Nurse-led mobile health symptom management in patients with head and neck cancer undergoing concurrent chemoradiotherapy: a randomized controlled trial

Abstract

Background

Early identification and timely intervention of radiotherapy‑related toxicities are crucial for maintaining treatment adherence and improving quality of life in patients with head and neck cancer (HNC) undergoing concurrent chemoradiotherapy (CCRT). To address this clinical need, a nurse‑led smartphone application, Sun Yat-sen Symptom Management (SYS-SM), was developed to integrate patient‑reported outcome (PRO) assessments with personalized self‑management guidance tailored to individual symptom profiles. The feasibility and preliminary efficacy of this application were evaluated in a randomized controlled trial (RCT).

Methods

Eligible patients were randomly assigned to an intervention group or a control group. During the 3-month intervention period, the intervention group received standard care combined with nurse-led personalized mobile symptom management, whereas the control group received standard care alone. The intervention included in-person discharge education, structured instruction on application use, and automated push notifications delivering individualized health management recommendations based on PRO data. When symptom scores exceeded predefined thresholds, alerts were automatically generated for nursing staff. Clinical outcomes were assessed at baseline (T0), 1 month (T1), and 3 months (T2).

Results

A total of 136 patients were enrolled (107 [78.7%] men; mean [SD] age, 50.88 [12.6] years), including 67 participants in the intervention group and 69 in the control group, with no significant differences in baseline characteristics. For health-related quality of life (HRQOL), the primary summary scores showed no significant overall group × time interaction (P = 0.108; T2: mean difference [MD] = 4.61; 95% CI, -1.02–10.24; effect size [ES] = 0.23). In the physical functioning domain demonstrated a significant interaction effect (P = 0.048; T2: MD = 8.25; 95% CI, 1.15–15.35), its effect size (ES = 0.35) remained below the clinical threshold (ES ≥ 0.40). Notably, significant improvements were observed in several secondary domains, specifically appetite loss (interaction P = 0.001; ES = -0.56) and global health status (interaction P = 0.009; ES = 0.46). Improvements in self-efficacy (SUPPH interaction P = 0.012) and psychological status (IDA interaction P < 0.01) were significantly greater in the intervention group. The SYS-SM application demonstrated high feasibility and acceptability.

Conclusions

A nurse-led personalized mobile health symptom management intervention did not significantly improve primary overall HRQOL summary scores, but it demonstrated targeted benefits in reducing specific symptoms (particularly appetite loss), enhancing self-management capacity, and alleviating negative affect in patients with HNC undergoing concurrent chemoradiotherapy. The application shows potential for broader clinical implementation.

Trial registration

The study was prospectively registered with the Chinese Clinical Trial Registry (ChiCTR2300072314) on June 9, 2023.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12912-026-04871-x.

Introduction

Head and neck cancer (HNC) represents the sixth most common malignancy worldwide and is associated with poor overall survival [1]. China, as a high-incidence region, accounts for approximately 15.2% of global new cases and 16.8% of cancer-related deaths according to GLOBOCAN 2022 estimates, ranking second globally in both incidence and mortality [2]. Within this clinical context, multidisciplinary treatment strategies integrating surgery, radiotherapy, and systemic therapy constitute the foundation of curative management [3]. Despite favorable oncologic outcomes, these treatments frequently result in substantial acute toxicities that markedly impair quality of life and compromise treatment adherence [4]. Preservation of dose intensity and treatment duration during concurrent chemoradiotherapy (CCRT) has been shown to be essential for optimal therapeutic efficacy [5, 6]. Accordingly, continuous supportive care throughout the treatment trajectory remains critical for effective control of treatment-related symptoms and toxicities, thereby maintaining treatment compliance and improving clinical outcomes [7, 8].

Traditionally, symptom monitoring in patients with HNC has followed a reactive model, relying primarily on symptom reporting during scheduled outpatient visits. Such an approach is limited by infrequent assessments and recall bias, often resulting in underreporting and delayed intervention. In China, continuous follow-up through home visits or call center-based services remains constrained by workforce limitations, wide rural dispersion, and the concentration of specialized oncologic care within a small number of tertiary cancer centers. As a result, many patients are required to travel long distances or reside away from home for extended periods, often without adequate primary care support. These structural constraints underscore a pressing need for proactive, accessible, and sustainable approaches to symptom surveillance. Unlike standard monitoring, which relies on retrospective symptom recall during scheduled visits and can delay necessary care, proactive symptom monitoring facilitates early detection. It enables regular patient-reported outcome collection and timely clinical feedback and intervention through digital health technologies, thereby enhancing real-time symptom detection and management [9–11].

Within this framework, nurse-led personalized care interventions have attracted increasing attention as a key component of supportive oncology practice. Personalized care refers to the systematic tailoring of interventions to individual patient needs, preferences, and contextual factors, with the aim of strengthening health beliefs, reducing behavioral barriers, and improving treatment adherence and clinical outcomes [12]. Owing to sustained and close clinical contact with patients, nursing professionals are well positioned to coordinate symptom surveillance, deliver patient education, and provide psychosocial support throughout the treatment course [13]. In the present study, a nurse-led intervention was defined as “an intervention led by nurses without excluding the presence or appropriate participation of other health care professionals” [14]. Although such approaches have demonstrated substantial benefits in chronic disease management, evidence supporting effectiveness in symptom monitoring among patients with HNC undergoing chemoradiotherapy remains limited.

The rapid advancement of mobile health (mHealth) technologies provides new opportunities to operationalize nurse-led personalized symptom management. The World Health Organization defined mHealth as “medical and public health practice supported by mobile devices such as applications and telephones” [15], allowing remote monitoring, continuous patient education, and real-time clinical feedback. Compared with conventional face-to-face care models, mHealth interventions offer greater accessibility, cost-efficiency, and scalability. Asynchronous platforms, including mobile applications or mini-programs, enable symptom reporting and access to educational resources at flexible times, whereas synchronous modalities, such as telephone communication, support timely feedback and individualized guidance. Evidence indicates that combining synchronous and asynchronous approaches enhances patient engagement, flexibility, and treatment adherence, thereby improving overall intervention effectiveness [16–18].

