Factors in the 4-week Acceptance of a Computer-Based,Chronic Disease Self-Monitoring System in Patients with Type 2 Diabetes Mellitus and/or Hypertension

Abstract

Background:

Health information technology (HIT) interventions developed to support patients' self-care for chronic diseases have become popular, but people may not always accept and sustain their use.

Introduction:

This study examined factors that affected patients' acceptance of a computer-based, chronic disease self-monitoring system over a 4-week period.

Materials and Methods:

A research model was developed to test the relationships between the perceived usefulness, perceived ease of use, attitude, and subjective norm and the patients' behavioral intention to use the system (i.e., acceptance). Data were collected with surveys of 42 patients with type 2 diabetes mellitus and/or hypertension at baseline, 2 weeks after implementation, and 4 weeks after implementation. Path analysis was used for model testing.

Results:

Perceived usefulness affected behavioral intention indirectly at 2 weeks and directly at 4 weeks; perceived ease of use affected behavioral intention indirectly at 2 and 4 weeks; attitude directly affected behavioral intention at 2 weeks; and subjective norm affected behavioral intention indirectly at 2 weeks and directly at baseline and at 4 weeks.

Conclusions:

Patients' acceptance of HIT is affected by the factors proposed in our research model. It is suggested that healthcare stakeholders consider and address the effects of these factors and their variations over time before implementing HIT.

Introduction

Health information technology (HIT) interventions, herein referred to as patient-focused, computer-based health applications,¹ have become popular for chronically ill patients because the technology can provide patients with easy access to resources that can augment their ability to engage in self-care activities, which in turn may lead to better patient health outcomes, well-being, and quality of life.^2

–5 However, patients do not always accept and sustain the use of this technology,^6
–8 which can be of significant concern because nonacceptance or nonuse of the technology may prevent them from realizing its full benefits and can lead to failure to implement the technology. To promote patients' acceptance and sustained use of HIT, it is essential to understand the factors that may influence acceptance of the technologies.

Many studies of such factors have been qualitative. For instance, Zhang et al.⁶ interviewed patients about their use of an e-appointment scheduling system and found that acceptance was associated with the system's functionality and the patient's characteristics. Hoaas et al.⁸ conducted semistructured focus groups and found that patients' perceived improvement in health benefits, self-efficacy, and emotional safety were the factors that affected their use of a telerehabilitation system to manage their chronic obstructive pulmonary disease. In Radhakrishnan et al.,⁹ patients who used a telehealth system to monitor heart failure were interviewed, and the results indicated that patients' use of the system was affected by their own health conditions and by the perceived usefulness of the system.

Quantitative studies of factors that affect patients' acceptance of HIT have also been performed; however, only a few were conducted in longitudinal contexts, and the research contained limitations. For example, Kelley et al.¹⁰ studied patients' use of a Web-based e-Health system over a 12-month period. Acceptance surveys were conducted when the patients entered the study and every 3 months thereafter. However, their research hypotheses were tested using data that were pooled across all time points, which overlooked the potential changes in patients' responses over time. In Klein,¹¹ patients' use of an Internet-based patient–physician communication application was examined for a 12-month period. However, their hypotheses were tested based on patients' responses collected at only one time point, which may not have been representative of the whole study period.

In this study, we set out to examine factors that may have affected patients' acceptance of a computer-based, chronic disease self-monitoring system over a 4-week period. A research model concerning patients' acceptance of HIT was developed and tested using data collected at three different time points during the study period.

Theoretical Background and Research Hypotheses

We developed our research model using Davis's technology acceptance model (TAM)¹² as a theoretical basis to specify which factors to study and the relationships among them because it provides a reliable foundation for examining individuals' HIT acceptance.^13,14 In this study, acceptance was measured as patients' behavioral intention to use the disease self-monitoring system. Based on TAM, two key factors were posited to predict acceptance: perceived usefulness, which refers to the degree to which patients believe that using the system will enhance their performance in managing chronic conditions, and perceived ease of use, which refers to the degree to which patients believe that using the system will require little effort. The effects of these two factors on acceptance were believed to be mediated by attitude, which refers to patients' positive feelings about using the system. Subjective norm, which refers to the degree to which patients perceive that people who are important to them think that they should use the system, was also included in our research model because the opinions of important others can influence patients' health-related behaviors.^15,16 Previous studies outside healthcare have shown the significant effects of subjective norm on perceived usefulness, perceived ease of use, attitude, and behavioral intention¹⁷; however, the pertinent evidence within the health informatics discipline is lacking. Further studies on how subjective norm influences patients' HIT acceptance and how the effects may change over time are needed. Based on theoretical considerations, the following hypotheses were posited in our research model (Figure 1):

Fig. 1.

