Current awareness and influencing factors regarding digital out-of-hospital health monitoring among patients after gastrointestinal surgery: A cross-sectional study

Abstract

BACKGROUND

Postoperative patients in gastrointestinal surgery often face interrupted professional care and insufficient rehabilitation guidance after discharge. Digital out-of-hospital health monitoring can provide remote monitoring and abnormality alerts, yet its effectiveness depends on patients' adequate cognition of the service. Currently, there is a lack of research on the cognitive status and influencing factors in this population.

AIM

To investigate the cognitive status of digital out-of-hospital health monitoring among postoperative patients in gastrointestinal surgery and analyze its influencing factors.

METHODS

A convenience sampling method was used to select 126 postoperative patients discharged from the gastrointestinal surgery department of a tertiary hospital. A self-designed cognitive questionnaire (4 dimensions, 15 items, Likert 5-point scale) was administered. A t test, analysis of variance, and multiple linear regression were used to analyze influencing factors.

RESULTS

The mean item score of cognition among the 126 patients was (3.02 ± 0.43), indicating a moderately low level. The value cognition dimension scored the highest, while the concern cognition dimension scored the lowest. Univariate analysis showed that age, education level, years of smartphone use, and health app use experience significantly affected the total cognitive score (all P < 0.05). Multiple regression analysis revealed that years of smartphone use (β = 0.203) and health app use experience (β = -0.176) were independent influencing factors.

CONCLUSION

Postoperative patients in gastrointestinal surgery exhibit generally low cognition of digital out-of-hospital health monitoring, characterized by a "value recognition preceding basic cognition" pattern and prevalent concerns. Digital contact experience predicts cognitive level better than demographic characteristics. Stratified health education should be implemented to strengthen service concept awareness, while technical optimization should be adopted to alleviate patient concerns.

Key Words: Gastrointestinal surgery; Postoperative patients; Digital out-of-hospital health monitoring; Cognition; Influencing factors; Cross-sectional study

Core Tip: This cross-sectional study of 126 postoperative gastrointestinal surgery patients found moderately low cognition of digital out-of-hospital health monitoring (mean score 3.02 ± 0.43), characterized by value cognition exceeding basic cognition. Smartphone experience and health app use independently predicted cognition, suggesting that digital contact experience outweighs demographic characteristics. Pervasive concerns regarding data accuracy and privacy security hindered the transition from perceived usefulness to actual utilization. Stratified health education and technology optimization are essential to bridge this gap.

INTRODUCTION

Postoperative patients in gastrointestinal surgery face multiple risks of complications such as incision infection, anastomotic leakage, intestinal obstruction, and malnutrition due to extensive surgical trauma and prolonged recovery periods, which seriously threaten their quality of rehabilitation and safety[1,2]. With the widespread adoption of minimally invasive surgical techniques and the enhanced recovery after surgery concept, the length of hospital stay for gastrointestinal surgery patients has been significantly shortened; however, most patients have not fully recovered at discharge and still require long-term home-based rehabilitation guidance and health monitoring[3]. Nevertheless, some patients often encounter interrupted professional care, insufficient rehabilitation guidance, and delayed identification of abnormal conditions after discharge, which may increase the incidence of complications and the risk of unplanned readmission, leading to poor prognosis and imposing a heavy burden on families and society[4,5].

As an emerging model integrating information technology with healthcare services, digital out-of-hospital health monitoring utilizes mobile terminals such as smartphones and wearable devices to collect and transmit patients' physiological data, including body weight, temperature, and blood pressure, to the medical team in real time, enabling remote monitoring, abnormality alerts, and personalized guidance, serving as an extension and expansion of transitional care services[6,7]. This model allows patients to receive continuous health monitoring and professional support after discharge, facilitating early detection of abnormal issues during postoperative recovery, reducing unnecessary outpatient visits, and lowering the risk of complication-related readmission[8,9]. However, the actual effectiveness of digital out-of-hospital health monitoring is highly dependent on patients' active participation and cooperation, and patients' willingness and behavior to participate are primarily based on adequate cognition of the service[10].