Building on this foundation, a nurse-led, mHealth-based personalized intervention was developed and implemented through the WeChat platform, integrating synchronous (telephone communication) and asynchronous (mini-program) components to support proactive symptom monitoring and self-management in patients with HNC undergoing CCRT. The SYS-SM mini-program (Supplementary File 1) was designed to facilitate real-time symptom reporting, quality-of-life assessment, and timely nursing feedback. A randomized controlled trial was conducted with the primary aim of evaluating the efficacy of this intervention in enhancing quality of life, self-management capacity, and psychological well-being. Secondary implementation-related outcomes, including feasibility and acceptability, were also assessed to provide complementary information regarding the practicality of the intervention in routine clinical settings. While a decline in physical and psychosocial functioning is clinically expected during the acute phase of CCRT, we postulated that the SYS-SM intervention would significantly mitigate this deterioration. By empowering patients through real-time feedback and self-management education, the app buffers psychological distress and anxiety. Furthermore, early detection of toxicities prevents them from escalating to debilitating levels, thereby preserving overall HRQOL to a greater extent than standard care during this critical 3-month window. Therefore, it was hypothesized that participants receiving the intervention would demonstrate superior overall clinical health outcomes compared with those receiving standard care.

Research hypotheses

H0

The nurse-led mobile health–based symptom management intervention has no significant effect on quality of life, self-management capacity, or psychological well-being in patients with head and neck cancer undergoing concurrent chemoradiotherapy compared with standard care.

Methods

Study design and setting

A single-centre, single-blind, two-arm, parallel-group randomized controlled trial was conducted to evaluate the effectiveness of a nurse-led, individualized mHealth-based symptom management intervention compared with standard care. The intervention and follow-up period lasted for 3 months. This timeframe was deliberately chosen to encompass both the 7-week active treatment phase and the immediate acute recovery phase, capturing the window when acute treatment-related toxicities typically peak and begin to resolve. The trial was performed at a tertiary hospital in Guangzhou, China. Ethical approval was obtained from the Institutional Review Board of Sun Yat-sen Memorial Hospital, Sun Yat-sen University, in March 2023 (approval number: SYSKY-2023-101-02). The study was prospectively registered with the Chinese Clinical Trial Registry (ChiCTR2300072314) and reported in accordance with the Consolidated Standards of Reporting Trials (CONSORT) 2010 guidelines (Supplementary file 2: Table S1).

Participants

Participants were recruited between June and December 2023, and follow-up assessments were completed in March 2024. Recruitment was conducted at a tertiary hospital in Guangzhou, China. Written informed consent was obtained from all participants prior to enrollment.

Eligibility criteria included age ≥ 18 years; histologically confirmed HNC; planned receipt of CCRT; expected survival time of more than three months; ability to use mobile applications and participate in telephone communication; and proficiency in reading and understanding Mandarin Chinese. Exclusion criteria included diagnosis of a second primary malignancy, presence of cognitive impairment or psychiatric disorders, and significant visual, auditory, or speech impairment.

Study procedures

Prior to trial initiation, all (MDT) members completed a standardized 4-hour face-to-face training program. The training consisted of a structured lecture on the intervention delivered by a researcher, followed by a hands-on workshop on mini-program operation conducted by a software engineer. Competency was verified through a qualification examination, with a minimum passing score of 80 out of 100 required for study participation.

Following completion of training, the randomized controlled trial was conducted at the study site. Two research nurses screened and recruited eligible patients, and written informed consent was obtained prior to enrollment. Baseline assessments (T0) were completed before randomization, after which participants were assigned to either the intervention group or the control group. Both groups received standard care, whereas the intervention group additionally received structured instruction on mini-program use and secure access via an embedded identity verification system. The intervention was delivered by a nurse-led MDT over a 3-month period following hospital discharge. Outcome assessments were conducted at 1 month (T1) and 3 months (T2) by a research assistant blinded to group allocation, ensuring objective evaluation of intervention effects.

Randomization, concealment, and blinding

Participants were randomly assigned to the intervention group or the control group in a 1:1 ratio using a stratified block randomization method, with block sizes of 4 or 6. The allocation sequence was generated using an online randomization tool (https://www.randomizer.org/), with a predefined seed value to ensure reproducibility. Sequence generation was performed independently by a research assistant not involved in study implementation. The resulting allocation sequence and corresponding group assignments were sealed in sequentially numbered, opaque envelopes, which were opened only after completion of baseline assessment and confirmation of enrollment. Personnel responsible for outcome assessment and data analysis remained blinded to group allocation throughout the study, thereby minimizing potential assessment and analytical bias.

Care as usual

Participants in both groups received standard routine care throughout the study period. Standard care included provision of the prescribed treatment regimen, detailed information on the therapeutic plan, access to emergency contact information for the management of severe symptoms and treatment-related adverse effects, and appropriate pharmacologic treatment when indicated. In addition, routine telephone follow-up was provided by nursing staff. Scheduled outpatient visits and related treatments were conducted in accordance with the attending physician’s clinical management plan. Between these scheduled contacts, symptom management was largely episodic and reactive, relying on patients to initiate contact if severe issues arose.

Study intervention: nurse-led individualized mHealth symptom management

Development of the nurse-led individualized mHealth symptom management

A WeChat-based mini-program, designated SYS-SM (Supplementary file 1), was developed by a MDT to support symptom monitoring and management in patients undergoing chemoradiotherapy. Embedded within the WeChat platform, the program functions similarly to a conventional mobile application while offering greater accessibility by eliminating the need for download or installation. An expert panel comprising 11 specialists in radiation oncology, dietetics, nutritional nursing, nursing management, and research methodology was convened to refine the preliminary intervention framework, with the calculated authority coefficient supporting methodological rigor. Preliminary feasibility was evaluated through a pilot study, which demonstrated that the intervention was both practical and potentially beneficial. Based on feedback from pilot participants, two key modifications were implemented: expansion and reorganization of symptom items within the questionnaire to improve clinical relevance and usability, and restructuring of the program into a nurse-led model to enhance patient adherence and continuity of care.

The SYS-SM platform enables continuous, real-time surveillance and management of treatment-related toxicities. A validated self-report instrument, the MD Anderson Symptom Inventory—Head and Neck module (MDASI-HN), is integrated into the program for routine symptom assessment. The questionnaire assesses 15 core symptoms, including pain, fatigue, nausea, sleep disturbance, distress, dyspnea, poor appetite, drowsiness, xerostomia, sadness, vomiting, numbness, oral discomfort, and dysphagia. In addition, body temperature measurement using a digital thermometer is required and recorded within the platform (Supplementary file 2: Table S2). Evidence-based, tailored self-management recommendations, adapted from established clinical guidelines and authoritative resources, are accessible on demand within the program interface (Supplementary file 2: Fig. S1).

