Technostress in older healthcare workers: Latent profiles and implications for care robot acceptance

Abstract

Background

The adoption of care technologies in long-term care settings is ongoing; while the aging of the care workforce continues, particularly in South Korea. This study aims to identify technostress types among older care workers in Korea, including their influencing factors, and examine differences in acceptance of transfer-assistive care robots.

Methods

The study analyzed data from a 2023 survey on 336 care workers aged 50 years and above. Latent profile analysis was conducted on five technostress dimensions: overload, invasion, complexity, privacy, and inclusion. The Three-Step method and the Bolck–Croon–Hagenaars command were used to examine the antecedents of profile membership and differences in technology acceptance variables, including self-efficacy, anxiety, attitude, perceived ease of use, perceived usefulness, and intention to use robots.

Results

Four technostress profiles were identified: low technostress, high complexity and inclusion, moderate technostress, and high technostress with low inclusion. Age, education, income, job position, type of facility, work experience, and digital competence were all significantly associated with technostress profile membership. The low technostress group demonstrated the most favorable outcomes across all technology acceptance variables, whereas the high complexity and inclusion group exhibited the least favorable outcomes, characterized by the highest levels of anxiety and lowest levels of self-efficacy, perceived usefulness, and behavioral intention to use.

Conclusions

Technostress profiles significantly influenced older healthcare workers’ acceptance of transfer-assistive robots. Tailored interventions to decrease technostress—particularly related to complexity and inclusion pressures—may enhance technology adoption and improve care delivery in healthcare settings.

Keywords

technostress healthcare worker care robot transfer-assistive robots care worker latent profile analysis

Introduction

The global trend of population aging poses a critical challenge to the sustainability of long-term care (LTC) systems. The OECD estimates that the number of LTC workers needs to increase by approximately 32% over the next decade.¹ South Korea represents an especially urgent case: in 2025, the population aged 65 years and older exceeded 10.6 million, which surpasses 20% of the total population and confirms the nation’s transition into a super-aged society.² This demographic transformation is intensifying labor shortage and workforce aging.³ Notably, South Korea has recorded the highest average age of healthcare workers among OECD countries—above 50 years as of 2016.³ Furthermore, the proportion of care workers aged 60 years and older increased from 40% in 2019 to 48% in 2022, with those aged 70 years and over nearly doubling from 8% to 15%–.^4,5 These patterns underscore the urgent need for systems that support older healthcare workers in maintaining employability and adapting to technological change.⁶ Along with these demographic shifts, the decreasing number of elderly care personnel raises the question of how robotics can contribute to maintaining or improving the quality of elderly care.⁷

Technological innovation is increasingly promoted as a potential solution to these workforce challenges. In particular, care robots have attracted scholarly attention due to their potential to address staff shortages while enhancing care quality.^8,9 Transfer-assistive robots represent one of the highest-priority categories, because they target the physically demanding task of patient transfers,^10–13 which are a leading cause of musculoskeletal disorders and injuries among care staff.^14–16 Scholars have demonstrated that such health deterioration accelerates workforce exit.¹⁷ By mitigating these risks, transfer-assistive robots hold the potential to extend the careers of older workers and safeguard care quality. However, physical relief alone is insufficient. Their successful integration into practice ultimately depends on user acceptance.¹⁸

Accordingly, recent studies have increasingly focused on the acceptance of care robots among healthcare workers. Drawing on Davis’s technology acceptance model (TAM),¹⁹ researchers have identified self-efficacy, anxiety, perceived usefulness, perceived ease of use, attitudes, and intention to use as key predictors of robot acceptance.^13,20,21 For instance, Yoon and Kim¹² found that self-efficacy, anxiety, perceived usefulness, and perceived ease of use significantly influenced the intention of Korean healthcare workers to use transfer-assistive robots. In Finland, Latikka et al.²⁰ examined healthcare workers’ acceptance of humanoid, telepresence, companion, and transfer-assistive robots, and consolidated nine factors—attitude, usefulness, ease of use, enjoyment, trust, adaptability, facilitating conditions, anxiety, and familiarity—into a single construct of acceptance. However, healthcare workers’ responses to robots may vary substantially depending on their professional roles and caregiving responsibilities, and even the same robot may elicit divergent reactions within the healthcare workforce, as demonstrated in a case study.²² This suggests the need for approaches that explicitly consider heterogeneity within a user group.

Apart from TAM-based predictors, external variables such as digital literacy and psychological factors related to technology have also attracted increased scholarly attention. Within this context, technostress—psychological strain caused by interaction with Information and Communication Technology(ICT)²³—has emerged as a significant barrier to technology adoption. Notably, technostress has also been observed at substantial levels in hospital settings; one hospital-based study reported that 41% of healthcare practitioners experienced moderate technostress and 36% experienced high technostress.²⁴ Initially conceptualized as a maladaptive response to technological change,²⁵ technostress is now widely understood through five dimensions: techno-overload, techno-complexity, techno-insecurity, techno-uncertainty, and techno-invasion.²³ Techno-overload refers to situations in which ICT use requires individuals to work faster and for longer periods than they can reasonably manage. Techno-complexity refers to situations in which the complexity associated with ICT leads individuals to feel inadequate in their technology-related skills and forces them to invest additional time and effort in learning and understanding ICT. Techno-insecurity is associated with situations in which individuals feel threatened about losing their jobs, either due to ICT-driven automation or to other people who have a better understanding of ICT. Techno-uncertainty refers to contexts in which continuous ICT changes and upgrades unsettle individuals and create uncertainty, requiring them to constantly learn and adapt to new ICT. Techno-invasion describes the intrusive impact of ICT in situations where individuals feel the need to be constantly connected and reachable, thereby blurring the boundaries between work and personal life.²⁶ Consistent with this multidimensional framework, the same study identified techno-uncertainty as the most pronounced contributor to technostress among healthcare practitioners, followed by techno-overload and techno-complexity.²⁴ Building on this concept, Nimrod²⁷ developed a technostress scale specifically for older adults that includes the following dimensions: techno-overload, techno-invasion, techno-complexity, techno-privacy, and techno-inclusion. The techno-inclusion represents a distinct threat, thus capturing feelings of inadequacy when comparing oneself to younger users. Accumulating evidence associates technostress with decreased technology adoption.^28–33 Gabbiadini et al.²⁹ integrated technostress into the TAM as a key external factor, reporting that increased technostress among Italian university faculty negatively influenced perceptions of ease of use. Similarly, Zhang³¹ demonstrated that technostress diminishes students’ self-efficacy, which, in turn, increases resistance to innovation. Wang and Yu³² found that technostress among art students exerted direct negative effects on perceived usefulness and continued usage intention. In a workplace setting, Chang et al.³³ reported that hindrance technology stressors impede the AI adoption intention of employees by triggering AI-related anxiety. In the healthcare context, Tacy et al.²⁸ reported that technostress among nurse faculty inhibited technology use. Similarly, Thunberg et al.³⁰ found that Swedish healthcare workers faced difficulties in adopting and using welfare technology due to various factors such as older age or negative attitudes toward technology. However, other participants neither reported uncertainty or anxiety regarding technology nor perceived it as extremely complex or overwhelming at the initial adoption stage. These findings imply that technostress is not a universal experience, thus highlighting its nuanced patterns and underscoring the importance of examining various contributing factors. In this context, technostress research suggests that experiences of ICT interaction can be categorized into at least three distinct user types.³⁴ Moreover, certain technostress subgroups have been found to show strong associations with psychological symptoms.³⁴ Taken together, these findings indicate that technostress should be understood not merely as a matter of overall intensity, but as a heterogeneous experience that may involve distinct patterns and psychological consequences.