Research model and hypotheses.

H1: Perceived ease of use will have a significant positive effect on perceived usefulness.

H2: Perceived usefulness will have a significant positive effect on attitude.

H3: Perceived ease of use will have a significant positive effect on attitude.

H4: Attitude will have a significant positive effect on behavioral intention.

H5: Perceived usefulness will have a significant positive effect on behavioral intention.

H6: Subjective norm will have a significant positive effect on behavioral intention.

H7: Subjective norm will have a significant positive effect on attitude.

H8: Subjective norm will have a significant positive effect on perceived usefulness.

H9: Subjective norm will have a significant positive effect on perceived ease of use.

Materials and Methods

Study intervention

The HIT involved in this study was a computer-based, chronic disease self-monitoring system that was based on a user-centered design process. The system comprised a two-in-one blood glucose and blood pressure monitor connected to a 10-inch touchscreen tablet computer that enabled patients to measure and track their blood glucose, blood pressure, and pulse. The system automatically stored the measurement records and indicated whether they were within a normal range. All measurement records were presented in structured tables and graphs in which normal values were displayed in black and abnormal values were displayed in orange or red based on the specific situation. The system also provided video-based educational materials that allowed the patients to learn proper self-management of their chronic conditions, such as how to measure blood glucose and blood pressure, collocate diet, and exercise.

Participants

We recruited participants from three community health centers of a nonprofit health service network. The participant inclusion criteria were as follows: (1) age of 18 years or above with a diagnosis of type 2 diabetes mellitus and/or hypertension for at least 3 months, (2) a prescription for oral hypoglycemic and/or antihypertensive medications, (3) normal (or corrected-to-normal) vision, (4) no cognitive or physical impairment, (5) the ability to learn and perform computer-based disease self-monitoring, and (6) the ability to understand written and spoken Chinese. Patients with unstable or life-threatening conditions were excluded.

Outcome measures

An HIT acceptance questionnaire that comprised measurement items adapted from previously validated scales^18
–20 was developed to measure the factors in the model. Interviews with 10 senior citizens (5 males and 5 females) were conducted to modify the measurement items to ensure that they could be comprehended as intended by the target population. We used 7-point Likert-type scales ranging from 1 (very strongly disagree) to 7 (very strongly agree) to assess the measurement items. Table 1 shows the factors and measurement items used in this study.

Table 1.

Factors and Measurement Items Used in This Study

FACTOR	MEASUREMENT ITEM	ADAPTED FROM
PU	PU1: Using the self-monitoring system improves your ability to manage your disease.	Davis et al.¹⁸
	PU2: Using the self-monitoring system helps you save time managing your disease.
	PU3: Using the self-monitoring system enhances your effectiveness in managing your disease.
	PU4: You find the self-monitoring system to be useful in managing your disease.
PEOU	PEOU1: Learning to use the self-monitoring system is easy for you.	Davis et al.¹⁸
	PEOU2: You find it easy to get the self-monitoring system to do what you want it to do.
	PEOU3: It is easy for you to become skillful at using the self-monitoring system.
	PEOU4: You find the self-monitoring system easy to use.
ATT	ATT1: Using the self-monitoring system is a good idea.	Taylor and Todd¹⁹
	ATT2: Using the self-monitoring system is a wise idea.
	ATT3: You like the idea of using the self-monitoring system.
	ATT4: Using the self-monitoring system is pleasant.
SN	SN1: People who influence your behavior would think that you should use the self-monitoring system.	Taylor and Todd¹⁹
SN	SN2: People who are important to you would think you should use the self-monitoring system.	Taylor and Todd¹⁹
BI	BI1: You intend to use the self-monitoring system.	Venkatesh et al.²⁰
	BI2: You predict you would use the self-monitoring system.
	BI3: You plan to use the self-monitoring system.