Description of the nurse-led individualized mHealth symptom management intervention

The nurse-led, personalized mHealth symptom management intervention consists of three core components: structured face-to-face education prior to hospital discharge, continuous use of the SYS-SM platform over a three-month period, and schedules biweekly telephone consultations. Implementation was conducted by a MDT led by nursing professionals, including an oncology specialist nurse, two junior nurses responsible for comprehensive patient care, two physicians, and a registered nutritionist. The oncology specialist nurse served as team leader and care coordinator, ensuring consistency and quality of intervention delivery through advanced clinical expertise. A detailed description of the intervention protocol is presented in Table 1.

Table 1.

Details of the nurse-led individualized mHealth multidisciplinary team intervention

Item	Facilitator	Functions	Content
1. Face-to-face education	Primary nurse	1.1 Applet usage	• Usage of SYS-SM applet.
2. SYS-SM applet	Multidisciplinary team (led by a oncology specialist nurse);	2.1 Management of symptoms in head and neck cancer radiotherapy and chemotherapy	• Acquire knowledge on symptom management for head and neck cancer radiotherapy and chemotherapy • Healthcare providers receive symptom alert notifications • Review the line chart of the most recent seven symptom reports
		2.2 Patient education center	• Nursing management of complications related to radiotherapy and chemotherapy • Precautions during radiotherapy and chemotherapy • Nutritional supplementation during radiotherapy and chemotherapy
		2.3 Nutritional diary	• Nutritional calculator to determine daily caloric intake • Log daily actual caloric consumption
		2.4 Nurse-patient online communication	• Patients can initiate medical and nursing consultations at any time • One-on-one confidential communication between healthcare providers and patients • Oncology specialty nurses directly address inquiries or refer questions beyond their scope to relevant multidisciplinary team members.
3. Biweekly symptom monitoring follow-up	The oncology specialist nurse	3.1 Assessment by phone calls	• Chemotherapy and radiotherapy-related complications, including pyrexia • Nutritional status (encompassing body weight and dietary recovery) • Gastrointestinal symptoms

Abbreviation: SYS-SM, Sun Yat-sen Symptom Management

Prior to discharge, junior nurses conducted structured face-to-face training sessions lasting approximately 15–20 min to familiarize participants with the core functions of the application. Training included the following procedures: system login; completion of the symptom questionnaire and review of historical symptom data; access to the dietary calculator and individualized nutritional targets; navigation of the self-management module; recording of daily dietary intake and body weight in the nutrition diary; use of the nurse-patient communication interface; access to educational materials and review of radiotherapy and chemotherapy precautions; consultation of evidence-based symptom management guidelines and documentation of current symptom status; and secure logout. Each function is practiced by participants under direct supervision of the junior nurse to ensure operational proficiency.

Participants were instructed to engage with the SYS-SM platform continuously over a three-month period. The system comprises four principal modules: symptom management for patients with HNC undergoing CCRT, a nutrition diary with an integrated calorie calculator, a health education resource center, and a secure communication portal for nurse-patient interaction. Detailed descriptions of system functionality are provided in Table 1 and Supplementary File 1.

The SYS-SM system operates through two integrated interfaces: a nurse portal and a patient portal. Symptom data entered by patients are automatically analyzed, generating one of three alert levels: green, yellow, or red. A green alert indicates a mild symptom burden and stable clinical status and does not trigger notification within the nurse portal. A yellow alert is generated in response to persistent mild-to-moderate symptoms, indicating the need for timely clinical attention to prevent further deterioration. A red alert reflects urgent oncology-related conditions, particularly complications associated with radiotherapy or chemotherapy. The predefined cut-off values for these symptom alerts (green, yellow, and red) were established based on the grading system of the Common Terminology Criteria for Adverse Events (CTCAE, Version 5.0) and multidisciplinary expert consensus [19]. Specifically, Grade 1 and 2 toxicities were configured to trigger green or yellow alerts, whereas Grade 3 or higher toxicities triggered red alerts. In parallel, the system generates a dynamic symptom trend chart, enabling continuous visualization of symptom trajectories over the preceding seven-day period.

Alerts are transmitted to nursing staff through the dedicated nurse portal. In response to a red alert, nursing personnel review reported symptoms, consult the attending physician for further clinical evaluation, and, when indicated, recommend immediate return to the hospital for medical management. For yellow alerts, direct patient communication is typically initiated to provide guidance or supportive interventions. Clinical actions taken in response to alerts, whether by nursing staff or physicians, are not constrained by study protocols and remain guided by professional clinical judgment.

Outcomes

Clinical outcomes

All clinical outcomes were assessed using standardized questionnaires at baseline (T0), 1 month (T1), and 3 months (T2).

Primary objective: clinical outcomes

Primary outcome

Health-related quality of life (HRQOL) was evaluated using the validated Chinese (version 3.0) of the 30-item European Organization for Research and Treatment of Cancer Quality of Life Questionnaire–Core 30 (EORTC QLQ-C30) [20, 21]. Overall HRQOL was represented by the summary score (SS), calculated as the mean of standardized scores across all functional and symptom subscales, excluding financial impact and global quality-of-life items [22]. Physical functioning was assessed using the five-item physical functioning (PF) subscale of the QLQ-C30. In accordance with EORTC scoring guidelines, all scale scores were linearly transformed to a 0-100 scale, with higher scores indicating better functional status. Mean change values were reported together with Cohen’s d effect size (ES) [23]. Effect sizes of 0.2, 0.5, and 0.8 are conventionally interpreted as small, medium, and large, respectively [24]. An ES ≥ 0.4 was predefined as the threshold for clinically meaningful superiority of the intervention over control, whereas an ES < 0.2 was considered to indicate no clinically relevant difference between groups.

Secondary clinical outcomes

Additional subscales of the QLQ-C30 were designated as secondary outcome measures. Within the functional domains, including role functioning, emotional functioning, cognitive functioning, and social functioning, as well as the global health status scale, higher scores reflect better performance. In contrast, within symptom domains, including fatigue, nausea and vomiting, pain, dyspnea, insomnia, appetite loss, constipation, diarrhea, and financial difficulties, higher scores indicate greater symptom burden [22].