Conceptually, technostress is not a unitary construct; rather, it can manifest through multiple sources of strain that co-occur and vary across individuals.^25,26 Technostress in healthcare is further associated with anxiety and reluctance to engage with digital systems,³⁵ which directly undermines technology adoption and care delivery. Older healthcare workers are especially vulnerable, because they frequently report low levels of digital literacy and elevated psychological barriers to learning new technologies.³⁶ This vulnerability is closely related to digital literacy—encompassing not only technical skills but also confidence and attitudes toward technology—and to the digital divide that persists even when access to digital tools is ensured, particularly among older adults, which may intensify psychological strain during technology use and increase susceptibility to technostress.³⁷

Although research on technostress and technology acceptance has been increasing, studies that specifically examine technostress in relation to the adoption of care robots among older healthcare workers remain limited. Meanwhile, digital healthcare automation using robotics and AI has been proposed as a potential contributor to the sustainability of long-term care and healthcare systems.³⁸ However, evidence from AI-enhanced healthcare and eldercare research suggests that, despite technological advances, older users often demonstrate ambivalence or resistance toward physical robots, highlighting the central role of psychological and acceptance-related factors in real-world implementation.³⁹

Moreover, the existing research designs are predominantly variable-centered, assuming population homogeneity and overlooking subgroup differences in stress experiences.^40,41 Nevertheless, recent findings demonstrate that individuals with similar overall levels of technostress may substantially differ in the relative intensity of the five dimensions.³⁰ To capture this heterogeneity, person-centered approaches are required.

Latent profile analysis (LPA) offers a suitable framework for this objective, as it enables researchers to identify distinct subgroups with unique patterns of technostress.^42,43 Integrating this perspective into technology acceptance frameworks is particularly valuable. While models such as the TAM and UTAUT highlight predictors (e.g., perceived usefulness, ease of use, and self-efficacy), recent evidence indicates that technostress not only moderates these pathways but also directly hinders adoption in the absence of psychological support.^44,45 Thus, examining technostress profiles can provide in-depth insights into the interaction between psychological barriers and established acceptance factors in the formation of adoption outcomes.

Against this background, the present study aims to advance the current understanding of technostress among older care workers in South Korea, a nation faced with one of the fastest rates of population aging worldwide. Specifically, the study pursues three objectives: (1) to classify technostress profiles among older care workers using LPA, thereby capturing heterogeneous stress patterns, (2) to examine sociodemographic predictors of profile membership, thus offering insights into which characteristics are associated with technostress patterns, and (3) to analyze differences in the acceptance of transfer-assistive robots across subgroups, identifying psychological barriers that may hinder adoption. Its focus on older healthcare workers enables this study to contribute to the literature theoretically by integrating technostress profiles into technology acceptance frameworks and practically by offering policy-relevant insights to support workforce sustainability in super-aged societies.

Methods

Data and participants

Conducted by the AgeTech Research Institute at Kyung Hee University, the survey was intended to assess perceptions of care and transfer-assistive robots, levels of technostress, and technology acceptance among healthcare workers. Data were collected through an online survey administered by a professional research firm using convenience sampling between October and November 2023. The firm has extensive experience in conducting nationwide online surveys and developed the survey platform. To recruit participants, the firm contacted healthcare workers from long-term care facilities, long-term care hospitals, and general hospitals across metropolitan and provincial regions in South Korea. Participants were recruited from institutions located in all 17 metropolitan and provincial administrative regions nationwide. To reduce institutional clustering, the number of respondents from each facility was limited to one to three individuals. Prior to participation, all participants provided informed consent electronically.

As transfer-assistive robots are currently in the demonstration stage in South Korea and are not yet widely implemented in care settings, the online questionnaire included a description and a demonstration video illustrating their use in care settings to ensure participants’ understanding. Therefore, this study assessed participants’ anticipated acceptance of these technologies rather than acceptance based on actual use. All questionnaires used in this study were employed with permission from the original authors when required, and the original sources were appropriately cited. The Institutional Review Board of Kyung Hee University (IRB No: KHGIRB-23-415) reviewed and approved the study protocol, including the questionnaire and survey procedures, thus ensuring ethical and methodological validity.

The participants were from LTC facilities and hospitals; integrated nursing care service hospitals; and general hospitals—institutions that primarily provide care for older adults across South Korea. A total of 542 healthcare workers completed the structured questionnaire. Responses were screened for eligibility, and participants younger than 50 years were excluded because the study focused on older healthcare workers. This age cutoff (≥50 years) was applied because the technostress scale⁴⁶ used in this study was developed for older adults and has been used in prior research focusing on individuals aged 50 years and older.^47,48 After applying this criterion, the final analytic sample consisted of 336 participants. No missing data were identified in the variables used for the analysis. The majority were female (92.6%), with a mean age of 57.4 years (SD = 5.07). Regarding education, 50.0%, 26.5%, 15.5%, 7.4%, and 0.6% graduated from high school, college, university, middle school, and elementary school, respectively. The majority (86.3%) resided in urban areas, and their average monthly income was 2.67 million Korean won (M = 2.67, SD = 0.58). In terms of workplace, 36.3%, 46.4%, and 17.3% were employed in LTC facilities, LTC hospitals, and general hospitals, respectively. By occupation, 21.4%, 19.9%, 33.6%, 14.6%, and 10.4% were registered nurses, licensed practical nurses, certified care workers, personal care assistants, and physical/occupational therapists/other related professions, respectively. On average, participants had been working for 14.9 years (SD = 9.81) in their current profession. Overall, the demographic composition of the sample appears to reflect the broader workforce structure. The demographic characteristics of the sample are largely consistent with national statistics on older healthcare workers in South Korea. Approximately 48.3% of older healthcare workers are in their 60s, 26.8% are in their 50s, and 14.6% are in their 70s. In addition, women account for 94.4% of the workforce.⁴⁹

Instruments and measurements

Technostress

Technostress refers to stress in response to ICT use. This study assessed technostress using the Technostress Scale developed by Nimrod.⁴⁶ It consists of five subdimensions, namely, Overload (excessive ICT-related demands), Invasion (intrusion of ICT into daily life), Complexity (difficulties in learning and use), Privacy (threat to personal information), and Inclusion (feelings of inferiority and pressure to cope with younger users). Items were rated using a four-point Likert-type scale (1 = Not at all, 2 = Somewhat disagree, 3 = Somewhat agree, 4 = Strongly agree). Although the original instrument employed a five-point Likert scale, this study used a four-point Likert scale to reduce neutral responses and encourage participants to provide more definitive evaluations. Scores were calculated by averaging item responses; high scores indicate high levels of technostress. Items T2 and T12 were reverse-coded to ensure that higher scores consistently reflected greater technostress. Table 1 presents the mean, standard deviation, and Cronbach’s α of technostress, while the detailed items are provided in the Supplementary Materials.

Table 1.

Descriptive statistics and reliability of the technostress scale (N=336).

Variable		M	SD	Cronbach’s alpha
Overload	T1	2.43	0.70	0.781
	T2	2.41	0.69
	T3	2.44	0.77
Invasion	T4	2.37	0.71
Invasion	T5	2.54	0.75
Complexity	T6	2.71	0.79
	T7	2.77	0.76
	T8	2.72	0.78
Privacy	T9	2.50	0.72
	T10	2.54	0.76
	T11	2.63	0.75
Inclusion	T12	2.67	0.83
	T13	2.76	0.70
	T14	2.60	0.64

Acceptance of transfer-assistive robots

The study assessed acceptance of transfer-assistive robots using variables widely applied by previous empirical studies.^50–52 Self-efficacy, anxiety, attitude, perceived usefulness, perceived ease of use, and behavioral intention have been identified as key predictors of acceptance of care robot use.^20,53,54 Because transfer-assistive robots were not currently implemented in participants’ workplaces, all items were framed to assess anticipated acceptance and intention regarding potential future use. Each construct was rated using a five-point Likert-type scale ranging from 1 = Strongly disagree to 5 = Strongly agree, and item scores were averaged. The reliability coefficients for the six constructs were high: Cronbach’s α = 0.769, 0.893, 0.824, 0.875, 0.922, and 0.898, respectively. Table 2 presents the mean, standard deviation, and Cronbach’s α of the acceptance constructs for transfer-assistive robots, while the detailed items and their sources are provided in Table S2 in the Supplementary Materials.