Postoperative patients in gastrointestinal surgery are predominantly middle-aged and elderly individuals with generally low digital literacy, who may face specific barriers to the cognition and acceptance of emerging technologies[11]. A comprehensive and accurate understanding of the cognitive status and influencing factors of digital out-of-hospital health monitoring among postoperative patients in gastrointestinal surgery is of great significance for guiding clinical practice in developing targeted health education strategies and promoting the implementation of this service in this population. Previous studies have explored the willingness to use mobile health services and its influencing factors in populations such as patients with chronic heart failure[12] and diabetes[13]. However, postoperative patients in gastrointestinal surgery are characterized by complex conditions, diverse treatment modalities, and individualized rehabilitation needs, and their cognitive characteristics may differ considerably from those of other disease populations.

This study aimed to investigate the cognitive status of digital out-of-hospital health monitoring among postoperative patients in gastrointestinal surgery and analyze its influencing factors, with a view to providing empirical evidence for developing more targeted health education programs and promoting the implementation of digital transitional care services in clinical practice.

MATERIALS AND METHODS

Participants

This study used a cross-sectional design. A convenience sampling method was used to recruit postoperative patients discharged from the gastrointestinal surgery department of Zhuhai Hospital of Integrated Traditional Chinese and Western Medicine between June 2024 and August 2024.

Inclusion criteria: (1) Aged ≥ 18 years; (2) Underwent surgical treatment in the gastrointestinal surgery department and were discharged; (3) Alert and oriented, with basic language comprehension and communication ability; (4) Able to use a smartphone; and (5) Provided informed consent and voluntarily participated.

Exclusion criteria: (1) Severe mental illness or cognitive impairment; (2) Critically ill or in a terminal state; or (3) Unable to complete the questionnaire due to visual, hearing, or other impairments.

A total of 146 patients were initially screened, of whom 9 were excluded prior to questionnaire administration. Consequently, 137 eligible patients completed the questionnaires. After excluding 11 invalid questionnaires, 126 valid questionnaires were obtained, yielding an effective response rate of 91.97% (Figure 1). This study was approved by the Ethics Committee of the Zhuhai Hospital of Integrated Traditional Chinese and Western Medicine (Approval No. 2024-03-038-E01).

Figure 1 Participant flow diagram of the cross-sectional study.

General information questionnaire

This questionnaire was self-designed by the researchers, covering sociodemographic variables (sex, age, education level, marital status, monthly household income), clinical variables (primary diagnosis, surgical approach, postoperative time, presence of stoma, presence of complications, postoperative chemotherapy), and technology exposure variables (years of smartphone use, prior use of health-related apps or mini-programs, average daily internet use time).

Digital health monitoring cognition questionnaire

This questionnaire was self-designed based on literature review and expert consultation, comprising 4 dimensions with a total of 15 items, scored on a 5-point Likert scale (1 = strongly disagree/strongly unfamiliar, 5 = strongly agree/strongly familiar).

Basic cognition dimension (items 1-3): Assessed patients' basic understanding of digital out-of-hospital health monitoring, including whether they had heard of the service, whether they understood its basic content and procedures, and whether they knew the channels through which to access it.

Functional cognition dimension (items 4-7): Evaluated patients' knowledge of specific functions, including data uploading, remote consultation, abnormality alerts, and data recording and storage.

Value cognition dimension (items 8-12): Assessed patients' recognition of the service's value, including early detection of problems, reduction of outpatient visits, reduction of readmission risk, enhanced sense of security, and improved patient-provider communication.

Concern cognition dimension (items 13-15): Evaluated patients' level of worry about potential risks, including privacy breaches, operational complexity, and information accuracy. This dimension was reverse-scored, with higher scores indicating fewer concerns and greater acceptance.