Self-care self-efficacy was assessed using the Strategies Used by People to Promote Health (SUPPH) instrument [25], a validated 29-item self-report scale developed based on self-efficacy theory to evaluated confidence in self-care behaviors. The SUPPH comprises three dimensions: positive attitude, stress reduction, and decision-making. Each item is rated on a 5-point Likert scale, with item scores ranging from 1 to 5. Dimension scores are calculated by summing item responses within each domain, and the total score is obtained by aggregating all domain scores (range: 29–145), with higher values indicating greater self-care self-efficacy.

Negative psychological status was assessed using the Irritability, Depression, and Anxiety (IDA) scale, a self-report instrument designed to quantify the severity of depression symptoms, anxiety, and both internal and external irritability [26]. The Chinese version usesd in the present study is a validated 18-item adaptation [27]. Items are rated on a 4-point Likert scale (ranging from 0 to 3). Subscale scores are calculated by summing the relevant items, with higher scores representing greater severity of negative psychological symptoms.

Secondary objective: feasibility and acceptability outcomes

An additional objective was to evaluate the feasibility and acceptability of SYS-SM in routine clinical practice. Feasibility was defined as the practicality of implementation under existing technological and human resource conditions, the degree of integration into standard clinical workflows, utilization of the SYS-SM platform and its associated guidance, and patient adherence to the intervention. Indicators included recruitment rate, refusal rate, attrition rate, reasons for refusal and withdrawal, willingness to accept randomization, and perceived clinical value among healthcare professionals. To assess feasibility and acceptability, a structured patient feedback questionnaire was developed to evaluate satisfaction with the SYS-SM system. An additional survey was administered to oncology clinicians to assess practicality and usability in clinical settings. All feasibility data were collected using the Questionnaire Star platform.

Sample size

Sample size was determined based on a two-sample comparison of independent means using a two-sided significance level (α) of 0.05 and statistical power (1 − β) of 80%. The sample size calculation was based on the primary outcome measure, the EORTC QLQ-C30 summary score. Effect size and variability estimates were derived from a previously published randomized controlled trial by Roma [28] et al., which reported a mean difference of 0.21 and standard deviations of 0.29 and 0.41 in the intervention and control groups, respectively (yielding an estimated standardized effect size, Cohen’s d, of approximately 0.59). Calculations were performed using PASS software (version 15; NCSS, LLC, Kaysville, UT, USA). On the basis of these parameters, the minimum required sample size was 45 participants per group. To account for an anticipated attrition rate of 10% during follow-up, the target sample size was increased to 50 participants per group to ensure adequate statistical power for the primary analysis.

Data analysis

All analyses were conducted according to the intention-to-treat (ITT) principle. Primary analyses compared outcomes between the intervention and control groups. Linear mixed-effects models were employed to evaluate the intervention effect on changes in HRQOL over time, including fixed effects for group assignment and time point, and a random intercept for individual participants to account for the intra-subject correlation of repeated measures over time. To adjust for potential clinical confounders, two models were specified: an unadjusted model (Model A) and an adjusted model (Model B). In Model B, age, cancer stage, and cancer type were explicitly entered as fixed-effect covariates into the linear mixed-effects models to statistically adjust for their potential confounding effects. Statistical analyses were performed using R software, with a two-sided significance level of α = 0.05 applied to all hypothesis tests.

Results

Analytical sample

A total of 231 individuals were screened for study participation. Among these, 139 participants were excluded: 75 (54%) did not meet eligibility criteria, 39 (28%) declined participation, and 25 (18%) were unreachable or unable to participate for other reasons, such as failure to receive chemoradiotherapy or death. The remaining 136 participants (58.8%) were enrolled and randomly assigned to the intervention group (n = 57) or the control group (n = 55), as depicted in Fig. 1.

Fig. 1.

Study flow diagram

Baseline characteristics of the study population

The final study cohort comprised 136 patients with HNC undergoing chemoradiotherapy, including 107 men (78.7%) and 29 women (21.3%), with a mean (SD) age of 50.88 (12.6) years. Most participants were married (124/136, 91.2%), had completed middle school education (68, 50.0%), resided with family members (127, 93.4%), and were unemployed (88, 64.7%). The majority presented with advanced disease, including stage IV (66, 48.5%) and stage III cancer (54, 39.7%), and most cases were non-metastatic (107, 78.7%). With respect to cancer subtype, nasopharyngeal carcinoma accounted for the largest proportion (71/136, 52.2%) (Table 2).

Table 2.

Baseline characteristics of participants

Characteristics	All patients (N = 136)	Intervention group (n = 67)	Control group (n = 69)	P value
Mean (SD) age	50.88 (12.6)	49.73 (12.0)	52.00 (13.2)	0.296
Female sex	29 (21.32)	11 (16.42)	18 (26.09)	0.243
Marital status:				0.393
Married	124 (91.18)	63 (94.03)	61 (88.41)
Other (unmarried/widowed/divorced)	12 (8.82)	4 (5.97)	8 (11.59)
Education				0.263
Below primary school	41 (30.15)	24 (35.82)	17 (24.64)
Middle school	68 (50.00)	29 (43.28)	39 (56.52)
Above high school	27 (19.85)	14 (20.90)	13 (18.84)
Living status				0.519
Alone	9 (6.62)	3 (4.48)	6 (8.70)
With another adult	127 (93.38)	64 (95.52)	63 (91.30)
Employment status				0.167
No	88 (64.71)	39 (58.21)	49 (71.01)
Yes	48 (35.29)	28 (41.79)	20 (28.99)
Cancer stage				0.479
I	5 (3.68)	4 (5.97)	1 (1.45)
II	11 (8.09)	5 (7.46)	6 (8.70)
III	54 (39.71)	28 (41.79)	26 (37.68)
IV	66 (48.53)	30 (44.78)	36 (52.17)
Tumor metastasis				0.113
No	107 (78.68)	57 (85.07)	50 (72.46)
Yes	29 (21.32)	10 (14.93)	19 (27.54)
Comorbidity
No	102 (75.0)	51 (76.1)	51 (73.9)	0.534
Yes	34 (25.0)	16 (23.9)	18 (26.1)
Smoking history				0.134
No	122 (89.7)	63 (94.0)	59 (85.5)
Yes	14 (10.3)	4 (6.0)	10 (14.5)
Alcohol use				0.231
No	124 (91.2)	63 (94.0)	61 (88.4)
Yes	12 (8.8)	4 (6.0)	8 (11.6)
Cancer type				0.01
Nasopharyngeal carcinoma	71 (52.21)	43 (64.18)	28 (40.58)
Other head and neck tumors (tongue, throat)	65 (47.79)	24 (35.82)	41 (59.42)