Table 2.

Descriptive Statistics and Reliability of the acceptance of transfer-assistive robot (N=336).

Variable		M	SD	Cronbach’s alpha
Self-efficacy	SE1	3.34	0.91	0.769
Self-efficacy	SE2	3.26	0.94	0.769
Anxiety	ANX1	3.24	1.03	0.893
Anxiety	ANX2	3.13	0.98	0.893
Attitude	AT1	3.65	0.78	0.824
Attitude	AT2	3.55	0.78	0.824
Perceived usefulness	PU1	3.65	0.75	0.875
	PU2	3.73	0.75
	PU3	3.64	0.68
Perceived ease of use	PEOU1	3.13	0.87	0.922
Perceived ease of use	PEOU2	3.08	0.87	0.922
Intention to use	BI1	3.30	0.84	0.898
	BI2	3.40	0.85
	BI3	3.29	0.78

Statistical analysis

The descriptive statistics of all variables were calculated using IBM SPSS Statistics (Version 28.0). LPA was then conducted on the 14 items of the Technostress Scale as profile indicators using Mplus (Version 8.10) with robust maximum likelihood estimation (MLR), thus exploring solutions ranging from one to five profiles. LPA identified the optimal number of subgroups and estimated the probability of individual membership in each profile.⁵⁵

The study determined the optimal number of latent profiles on the basis of four widely used statistical criteria:

(1) Information indices, including AIC,⁵⁶ BIC,⁵⁷ and sample-size adjusted Bayesian Information Criterion (SABIC),⁵⁸ with low values indicating better fit;

(2) Model comparison tests, including the Lo–Mendell–Rubin (LMR) adjusted likelihood ratio test (LRT; presented as LMR-LRT) and the parametric bootstrapped LRT (BLRT), which compare models with k versus k − 1 profiles, in which p-value < .05 supports the k-profile model;

(3) Classification quality, assessed using entropy values ranging from 0 to 1; values close to 1 indicate greater accuracy. Specifically, an entropy value of approximately 0.6 corresponds to roughly 80% classification accuracy, while a value of 0.8 or higher indicates an accuracy of 90% or greater⁵⁹ and

(4) Subgroup size and proportion; each class is required to include at least 5% of the sample.

The final number of profiles was selected by jointly considering all statistical criteria and the interpretability of the solution; labels for each technostress typology were assigned according to subgroup characteristics.

The study conducted multinomial logistic regression using the three-step method (R3STEP) procedure⁴² in Mplus to examine whether antecedent variables could predict profile membership. The antecedent variables included demographic factors (age, gender, level of education, residential area, and monthly income), organizational factors (job position, type of facility, work experience, and musculoskeletal symptoms), and digital competence.⁶⁰ Finally, the Bolck–Croon–Hagenaars (BCH) approach⁶¹ was applied to examine differences in acceptance of transfer-assistive robots as a distal outcome across technostress profiles while accounting for classification uncertainty.

Results

Model selection

To examine technostress among the 336 older healthcare workers, latent profile models ranging from one to five profiles were estimated. Table 3 indicates that the four-profile model is the optimal solution. This decision was supported by (a) decreasing AIC, BIC, and SABIC values that plateaued after the two-profile model, (b) statistically significant results of the LMR-LRT and BLRT tests (p < 0.05) for models up to the four-profile solution. Importantly, the LMR-LRT for the five-profile model was not significant (p = 0.5895), indicating that the five-profile model did not provide a better fit than the four-profile model. (c) An entropy value more than 0.80, which indicates high classification accuracy (90%), and (d) the smallest class, comprising more than 5% of the sample. Considering statistical fit and practical interpretability, the four-profile model was determined to be most appropriate.

Table 3.

Fit statistics of latent profile analysis.

Model	AIC	BIC	SABIC	LMR-LRT	BLRT	Entropy	Smallest class proportion
1	10541.648	10648.527	10559.71	NA	NA	NA	100
2	9997.923	10162.058	10025.66	<.001	<.001	0.896	23.42
3	9749.951	9971.344	9787.361	0.0387	<.001	0.872	18.74
4	9635.647	9914.296	9682.73	0.0295	<.001	0.882	7.85
5	9563.36	9899.266	9620.119	0.5895	<.001	0.869	7.57

Note. BIC: Bayesian information criterion; AIC: Akaike information criterion; SABIC: Sample-size adjusted Bayesian information criterion; LMR-LRT: Lo–Mendell–Rubin likelihood ratio test; BLRT: bootstrap likelihood ratio test.

Characteristics and naming of latent profiles of technostress

Figure 1 and Table 4 present the four latent profiles of older healthcare workers identified based on their response patterns across the five dimensions of technostress. The profiles were labeled according to their distinguishing characteristics as follows.

Profile 1 Low technostress group (n = 64, 19.0%)

This group showed consistently low scores across all five technostress dimensions, indicating that these workers experienced relatively minimal ICT-related stress.

Profile 2 High complexity and inclusion pressure group (n = 26, 7.7%)

This was the smallest group and was characterized by the highest scores on the complexity and inclusion dimensions, while showing moderate levels of overload, invasion, and privacy concerns. These workers perceived ICT as difficult to learn and use and reported strong pressure to keep up with younger users and remain included in the technological environment. However, they did not report particularly high levels of excessive ICT demands, intrusion into daily life, or privacy threats.

Profile 3 Moderate technostress group (n = 183, 54.5%)

This was the largest group and displayed moderate scores across all five technostress dimensions. Compared with Profile 2, this group reported lower levels of perceived technological complexity and inclusion pressure, reflecting a more balanced and moderate experience of technostress.

Profile 4 High technostress with low inclusion pressure group (n = 63, 18.8%)

This group was characterized by relatively high levels of invasion, complexity, and privacy concerns, while reporting comparatively lower levels of inclusion pressure. In other words, these workers experienced substantial stress related to ICT intrusion into daily life, difficulty in using technology, and privacy risks, while feeling less pressure to keep up with younger users in the technological environment. Although overload was not particularly elevated in this group, the overall pattern suggests a form of technostress driven more by perceived technological burden and intrusion than by competitive pressure to remain technologically included. This pattern distinguishes Profile 4 from Profile 2, where perceived inclusion pressure was considerably higher.

Figure 1.

Latent profiles of technostress.

Table 4.

Subfactor means and standard errors by latent profile.

		Profile 1 (n=64, 19.0%)		Profile 2 (n=26, 7.7%)		Profile 3 (n=183, 54.5%)		Profile 4 (n=63, 18.8%)
		M	SD	M	SD	M	SD	M	SD
Overload	t1	1.91	0.09	2.28	0.15	2.50	0.07	2.87	0.13
	t2	2.51	0.12	2.51	0.13	2.47	0.05	2.10	0.09
	t3	1.68	0.10	2.49	0.12	2.45	0.07	3.16	0.09
Invasion	t4	1.81	0.09	2.09	0.08	2.36	0.08	3.12	0.10
Invasion	t5	1.92	0.08	2.63	0.11	2.41	0.05	3.52	0.19
Complexity	t6	1.82	0.10	3.50	0.10	2.78	0.05	3.12	0.12
	t7	1.88	0.08	3.31	0.11	2.89	0.05	3.13	0.15
	t8	1.80	0.08	3.37	0.10	2.88	0.06	2.96	0.18
Privacy	t9	1.84	0.08	2.46	0.12	2.57	0.06	3.01	0.24
	t10	2.06	0.09	2.17	0.08	2.55	0.06	3.18	0.15
	t11	2.41	0.10	2.48	0.11	2.59	0.06	3.08	0.17
Inclusion	t12	2.54	0.11	3.71	0.13	2.84	0.09	1.84	0.11
	t13	2.29	0.09	3.95	0.04	2.68	0.05	3.02	0.11
	t14	2.11	0.08	3.59	0.13	2.56	0.05	2.81	0.08

Analysis of antecedents of latent profiles of technostress

Table 5 presents the results of multinomial logistic regression using the R3STEP method.

