Attitudes Toward Telemedicine in Urban,Rural,and Highly Rural Communities

Abstract

Introduction: The rate of telemedicine adoption using interactive video between patient and provider has not met expectations. Technology, regulations, and physician buy-in are cited reasons, but patient acceptance has not received much consideration. We examine attitudes regarding telemedicine to better understand the subjective definitions of its acceptability and utility that shape patients' willingness to use telemedicine. Materials and Methods: Using the Montana Health Matters study (a random, statewide survey [n=3,512]), we use latent class analysis to identify groups with similar patterns of attitudes toward telemedicine followed by multinomial logistic regression to estimate predictors of group membership. Results: Although only 5% are amenable to telemedicine regardless of circumstance, 23% would be comfortable if it could be convenient, whereas 29% would be situationally amenable but uncomfortable using telemedicine. Still, a substantial percentage (43%) is unequivocally averse to telemedicine despite the inconvenience of in-person visits. Educational attainment, prior Internet use, and rural residence are main predictors that increase the likelihood of being in an amenable group. Conclusions: From the patient's perspective, the advantages of reduced travel and convenience are recognized, but questions remain about the equivalence to physician visits. Many people are averse to telemedicine, indicating a perceived incompatibility with patient needs. Only 1.7% of the respondents reported using telemedicine in the previous year; about half were veterans. Hence, few have used telemedicine, and key innovation adoption criteria—trialability and observability—are low. Increased attention to public awareness in the adoption process is needed to increase willingness to embrace telemedicine as a convenient way to obtain quality healthcare services.

Introduction

The rapid increase in Internet adoption and the expansion of mobile connectivity over the last decade have altered global communications, access to information, commerce, and the delivery of services. Although this shift has become almost ubiquitous, the financial sustainability of telemedicine and the scope of adoption into the mainstream healthcare system remain uncertain.¹ Significant advances have occurred in the underlying technology supporting online healthcare services, resulting in evidence of telemedicine's feasibility as a clinically effective equivalent to in-person medical care for many health problems.^2
–4 Yet there remain nontechnological barriers (e.g., implementation leadership, operating costs, physician buy-in, regulatory/legal restrictions, community support, and patient acceptance) that are frequently cited as undermining the adoption of telemedicine.^5,6 Consequently, despite the benefits of telemedicine, its adoption has been slow and below expectations.^1,7

Physicians, hospital administrators, and healthcare payers are considered the largest barriers to telemedicine adoption, with patients as largely passive participants in the process.⁸ Many companies have developed innovative solutions to address a perceived need in the marketplace only to discover that they misjudged consumer willingness to embrace it. Because telemedicine is a contingent process, physicians and hospitals may provide telemedicine as a service-delivery option, but the general public (patients/customers) can use other providers if telemedicine's quality of service does not match in-person visits. Despite the convenience of telemedicine, people consistently report that they prefer seeing their doctor in-person.^7,9 Thus hospitals and physicians may not invest financial resources into a product that patients might reject. More consideration must be given to the public's role as the “adopter” in the innovation-adoption process.^10,11

Broadly defined, telemedicine involves the use of various communication technologies to deliver or support direct medical care to patients and to promote collaboration between physicians regardless of physical distance.⁸ In this study, telemedicine is limited to direct physician–patient consultations using video communication over the Internet. Although many studies suggest that patients who use telemedicine are satisfied with it¹² and that exposure to telemedicine influences patients to have more favorable attitudes toward it,¹³ little existing research sheds light on perceptions about telemedicine among the general or minority populations.¹⁴ With a few notable exceptions,^9,15 studies tend to include small samples with limited population coverage, and study designs limit the generalizability of findings.¹⁶ The lack of understanding of public perceptions of telemedicine may be an important impediment to its broader acceptance.^7,17 We addressed this gap in the literature by examining public perceptions of telemedicine using latent class analysis (LCA) in a state-representative sample of Montana.

Materials and Methods

This study was based on survey data from the 2010–2012 Montana Health Matters study, a state-representative survey of Montana. These data are highly suited for this study of telemedicine as Montana has the fourth highest proportion of its population living in rural areas.¹⁸ Furthermore, Montana has a long history of telemedicine use. The Montana Healthcare Telecommunications Alliance was formed in 1995 and has four active telemedicine networks.^19,20 In addition, the Veterans Health Administration (VHA) has a telemedicine network for enrolled veterans.