Pilot survey and reliability and validity test

Prior to the formal survey, a pilot study was conducted with 30 postoperative patients discharged from the gastrointestinal surgery department. Three clinical experts in gastrointestinal surgery and two nursing experts were invited to evaluate the content validity. The scale-level content validity index was 0.864, and the item-level content validity index ranged from 0.80 to 1.00. The pilot data showed that the overall Cronbach's α coefficient of the questionnaire was 0.908, with the α coefficients for each dimension ranging from 0.758 to 0.896, indicating satisfactory reliability and validity.

Statistical analysis

Data were analyzed using SPSS 26.0. Categorical data were described as n (%). Continuous data were tested for normality using the Kolmogorov-Smirnov test and conformed to a normal distribution, expressed as mean ± SD. An independent-samples t test was used for comparisons between two groups, and one-way analysis of variance (ANOVA) was used for comparisons among multiple groups. Variables with P < 0.05 in univariate analysis were included in a multiple linear regression model to identify independent factors influencing the cognitive level of digital out-of-hospital health monitoring. The significance level was set at α = 0.05, and P < 0.05 was considered statistically significant.

RESULTS

General characteristics of participants

Among the 126 patients, 66 were male (52.4%) and 60 were female (47.6%). The majority were aged over 45 years (62.7%). Most patients underwent laparoscopic surgery (73.0%), and 59.52% were more than 12 weeks post-surgery. Other general characteristics are shown in Table 1.

Table 1 General characteristics of participants.

Item	Category	n	Percentage (%)
Sex	Male	66	52.4
	Female	60	47.6
Age (years)	< 45	47	37.3
	45-60	42	33.3
	> 60	37	29.4
Education level	Primary school or below	7	5.6
	Junior high school	28	22.2
	Senior high school or secondary vocational	34	27.0
	College or above	57	45.2
Marital status	Married	93	73.8
	Unmarried	15	11.9
	Divorced or widowed	18	14.3
Monthly household income (yuan)	< 6000	54	42.9
	> 6000	72	57.1
Primary diagnosis	Colorectal cancer	30	23.8
	Intestinal obstruction	4	3.2
	Appendicitis	30	23.8
	Gallbladder stones/polyps	55	43.7
	Other	7	5.6
Surgical approach	Laparoscopic	92	73.0
	Open	3	2.4
	Other	31	24.6
Postoperative time (weeks)	< 4	22	17.5
	4-12	29	23.0
	> 12	75	59.5
Presence of stoma	Yes	5	4.0
	No	121	96.0
Complications	Yes	9	7.1
	No	117	92.9
Postoperative chemotherapy	Yes	31	24.6
	No	95	75.4
Years of smartphone use	< 3	26	20.6
	3-5	36	28.6
	> 5	64	50.8
Prior use of health-related apps	Yes	57	45.2
	No	69	54.8
Average daily internet use time (hours)	< 3	37	29.4
	3-5	39	31.0
	> 5	50	39.7

Cognitive scores of digital out-of-hospital health monitoring

The mean item score of cognition among the 126 patients was (3.02 ± 0.43), indicating a moderately low overall level. Scores across dimensions were similar, with value cognition scoring slightly higher (3.08), while basic cognition and concern cognition scored the lowest (both below 3) (Table 2).

Table 2 Scores of each dimension of cognition of digital out-of-hospital health monitoring among postoperative patients in gastrointestinal surgery.

Dimension	Number of items	Theoretical score range	Score (mean ± SD)	Item mean score	Actual minimum	Actual maximum	Standardized score rate (%)
Basic cognition	3	3-15	8.67 ± 2.748	2.89	3	15	57.8
Functional cognition	4	4-20	12.00 ± 3.129	3.00	4	19	60.0
Value cognition	5	5-25	15.39 ± 3.075	3.08	9	25	61.6
Concern cognition	3	3-15	9.36 ± 2.515	2.88	4	15	57.6
Total scale	15	15-75	45.41 ± 6.427	3.02	28	61	60.4

As shown in Table 3, the three highest-scoring items focused on complication prevention and early warning functions, whereas the three lowest-scoring items centered on service awareness and operational cognition. All high-scoring items fell within the value cognition and functional cognition dimensions, while low-scoring items were mainly in the basic cognition dimension, suggesting that patients had preliminary perceptions of the potential value of digital monitoring, yet still lacked fundamental understanding of the service content, operational procedures, and access channels. Furthermore, the mean score of all items did not reach 4, and the gap between the highest and lowest items was not large, indicating that cognitive weakness was a global phenomenon rather than a problem confined to specific dimensions. The original mean scores of all items in the concern dimension exceeded 3, suggesting that concerns were prevalent and constituted a significant barrier to the translation of cognition into behavior.