Table 5.

Multinomial logistic regression results for predictors of latent technostress profiles.

		Profile 1 vs profile 4			Profile 2 vs profile 4			Profile 3 vs profile 4			Profile 2 vs profile 1			Profile 3 vs profile 1			Profile 3 vs profile 2
		Estimate	S.E.	Or (95% CI)	Estimate	S.E.	Or (95% CI)	Estimate	S.E.	Or (95% CI)	Estimate	S.E.	Or (95% CI)	Estimate	S.E.	Or (95% CI)	Estimate	S.E.	Or (95% CI)
Gender		1.19	2.11	3.30	-0.03	3.19	0.97	-0.20	2.43	0.82	-1.22	2.10	0.30	-1.39	1.06	0.25	-0.17	2.21	0.84
Age		-0.18	0.12	0.84	0.01	0.13	1.01	-0.22	0.11	0.81	0.18*	0.09	1.20	-0.04	0.06	0.96	-0.23**	0.07	0.80
Educational attainment		-1.25	0.74	0.29	-3.92*	1.56	0.02	-0.83	0.75	0.44	-2.67	1.38	0.07	0.42	0.36	1.52	3.09*	1.35	21.94
Monthly personal income		0.01	0.01	1.01	-0.03	0.01	0.97	-0.01	0.01	1.00	-0.03**	0.01	0.97	-0.01	0.01	0.99	0.02	0.01	1.02
Residential region		-0.16	1.06	0.86	-0.11	1.37	0.90	-0.65	1.05	0.52	0.05	0.99	1.05	-0.50	0.62	0.61	-0.54	0.93	0.58
Job position	Registered nurses	2.46	1.47	11.74	0.76	2.80	2.15	0.50	1.13	1.65	-1.70	2.41	0.18	-1.96	1.03	0.14	-0.26	2.42	0.77
	Licensed practical nurses	2.37	1.41	10.66	-3.79	2.72	0.02	0.72	0.99	2.05	-6.16*	2.70	0.00	-1.65	1.12	0.19	4.51	2.54	90.91
	Certified care workers	0.87	1.62	2.38	-3.41	2.18	0.03	0.51	1.27	1.66	-4.28*	1.97	0.01	-0.36	1.10	0.70	3.92*	1.78	50.19
	Personal care assistants	-0.57	2.00	0.57	-4.22	2.54	0.02	-1.73	1.79	0.18	-3.66	1.91	0.03	-1.17	1.17	0.31	2.49	1.75	12.06
	Others (Reference)
Type of facility	LTC facilities	29.40***	2.65	NA	25.20***	0.94	NA	26.05***	0.00	NA	-4.21	2.69	0.02	-3.35	2.65	0.04	0.85	0.94	2.34
	LTC hospitals	-1.01	2.52	0.37	-9.44**	3.10	NA	-4.52***	1.23	0.01	-8.43*	3.76	NA	-3.51	2.58	0.03	4.92	2.80	137.12
	General hospital (Reference)
Work experience (years)		0.11*	0.06	1.12	0.20*	0.09	1.22	0.12*	0.05	1.13	0.09	0.08	1.09	0.01	0.03	1.01	-0.08	0.07	0.93
Musculoskeletal pain (Yes/No)		0.84	0.71	2.33	-0.08	0.96	0.92	0.61	0.69	1.84	-0.93	0.81	0.40	-0.23	0.43	0.79	0.69	0.71	2.00
Digital competence		-0.65	0.75	0.52	-1.10	0.81	0.33	-1.43	0.74	0.24	-0.46	0.44	0.64	-0.78*	0.31	0.46	-0.32	0.36	0.72

Note. The values represent adjusted logit differences between latent profiles (e.g., Profile 1 vs. Profile 4) derived using the R3STEP method, which corrects for classification errors in latent profile analysis. Positive values indicate that the logit value for the comparison group (e.g., Profile 1) is higher than the reference group (e.g., Profile 4), while negative values indicate the opposite. “Reference” denotes the comparator category. “NA” indicates estimates that were not obtainable due to sparse data.

∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Regarding basic demographic factors, among healthcare workers, older age was associated with a higher likelihood of belonging to the High complexity and inclusion group (Profile 2) compared with both the Low technostress group (Profile 1) and the Moderate technostress group (Profile 3). Lower levels of education were associated with a higher likelihood of belonging to the High complexity and inclusion group (Profile 2) rather than the Moderate technostress group (Profile 3) or the High technostress with low inclusion group (Profile 4). In addition, lower income levels increased the likelihood of belonging to the High complexity and inclusion group (Profile 2) compared with the Low technostress group (Profile 1).

For organizational factors, job position showed significant differences across profiles. Compared with licensed practical nurses, individuals classified as others (physical therapists, occupational therapists, and other related professions) were more likely to belong to the High complexity and inclusion group (Profile 2) than to the Low technostress group (Profile 1). Similarly, compared with certified care workers, those in the other professions category were more likely to belong to the High complexity and inclusion group (Profile 2) rather than the Low technostress group (Profile 1) or the Moderate technostress group (Profile 3).

With regard to the type of facility, healthcare workers employed in general hospitals were more likely to belong to the High technostress with low inclusion group (Profile 4) than to other profiles compared with those working in long-term care (LTC) facilities. In addition, compared with workers in LTC hospitals, those employed in general hospitals showed a higher likelihood of belonging to the High technostress with low inclusion group (Profile 4) rather than Profile 2 or Profile 3. Furthermore, shorter work experience in the current occupation increased the likelihood of belonging to the High technostress with low inclusion group (Profile 4) compared with the Low technostress group (Profile 1). Due to small cell counts in some profile groups, several coefficients related to job position and type of facility were relatively large and should therefore be interpreted cautiously.

Finally, regarding digital competence, lower levels of digital competence were associated with a higher likelihood of belonging to the Moderate technostress group (Profile 3) compared with the Low technostress group (Profile 1).

Differences in acceptance levels of transfer-assistive robots according to technostress profiles

The BCH command was used to examine differences in the acceptance levels of transfer-assistive robots according to technostress profiles. In Table 6, significant differences were found in self-efficacy, anxiety, attitude, perceived ease of use, perceived usefulness, and intention to use.

Table 6.

BCH results for the differences on the acceptance of transfer assistive robot across latent profiles.