The sampling frame included all Montana households in the U.S. Postal Service's computerized Delivery Sequence File. The target sample was stratified using the VHA urban, rural, and highly rural designations.²¹ Zip codes within each stratum were randomly selected with replacement, and 100 households were randomly selected without replacement to achieve the desired stratum size (1,000 urban, 2,000 rural, and 2,000 highly rural). We used a multimethod, five-wave mail/telephone survey protocol,²² and a $2 honorarium was included in the original mailing to maximize survey response. Of the 5,000 in the target sample, 3,512 responded to the 2010 survey (493 urban [comparison group], 1,391 rural, and 1,628 in highly rural). About 18% of the respondents were veterans. Telemedicine attitude questions come from a 1-year follow-up survey that had a mortality-adjusted 77% response rate (n=2,659; 355 urban, 1,042 rural, and 1,262 highly rural); all other measures come from the original 2010 survey. Only respondents who had at least five nonmissing responses to the eight questions (n=2,399; 312 urban, 947 rural, and 1,142 highly rural) were included in these analyses. Weights used take into account nonresidential addresses in the Delivery Sequence File, the multistage cluster sampling design, and survey nonresponse so results are representative of the state of Montana. The study was approved by Brigham Young University's Institutional Review Board.

Measures

Telemedicine attitudes were assessed using eight survey items. The survey defined “telehealth” as “an Internet service that uses video cameras and specialized equipment to permit physicians and medical specialists to consult with patients and to conduct some medical examinations over the Internet. Physicians and patients can talk and see each other without the patient having to travel to the doctor's office.” This description most accurately defines telemedicine, but the survey applied the term “telehealth” because of its broader public recognition. Respondents were asked to indicate their level of agreement or disagreement with the following items:

a. I prefer to see my doctor in person rather than using a video system on the Internet.

b. Using a telehealth system to meet with my doctor would be better than traveling long distances to see my doctor.

c. Using a telehealth system to meet with my doctor would be better than traveling during bad weather conditions to see my doctor.

d. People are likely to receive better-quality care when they see their doctor in-person rather than over an interactive video system.

e. I would use telehealth if it allowed me to significantly reduce the time I spend traveling to other communities to see my doctor.

f. I would prefer a telehealth visit with my own specialist over an in-person visit with another physician.

g. Having telehealth services in my community would mean that I would miss fewer appointments.

h. I would feel comfortable having telehealth visits with my doctor.

Response options were 1=“strongly agree,” 2=“agree,” 3=“neither agree nor disagree,” 4=“disagree,” and 5=“strongly disagree.” “Strongly agree” and “agree” were coded as 1, and all other response options were coded as 0. This not only allows a clear substantive interpretation of the analyses but also facilitates the LCA model estimates.¹⁷

LCA: Model Selection

LCA estimates probabilities of membership in unobserved classes or groups. Because the groups estimated in LCA are not observed in the population, identifying them is a nontrivial issue. In many statistical applications, likelihood ratio tests can be used to identify an appropriate model. However, models that estimate different numbers of groups are not nested. The Bayesian information criterion (BIC) is a common alternative to compare non-nested models. Smaller BIC values indicate better model fit. The final consideration is the ability of the estimated models to identify unique patterns in the data.²³ If an estimate with additional groups produces substantively similar patterns in responses to the measured items used to estimate the classes, the model with fewer classes is the one selected based on the principle of parsimony.¹⁷ Missing responses on the eight telehealth questions were treated in the estimation of the LCA model as missing at random using full information maximum likelihood.²⁴

Based on the BIC, the optimal number of groups of attitudes was five, but two groups were substantively similar. Consequently, we present results for the four-group solution. Estimates of gamma, the proportion of the sample in each group, and rho, the proportion of respondents in each group who agreed with a particular attitude, are presented in Table 1.

Table 1.