Primary objective: clinical outcomes

Superiority analyses: control group versus intervention group

Between-group changes from baseline in EORTC QLQ-C30 summary score (SS) and physical functioning (PF) are presented in Table 3; Fig. 2. Based on adjusted linear mixed-effects models, the primary outcome of summary score did not show a significant overall group × time interaction (P = 0.108). While physical functioning reached statistical significance for the interaction (P = 0.048), T2 effect sizes for both domains (SS: ES = 0.23; PF: ES = 0.35) fell below the clinical threshold (ES ≥ 0.40). Between-group differences in SS remained modest (T1: MD = 4.65; T2: MD = 4.61). PF showed a significant difference at T2 (MD = 8.25; 95% CI, 1.15–15.35) but not T1 (MD = 5.45). Notably, significant interaction effects and clinically meaningful improvements (ES ≥ 0.40) were observed in secondary domains, specifically appetite loss (Interaction P = 0.001; T2: MD = -18.25; ES = 0.56) and global health status (Interaction P = 0.009; ES = 0.46). Unadjusted Model A results remained consistent with those of the adjusted models (Supplement File 2: Table S3).

Table 3.

Linear mixed-effects model results for health-related quality of life: intervention vs. control group over time, adjusted for confounders

Scale domain and time point	Mean estimate (SE)		Mean difference from baseline between groups (95% CI)	P-value (post-hoc)	Cohen d effect size	Interaction P (Group*Time)
	Intervention group	Control group
Primary EORTC-QLQ-C30 domain
Summary score						0.108
Baseline	76.55 (1.92)	78.32 (1.88)	NA	NA	NA
1 mo	79.72 (1.98)	76.75 (1.95)	4.65 (-0.85- 10.15)	0.098	0.24
3 mo	80.58 (2.02)	77.65 (2.05)	4.61 (-1.02- 10.24)	0.105	0.23
Physical functioning						0.048 *
Baseline	85.35 (2.45)	83.85 (2.41)	NA	NA	NA
1 mo	85.52 (2.55)	78.42 (2.52)	5.45 (-1.25- 12.15)	0.112	0.23
3 mo	91.15 (2.58)	81.25 (2.61)	8.25 (1.15–15.35)	0.022 *	0.35
Global health status						0.009 *
Baseline	50.85 (2.95)	54.55 (2.90)	NA	NA	NA
1 mo	57.65 (3.05)	51.42 (3.02)	9.85 (-0.15- 19.85)	0.054	0.31
3 mo	66.42 (3.12)	55.25 (3.15)	14.85 (5.15–24.55)	0.003 *	0.46
Role functioning						0.051
Baseline	78.15 (3.12)	79.42 (3.08)	NA	NA	NA
1 mo	82.45 (3.25)	77.15 (3.18)	5.30 (-3.65- 14.25)	0.245	0.18
3 mo	88.22 (3.35)	78.85 (3.42)	9.37 (0.05–18.69)	0.048 *	0.32
Emotional functioning						0.042 *
Baseline	72.14 (2.85)	74.35 (2.82)	NA	NA	NA
1 mo	78.55 (2.92)	73.12 (2.88)	5.43 (-2.65- 13.51)	0.186	0.25
3 mo	84.65 (3.05)	75.42 (3.12)	9.23 (1.05–17.41)	0.027 *	0.40
Cognitive functioning						0.058
Baseline	84.15 (2.65)	85.32 (2.61)	NA	NA	NA
1 mo	85.22 (2.75)	84.15 (2.72)	1.07 (-6.55- 8.69)	0.785	0.05
3 mo	91.05 (2.82)	83.25 (2.85)	7.80 (0.12–15.48)	0.046 *	0.32
Social functioning						0.038 *
Baseline	75.38 (3.05)	77.12 (3.01)	NA	NA	NA
1 mo	80.12 (3.15)	75.45 (3.12)	4.67 (-4.12- 13.46)	0.295	0.20
3 mo	89.45 (3.28)	76.22 (3.35)	13.23 (3.85–22.61)	0.006 *	0.45
Secondary EORTC-QLQ-C30 domain
Fatigue
Baseline	26.42 (2.85)	24.85 (2.82)	NA	NA	NA	0.035 *
1 mo	21.15 (2.75)	28.25 (2.78)	-8.85 (-16.95- -0.75)	0.033 *	0.38
3 mo	17.85 (2.68)	26.42 (2.81)	-10.65 (-18.95- -2.35)	0.012 *	0.48
Nausea and vomiting						0.542
Baseline	12.10 (2.92)	14.15 (2.88)	NA	NA	NA
1 mo	11.95 (3.02)	14.05 (2.98)	-2.55 (-11.45- 6.35)	0.582	0.10
3 mo	16.42 (3.12)	17.85 (3.15)	-1.85 (-10.95- 7.25)	0.695	0.07
Pain						0.088
Baseline	24.05 (3.45)	22.15 (3.42)	NA	NA	NA
1 mo	20.15 (3.35)	26.42 (3.32)	-8.55 (-18.15- 1.05)	0.081	0.33
3 mo	17.85 (3.28)	25.15 (3.35)	-9.45 (-19.25- 0.35)	0.058	0.36
Dyspnea						0.155
Baseline	15.65 (3.75)	16.72 (3.70)	NA	NA	NA
1 mo	12.05 (3.65)	18.05 (3.62)	-7.35 (-17.75- 3.05)	0.165	0.27
3 mo	10.25 (3.58)	19.42 (3.65)	-10.55 (-21.15- 0.05)	0.051	0.40
Insomnia						0.022 *
Baseline	35.15 (4.65)	32.25 (4.60)	NA	NA	NA
1 mo	28.25 (4.55)	36.25 (4.52)	-11.15 (-23.95- 1.65)	0.086	0.35
3 mo	24.85 (4.48)	38.52 (4.65)	-16.85 (-29.95- -3.75)	0.012 *	0.51
Appetite loss						0.001 *
Baseline	34.65 (3.25)	32.05 (3.22)	NA	NA	NA
1 mo	29.85 (3.45)	40.25 (3.42)	-13.15 (-24.95- -1.35)	0.028 *	0.41
3 mo	24.25 (3.52)	39.85 (3.58)	-18.25 (-30.45- -6.05)	< 0.001 *	0.56
Constipation
Baseline	28.35 (3.75)	24.15 (3.72)	NA	NA	NA	0.045 *
1 mo	21.15 (3.95)	23.85 (3.92)	-6.95 (-21.15- 7.25)	0.335	0.21
3 mo	22.65 (4.02)	30.75 (4.12)	-12.45 (-26.25- 1.35)	0.076	0.37
Diarrhea						0.752
Baseline	8.15 (1.95)	7.35 (1.98)	NA	NA	NA
1 mo	4.55 (1.90)	3.12 (1.92)	0.85 (-4.55- 6.25)	0.755	0.04
3 mo	7.15 (1.92)	7.42 (2.05)	-1.25 (-6.85- 4.35)	0.655	0.06
Financial impact						0.138
Baseline	30.55 (4.75)	23.55 (4.72)	NA	NA	NA
1 mo	40.25 (4.85)	32.25 (4.82)	1.55 (-11.95- 15.05)	0.815	0.07
3 mo	49.35 (5.25)	29.85 (4.95)	12.85 (-1.15- 26.85)	0.072	0.13