	Profile 1 (n=64, 19.0%)	Profile 2 (n=26, 7.7%)	Profile 3 (n=183, 54.5%)	Profile 4 (n=63, 18.8%)	BCHχ2	Comparison of different profiles
	M(SE)	M(SE)	M(SE)	M(SE)	BCHχ2	Comparison of different profiles
Self-efficacy	3.92(0.09)	2.49(0.15)	3.26(0.07)	3.11(0.09)	84.366***	1>3=4>2
Anxiety	2.47(0.13)	3.91(0.14)	3.14(0.07)	3.76(0.10)	84.291***	2=4>3>1
Attitude	3.97(0.09)	3.27(0.12)	3.47(0.06)	3.73(0.10)	31.138***	1=4>3=2
Perceived ease of use	4.05(0.08)	3.57(0.12)	3.54(0.05)	3.72(0.09)	30.695***	1>4=2=3
Perceived usefulness	3.68(0.09)	2.22(0.16)	2.91(0.06)	3.47(0.09)	90.761***	1=4>3>2
Intention to use	3.84(0.10)	2.94(0.13)	3.23(0.06)	3.28(0.09)	38.279***	1>4=3>2

First, self-efficacy was highest in the low technostress group followed by the moderate technostress group and high technostress with low inclusion pressure group (they demonstrated similar levels), while it was lowest in the high complexity and inclusion pressure group. Anxiety was highest in the high complexity and inclusion pressure group and high technostress with low inclusion pressure group (similar levels) followed by the moderate technostress group, whereas it was lowest in the low technostress group. Attitude was highest in the low technostress group and high technostress with low inclusion pressure group (similar levels) followed by the moderate technostress group; it was lowest in the high complexity and inclusion pressure group (similar levels). Perceived ease of use was highest in the low technostress group, with the remaining groups exhibiting similar levels. Perceived usefulness was highest in the low technostress group and high technostress with low inclusion pressure group (similar levels) followed by the moderate technostress group; meanwhile, it was lowest in the high complexity and inclusion pressure group. Finally, intention to use was highest in the low technostress group followed by the high technostress with low inclusion pressure group and moderate technostress group (similar levels) but lowest in the high complexity and inclusion pressure group.

Discussion

This study aimed to identify the types of technostress experienced by 336 healthcare workers aged 50 years and older in Korea, explore the factors that influence these types, and examine differences in technology acceptance model-related variables for transfer-assistive robots across the identified technostress profiles.

The study identified four distinct latent profiles of technostress, and each reflects unique patterns across the five dimensions of technostress. The largest proportion of participants belonged to the moderate technostress (profile 3) group followed by the low technostress group (profile 1), high technostress with low inclusion pressure group (profile 4), and the high complexity and inclusion pressure group (profile 2). Notably, Profiles 2 and 4 represent two distinct forms of technostress. Profile 2 is characterized by strong pressure to keep up with technological change, whereas Profile 4 reflects elevated technological burden accompanied by relatively low inclusion pressure. These patterns suggest that technostress among older healthcare workers may manifest in different forms, including pressure-driven technostress (Profile 2) and burden-driven technostress (Profile 4). These findings highlight the heterogeneity of technostress among healthcare workers, which indicates that the distribution of technostress across subgroups substantially varies. A similar classification was reported by Rašticová et al.,⁶² who identified four technostress clusters among European workers aged 50 years and older. However, the present findings suggest that technostress patterns among Korean healthcare workers differ in their configuration, likely reflecting contextual and occupational differences. This suggests that their technostress is driven more by perceived technological intrusion and risk than by competitive pressure, highlighting the importance of improving system usability, minimizing unnecessary digital interruptions, and strengthening data security measures. Overall, These findings highlight the need for profile-specific strategies rather than one-size-fits-all approaches in managing technostress in digital healthcare environments.

The study further explored the antecedents of latent technostress profiles using multinomial logistic regression, examining demographic factors, organizational characteristics, and digital competence as predictors. The results revealed that several individual characteristics were significantly associated with profile membership. Among healthcare workers aged 50 years and older, individuals who were older, had lower levels of education, and reported lower income were more likely to be classified into the high complexity and inclusion pressure group compared with the other technostress profiles. These findings suggest that sociodemographic disadvantage may increase vulnerability to specific dimensions of technostress, particularly those related to difficulties in learning and using ICT (complexity) and perceived pressures associated with technological inclusion. Older age may be associated with greater perceived difficulty in adapting to new technologies,⁶³ while lower educational attainment⁶⁴ and income⁶⁵ may reflect limited access to digital resources or fewer opportunities for digital skill development. Together, these factors may increase susceptibility to technology-related stress in increasingly digitalized healthcare environments. These findings have important practical implications. Because older workers with lower levels of education and income are more likely to experience technostress characterized by complexity and inclusion pressure, targeted interventions should be prioritized for these vulnerable groups. In particular, tailored training programs that focus on strengthening basic digital skills and reducing perceived technological complexity may help alleviate technostress among these workers. In addition, organizations should consider reducing inclusion-related pressure by fostering a supportive and non-competitive technological environment, where employees are not implicitly compared based on their digital proficiency. Clarifying expectations regarding technology use and establishing clear role boundaries may further help prevent workers from feeling pressured to constantly keep up with technological changes.

With regard to organizational characteristics, several factors were significantly associated with technostress profile membership. First, in terms of job position, workers categorized as physical or occupational therapists and other related professions were more likely to be classified into the high complexity and inclusion pressure group compared to other occupational groups. Second, employees working in general hospitals were more likely to belong to the high technostress with low inclusion pressure group. Finally, workers with shorter work experience were more likely to be classified into the high technostress with low inclusion pressure group. In particular, certain work settings and job roles may involve greater exposure to digital systems or more complex technological demands, which can increase the likelihood of experiencing technostress. Previous studies have shown that healthcare work increasingly requires the use of multiple ICT systems, such as electronic medical records, administrative platforms, and communication tools, and that the level of ICT demand may vary depending on occupational roles.⁶⁶ In addition, limited work experience may be associated with lower familiarity with organizational technologies and workflows, which may increase perceived difficulty in managing technology-related tasks. Therefore, further research is needed to better understand how organizational context and job-related demands interact to shape distinct technostress profiles.

Finally, lower levels of digital competence were associated with a higher likelihood of belonging to the moderate technostress group rather than the low technostress group. This finding suggests that digital competence may function as a protective resource, as individuals with lower competence may be more likely to experience technostress instead of maintaining a low-stress state. This interpretation is consistent with prior research showing that higher digital competence is associated with lower levels of technostress among healthcare professionals and other occupational groups.^66–68 Therefore, enhancing the digital competence of healthcare workers may represent an important strategy for mitigating technostress, particularly for older workers who may be less familiar with digital systems. In particular, targeted training programs that strengthen foundational digital skills and practical system use may help reduce perceived technological complexity. In addition, improving the usability and interface design of digital systems may further reduce the cognitive burden associated with learning and using new technologies. Furthermore, ongoing organizational support, such as continuous learning opportunities and accessible technical assistance, may facilitate adaptation to digital environments and prevent the accumulation of technostress.⁶⁸

This study further examined the relationship between technostress profiles and acceptance of transfer-assistive robots. The findings indicate that technostress is not only a multidimensional construct but also a critical factor that influences technology adoption in care settings. The low technostress group displayed the most favorable outcomes across the acceptance-related variables, including the highest self-efficacy, lowest anxiety, most positive attitudes, highest levels of perceived ease of use and usefulness, and strongest behavioral intention to adopt transfer-assistive robots. These results are consistent with prior research suggesting that technostress may hinder healthcare professionals’ acceptance of emerging technologies by increasing anxiety and resistance toward complex digital systems, including robotic technologies used in clinical settings.^24,69 Previous studies have also indicated that technology-related stress can influence users’ perceptions of usability and acceptance during the early stages of technology adoption. For example, a longitudinal pilot study⁷⁰ involving caregivers and older adult users reported that technostress was associated with perceived ease of use and acceptance during initial use; however, technology-related anxiety and stress decreased as users became more familiar with the system over time. These findings suggest that reducing technostress may be particularly important during the early stages of introducing new technologies, such as transfer-assistive robots, and may support more positive perceptions and intentions toward technology use in care settings.