Rho and Gamma Values for Groups Derived from Latent Class Analysis Using Telemedicine Attitudes

	AVERSE	AMENABLE	SITUATIONALLY COMFORTABLE	SITUATIONALLY UNCOMFORTABLE
Gamma	0.43	0.05	0.23	0.29
Rho
I prefer to see my doctor in-person.	0.97	0.28	0.51	0.92
People get better care in-person.	0.90	0.00	0.46	0.83
Telemedicine is better than traveling long distances.	0.03	0.00	0.86	0.72
Telemedicine is better than traveling in weather.	0.23	0.10	0.96	0.95
I would use telemedicine to save traveling time.	0.05	0.06	0.92	0.55
I prefer telemedicine with my specialist than with another physician in person.	0.19	0.05	0.72	0.43
With telemedicine I would miss fewer appointments.	0.03	0.00	0.44	0.12
I would be comfortable with telemedicine visits.	0.04	0.18	0.85	0.18

Data are from the 2010–2012 Montana Health Matters study (n=2,399). Gamma is the estimated proportion of the sample in each group. Rho is the proportion of group members who agreed with the telemedicine question.

Because the groups are unobserved, each respondent has a nonzero probability of belonging in each group but for convenience is “assigned” to the group for which his or her probability of membership is greatest. Groups are then given substantive names based on the patterns of responses to the telemedicine items for respondents classified in each group. Only a small proportion of the sample has attitudes generally “amenable” to telemedicine use (5%); this group is less predisposed to say they want to see their doctor in person than are the other groups. Another 23% of the sample is amenable to telemedicine when in-person visits are perceived as inconvenient. We refer to this group as being “situationally comfortable” with telemedicine. In contrast, the “situationally uncomfortable” group (29%) reports a willingness to use telemedicine when in-person visits are inconvenient, but they feel uncomfortable with telemedicine. The remainder of the sample (43%) appears totally “averse” to using telemedicine services regardless of its potential convenience.

Results

Missing values on independent variables were treated using multiple imputation. Given the large number of variables in this exploratory analysis, we examined for multicolinearity but found no evidence for it. The means and proportions for study variables for each telemedicine group are presented in Table 2.

Table 2.

Descriptive Profiles of Groups of Telemedicine Attitudes: Means and Proportions

	AVERSE	AMENABLE	SITUATIONALLY COMFORTABLE	SITUATIONALLY UNCOMFORTABLE
Demographics
Female^a	0.54	0.52	0.64	0.58
Age (years)	58.40	59.14	56.65	58.57
Income (in $1,000)	56.65	53.02	62.49	61.20
Educational attainment	2.99	3.18	3.21	3.24
Marital status^a
Married	0.77	0.69	0.73	0.74
Divorced or separated	0.09	0.12	0.12	0.12
Widowed	0.07	0.13	0.08	0.06
Never married	0.07	0.06	0.07	0.07
Veteran status^a
Nonveteran	0.78	0.80	0.87	0.82
Enrolled veteran	0.07	0.09	0.06	0.07
Nonenrolled veteran	0.15	0.11	0.07	0.11
Has Internet at home	0.81	0.78	0.86	0.84
Health
Self-rated health	3.53	3.66	3.59	3.57
Number of chronic conditions	4.05	4.32	4.03	3.69
SF-12 physical	47.69	48.68	47.97	48.31
SF-12 mental	44.86	44.89	43.62	44.32
Depression	3.18	3.69	4.27	3.43
Pain	1.59	1.50	1.66	1.60
Doctor visits in last 60 days	1.09	0.93	1.08	0.98
Personality
Internet use	1.28	1.60	1.76	1.53
Prior telehealth use	0.01	0.02	0.04	0.01
External locus of control	2.29	2.42	2.36	2.38
Community satisfaction	5.74	5.56	5.67	5.67
Accessibility
Rurality^a
Urban	0.12	0.17	0.08	0.18
Rural	0.41	0.35	0.38	0.38
Highly rural	0.46	0.48	0.54	0.44
Distance to PCP	19.30	20.20	20.59	16.96
Transportation is easy	6.57	6.20	6.50	6.52
Drive self to PCP	0.86	0.83	0.88	0.88
Prescriptions in community	0.60	0.50	0.57	0.61
Availability
PCP availability	1.57	1.52	1.53	1.55
Overnight visits	2.41	2.49	2.64	2.55
Doesn't stay overnight^a	0.31	0.35	0.28	0.36
Accommodation
Appointments with PCP	5.61	5.58	5.50	5.54
Appointments with specialist	5.21	4.97	5.10	5.07
Fill prescriptions	6.57	6.32	6.46	6.55
Complete paperwork	5.80	5.54	5.62	5.73
Affordability
Worries about finances	2.90	2.61	2.93	2.89
Money for care	2.54	2.71	2.49	2.57
Has insurance^a	0.90	0.86	0.89	0.91

Data are from the 2010–2012 Montana Health Matters study (n=2,399). Variables were coded so that higher values represent more of the concept. Groups were identified using latent class analysis.