Abbreviations: EORTC-QLQ-C30, European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-Core 30; NA, not applicable

^a P-values for post-hoc between-group comparisons based on estimated marginal means (EMMeans). * Indicates p < 0.05

^b Model B was adjusted for age, cancer type, and cancer stage. Full unadjusted results are available in Supplementary 2 Table S2

^c Interaction P-values evaluate the significance of the group-by-time interaction effect in the overall mixed-effects model

^d For EORTC quality-of-life score and functional scales, higher scores indicate better quality of life and functioning; for EORTC symptom scales, higher scores indicate greater symptom burden

Fig. 2.

Estimated marginal means with 95% confidence intervals for EORTC QLQ-C30 summary score (A) and physical functioning (B) at each time point, derived from linear mixed-effects models

Secondary clinical outcomes: SUPPH and IDA scales

Significant improvements favoring the intervention group were observed in self-efficacy and psychological outcomes (Table 4). Linear mixed-effects model analysis revealed significant overall interactions for total SUPPH (P = 0.012; T1: P = 0.016, T2: P = 0.028) and the self-decision subscale (P < 0.001). Regarding psychological outcomes, significant overall interactions were found for IDA total (P = 0.005), anxiety (P = 0.018), and depression (P = 0.008). Specifically, depression improved at both (P = 0.037) and 3 mo (P < 0.001), while IDA total and anxiety showed significant improvements at 3 mo (both P < 0.005). Other subscales remained non-significant (P > 0.05).

Table 4.

Linear mixed-effects model results for SUPPH and IDA scale scores from baseline to each time point between the intervention and control groups, adjusted for confounders

Variable	Time point	Intervention group mean estimate	Control group mean estimate	Mean difference from baseline between groups (95% CI)	P value
SUPPH total	Overall Group × Time Interaction (LRT Analysis):				0.012*
	Baseline	110.93 (2.42)	113.29 (2.25)	-	-
	1 mo	100.52 (2.41)	90.17 (2.89)	12.70 (2.35–23.05)	0.016*
	3 mo	106.74 (2.55)	96.68 (3.15)	12.42 (1.35–23.49)	0.028*
SUPPH - self-stress reduction	Overall Group × Time Interaction (LRT Analysis):				0.068
	Baseline	36.87 (1.15)	36.65 (1.08)	-	-
	1 mo	38.11 (0.98)	33.51 (1.15)	4.37 (-1.50- 10.24)	0.142
	3 mo	40.06 (1.02)	34.93 (1.18)	4.89 (-0.45- 10.23)	0.071
SUPPH -self-decision	Overall Group × Time Interaction (LRT Analysis):				< 0.001*
	Baseline	10.04 (0.43)	10.58 (0.45)	-	-
	1 mo	11.53 (0.28)	10.03 (0.35)	2.10 (0.58–3.62)	< 0.001*
	3 mo	12.43 (0.31)	10.15 (0.38)	2.82 (1.15–4.45)	< 0.001*
SUPPH - positive attitude	Overall Group × Time Interaction (LRT Analysis):				0.185
	Baseline	64.02 (1.28)	66.06 (1.35)	-	-
	1 mo	50.88 (1.58)	46.63 (1.78)	6.22 (-1.45- 13.89)	0.115
	3 mo	54.25 (1.65)	51.60 (1.92)	4.67 (-3.35- 12.69)	0.254
IDA total	Overall Group × Time Interaction (LRT Analysis):				0.005*
	Baseline	20.87 (1.18)	21.03 (1.25)	-	-
	1 mo	16.45 (1.12)	19.32 (1.31)	-2.71 (-5.85- 0.43)	0.089
	3 mo	15.12 (1.08)	20.15 (1.35)	-5.19 (-8.52- -1.86)	0.002*
IDA anxiety	Overall Group × Time Interaction (LRT Analysis):				0.018*
	Baseline	7.12 (0.45)	7.25 (0.42)	-	-
	1 mo	5.85 (0.38)	6.92 (0.41)	-0.94 (-1.95- 0.07)	0.068
	3 mo	5.21 (0.35)	7.38 (0.43)	-2.30 (-3.42- -1.18)	< 0.001*
IDA depression	Overall Group × Time Interaction (LRT Analysis):				0.008*
	Baseline	8.42 (0.52)	8.38 (0.55)	-	-
	1 mo	6.12 (0.48)	7.55 (0.51)	-1.47 (-2.85- -0.09)	0.037*
	3 mo	5.34 (0.45)	7.92 (0.58)	-2.62 (-4.12- -1.12)	< 0.001*
IDA outward irritation	Overall Group × Time Interaction (LRT Analysis):				0.245
	Baseline	3.15 (0.28)	3.22 (0.31)	-	-
	1 mo	2.82 (0.25)	2.95 (0.28)	-0.06 (-0.82- 0.70)	0.875
	3 mo	2.65 (0.22)	3.12 (0.30)	-0.40 (-1.15- 0.35)	0.294
IDA inward irritation	Overall Group × Time Interaction (LRT Analysis):				0.062
	Baseline	2.18 (0.19)	2.18 (0.21)	-	-
	1 mo	1.66 (0.15)	1.90 (0.18)	-0.24 (-0.68- 0.20)	0.285
	3 mo	1.92 (0.18)	1.73 (0.16)	0.19 (-0.32- 0.70)	0.452