ATT, attitude; BI, behavioral intention; PEOU, perceived ease of use; PU, perceived usefulness; SN, subjective norm.

Procedures

This study examined the acceptance of HIT for a 4-week period, during which the participants were provided with the disease self-monitoring system, blood glucose test strips, and lancets at no cost. The participants were trained and encouraged to use the system to monitor their chronic conditions at home while continuing with the usual care recommended by their healthcare providers. The acceptance questionnaire was administrated to participants during home visits conducted at baseline and 2 weeks and 4 weeks after implementation. The participants' demographic information, including age, sex, education level, and self-rated computer experience, was collected.

Data analysis

A power analysis was performed using G*Power to determine the number of participants for model testing. The internal consistency reliability was assessed for each factor in the model by examining Cronbach's alpha, of which a value higher than 0.7 was considered acceptable.²¹ Curve estimations were performed to test the assumption of linearity. Multicollinearity was tested by computing the variance inflation factor (VIF) of the independent variables in each subpart of the model. A VIF of less than 10 indicated an absence of multicollinearity.^22,23 Path analysis was used to examine the relationships among the factors at each time point. Maximum likelihood estimation was used for model fitting and parameter estimation. The path coefficients indicated by standardized regression weights and squared multiple correlations (R ²) for dependent variables in the model were calculated using SPSS Amos for Windows. Four commonly used goodness-of-fit indices were used to assess the overall fit of the model; a good model fit was indicated by the ratio of χ² to degrees of freedom (χ²/df <3),²² the comparative fit index (CFI ≥0.95),^24,25 the Tucker–Lewis index (TLI ≥0.95),^24,26 and the root mean square error of approximation (RMSEA <0.06; although RMSEA values as high as 0.08 are considered acceptable).^24,27

Ethics

This study received ethical approval from the institutional review board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster and informed consent was obtained from all participants.

Results

Sample size and participant characteristics

Our power analysis indicated that 36 observations were needed to detect an effect size f ² of 0.35 with 80% statistical power at the 0.05 significance level. In this study, 42 participants completed the questionnaire at baseline. Because of withdrawal and loss to follow-up, the data of four participants (two males and two females) were not collected at 2 weeks and those of another four participants (three males and one female) were not collected at 4 weeks. Therefore, the data from 42, 38, and 38 participants were used in the analyses for baseline, 2 weeks, and 4 weeks, respectively. Table 2 presents the participants' demographic characteristics.

Table 2.

Demographic Characteristics of the Study Sample

	BASELINE (n = 42)	2 WEEKS (n = 38)	4 WEEKS (n = 38)
Mean age (standard deviation)	70.1 (9.7) years	69.2 (9.4) years	70.0 (9.7) years
Sex, n (%)
Male	17 (40.5)	15 (39.5)	14 (36.8)
Female	25 (59.5)	23 (60.5)	24 (63.2)
Education, n (%)
Primary school or below	25 (59.5)	22 (57.9)	22 (57.9)
Secondary school	13 (31.0)	12 (31.6)	12 (31.6)
College or above	4 (9.5)	4 (10.5)	4 (10.5)
Self-rated computer experience, n (%)
None	27 (64.3)	23 (60.5)	24 (63.1)
Beginner	10 (23.8)	10 (26.3)	9 (23.7)
Competent	5 (11.9)	5 (13.2)	5 (13.2)

Internal consistency reliability

Table 3 shows the means, standard deviations, and Cronbach's alphas for the factors at baseline, 2 weeks, and 4 weeks. The measurement scales showed good internal consistency reliability, except that Cronbach's alpha for subjective norm at baseline was slightly lower than 0.7.

Table 3.