Proportions.

PCP, primary care provider; SF-12, 12-item Short Form.

Being averse to telemedicine is not a result of the absence of home Internet—there is very little difference in Internet access across the four groups (Table 2). The averse group, however, is less likely to use the Internet across a range of common uses (Table 3). Subsequent multivariate analyses confirm that the propensity to not use the Internet for everyday activities does impact attitudes about telemedicine.

Table 3.

Internet Use (Days During the Week) in Full Sample and Telemedicine Groups

	FULL	AVERSE	AMENABLE	SITUATIONALLY COMFORTABLE	SITUATIONALLY UNCOMFORTABLE
Online banking	1.55	1.41^c	1.82	1.87^a,d	1.47^c
Local businesses	0.19	0.15^b,c	0.36^a,d	0.25^a	0.19^b
Nonlocal businesses	1.13	0.98^c	1.25	1.46^a,d	1.10^c
Groceries, etc.	0.14	0.10^b,c	0.30^a,d	0.24^a,c	0.10^b,d
Check news	3.18	2.87^c,d	3.52	3.47^a	3.39^a
Local events	1.18	0.99^b –d	1.55^a	1.31^a	1.33^a
Nonlocal events	1.91	1.66^b –d	2.33^a	2.23^a	2.01^a
General information	3.86	3.46^b –d	4.24^a	4.31^a	4.07^a
Health information	1.45	1.17^b –d	1.80^a	1.91^a,d	1.47^a,c
Social networking	2.01	1.73^c,d	1.90	2.45^a	2.12^a
Gaming	1.05	1.03	1.23	0.95	1.12
Movies or TV	0.67	0.55^c	0.86	0.90^a	0.65

Data are from the 2010–2012 Montana Health Matters study (n=2,399). Groups were identified using latent class analysis.

Significantly different than the Averse group at p<0.05.

Significantly different than the Amenable group at p<0.05.

Significantly different than the Situationally Comfortable group at p<0.05.

Significantly different than the Situationally Uncomfortable group at p<0.05.

Relative risk ratios from multinomial logistic regression analyses are presented in Table 4. The averse group is omitted, serving as a referent for each of the remaining three groups. The first column compares the amenable to the averse group. Using the Internet for daily activities (1.40, p<0.01) and having an external locus of control (1.52, p<0.05) modestly increase the likelihood of being amenable to telemedicine rather than averse. Worrying about finances reduces the likelihood (0.78, p<0.05) and having money to pay for healthcare increases the likelihood (2.12, p<0.01) of being amenable. Having health insurance (0.28, p<0.01) and easily obtaining transportation for doctor appointments (0.77, p<0.05) and prescriptions in one's community (0.58, p<0.05) reduce the likelihood of being amenable to telemedicine.

Table 4.

Predictors of Membership in Groups of Telehealth Attitudes: Relative Risk Ratios from Multinomial Logistic Regression