Table 3 Top 3 and bottom 3 scoring items.

Ranking	Item	Dimension	Score (mean ± SD)
Top 3 items
1	I think digital out-of-hospital health monitoring can help reduce the risk of readmission due to complications	Value cognition	3.14 ± 1.43
2	I know that when uploaded data are abnormal, the system can automatically alert the medical team	Functional cognition	3.12 ± 1.40
3	I think digital out-of-hospital health monitoring can improve communication between me and healthcare staff	Value cognition	3.11 ± 1.49
Bottom 3 items
1	I understand the basic content and procedures of digital out-of-hospital health monitoring	Basic cognition	2.84 ± 1.41
2	I know that I can upload health data such as weight and body temperature to the medical team via mobile phone	Functional cognition	2.87 ± 1.35
3	I have heard of "digital out-of-hospital health monitoring" or similar out-of-hospital health management services	Basic cognition	2.90 ± 1.47

Univariate analysis of cognitive scores by patient characteristics

Univariate analysis was conducted with the total cognitive score as the dependent variable across different demographic and clinical characteristics. The results showed that age, education level, years of smartphone use, and prior use of health-related apps had statistically significant effects on the total cognitive score (P < 0.05), as shown in Table 4.

Table 4 Univariate analysis of total cognitive scores among postoperative patients in gastrointestinal surgery by different characteristics.

Item	Category	Total cognitive score (mean ± SD)	Statistic	P value
Age (years)			F = 5.304	0.006
	< 45	47.55 ± 6.55
	45-60	45.00 ± 5.25
	> 60	43.16 ± 6.76
Education level			F = 4.472	0.005
	Primary school or below	37.86 ± 7.20
	Junior high school	45.11 ± 6.45
	Senior high school or secondary vocational	44.97 ± 5.18
	College or above	46.75 ± 6.45
Years of smartphone use			F = 4.807	0.010
	< 3	42.50 ± 6.74
	3-5	44.86 ± 5.61
	> 5	46.91 ± 6.36
Prior use of health-related apps			t = 2.818	0.006
	Yes	47.14 ± 5.87
	No	43.99 ± 6.56

Sex, marital status, monthly household income, primary diagnosis, surgical approach, postoperative time, presence of stoma, complications, postoperative chemotherapy, and average daily internet use time showed no statistically significant differences (all P > 0.05) and are not listed in the table.

Multiple linear regression analysis of factors influencing cognition

With the total cognitive score as the dependent variable, the four variables that were statistically significant in univariate analysis were entered into multiple linear regression. The results showed that years of smartphone use (β = 0.203, P = 0.018) and prior use of health-related apps (β = -0.176, P = 0.039) were independent factors influencing cognitive level, as shown in Table 5.

Table 5 Multiple linear regression analysis of factors influencing cognition.

Variable	B	SE	β	t value	P value
Constant	44.713	3.864	-	11.571	< 0.001
Years of smartphone use	1.644	0.684	0.203	2.402	0.018
Prior use of health-related apps	-2.263	1.087	-0.176	-2.082	0.039
Age group	-1.271	0.761	-0.161	-1.670	0.098
Education level	0.917	0.653	0.135	1.404	0.163

Variable assignment: Prior use of health-related apps: Yes = 1, No = 2. Age group: < 45 = 1, 45-60 = 2, > 60 = 3. Education level: Primary school or below = 1, Junior high school = 2, Senior high school or secondary vocational = 3, College or above = 4. Years of smartphone use: < 3 = 1, 3-5 = 2, > 5 = 3. For the variable "prior use of health-related apps" (coded 1 = Yes, 2 = No), the negative β indicates that patients without prior app experience had lower cognitive scores than those with prior experience.