Means, Standard Deviations, and Cronbach's Alphas for the Factors at Baseline, 2 Weeks, and 4 Weeks

FACTOR	MEAN	STANDARD DEVIATION	CRONBACH'S ALPHA
Perceived usefulness
Baseline	5.70	0.71	0.86
2 weeks	5.59	0.70	0.88
4 weeks	5.42	0.80	0.91
Perceived ease of use
Baseline	5.47	0.69	0.83
2 weeks	5.47	0.82	0.91
4 weeks	5.50	0.65	0.91
Attitude
Baseline	5.89	0.67	0.84
2 weeks	5.57	0.71	0.90
4 weeks	5.55	0.68	0.88
Subjective norm
Baseline	5.51	0.80	0.65
2 weeks	5.32	0.72	0.86
4 weeks	5.32	0.81	0.85
Behavioral intention
Baseline	5.85	0.68	0.91
2 weeks	5.50	0.83	0.97
4 weeks	5.40	0.85	0.97

Testing of linearity and multicollinearity

The linearity assumption for path analysis model testing was fulfilled as all paths in our model indicated linear relationships at the significance level of 0.01 (see Table 4). Table 5 shows that all VIF values were below 10, indicating the absence of multicollinearity in the model.

Table 4.

Curve Estimations for Linearity Assumption Test

HYPOTHESIS	BASELINE			2 WEEKS			4 WEEKS
HYPOTHESIS	R ²	F-VALUE	p	R ²	F-VALUE	p	R ²	F-VALUE	p
H1	0.58	54.92	0.00	0.41	24.72	0.00	0.60	53.65	0.00
H2	0.45	32.49	0.00	0.61	55.27	0.00	0.71	87.32	0.00
H3	0.28	15.77	0.00	0.73	99.04	0.00	0.79	138.23	0.00
H4	0.37	23.29	0.00	0.45	29.90	0.00	0.59	51.44	0.00
H5	0.35	21.07	0.00	0.31	16.13	0.00	0.62	59.80	0.00
H6	0.48	36.60	0.00	0.26	12.44	0.00	0.57	48.29	0.00
H7	0.58	55.28	0.00	0.57	47.87	0.00	0.41	25.29	0.00
H8	0.46	34.47	0.00	0.43	27.62	0.00	0.35	18.96	0.00
H9	0.35	21.11	0.00	0.32	16.78	0.00	0.54	41.72	0.00

Table 5.

Results of Multicollinearity Tests

	BASELINE		2 WEEKS		4 WEEKS
	TOLERANCE	VIF	TOLERANCE	VIF	TOLERANCE	VIF
Dependent variable: Behaviorial intention
Independent variable
Attitude	0.38	2.65	0.29	3.45	0.26	3.87
Perceived usefulness	0.48	2.07	0.38	2.61	0.29	3.47
Subjective norm	0.37	2.72	0.42	2.40	0.58	1.73
Dependent variable: Attitude
Independent variable
Perceived ease of use	0.41	2.43	0.56	1.80	0.28	3.53
Perceived usefulness	0.34	2.96	0.46	2.17	0.40	2.50
Subjective norm	0.53	1.90	0.53	1.88	0.46	2.16
Dependent variable: Perceived usefulness
Independent variable
Perceived ease of use	0.66	1.53	0.68	1.47	0.46	2.16
Subjective norm	0.66	1.53	0.68	1.47	0.46	2.16

VIF, variance inflation factor.

Model testing

The fit indices showed an adequate fit between the model and the data for 2 weeks and 4 weeks, but less of a fit for baseline (χ²/df = 5.719, CFI = 0.963, TLI = 0.628, and RMSEA = 0.339 for baseline; χ²/df = 1.360, CFI = 0.997, TLI = 0.973, and RMSEA = 0.099 for 2 weeks; and χ²/df = 1.046, CFI = 1.000, TLI = 0.997, and RMSEA = 0.035 for 4 weeks).

As shown in Figures 2 –4, H1, H6, H7, H8, and H9 were supported at baseline; H1, H2, H3, H4, H7, H8, and H9 were supported at 2 weeks; and H1, H2, H3, H5, H6, and H9 were supported at 4 weeks. Overall, our model was able to account for 51%, 46%, and 76% of the variance in patients' behavioral intention to use the disease self-monitoring system at baseline, 2 weeks, and 4 weeks, respectively.

Fig. 2.

Path diagram for baseline (**p < 0.01; ***p < 0.001).

Fig. 3.

Path diagram for 2 weeks (**p < 0.01; ***p < 0.001).

Fig. 4.

Path diagram for 4 weeks (**p < 0.01; ***p < 0.001).