	AMENABLE	SITUATIONALLY COMFORTABLE	SITUATIONALLY UNCOMFORTABLE
Demographics
Female	0.77	1.31^a	1.09
Age (years)	1.02	1.01	1.02^c
Income (in $1,000)	1.00	1.00	1.00
Educational attainment	1.26	1.13^a	1.17^b
Marital status
Married	—	—	—
Divorced or separated	1.20	1.26	1.23
Widowed	1.88	1.31	0.97
Never married	0.50	1.09	1.15
Veteran status
Nonveteran	—	—	—
Enrolled veteran	0.98	0.51^a	0.89
Nonenrolled veteran	0.58	0.52^b	0.72
Home Internet access	0.86	1.23	1.06
Health
Self-rated health	1.10	1.10	0.94
Number of chronic conditions	1.00	1.00	0.99
RAND-12 physical	1.02	1.00	1.02
RAND-12 mental	1.01	0.97^a	0.99
Depression	1.04	1.03	1.00
Pain	1.05	1.07	1.06
Doctor visits in last 60 days	0.91	0.97	0.94
Personality
Internet use	1.40^b	1.40^c	1.24^c
Prior telemedicine use	1.16	6.22^c	1.09
External locus of control	1.52^a	1.20	1.30^b
Community satisfaction	0.97	1.04	1.00
Accessibility
Rurality
Urban	—	—	—
Rural	0.67	1.49	0.61^b
Highly rural	0.82	1.97^b	0.71^a
Distance to PCP	1.00	1.00	1.00
Transportation is easy	0.75^b	1.01	0.98
Drive self to PCP	0.87	1.11	1.08
Prescriptions in community	0.58^a	0.81	0.91
Availability
PCP availability	1.15	0.98	0.96
Overnight visits	1.12	1.08	1.08
Doesn't stay overnight	1.50	1.05	1.21
Accommodation
Appointments with PCP	1.19	1.00	1.06
Appointments with specialist	0.84	1.02	0.93
Fill prescriptions	0.92	0.95	1.02
Complete paperwork	1.00	0.98	0.98
Affordability
Worries about finances	0.78^a	0.91	0.98
Money for care	2.12^b	0.94	1.03
Has insurance	0.28^b	0.89	1.06

Data are from the 2010–2012 Montana Health Matters study (n=2,399). Groups were identified using latent class analysis. Results are from 20 multiple imputation datasets and are relative risk ratios are the risk compared with the Averse group. Dashes indicate the reference group.

p<0.05, ^b p<0.01, ^c p<0.001.

In the second column of Table 4, the situationally comfortable group is compared with the averse group. Females are more likely to be situationally comfortable with telemedicine (1.31, p<0.05). Higher levels of education (1.13, p<0.05) and increased Internet use (1.40, p<0.001) modestly increase the likelihood of being situationally comfortable. Living in highly rural communities doubles the likelihood of membership in this group (1.97, p<0.01). Better mental functioning (RAND-12 mental subscale) decreases the likelihood of being situationally comfortable relative to the averse group (0.97, p<0.05). Being a nonenrolled veteran also decreases the likelihood of being situationally comfortable with telemedicine (0.50, p<0.01). Prior telemedicine use had a strong positive relationship with membership in the situationally comfortable group (6.22, p<0.001).

The final column in Table 4 compares the situationally uncomfortable group with the averse group. Increased education (1.17, p<0.01), daily Internet use (1.24, p<0.001), and increased external locus of control (1.30, p<0.01) increase the odds of being in this group relative to the averse group. Being older increases the likelihood of being situationally uncomfortable (1.02, p<0.001). Living in rural (0.61, p<0.01) and highly rural (0.71, p<0.05) communities decreases the likelihood of being in the situationally uncomfortable group relative to the averse group.

Discussion

After a half century of telemedicine development, the realization of its benefits remains elusive. Although reasons for its slow adoption are not clear, insurance coverage, doctor reimbursement, regulatory restrictions, and technology issues are common explanations.¹ Despite a substantial focus on developing technologies and enhancing physician buy-in, the current analysis suggests that consideration of public acceptance is needed for telemedicine to receive less conditional and more widespread acceptance among the general public. A key consideration is that telemedicine is a mode of service delivery and is potentially more sensitive to patients' preferences than a medical-treatment option where patients might rely more heavily on physicians' judgments. Although some funders of medical care, like the VHA or health maintenance organizations, can dictate how care is delivered, most patients choose their physicians and their method of accessing healthcare. Because patient satisfaction determines the financial viability of a doctor's practice, the adoption of telemedicine is a collective decision where both doctors and patients must favor the innovation.¹⁰

Results from multinomial logistic regression analysis indicated that the small amenable group has the Internet skills to use telemedicine and the financial resources to obtain the care they want and are less inclined to feel that quality care requires face-to-face care. Like the amenable group, people in the situationally comfortable group use Internet-based services more extensively and have higher levels of education. This group seems to appreciate the convenience of telemedicine in part because of physician access issues that come from living in highly rural communities. Why being a nonenrolled veteran lowers the odds of being in the situationally comfortable group and no others is unclear and requires more in-depth investigation. This finding is unexpected given that the VHA is not limited by private-sector physician reimbursement issues, has developed an extensive telemedicine network,²⁵ and has aggressively developed telemedicine services for physical and mental healthcare of rural veterans.^26
–28