In contrast, the high complexity and inclusion pressure group exhibited the least favorable outcomes, including the highest levels of anxiety and the lowest levels of self-efficacy, perceived usefulness, and behavioral intention to use. This profile may reflect a combination of technological challenges and psychological pressures related to feelings of relative disadvantage and the need to keep pace with technologically proficient colleagues. In LTC facilities, the introduction of increasingly complex technologies continues to increase the level of technological knowledge required from healthcare workers, which may intensify the pressure they experience.³⁰ Previous studies have reported that care staff often perceive the use of care robots as requiring additional technological knowledge and skills, which can act as a barrier to adoption.⁷¹ Therefore, reducing technological complexity—for example, through user-friendly system design or targeted training—may help facilitate the acceptance of transfer-assistive robots. In addition, the pressure related to technological inclusion identified in this study represents an important contextual factor. Workers who feel compelled to keep up with rapidly evolving technologies or with more technologically proficient colleagues may experience heightened stress, which may negatively influence their self-reported intentions or attitudes toward adopting new technologies. These results may help explain the relatively lower self-reported acceptance observed in this group and highlight that both technological complexity and the social–organizational context of inclusion are key determinants of care robot acceptance among older healthcare workers.

Interestingly, the high technostress with low inclusion pressure group demonstrated a relatively moderate-to-high level of acceptance. This finding suggests that lower inclusion-related stress may be associated with relatively better acceptance, even under conditions of high overall technostress. Moreover, the moderate technostress group, the largest subgroup, reported moderate-to-low levels of acceptance, particularly in terms of perceived ease of use—at times even lower than that of the high technostress with low inclusion group. This pattern underscores that pressure related to inclusion may be a particularly critical barrier to adoption and may even be more influential than overall technostress levels. One possible explanation is that inclusion-related pressure may intensify social comparison and perceived expectations regarding technological competence, particularly among older healthcare workers. When employees feel that they must keep pace with rapidly evolving technologies or with more technologically proficient colleagues, they may experience increased anxiety and reduced confidence in their ability to use new technologies. Such psychological pressure may discourage engagement with unfamiliar technologies and reduce willingness to adopt new systems, including transfer-assistive robots.

These findings highlight the importance of addressing technostress when introducing robotic technologies into healthcare environments, particularly among older healthcare workers. They also underscore the need for profile-specific strategies to mitigate technostress and promote successful adoption of care robots in healthcare settings.

Conclusions

As long-term care settings in South Korea increasingly adopt care technologies such as transfer-assistive robots to enhance productivity and operational efficiency, and as the proportion of older healthcare workers continues to rise, understanding technostress and technology acceptance among this population is essential. This study identified four distinct technostress profiles, revealing heterogeneous patterns that would not be captured by traditional variable-centered approaches. Notably, older workers experiencing high complexity and inclusion pressures reported lower self-efficacy, higher anxiety, and reduced intentions to adopt care robots, whereas other profiles exhibited moderate technostress but relatively higher acceptance.

These findings have several practical implications. Reducing technostress—particularly related to technological complexity and inclusion pressure—is critical for fostering positive perceptions and intentions toward care robot adoption. Targeted interventions such as peer mentoring, stepwise hands-on training, and gradual exposure to robot-assisted tasks, tailored to the specific needs of each technostress profile, may enhance self-efficacy, reduce anxiety, and improve perceived usefulness and behavioral intention.

From a theoretical perspective, this study highlights the multidimensional and heterogeneous nature of technostress and its influence on technology acceptance. While traditional variable-centered approaches can identify general relationships (e.g., higher technostress is associated with lower acceptance), they do not reveal the distinct subgroups of workers experiencing different patterns of technostress across its dimensions. Latent profile analysis (LPA), a person-centered approach, allows us to identify these subgroups, providing new theoretical insight into how combinations of technostress dimensions—such as high complexity with high inclusion pressure versus high complexity with low inclusion pressure—differentially affect acceptance of care robots. This distinction enables more targeted, profile-specific strategies and advances understanding of the mechanisms underlying technology adoption among older healthcare workers.

Finally, alleviating technostress may not only facilitate favorable perceptions and intentions toward transfer-assistive robots but also support well-being, reduce psychological burdens, and enhance care quality, ultimately contributing to the sustainability of LTC systems.

Limitations and future directions

This study has several limitations that should be considered when interpreting the findings. First, the study focused on healthcare workers in South Korea, which may limit the generalizability of the results to other occupational or cultural contexts. LTC policies, workforce composition, and digital infrastructure in Korea may differ from those in other countries, potentially affecting technostress experiences and technology acceptance. Although participants were recruited from all 17 administrative regions nationwide, the proportion of urban respondents was relatively high, reflecting the concentration of healthcare institutions in metropolitan areas. Future cross-national studies are warranted to examine how technostress profiles and their impact on technology acceptance may vary across cultural and healthcare system contexts. Second, the cross-sectional design precludes causal inferences. While associations between technostress profiles, individual characteristics, and technology acceptance were observed, causal relationships cannot be established. Future longitudinal research is needed to track changes in technostress and technology adoption over time, as well as to evaluate the predictive validity of technostress profiles on long-term outcomes such as care robot implementation and patient care quality. Third, the use of an online survey with convenience sampling introduces the possibility of self-selection bias. Participants who chose to respond may differ systematically from non-respondents; for example, those experiencing higher technostress might have been either more motivated to participate or less willing to engage with an online survey. Fourth, sparse data in certain profile–predictor combinations led to relatively large coefficients in some multinomial logistic regression estimates, which may reflect unstable or inflated estimates. In addition, several potentially relevant predictors—such as staffing ratios, prior digital literacy training, and organizational support—were not included in the survey and should be considered in future studies to provide a more comprehensive understanding of factors influencing technostress and technology acceptance. Finally, while this study examined antecedents and associations, it relied on self-reported perceptions, attitudes, and intentions rather than actual usage data. Future intervention-based studies are needed to test whether tailored training and support programs can effectively reduce technostress and enhance the adoption of assistive technologies among older healthcare workers.

Supplemental Material

Supplemental Material - Technostress in older healthcare workers: Latent profiles and implications for care robot acceptance

Supplemental Material for Technostress in older healthcare workers: Latent profiles and implications for care robot acceptance by Heejeong Yoon, Jungwan Lee, Namhwa Kim, Siwoo Ban, Hyeri Shin, Youngsun Kim in Digital Health

Footnotes

Acknowledgments

The authors used AI-assisted tools for language editing and text refinement. All interpretations, analyses, and conclusions were solely developed and verified by the authors.

ORCID iDs

Heejeong Yoon

Jungwan Lee

Hyeri Shin

Youngsun Kim

Ethical considerations

The Institutional Review Board of Kyung Hee University (IRB No: KHGIRB-23-415) reviewed and approved the study protocol, including the questionnaire and survey procedures, thus ensuring ethical and methodological validity.

Consent to participate

Participation in this study was voluntary and informed consent was obtained from all respondents.

Author contributions

Heejeong Yoon: Conceptualization, Methodology, Formal analysis, Writing – Original draft.

Jungwan Lee: Writing – Original draft.

Namhwa Kim: Writing – Original draft.

Siwoo Ban: Writing – Original draft.

Hyeri Shin: Project administration, Writing – Review & Editing.

Youngsun Kim: Supervision, Funding acquisition, Writing – Review & Editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the National Research Foundation of Korea grant funded by the ministry of education in 2021(NRF-2021S1A3A2A01096346).

Declaration of conflicting interests

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

Data Availability Statement

The datasets used during the current study are available from the corresponding author on reasonable request.*

Supplemental material

Supplemental material for this article is available online.

References

Organization for Economic Co-operation and Development . Beyond applause? Improving working conditions in long-term care. https://doi.org/10.1787/27d33ab3-en (2023, accessed on 12 September 2025).

Statistics Korea . Search results for “Elderly population. KOSIS Korean Statistical Information Service. https://kosis.kr/search/search.do (accessed on 12 September 2025).

Organization for Economic Co-operation and Development . Who cares? Attracting and retaining care workers for the elderly. OECD Health Policy Studies. https://doi.org/10.1787/92c0ef68-en (accessed on 12 September 2025).