Abbreviations: SE, standard error; CI, confidence interval; SUPPH, Strategies Used by People to Promote Health; IDA, Irritability, Depression and Anxiety Scale

^a P-value for the overall Group × Time interaction effect derived from the Likelihood Ratio Test (LRT) of the linear mixed-effects model. This assesses whether the trajectories of change differ significantly between the groups over the 3-month period

^b Mean difference from baseline between groups represents the “difference-in-differences” estimate at each post-baseline time point (1 or 3 mo) relative to baseline (0 months), with 95% confidence intervals

^c Models were adjusted for baseline age, cancer stage (III vs. IV), and cancer type as covariates

Secondary objective: feasibility and acceptability outcomes

Patients’ evaluation

At post-intervention assessment, the SYS-SM platform was widely reported to be easy to use and understand. Most participants indicated that the system improved overall treatment quality, facilitated timely identification and management of treatment-related side effects, and enhanced physician–patient communication. More than 90% of participants reported willingness to recommend the system to others, and nearly all expressed willingness to continue using SYS-SM (Table 5).

Table 5.

Patients’ evaluation of the Sun Yat-sen Symptom Management system at post-intervention

Item	Strongly Disagree, No (%)	Disagree, No (%)	Agree, No (%)	Strongly Agree, No (%)
SYS-SM was easy to complete	0 (0)	0 (0)	7 (12.28)	50 (87.72)
SYS-SM was easy to understand	0 (0)	2 (3.51)	5 (8.77)	50 (87.72)
SYS-SM helps me identify and better describe the side effects of treatment	0 (0)	2 (3.51)	8 (14.04)	47 (82.46)
SYS-SM improved the quality of my treatment	0 (0)	2 (3.51)	6 (10.53)	49 (85.96)
SYS-SM helps me prevent and manage side effects	0 (0)	2 (3.51)	7 (12.28)	48 (84.21)
SYS-SM improved the communication with my physician	0 (0)	7 (12.28)	50 (87.72)	0 (0)
I would recommend SYS-SM to other patients	0 (0)	5 (8.77)	52 (91.23)	0 (0)
I would like to use SYS-SM further	0 (0)	2 (3.51)	6 (10.53)	49 (85.96)

Abbreviation: SYS-SM, Sun Yat-sen Symptom Management

Note: n = 57

Physicians’ evaluation

All participating physicians reported that the SYS-SM system was easy to operate. More than two-thirds strongly agreed that the platform improved clinical understanding of patient status, and over 80% recognized its value in symptom monitoring, clinical communication, and treatment support. Nearly 90% of physicians reported frequent use of the system in routine clinical practice (Supplement File 2: Table S4).

Discussion

The present study evaluated a nurse-led, personalized mHealth-based symptom management intervention delivered through integration of a mobile application and scheduled telephone follow-up, with the aim of providing continuous symptom surveillance, structured self-management support, and targeted health education for patients with HNC undergoing CCRT. Findings indicate that while the intervention did not significantly improve the primary overall health-related quality of life summary score, it demonstrated targeted benefits in specific symptom domains (e.g., appetite loss), enhanced self-management capacity, and reduced negative psychological symptoms, including anxiety and depression. Patient feedback further suggests that the intervention was feasible and well accepted within routine clinical practice.

Effect of the intervention on HRQOL

Compared with standard care, the intervention group demonstrated favorable trends in overall HRQOL. However, these differences did not reach statistical significance, and therefore should not be interpreted as definitive evidence of intervention effectiveness. These findings are consistent with previous randomized trials integrating symptom monitoring with automated alert systems [4, 8–10, 27]. Although improvements in overall HRQOL did not reach statistical significance, the observed mean differences in this pragmatic trial exceeded those reported by Bash et al. [29]. However, corresponding effect sizes did not consistently meet the predefined threshold of 0.4, likely reflecting heterogeneity in disease characteristics and prognostic profiles across cancer populations [11]. Among patients undergoing CCRT for HNC, persistent treatment-related toxicities, including dysphagia, xerostomia, and dysgeusia, remain prevalent and substantially impair quality of life [9]. Even with effective intervention, the magnitude of HRQOL improvement is therefore limited. Under such conditions, the predefined superiority threshold may be overly stringent, as achieving this level of improvement in HRQOL may not be feasible for a substantial proportion of patients.

In addition to global HRQOL outcomes, the intervention demonstrated a beneficial effect on appetite loss. Appetite loss is a common and clinically significant symptom during and after cancer treatment, particularly in patients receiving chemoradiotherapy, and is closely associated with treatment tolerance and nutritional decline [30]. The observed improvement is likely attributable to timely symptom monitoring, individualized dietary guidance, and enhanced nurse–patient communication facilitated by the mHealth platform [31]. These mechanisms may enable earlier identification and management of nutrition-related symptoms and improve adherence to dietary recommendations. Although appetite loss was the only symptom domain reaching statistical significance, this finding supports the value of targeted symptom management within comprehensive supportive care strategies.

Effect of the intervention on secondary outcomes

Patients with HNC commonly undergo multimodal treatment, including surgery, radiotherapy, and chemotherapy. Limited understanding of treatment procedures and potential adverse effects often contributes to heightened psychological distress, particularly anxiety and fear, underscoring the importance of structured education on disease knowledge and self-care strategies to facilitate treatment adaptation and lifestyle adjustment [32, 33]. These findings suggest a possible relationship between improved self-management capacity and better psychological outcomes, although the mechanisms underlying this association warrant further investigation, as reflected by improvements in overall self-efficacy and decision-making domains, while concurrently alleviating negative psychological symptoms, including overall psychological distress and anxiety. These findings suggest that strengthening symptom management competence is closely associated with improved psychological well-being.

In contrast to previous interventions that primarily focused on symptom surveillance or physiological parameters [34, 35], the current model integrates continuous nurse-led monitoring with individualized guidance. This approach not only reinforces self-efficacy but also incorporates therapeutic communication and ongoing feedback, enabling early identification and regulation of negative emotional states. These mechanisms may contribute to the observed improvements in psychological outcomes and decision-making capacity.

No statistically significant improvements were observed in the positive attitude domain of self-management or in the external irritability dimension of psychological distress, likely reflecting the influence of multiple contributing factors [36]. The intervention was primarily designed to strengthen symptom recognition, self-management skills, and emotional support, with comparatively limited emphasis on long-term attitudinal change or modification of external environmental conditions. Consequently, measurable short-term changes within these domains may be difficult to achieve.