Although the amenable and situationally comfortable groups have similar levels of education and Internet use, the situationally uncomfortable group has some specific differences. First, being older somewhat increases the odds of being in this group. Older people may be accustomed to seeing a doctor in-person and less comfortable with technology.¹⁵ Or, older respondents' attitudes could be informed by existing physical limitations, such as impaired vision or manual dexterity.²⁹ Second, living in rural and highly rural communities significantly reduces the odds of being in the situationally uncomfortable group. Thus, living in rural and highly rural areas produces a context where people will be situationally amenable to telemedicine given the difficulty of face-to-face doctor visits, but one group (i.e., younger people) will be more comfortable than the other using the technology.

Although Montana may be very different from other states in terms of economic conditions, physician availability, and rurality, these factors should have promoted favorable attitudes toward telemedicine use, particularly given the greater physical and mental healthcare needs in rural areas.³⁰ But they did not. Thus, there remains a significant divide between the perceptions of benefits by telemedicine advocates and attitudes toward this medical innovation by the general public.

Consequently, our results provide important context for discussions of the benefits of the technology, clinical applications, and changes in business models that occur in the telemedicine literature because little emphasis is placed on key human dimensions that impact telemedicine adoption by patients. Contingent innovation decisions are always slow,¹⁰ and the relative advantage of telemedicine versus seeing a doctor in person is difficult to demonstrate to the public. The convenience of telemedicine enhances its public perception, but this is largely limited to those who have access barriers. Most people (70%) are still concerned about the quality of care available through telemedicine; part of this concern may be discomfort with virtual relationships. As Skype™ (Google, Mountain View, CA) and FaceTime™ (Apple, Cupertino, CA) gain more acceptance for personal contacts among family members, discomfort with virtual “visits” should lessen, just as e-mail and Facebook^© (Menlo Park, CA) innovations have altered family communication patterns.

To enhance acceptance, more emphasis must be placed on increasing experience with telemedicine and communicating information to the public documenting its quality, convenience, and ease of use.²⁶ Media efforts could specifically address two of the most critical issues that affect the adoption rate of new technologies—observability (being able to see telemedicine in use) and trialability (being able to talk to people who have used it, or try it themselves). Policies must allow the public to assess the quality of telemedicine relative to actual physician visits. For example, expanding telemedicine to the entire Veterans Administration healthcare system has increased trialability and observability, although this is limited to veterans.³¹ In addition, pilot projects in highly rural areas that engage the community-as-a-partner to facilitate the provision of telemedicine services not only meet the needs of veterans living in the local area but also provide other community members the ability to directly observe the benefits of telemedicine.²⁶ Perhaps most important is that telemedicine could be mainstreamed as a service-delivery option for follow-up and other routine office visits, highlighting its convenience in a context where direct patient–doctor contact is deemed less critical. This could dispel perceptions of telemedicine as inferior care that is only useful for accommodating patients with extensive travel or other access issues.

In sum, although physician acceptance and other financial and technical barriers to telemedicine undoubtedly remain, extensive telemedicine networks have been in place in some areas for over a decade. Yet the sustainability and future of telemedicine are still uncertain.¹ In this study, perceptions of the general public toward telemedicine are largely negative—most still want to see their doctor in-person despite the convenience of telemedicine and the ubiquitous use of the Internet to obtain health information.^27,32 Consequently, we suggest, as does a recent study in Germany,⁹ that a neglected consideration in the slow telemedicine adoption process is public acceptance. Most other digital patient-engagement tools are facing similar adoption problems, leading marketing specialists to suggest that the major problem is that the various technologies may not be addressing the right patient-focused questions or needs.³³ Does telemedicine allow patients to feel as informed and confident about their health status as they do talking face to face with their physician? Do patients feel that online experience provides the same level of service quality they would receive in their doctor's office? Under what conditions and to what extent is telemedicine acceptable for various types of primary care and specialized medical services? Although past research finds that patients who use various telemedicine applications are satisfied with it¹² and our results indicate they are more comfortable with it, the overwhelming majority of the general public has yet to be convinced about the quality of telemedicine compared with an actual physician visit. Making the general public more aware of its benefits should increase customer willingness to embrace telemedicine as a convenient way to obtain quality healthcare services.