DISCUSSION

This study showed that the mean item score of cognition of digital out-of-hospital health monitoring among 126 postoperative patients in gastrointestinal surgery was (3.02 ± 0.43), indicating a moderately low overall level, with standardized score rates for each dimension ranging from 57.6% to 61.6%, and no item reaching a mean score of 4, suggesting generally weak cognition of this service in this population. This finding is consistent with the systematic review by De La Cruz Monroy and Mosahebi[14], which reported that although surgical patients' overall impression of perioperative mobile health applications was positive, significant barriers remained in terms of confidence in use and digital health literacy, which may leave most patients in a state of "willingness but insufficient knowledge". Notably, in this study, the value cognition dimension scored slightly higher (3.08), while basic cognition (2.89) and concern cognition (2.88, reverse-scored) were both below 3, presenting a typical cognitive structure characterized by "value recognition preceding basic cognition". In terms of specific items, the three highest-scoring items focused on complication early warning and safety assurance functions ("reducing readmission risk" 3.14, "system automatically alerting the medical team" 3.12, "improving patient-provider communication" 3.11), whereas the three lowest-scoring items all fell within the basic cognition and functional cognition dimensions ("understanding basic content and procedures" 2.84, "knowing how to upload data" 2.87, "having heard of the service" 2.90). This "inverted" cognitive structure reveals an important practical implication: Patients hold clear expectations for the clinical value of digital out-of-hospital health monitoring, particularly showing strong interest in functions related to postoperative complication prevention and control, yet they still lack basic understanding of the concept, operational procedures, and access channels of the service itself. Chu et al[15] similarly found in their qualitative study of a remote hypertension monitoring program that physician recommendation and willingness to manage one's condition were the primary motivators for patient participation. In this study, the highest-scoring items also centered on complication early warning functions, indicating that postoperative patients also have a clear motivation for health management. However, the notable gap between management motivation and cognitive level suggests that willingness to participate alone is insufficient to drive actual use, and it is necessary to simultaneously strengthen cognitive education on service concepts and operational pathways. Sinyard et al[16], in their qualitative study on the acceptability of mobile health technology among elective surgery patients, also found that concerns about postoperative complications were the core motivation driving patients to adopt remote monitoring tools, and patients were more inclined to accept digital functions that directly address their safety concerns, a conclusion that corroborates the high recognition of early warning functions observed in this study.

In terms of influencing factors, univariate analysis showed that age, education level, years of smartphone use, and prior health app experience were significant factors affecting the total cognitive score (P < 0.05), with cognitive level decreasing with increasing age and increasing with higher education levels and greater digital experience. Among these, the between-group difference in education level was the most pronounced, with the total cognitive score of the primary school or below group (37.86 ± 7.20) nearly 9 points lower than that of the college or above group (46.75 ± 6.45), suggesting that patients with lower education levels face a greater cognitive gap in understanding digital health monitoring. Multiple linear regression further revealed that years of smartphone use (β = 0.203, P = 0.018) and prior health app experience (β = -0.176, P = 0.039) were independent factors influencing cognitive level, whereas age and education level did not enter the regression model after controlling for other variables. This finding suggests that, compared with static sociodemographic characteristics, dynamically accumulated digital technology contact experience better predicts patients' cognitive level of digital out-of-hospital health monitoring. Buawangpong et al[17] found in their analysis of willingness to use mHealth among middle-aged and older Thai adults that, among those who had never used mobile health applications, smartphone ownership was the strongest predictor of willingness to use, a finding consistent with the regression results of this study, namely that digital contact experience outweighs sociodemographic characteristics in predicting patients' cognition and acceptance of mobile health tools. Xu et al[11], in their qualitative study on electronic health literacy among colorectal cancer patients, also noted that older patients face particular barriers to the cognition and acceptance of emerging technologies due to low digital literacy, and that such barriers can be gradually ameliorated through increased technological contact and experience. Gillie et al[18] recommended that, when promoting remote monitoring, target populations should not be presupposed based solely on age or education level; rather, simple digital literacy screening should be used to identify patients genuinely requiring additional support, a suggestion that holds similar applicability in this study. Notably, even in the highest-scoring subgroups (college or above, years of smartphone use > 5 years), the total cognitive score was only approximately 47 points, still considerably lower than the full score of 75, indicating that cognitive inadequacy is a common problem across all subgroups rather than being confined to older adults or those with lower education levels.