The feasibility of integrating the SYS-SM intervention into routine clinical practice was also evaluated. A recruitment rate of 59% was observed, comparable to or exceeding rates reported in previous electronic health interventions involving symptom monitoring systems [4, 9, 29]. The primary reason for non-participation was perceived research burden, accounting for approximately 13.7% of refusals. Given the advanced disease stage of most participants, this recruitment rate may be considered favorable. Patient feedback further indicated that more than 80% of participants in the intervention group perceived SYS-SM as beneficial for identifying and accurately describing treatment-related side effects, reflecting substantial perceived clinical value. This positive response is likely attributable to comprehensive pre-discharge training, optimized platform functionality, and reinforced education supporting sustained system use. Core components of the intervention, including structured application training, integration of synchronous and asynchronous mHealth support, and biweekly personalized guidance provided by nursing staff, demonstrated both feasibility and high acceptability [18]. Compared with previous studies, higher participant adherence observed in this study may have increased exposure to the intervention and facilitated participant engagement, although its impact on clinical outcomes requires further investigation [37].

Implications for clinical practice

The nurse-led, personalized mHealth symptom management intervention addresses a critical gap in continuous supportive care for patients with HNC undergoing CCRT, particularly by facilitating symptom monitoring and self-management during the post-discharge home phase. Sustained and proactive symptom surveillance may facilitate earlier symptom recognition and management, which could potentially support treatment adherence and patient engagement during the home-care phase. Consistent with these observations, previous studies have shown that active symptom monitoring can prolong overall survival among patients receiving palliative chemotherapy [38]. Reinforcing dynamic communication between patients and clinicians therefore carries substantial clinical value.

Future applications may further expand system functionality, including integration of toxicity reporting based on the Common Terminology Criteria for Adverse Events (CTCAE) within electronic case report forms (eCRFs) [39], enabling more precise and objective data collection, such as changes in body weight and limitations in daily activities. As oncologic treatment paradigms continue to evolve, particularly with increasing incorporation of immunotherapy, symptom management will encounter new challenges and opportunities. Although immunotherapy offers significant therapeutic benefits, it may also be associated with rare but potentially severe adverse events. Digital platforms that enhance patient awareness of early warning symptoms and provide real-time risk alerts may play a critical role in preventing serious complications and reducing unnecessary treatment discontinuation.

When implementing nurse-led mHealth interventions, systematic consideration of institutional capacity, human resources, and technological readiness is essential [40]. At the institutional level, tertiary and urban cancer centers with established infrastructure are well positioned to integrate such interventions into discharge planning and follow-up workflows through standardized patient education, routine remote monitoring, and structured multidisciplinary collaboration. In resource-constrained settings, successful implementation depends on supportive policy frameworks and optimized care processes to ensure long-term sustainability [41]. From a human resource perspective, competency-based digital health training can expand nursing roles, strengthen interdisciplinary collaboration, and clarify responsibilities, thereby promoting continuity and consistency of care delivery. Technologically, robust digital infrastructure and culturally adapted system design are crucial, with careful consideration of regional disparities in digital literacy and access [42]. Integration of real-time telephonic or video-based follow-up may further improve patient adherence and engagement. In addition, cost-efficient automation technologies, such as push notifications and interactive voice response systems, although requiring initial investment and validation, hold considerable potential to support sustainability and scalability [43].

Overall, the findings support the feasibility and acceptability of a nurse-led mHealth symptom management intervention. Further adequately powered studies are needed to determine its effectiveness and clinical value across diverse oncology settings. Future research is warranted to validate effectiveness and scalability across diverse healthcare settings, thereby supporting broader, sustainable, and equitable implementation.

Although favorable trends were observed across several outcomes, the findings should be interpreted cautiously. The intervention did not result in statistically significant improvement in the primary HRQOL outcome, and the clinical meaningfulness of the observed changes remains uncertain. Several factors may explain these findings, including the pilot nature of the study, limited sample size, baseline heterogeneity, and the relatively short follow-up period. Similar observations have been reported in previous mHealth and supportive care studies, where promising trends were identified but larger trials were required to establish definitive effectiveness. Future studies with sufficient statistical power and longer follow-up are warranted to determine whether these preliminary findings translate into clinically meaningful benefits.

Limitations

Several limitations should be considered. First, recruitment was conducted at a single tertiary hospital in Guangdong, China, which may limit generalizability. Multicenter randomized controlled trials with larger sample sizes are required to provide more robust evidence. Second, the follow-up period of three months was relatively short; longer-term follow-up is necessary to evaluate the durability of intervention effects. Third, although the primary objective focused on intervention effectiveness, comprehensive cost analysis was beyond the scope of the study and warrants investigation in future research. Fourth, the digital delivery format may have introduced selection bias, as individuals without internet access were excluded, highlighting the need for strategies to improve accessibility among underrepresented populations. Finally, further optimization of the SYS-SM platform is needed to support long-term monitoring and implementation, including refinement of alert algorithms to minimize false-positive notifications and expansion of patient input options to better contextualize decisions regarding perceived need for clinical contact.

Conclusions

The present study demonstrates that a nurse‑led, personalized mHealth symptom management intervention improves quality of life, enhances self‑management capacity, and reduces negative affect in patients with HNC undergoing CCRT. These findings support the integration of nurse‑led, personalized interventions that combine systematic symptom monitoring with mHealth technologies as an effective strategy for optimizing patient outcomes in oncologic care.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

HY, JC, and SZ: Methodology, Data Curation, Formal Analysis, Investigation, Validation, Visualization, Writing-Original Draft Preparation; ZX, ZHX, and DW: Formal Analysis, Validation, Resources, Investigation; ZB and YQ: Conceptualization, Design, Supervision, Project Administration, Funding acquisition, Writing-Review and Editing. HY, JC, and SZ contributed equally to this work. The authors read and approved the final manuscript.

Funding

This work was supported by the Nurturing Funds for Nursing Young Talents of the Guangdong Province Medical Research Fund (Grant No. A2025109) and Yat-sen Scholarship for Young Nursing Scientists (Grant No. YXHL202103). All authors approved this submission.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethical approval

The study was approved by the Ethics Committee of Sun Yat-sen Memorial Hospital (SYSKY-2023-101-02) and conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.