Discussion

E-health initiatives focusing on the integration of HITs into patients' chronic disease self-care have been developed because they can improve patients' access to care resources and enable remote monitoring. However, some studies have shown a decline in patients' use of HIT, especially during the first few weeks.^28,29 This finding underscores the importance of studying patients' acceptance of HIT in the preliminary implementation stages. Our study examined patients' acceptance of a computer-based, chronic disease self-monitoring system at baseline, 2 weeks, and 4 weeks to identify the relationships between the perceived usefulness, perceived ease of use, attitude, and subjective norm and behavioral intention and to test how the relationships changed over a 4-week period. Overall, as indicated by the model fit indices, our research model fit the data well at 2 and 4 weeks, but less well at baseline. This suggests that this HIT acceptance model may apply better during the postimplementation stages, which could be attributed to variations in the concerns that patients may have regarding acceptance over time.^30,31

Our study found that perceived usefulness had no significant effect on behavioral intention at baseline, an indirect effect mediated by attitude on behavioral intention at 2 weeks, and a direct effect on behavioral intention at 4 weeks. The changes may be due to the process whereby patients familiarize themselves with the use of the system. Patients may have limited ideas about the usefulness of the system at the initial stage, but they may experience benefits with increased use over time. Therefore, the effect of perceived usefulness on behavioral intention emerged over the study period. Proper training delivered before implementation to familiarize patients with the use of the technology could be one of the principal methods to help patients shorten the process of recognizing the usefulness of HIT and realizing its benefits.^32,33 In addition, it is suggested that the intended target users be actively involved in the process of technology design and development to study their requirements and preferences to provide them with the desired functions and features.

Like previous studies,^11,34

–37 our study showed a significant effect of perceived ease of use on perceived usefulness. Perceived ease of use was also deemed to be important in acceptance because it showed indirect effects mediated by perceived usefulness and/or attitude on behavioral intention at 2 and 4 weeks. This indicates that patients who perceive that HIT is easy to use would be more likely to perceive the technology as useful, which would in turn generate positive feelings toward acceptance. Therefore, HITs should be designed to be user-friendly for the end users. Achieving this may require iterative usability evaluations, followed by careful attention to human factors and principles, both before and after implementation. One should also consider designs that are tailored for older adults, such as avoiding the use of small font sizes in the system interface, because they may be more likely to suffer from chronic conditions and functional decline or disability than younger populations.^38

–42

Patients' attitude was found to influence their behavioral intention to use the disease self-monitoring system only at 2 weeks. Thus, the effects of perceived usefulness, perceived ease of use, and subjective norm on behavioral intention were mediated by attitude at 2 weeks, but not at baseline and 4 weeks. This seems to agree with the argument of some studies that attitude may not be an appropriate mediator between the perceived beliefs and behavioral intention.^18,19,30,43 Moreover, it could be noted that the direct effects of perceived usefulness on behavioral intention appeared to grow over the mediation effects of attitude over time, implying that the determinant role of attitude on behavioral intention might become less important when other factors such as usefulness are independently taken into account (p. 166).¹⁹ More research is needed to further examine the role of attitude on patients' acceptance of HIT over a longer period.

In this study, subjective norm was found to influence behavioral intention either directly (at baseline and 4 weeks) or indirectly (at 2 weeks). Although subjective norm exerted a consistent influence on perceived ease of use over the study period, its effects on attitude and perceived usefulness appeared to diminish over time. This implies that patients' initial attitudes toward acceptance and perceptions about the usefulness of the technology may rely substantially on other referents' opinions, which may, along with increased use over time, become less important to patients as they gain hands-on experience with the technology. Therefore, it is suggested that patients' important others (e.g., family members, close friends, and caregivers) consider means to cultivate a positive social environment that helps to develop the patients' initial acceptance of HIT, such as by emphasizing to patients the usefulness of the technology in managing chronic conditions and providing prompt help when needed.

This study had several limitations. First, the small sample size restricted us from including more factors in the research model. A larger sample would have allowed the examination of more factors that may be relevant to HIT acceptance. Second, although our study provides important information about factors that affect patients' acceptance of HIT in the preliminary 4-week period, it would be meaningful to study acceptance over a longer period, particularly when dealing with chronic diseases that require long-term care. Third, we used behavioral intention to measure acceptance because it is a common indicator of the variable.⁴⁴ However, actual usage of the technology can also be considered for the measurement as it represents an objective approach to understand acceptance. The relationship between patients' behavioral intention to use the HIT and their actual usage behavior awaits further examination.