In terms of concerns, the concern cognition dimension in this study scored only 2.88 after reverse scoring, suggesting that patients harbor widespread worries about the potential risks of digital out-of-hospital health monitoring, with data accuracy and privacy security being the primary sources of concern. This finding is consistent with the qualitative study by Wood et al[19] on the remote self-monitoring experiences of patients with chronic cardiopulmonary diseases, which found that patients had significant concerns about the accuracy of self-measured data, especially regarding the assessment of complex conditions, where they believed that devices could hardly reflect their true health status accurately. Such concerns may stem from the occasional data loss and high false-positive rates observed in remote monitoring systems in clinical practice[20]. Furthermore, Seel et al[21], in their controlled study of day surgery patients, found that those who used a structured web-based education platform showed significant improvement in health knowledge application ability (P < 0.05), whereas the self-directed search group showed no significant change, suggesting that cognitive gaps can be effectively bridged through structured education. Walker et al[22] found in their systematic review of patient experiences with remote monitoring for chronic diseases that one of the core benefits perceived by patients was the ability to trigger timely clinical assessment and treatment. This expectation of "timely access to care" makes patients highly sensitive to feedback speed. If uploaded data do not elicit timely responses, patients may doubt the reliability of the system. Therefore, incorporating automatic verification and immediate feedback functions into platform design is of great significance for maintaining patient trust and sustained use intention. At the practical level, health education should therefore target all discharged patients, with emphasis on strengthening the popularization of service concepts and the introduction of operational pathways, while simultaneously addressing patients' concerns through technical measures such as data verification feedback and privacy encryption explanations. Collectively, these strategies aim to bridge the critical gap from "perceived usefulness" to "actual use", and these challenges suggest that multi-level interventions are needed. To guide future interventional studies, we propose a hypothetical logic model (Figure 2) linking the key barriers identified in our study to modifiable intervention strategies. This framework is intended as a starting point for future testing and validation.

Figure 2 A Hypothetical framework for targeted interventions to improve cognition and uptake of digital out-of-hospital health monitoring among postoperative gastrointestinal surgery patients. UI: User interface.

Regarding the cognitive structure, it is important to recognize that the total score might obscure differences among specific dimensions. As observed in our results, patients scored highest in value cognition but lowest in basic and concern cognition. This suggests that different factors may influence these dimensions differently; for instance, while digital contact experience might improve patients' recognition of the service's value, it may not effectively bridge their fundamental knowledge gaps or alleviate their concerns about privacy and data accuracy. Therefore, treating value, basic knowledge, functional knowledge, and concerns as distinct constructs provides a more comprehensive understanding of patient barriers than a summed total score, which should be a focus for future research.

Based on the above analysis, this study identified multiple challenges faced by postoperative patients in gastrointestinal surgery regarding the cognition and use of digital out-of-hospital health monitoring. First, cognitive foundations are weak: Patients lack basic understanding of the service concept, operational procedures, and access channels, presenting a cognitive gap characterized by "value recognition preceding basic cognition". Second, individual differences are significant: Patients with lower education levels face greater difficulties in understanding and using digital monitoring services due to insufficient digital literacy. Studies have shown that education level is a fundamental determinant of digital health literacy, and patients with lower educational attainment face systematic barriers in text comprehension and digital navigation[23]. Some patients exhibit excessive dependence on healthcare professionals, with insufficient self-efficacy and a lack of confidence to actively try digital monitoring tools. Research has found that low self-efficacy is an important psychological factor hindering patients' adoption of mobile health technologies[24,25]. Third, concerns are widespread: Worries about data accuracy and privacy security constitute key barriers to the transition from "perceived usefulness" to "actual use". Data quality issues and privacy risks in remote monitoring have been confirmed as major obstacles to patients' willingness to use[26]. Fourth, the digital divide is prominent: Patients with limited smartphone experience have markedly lower cognitive levels, and cognitive inadequacy is not confined to older adults or those with lower education levels but is a common problem across all subgroups. These challenges may also vary across different clinical subgroups. Our sample predominantly included patients with benign diseases and minor-to-moderate procedures. Compared to these patients, those with malignant tumors or undergoing major surgeries often experience greater surgical trauma and psychological stress, and thus may have a more urgent need for continuous monitoring; however, their cognitive barriers may be distinct. Future studies should explore these subgroup-specific differences.

