table 2). Baseline assessments revealed no statistically significant differences in physical function, symptom scores (pain, fatigue, and nausea/vomiting), or overall QoL (P > 0.05). Following the nursing intervention, both groups demonstrated improvements in physical function and overall QoL, with the observation group outperforming the control group (P < 0.05). Further, symptom severity decreased significantly, with the observation group reporting lower scores in all symptom categories (P < 0.05).

Table 2 Quality of life of patients in two groups, mean ± SD.

Indicators	Observation group (n = 65)	Control group (n = 50)	t value	P value
Physical function
Before nursing	36.11 ± 10.14	38.52 ± 12.01	1.166	0.246
After nursing	56.89 ± 14.80b	46.26 ± 14.13a	3.894	< 0.001
Symptom scores
Pain
Before nursing	61.31 ± 15.98	65.40 ± 13.71	1.446	0.151
After nursing	37.48 ± 11.42b	48.68 ± 16.97a	4.223	< 0.001
Fatigue
Before nursing	66.43 ± 15.17	66.80 ± 14.45	0.132	0.895
After nursing	51.45 ± 13.12b	59.26 ± 12.92a	3.186	0.002
Nausea/vomiting
Before nursing	51.29 ± 14.21	49.18 ± 16.75	0.730	0.467
After nursing	18.82 ± 9.28b	37.78 ± 14.64a	8.467	< 0.001
Overall quality of life
Before nursing	33.65 ± 11.41	34.68 ± 10.74	0.492	0.624
After nursing	54.49 ± 13.63b	40.62 ± 11.42a	5.797	< 0.001

^aP < 0.05, within the same group compared with before nursing care.

^bP < 0.01, within the same group compared with before nursing care.

Comparative assessment of QoD

The C-QODD-ICU scale was used to measure QoD of the two groups (Table 3). Significant differences favoring the observation group were identified in domains, including personal autonomy, treatment preferences, social connections, perceptions of death and dignity, spiritual needs, use of life-prolonging interventions, overall dying quality, quality of care provided by healthcare professionals, and support from involved caregivers (P < 0.05).

Table 3 Quality of death comparison, mean ± SD.

Score (points)	Observation group (n = 65)	Control group (n = 50)	t value	P value
Personal autonomy	7.45 ± 1.71	5.86 ± 1.75	4.893	< 0.001
Treatment preferences	7.48 ± 1.82	5.84 ± 2.20	4.373	< 0.001
Social connections	6.89 ± 1.73	6.18 ± 1.79	2.149	0.034
Perceptions of death and dignity	7.72 ± 1.71	6.76 ± 1.81	2.910	0.004
Spiritual needs	6.88 ± 1.75	5.96 ± 2.17	2.517	0.013
Use of life-prolonging interventions	7.08 ± 1.80	4.74 ± 1.58	7.283	< 0.001
Overall dying quality	7.65 ± 1.65	6.30 ± 1.71	4.281	< 0.001
Quality of care provided by healthcare professionals	8.09 ± 1.45	7.16 ± 1.83	3.041	0.003
Support from involved caregivers	7.82 ± 1.50	6.54 ± 1.66	4.330	< 0.001

Inter-group comparison of life attitudes

Figure 2 illustrates DAP-R scale results assessing patient life attitudes. Analysis revealed that the observation group scored statistically higher than controls in terms of natural acceptance and approach acceptance, while it obtained notably diminished scores in escape acceptance, fear of death, and death avoidance (P < 0.05).

Figure 2 Revised Death Attitude Profile scale evaluation of patient life attitudes. A: Natural acceptance scores; B: Approaching acceptance scores; C: Escape acceptance scores; D: Fear of death assessment; E: Death avoidance evaluation. ^aP < 0.05, ^bP < 0.01.

Comparison of nursing care satisfaction

Table 4 shows nursing satisfaction outcomes measured using a custom questionnaire. The observation group demonstrated substantially higher overall satisfaction (92.31%) compared with the control group (64.00%, P < 0.001).

Table 4 Nursing satisfaction comparison, n (%).

Indicators	Observation group (n = 65)	Control group (n = 50)	χ2	P value
Very satisfied	42 (64.62)	20 (40.00)
Satisfied	18 (27.69)	12 (24.00)
Fair	5 (7.69)	10 (20.00)
Dissatisfied	0 (0.00)	6 (12.00)
Very dissatisfied	0 (0.00)	2 (4.00)
Overall satisfaction	60 (92.31)	32 (64.00)	14.154	< 0.001

DISCUSSION

This investigation enrolled 115 patients with terminal GIC in the ICU. A novel ICU-EOLC model, structured on the Kano concept, was implemented and assessed against conventional care. Psychological well-being, QoL, QoD, death-related attitudes, and care satisfaction assessments revealed that the Kano-driven approach outperformed standard methods, demonstrating its clinical superiority.

Patients with end-stage GIC admitted to the ICU frequently suffer from severe and incapacitating symptoms, which exacerbate psychological strain and foster negative emotional responses[20]. In this study, Kano model-driven ICU-EOLC interventions were highly effective in reducing anxiety and depressive symptoms in patients with end-stage GIC in the ICU, evidenced by marked reductions in HADS scores (anxiety and depression) post-nursing. This may result from the targeted pain relief and emotional counseling to patients with end-stage GIC in ICU settings under the Kano model-based nursing approach, which promoted emotional well-being by mitigating both somatic symptoms and mental anguish[21].

To accommodate the functional constraints of patients with critically ill patients with end-stage GIC, our study used the EORTC QLQ-C15-PAL for QoL evaluation. This tool’s streamlined design (5-8 minutes administration time) preserved assessment of key symptom clusters (pain, fatigue, and nausea/vomiting) while including global QoL measurement[22]. Subsequent analysis revealed that the Kano-model-driven care protocol effectively attenuated these distressing symptoms, thereby improving overall QoL for patients with terminal GIC in the ICU.

Furthermore, our study on QoD revealed that patients with terminal GIC in ICU experienced better end-of-life outcomes when provided with nursing interventions grounded in the Kano model, with improvements in personal autonomy, treatment preferences, social connections, death and dignity perceptions, spiritual needs, use of life-prolonging interventions, overall dying quality, quality of care provided by healthcare professionals, and support from involved caregivers. This is associated with the precise identification of patient needs through the Kano model-driven ICU-EOLC protocol, which addresses three key demand categories. Results on death-related attitudes indicate that patients with terminal GIC in the ICU exhibit a more positive outlook toward death after receiving this intervention. A possible explanation is the continuous family support (24/7) facilitated by the Kano model, whereby relatives’ cooperation and emotional encouragement help patients cope more effectively with mortality[23,24].

Furthermore, implementation of Kano model-driven ICU-EOLC interventions increased satisfaction rates from 64% to 92.31% among patients with end-stage GIC, demonstrating superior acceptability and recognition compared with conventional care. This improvement may be attributed to the Kano model’s ability to address hierarchical patient needs – ranging from pain control and basic life care support to psychological counseling, family-involved caregiving, personalized wish fulfillment, spiritual care, and memorial services – thereby fostering a sense of being valued and respected[25].

CONCLUSION

To conclude, the use of Kano model-driven ICU-EOLC strategies for patients with terminal GIC contributes to anxiety/depression mitigation, QoL and QoD improvements, higher satisfaction rates, and positive shifts in death acceptance. These merits justify its wider integration into medical practice.