Conclusions

This study shows that patients' acceptance of HIT was affected by the factors proposed in our research model. These effects varied over the 4-week period. The effect of perceived usefulness on behavioral intention tended to increase over time. Perceived ease of use affected behavioral intention indirectly at 2 and 4 weeks. Attitude affected behavioral intention only at 2 weeks. Subjective norm was shown to be a prominent factor in behavioral intention and other perceived beliefs both before and after implementation. It is hoped that the likelihood of patients' acceptance of HIT can be increased by applying the findings of this study.

Footnotes

Acknowledgment

The study was conducted with the support of the General Research Fund of the University Grants Council of Hong Kong (PI: Calvin Or; project #: 716211).

Disclosure Statement

No competing financial interests exist.

References

CKL

, Karsh

B-T

, Severtson

, et al. Factors affecting home care patients' acceptance of a web-based interactive self-management technology. J Am Med Inform Assoc, 2011; 18:51–59.

Kashem

, Droogan

, Santamore

, et al. Web-based Internet telemedicine management of patients with heart failure. Telemed J E Health, 2006; 12:439–447.

Gustafson

, Hawkins

, Boberg

, et al. Impact of a patient-centered, computer-based health information/support system. Am J Prev Med, 1999; 16:1–9.

Tao

, Or

CKL

. Effects of self-management health information technology on glycaemic control for patients with diabetes: A meta-analysis of randomized controlled trials. J Telemed Telecare, 2013; 19:133–143.

CKL

, Tao

. Does the use of consumer health information technology improve outcomes in the patient self-management of diabetes? A meta-analysis and narrative review of randomized controlled trials. Int J Med Inform, 2014; 83:320–329.

Zhang

, Yu

, Yan

, Spil

ITA

. Using diffusion of innovation theory to understand the factors impacting patient acceptance and use of consumer e-health innovations: A case study in a primary care clinic. BMC Health Serv Res, 2015; 15:71–85.

Liddy

, Dusseault

, Dahrouge

, et al. Telehomecare for patients with multiple chronic illnesses: Pilot study. Can Fam Physician, 2008; 54:58–65.

Hoaas

, Andreassen

, Lien

, et al. Adherence and factors affecting satisfaction in long-term telerehabilitation for patients with chronic obstructive pulmonary disease: A mixed methods study. BMC Med Inform Decis Mak, 2016; 16:1–14.

Radhakrishnan

, Jacelon

, Roche

. Perceptions on the use of telehealth by homecare nurses and patients with heart failure: A mixed method study. Home Healthcare Manag Pract, 2012; 24:175–181.

10.

Kelley

, Chiasson

, Downey

, Pacaud

. The clinical impact of eHealth on the self-management of diabetes: A double adoption perspective. J Assoc Inf Syst, 2011; 12:208–234.

11.

Klein

. Internet-based patient-physician electronic communication applications: Patient acceptance and trust. E Service J, 2007; 5:27–51.

12.

Davis

. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quart, 1989; 13:319–340.

13.

King

, He

. A meta-analysis of the technology acceptance model. Inf Manag, 2006; 43:740–755.

14.

Holden

, Karsh

B-T

. The technology acceptance model: Its past and its future in healthcare. J Biomed Inform, 2010; 43:159–172.

15.

Finlay

, Trafimow

, Moroi

. The importance of subjective norms on intentions to perform health behaviors. J Appl Soc Psychol, 1999; 29:2381–2393.

16.

Finlay

, Trafimow

, Jones

. Predicting health behaviors from attitudes and subjective norms: Between-subjects and within-subjects analyses. J Appl Soc Psychol, 1997; 27:2015–2031.

17.

Schepers

, Wetzels

. A meta-analysis of the technology acceptance model: Investigating subjective norm and moderation effects. Inf Manag, 2007; 44:90–103.

18.

Davis

, Bagozzi

, Warshaw

. User acceptance of computer technology: A comparison of two theoretical models. Manag Sci, 1989; 35:982–1003.