To address these challenges, targeted strategies are recommended at the following three levels. At the health education level, the introduction of the concept and operational training of digital out-of-hospital health monitoring should be incorporated into the standardized discharge preparation process, with stratified education provided for all discharged patients. For patients with low education levels and older adults, plain language, illustrated demonstrations, and bedside hands-on practice can be adopted to lower the comprehension threshold; pictorial health materials have been proven effective in enhancing the understanding and engagement of patients with low health literacy[27,28]. At the technology optimization level, platform design should fully consider the usage habits of middle-aged and older patients by simplifying interface interactions, integrating age-friendly features such as voice input and one-click upload, while strengthening automatic data verification and intelligent abnormal value filtering mechanisms to reduce the erosion of patient trust caused by false-positive alerts, and addressing security concerns through transparent measures such as encrypted data transmission and privacy authorization statements. At the service transition level, patients should complete their first data upload and master basic operations before discharge, achieving a seamless transition from in-hospital education to out-of-hospital monitoring, while establishing a regular feedback mechanism by the medical team so that patients genuinely feel that their monitoring data receive timely attention and response.

This study has several limitations. First, the regression model explained only 15.0% of the variance in cognitive scores, and the effect sizes of the included predictors (β = 0.203 and β = -0.176) were modest, indicating limited predictive power. Other potentially important factors, such as health literacy, technology anxiety, patient-provider trust, and illness perception, were not included. Future research should adopt theory-driven frameworks and validated instruments to improve explanatory power. Second, convenience sampling may have introduced self-selection bias, as patients willing to participate may have differed systematically from those who declined. Baseline characteristics of responders vs non-responders were not compared, and future studies should adopt random sampling and record non-responder characteristics to assess the direction and magnitude of selection bias. Third, the sample predominantly comprised patients with benign diseases and minor-to-moderate procedures; pooling these distinct clinical groups may have obscured subgroup-specific cognitive characteristics, and subgroup analysis was not feasible due to limited numbers. Fourth, the cross-sectional design precludes causal inference. Fifth, the self-designed questionnaire demonstrated acceptable reliability and content validity in the pilot test, but its construct validity has not been established through exploratory or confirmatory factor analysis. Although we reported the scores for each dimension, the regression analysis was performed on the total score rather than modeling the four dimensions as distinct constructs. Future studies should explore dimension-level associations to capture nuanced differences. In addition, prior exposure to digital health monitoring was not assessed in a standardized manner; future studies should include such an assessment. Future research should validate the instrument in a larger, independent sample.

CONCLUSION

This study found that postoperative gastrointestinal surgery patients had moderately low cognition of digital out-of-hospital health monitoring. However, the regression model explained only 15.0% of the variance, suggesting that most individual differences remain unexplained and require further investigation. Scores across all dimensions were similarly low, revealing a structural imbalance where value cognition exceeded basic knowledge, while concerns about data accuracy and privacy persisted. Smartphone experience and prior health app use were independent predictors, suggesting that digital contact experience outweighs demographic characteristics in this sample. Therefore, stratified education should target all discharged patients to strengthen their basic understanding of the service concept and operational procedures, while technical optimization, including automatic verification, immediate feedback, and privacy protection, is needed to alleviate concerns and bridge the gap between perceived usefulness and actual use.