Lee

Kang

, et al. Long-term care survey. Korea Institute for Health and Social Affairs, 2022. https://repository.kihasa.re.kr/handle/201002/43743 (accessed on 12 September 2025).

Kang

Lee

Lim

, et al. 2019 long-term care survey. Korea Institute for Health and Social Affairs, 2019. https://repository.kihasa.re.kr/handle/201002/36930 (accessed on 12 September 2025).

Kim

Shin

Lee

, et al. Changes in technology acceptance of smart care beds among long-term care workers in Korea. Healthcare (Basel) 2024; 12(21): 2195. https://doi.org/10.3390/healthcare12212195

Khaksar

Lindblom

Bygrave

, et al. Robotics in elderly healthcare: a qualitative analysis of 20 recent European research projects. ACM Trans Hum Robot Interact 2025; 14(2): 1–38. https://doi.org/10.1145/3711936

Wong

Ren

, et al. Care partner experience with telepresence robots in long-term care during COVID-19 pandemic. Digit Health 2025; 11: 20552076251319820. https://doi.org/10.1177/20552076251319820

Wong

KLY

Hung

Wong

, et al. Adoption of artificial intelligence–enabled robots in long-term care homes by health care providers: scoping review. JMIR Aging 2024; 7(1): e55257. https://doi.org/10.2196/55257

10.

Lim

Song

Kweon

, et al. Care robot research and development plan for disability and aged care in Korea: a mixed-methods user participation study. Assist Technol 2023; 35(4): 292–301. https://doi.org/10.1080/10400435.2022.2038307

11.

Kim

Shin

Yoon

, et al. Usability study of a smart transfer-assistive robot with dual arms for care workers. Disabil Rehabil Assist Technol 2023; 19(5): 1864–1870. https://doi.org/10.1080/17483107.2023.2285427

12.

Kato

Yoshimi

Aimoto

, et al. Reduction of multiple-caregiver assistance through the long-term use of a transfer support robot in a nursing facility. Assist Technol 2023; 35(3): 271–278. https://doi.org/10.1080/10400435.2022.2039324

13.

Yoon

Kim

. A study on factors affecting the intention to use care robot among institutional elderly care professionals: focusing on transfer assist robot. J Korean Gerontol Soc 2021; 41(5): 715–736. https://doi.org/10.31888/jkgs.2021.41.5.715

14.

Minghelli

Inácio Guerreiro

Pinho

, et al. Work-related musculoskeletal injuries in formal caregivers of Portuguese rest home of elderly. Rev Assoc Med Bras 1992) 2025; 71(2): e20241034. https://doi.org/10.1590/1806-9282.20241034

15.

Richarz

Tamayo

Rahmig

, et al. The impact of mechanical devices for lifting and transferring of patients on low back pain and musculoskeletal injuries in health care personnel: a systematic review and meta-analysis. J Occup Health 2023; 65(1): e12423. https://doi.org/10.1002/1348-9585.12423

16.

Kong

Choi

Shim

, et al. Development of workload assessment system for comparison of manual care and robot-aided care loads in the transfer assistive care. J Ergon Soc Korea 2021; 40(5): 313–329. https://doi.org/10.5143/jesk.2021.40.5.313

17.

Jensen

Ryom

Christensen

, et al. Differences in risk factors for voluntary early retirement and disability pension: a 15-year follow-up in a cohort of nurses’ aides. BMJ Open 2012; 2(6): e000991. https://doi.org/10.1136/bmjopen-2012-000991

18.

Servaty

Kersten

Brukamp

, et al. Implementation of robotic devices in nursing care: barriers and facilitators: an integrative review. BMJ Open 2020; 10(9): e038650. https://doi.org/10.1136/bmjopen-2020-038650

19.

Davis

. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Q 1989; 13(3): 319–340. https://doi.org/10.2307/249008

20.

Latikka

Turja

Oksanen

. Self-efficacy and acceptance of robots. Comput Human Behav 2019; 93: 157–163. https://doi.org/10.1016/j.chb.2018.12.017

21.

Turja

Saurio

Katila

, et al. Intention to use exoskeletons in geriatric care work: Need for ergonomic and social design. Ergonom Des 2022; 30(2): 13–16. https://doi.org/10.1177/1064804620961577

22.

Ahlin

Mann

. Ambiguous animals, ambivalent carers and arbitrary care collectives: re-theorizing resistance to social robots in healthcare. Soc Sci Med 2025; 365: 117587. https://doi.org/10.1016/j.socscimed.2024.117587

23.

Tarafdar

Ragu-Nathan

, et al. The impact of technostress on role stress and productivity. J Manag Inf Syst 2007; 24(1): 301–328. https://doi.org/10.2753/mis0742-1222240109

24.

Keshavarz

Saeidnia

Wang

. Navigating technostress: a deep dive into health practitioners' technological challenges in hospital settings. BMC Health Serv Res 2025; 25(1): 18.18. https://doi.org/10.1186/s12913-024-12196-1

25.

Brod

. Technostress: the human cost of the computer revolution. Addison-Wesley, 1984.

26.

Ragu-Nathan

Tarafdar

Ragu-Nathan

, et al. The consequences of technostress for end users in organizations: conceptual development and empirical validation. Inf Syst Res 2008; 19(4): 417–433. https://doi.org/10.1287/isre.1070.0165

27.

Nimrod

. Technostress in later life: A multinational study. Educ Gerontol 2025; 1–6. https://doi.org/10.1080/03601277.2025.2521805

28.

Tacy

Northam

Wieck

. Understanding the Effects of Technology Acceptance in Nursing Faculty: A Hierarchical Regression. 2016. https://www.proquest.com/openview/71bdbb917c3d996344c7d88e1c24b64d/1?pq-origsite=gscholar&cbl=2034896. (accessed on 8 May 2026).

29.

Gabbiadini

Paganin

Simbula

. Teaching after the pandemic: The role of technostress and organizational support on intentions to adopt remote teaching technologies. Acta Psychol (Amst) 2023; 236: 103936. https://doi.org/10.1016/j.actpsy.2023.103936

30.

Thunberg

Johnson

Ziemke

. Investigating healthcare workers' technostress when welfare technology is introduced in long-term care facilities. Behav Inf Technol 2024; 43(13): 3288–3300. https://doi.org/10.1080/0144929x.2023.2276802

31.

Zhang

. Technostress, academic self-efficacy, and resistance to innovation: Buffering roles of knowledge sharing culture and constructive deviant behavior. Psychol Res Behav Manag 2023; 16: 3867–3881. https://doi.org/10.2147/PRBM.S424396

32.

Wang

. Art students’ technostress, perceived usefulness, satisfaction, and continuance intention to use Mobile educational applications. SAGE Open 2024; 14(2): 21582440241260206. https://doi.org/10.1177/21582440241260206

33.

Chang

Zhang

Cai

, et al. Does AI-driven technostress promote or hinder employees’ artificial intelligence adoption intention? A moderated mediation model of affective reactions and technical self-efficacy. Psychol Res Behav Manag 2024; 17: 413–427. https://doi.org/10.2147/PRBM.S441444

34.

Rosa

Miranda

Pascoal

. Uncovering latent profiles of ICT users and its relation to technostress and mental health in an adult sample: contributions of a transdiagnostic approach. Int J Hum Comput Interact 2025; 41: 1–14. https://doi.org/10.1080/10447318.2024.2439637

35.

Sınmaz

Kara

Seyhan Ak

, et al. Experiences of robotic surgery nurses regarding technostress: a qualitative study. J Robot Surg 2025; 19: 154. https://doi.org/10.1007/s11701-025-02320-6.

36.

Yoon

Lee

Ban

, et al. Factors affecting the acceptance of care robots by older adults: focusing on technical factors. Innov Aging 2023; 7(S1): 1181. https://doi.org/10.1093/geroni/igad104.3784

37.