19.

Taylor

, Todd

. Understanding information technology usage: A test of competing models. Inf Syst Res, 1995; 6:144–176.

20.

Venkatesh

, Morris

, Davis

. User acceptance of information technology: Toward a unified view. MIS Quart, 2003; 27:425–478.

21.

Nunnally

, Bernstein

. Psychometric Theory. New York, NY: McGraw-Hill Education, 1994.

22.

Kline

. Principles and Practice of Structural Equation Modeling, 3rd ed. New York, NY: Guilford Press, 2011.

23.

Kutner

, Nachtsheim

, Neter

. Applied Linear Regression Models. New York, NY: McGraw-Hill Education, 2004.

24.

, Bentler

. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct Equ Modeling, 1999; 6:1–55.

25.

Bentler

. Comparative fit indexes in structural models. Psychol Bull, 1990; 107:238–246.

26.

Tucker

, Lewis

. A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 1973; 38:1–10.

27.

MacCallum

, Browne

, Sugawara

. Power analysis and determination of sample size for covariance structure modeling. Psychol Methods, 1996; 1:130–149.

28.

Boberg

, Gustafson

, Hawkins

, et al. Development, acceptance, and use patterns of a computer-based education and social support system for people living with AIDS/HIV infection. Comput Human Behav, 1995; 11:289–311.

29.

Guzman-Clark

JRS

, van Servellen

, Chang

, et al. Predictors and outcomes of early adherence to the use of a home telehealth device by older veterans with heart failure. Telemed J E Health, 2013; 19:217–223.

30.

Taylor

, Todd

. Assessing IT usage: The role of prior experience. MIS Quart, 1995; 19:561–570.

31.

Karahanna

, Straub

, Chervany

. Information technology adoption across time: A cross-sectional comparison of pre-adoption and post-adoption beliefs. MIS Quart, 1999; 23:183–213.

32.

Czaja

, Sharit

, Lee

, et al. Factors influencing use of an e-health website in a community sample of older adults. J Am Med Inform Assoc, 2013; 20:277–284.

33.

Evangelista

, Strömberg

, Westlake

, et al. Developing a web‐based education and counseling program for heart failure patients. Prog Cardiovasc Nurs, 2006; 21:196–201.

34.

Chen

, Chou

S-W

. Measuring patients' perceptions and social influence on home telecare management system acceptance. Int J Healthc Inf Syst Inform, 2010; 5:44–68.

35.

Lai

T-Y

, Larson

, Rockoff

, Bakken

. User acceptance of HIV TIDES-tailored interventions for management of depressive symptoms in persons living with HIV/AIDS. J Am Med Inform Assoc, 2008; 15:217–226.

36.

Shareef

, Kumar

. Predicting mobile health adoption behaviour: A demand side perspective. J Custom Behav, 2014; 13:187–205.

37.

Wilson

, Lankton

. Modeling patients' acceptance of provider-delivered e-health. J Am Med Inform Assoc, 2004; 11:241–248.

38.

Rogers

, Fisk

. Human Factors Interventions for the Healthcare of Older Adults. Boca Raton, FL: CRC Press, 2001.

39.

Fisk

, Rogers

. Handbook of Human Factors and the Older Adult. San Diego, CA: Academic Press, 1997.

40.

Fisk

, Rogers

, Charness

, et al. Designing for Older Adults: Principles and Creative Human Factors Approaches. Boca Raton, FL: CRC Press, 2009.

41.

Rogers

, Meyer

, Walker

, Fisk

. Functional limitations to daily living tasks in the aged: A focus group analysis. Hum Factors, 1998; 40:111–125.

42.

Stroetmann

, Husing

, Kubitschke

, Stroetmann

. The attitudes, expectations and needs of elderly people in relation to e-health applications: Results from a European survey. J Telemed Telecare, 2002; 8:82–84.

43.

Jackson

, Chow

, Leitch

. Toward an understanding of the behavioral intention to use an information system. Decis Sci, 1997; 28:357–389.

44.

CKL

, Karsh

B-T

. A systematic review of patient acceptance of consumer health information technology. J Am Med Inform Assoc, 2009; 16:550–560.