Taiebine

Keegan

. Telepractice for bilingual and multilingual people with aphasia: a scoping review. Aphasiology 2024; 40: 1–23. https://doi.org/10.1080/02687038.2024.2373434

38.

Javaid

Haleem

Pratap Singh

, et al. Utilization of robotics for healthcare: a scoping review. J Ind Integr Manag 2025; 10(1): 43–65. https://doi.org/10.1142/s2424862222500154

39.

Karayaneva

Sharifzadeh

Szczepura

, et al. AI-enhanced healthcare and eldercare delivery using non-intrusive sensors and physical robots: an exploratory study. In: Impact of digital solutions for improved healthcare delivery. IGI Global, 2025, pp. 1–34.

40.

Yamashita

Narine

Ojomo

, et al. Digital skills use profiles among older workers in the United States: a person-centered approach. Int J Lifelong Educ 2024; 44: 1–19. https://doi.org/10.1080/02601370.2024.2443497

41.

, et al. Latent profile analysis of positive solitude during the recurrent outbreak of COVID-19. Front Public Health 2022; 10: 872128. https://doi.org/10.3389/fpubh.2022.872128

42.

Asparouhov

Muthén

. Auxiliary variables in mixture modeling: Three-step approaches using M plus. Struct Equ Modeling 2014; 21(3): 329–341. https://doi.org/10.1080/10705511.2014.915181

43.

McCutcheon

. Latent class analysis. Sage, 1987, vol 64.

44.

Turja

. Uncertainties about accepting care robots. Front Digit Health 2023; 5: 1092974. https://doi.org/10.3389/fdgth.2023.1092974

45.

Pekkarinen

Hennala

Melkas

. Towards technology domestication in a care organization? Facilitating and hindering factors in care robot use. Int J Soc Robot 2025; 17: 655–674. https://doi.org/10.1007/s12369-025-01226-5

46.

Nimrod

. Technostress: measuring a new threat to well-being in later life. Aging Ment Health 2018; 22(8): 1086–1093. https://doi.org/10.1080/13607863.2017.1334037

47.

Rašticová

Šácha

Lakomý

, et al. Technostress among older workers: a Central European perspective. Psychol Res Behav Manag 2025; 18: 1211–1225. https://doi.org/10.2147/PRBM.S508500

48.

Comotti

Fattori

Di Tecco

, et al. Cross-context validation of a technostress scale for the aging workforce. J Occup Environ Med 2025; 67(5): e347–e351. https://doi.org/10.1097/JOM.0000000000003349

49.

Ministry of Health and Welfare and Korea Institute for Health and Social Affairs Lee

Lee

Kang

, et al. 2022 Long-Term Care Survey, 2022. https://repository.kihasa.re.kr/handle/201002/43743

50.

Chen

Chan

AHS

. Gerontechnology acceptance by elderly Hong Kong Chinese: A senior technology acceptance model (STAM). Ergonomics 2014; 57(5): 635–652. https://doi.org/10.1080/00140139.2014.895855

51.

Davis

Bagozzi

Warshaw

. User acceptance of computer technology: A comparison of two theoretical models. Manag Sci 1989; 35(8): 982–1003. https://doi.org/10.1287/mnsc.35.8.982

52.

Venkatesh

Morris

Davis

, et al. User acceptance of information technology: Toward a unified view. MIS Q 2003; 27(3): 425–478. https://doi.org/10.2307/30036540

53.

Mostaghel

Oghazi

. Elderly and technology tools: a fuzzy-set qualitative comparative analysis. Qual Quant 2017; 51(5): 1969–1982. https://doi.org/10.1007/s11135-016-0390-6

54.

Tsai

Cheng

Tsai

, et al. Acceptance and resistance of telehealth: The perspective of dual-factor concepts in technology adoption. Int J Inf Manag 2019; 49: 34–44. https://doi.org/10.1016/j.ijinfomgt.2019.03.003

55.

Vermunt

Magidson

. Latent class cluster analysis. In: Hagenaars

McCutcheon

(eds). Applied latent class analysis. Cambridge University Press, 2002, pp. 89–106.

56.

Akaike

. A new look at the statistical model identification. IEEE Trans Automat Contr 1974; 19(6): 716–723. https://doi.org/10.1109/tac.1974.1100705

57.

Schwarz

. Estimating the dimension of a model. Ann Stat 1978; 6(2): 461–464. https://doi.org/10.1214/aos/1176344136

58.

Sclove

. Application of model-selection criteria to some problems in multivariate analysis. Psychometrika 1987; 52(3): 333–343. https://doi.org/10.1007/bf02294360

59.

Muthén

. Latent variable analysis. In: Kaplan

(ed). The Sage handbook of quantitative methodology for the social sciences. Sage, 2004.

60.

Lepkowsky

Arndt

. Functional assessment of currently employed technology scale (FACETS): reliability and validity. Int J Med Sci Clin Invent 2018; 5(9): 4064–4068.

61.

Bakk

Vermunt

. Robustness of stepwise latent class modeling with continuous distal outcomes. Struct Equ Modeling 2016; 23(1): 20–31. https://doi.org/10.1080/10705511.2014.955104

62.

Rašticová

Šácha

Lakomý

, et al. Technostress among older workers: A Central European perspective. Psychol Res Behav Manag 2025; 18: 1211–1225. https://doi.org/10.2147/PRBM.S508500

63.

Schlomann

Even

Hammann

. How older adults learn ICT—guided and self-regulated learning in individuals with and without disabilities. Front Comput Sci 2022; 3: 803740. https://doi.org/10.3389/fcomp.2021.803740

64.

Zhu

Wan

, et al. Older adults’ self-perception, technology anxiety, and intention to use digital public services. BMC Public Health 2024; 24(1): 3533. https://doi.org/10.1186/s12889-024-21088-2

65.

Chan

DYL

Lee

SWH

Teh

. Factors influencing technology use among low-income older adults: A systematic review. Heliyon 2023; 9(9): e20111. https://doi.org/10.1016/j.heliyon.2023.e20111

66.

Stadin

Nordin

Broström

, et al. Technostress operationalised as information and communication technology (ICT) demands among managers and other occupational groups—results from the Swedish Longitudinal Occupational Survey of Health (SLOSH). Comput Human Behav 2021; 114: 106486. https://doi.org/10.1016/j.chb.2020.106486

67.

Jurić

Podrug Krstulović

Mišurac

. Professional well-being of teachers in the digital age: the role of digital competences and technostress. Educ Sci 2026; 16(1): 130. https://doi.org/10.3390/educsci16010130

68.

Golz

Peter

Müller

, et al. Technostress and digital competence among health professionals in Swiss psychiatric hospitals: cross-sectional study. JMIR Ment Health 2021; 8(11): e31408. https://doi.org/10.2196/31408

69.

Di Stefano

Salerno

Matranga

, et al. Artificial intelligence perceptions and technostress in staff radiologists: The mediating role of artificial intelligence acceptance and the moderating role of self-efficacy. Behav Sci (Basel) 2025; 15(9): 1276. https://doi.org/10.3390/bs15091276

70.

Pni

Lorusso

Toccafondi

, et al. How time, living situation, and stress related to technology influence user acceptance and usability of a socialization service for older adults and their formal and informal caregivers: Six-month pilot study. JMIR Aging 2024; 7: e54736. https://doi.org/10.2196/54736

71.

Johansson-Pajala

Gustafsson

. Significant challenges when introducing care robots in Swedish elder care. Disabil Rehabil Assist Technol 2022; 17(2): 166–176. https://doi.org/10.1080/17483107.2020.1773549

Supplementary Material

Please find the following supplemental material available below.

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

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

0.23 MB

0.00 MB