Footnotes

Acknowledgments

The Veterans Health Administration Office of Rural Health reviewed this manuscript prior to submission to Telemedicine and e-Health. This research was supported by the Veterans Health Administration Office of Rural Health and Brigham Young University.

Disclosure Statement

No competing financial interests exist.

References

Bashshur

, Shannon

, Krupinski

, Grigsby

. Sustaining and realizing the promise of telemedicine. Telemed J E Health, 2013; 19:339–345.

Tuerk

, Yoder

, Ruggiero

, Gros

, Acierno

. A pilot study of prolonged exposure therapy for posttraumatic stress disorder delivered via telehealth technology. J Trauma Stress, 2010; 23:116–123.

Gómez

, Hernando

, García

, et al. Telemedicine as a tool for intensive management of diabetes: The DIABTel experience. Comput Methods Programs Biomed, 2002; 69:163–177.

Ciemins

, Coon

, Peck

, Holloway

, Min

S-J

. Using telehealth to provide diabetes care to patients in rural Montana: Findings from the promoting realistic individual self-management program. Telemed J E Health, 2011; 17:596–602.

Stumpf

, Zalunardo

, Chen

. Barriers to telemedicine implementation: Usually it's not technology issues that undermine a project—It's everything else. Healthcare Informatics. 2002. Available at http://asclepieion.mpl.uoa.gr/pubASPIS/Article%20on%20Barriers%20to%20Telemedicine%20Implementation.htm (last accessed May 23, 2013 ).

Helitzer

, Heath

, Maltrud

, Sullivan

, Alverson

. Assessing or predicting adoption of telehealth using the diffusion of innovations theory: A practical example from a rural program in new mexico. Telemed J E Health, 2003; 9:179–187.

Turner

, Thomas

, Reinsch

. Willingness to try a new communication technology: Perceptual factors and task situations in a health care context. J Bus Commun, 2004; 41:5–26.

Menachemi

, Burke

, Ayers

. Factors affecting the adoption of telemedicine—A multiple adopter perspective. J Med Syst, 2004; 28:617–632.

Bürmann Genannt Siggemann

, Mensing

, Classen

, Hornberg

, Terschuren

. Specific health status has an impact on the willingness to use telemonitoring: Data from a 2009 health survey in North Rhine-Westphalia, Germany. Telemed J E Health, 2013; 19:692–698.

10.

Rogers

, Shoemaker

. Communication of innovations: A cross-cultural approach, 2nd ed. New York: Free Press, 1971 .

11.

Davis

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

12.

Mair

, Whitten

. Systematic review of studies of patient satisfaction with telemedicine. BMJ, 2000; 320:1517–1520.

13.

Cranen

, Veld

, Ijzerman

, Vollenbroek-Hutten

. Change of patients' perceptions of telemedicine after brief use. Telemed J E Health, 2011; 17:530–535.

14.

George

, Hamilton

, Baker

. How do low-income urban African Americans and Latinos feel about telemedicine? A diffusion of innovation analysis. Int J Telemed Appl, 2012; 2012:715194.

15.

Miller

, West

. Where's the revolution? Digital technology and health care in the internet age. J Health Polit Policy Law, 2009; 34:261–284.

16.

Whitten

, Love

. Patient and provider satisfaction with the use of telemedicine: Overview and rational for cautious enthusiasm. J Postgrad Med, 2005; 51:294–300.

17.

Collins

, Lanza

. Latent class and latent transition analysis: With applications in the social, behavioral, and health sciences. Hoboken, NJ: John Wiley & Sons, Inc., 2010 .

18.

Bishop

. How rural are the states?. Daily Yonder. April 3, 2012. Available at www.dailyyonder.com/how-rural-are-states/2012/04/02/3847 (last accessed March 28, 2015 ).

19.

Zauher

. Telehealth in Montana. Paper presented at Rural Health Care in Japan and the United States: Shared Challenges and Solutions, held July 1, 2001, in Bozeman, MT.

20.

Billings Gazette. Montana telehealth alliance. February 15, 2013. Available at http://billingsgazette.com/montana-telehealth-alliance/pdf_85b59c74-0173-517e-9679-6ca8b4e80212.html (last accessed March 28, 2015 ).

21.

West

, Lee

, Shambaugh-Miller

, et al. Defining "rural" for Veterans' health care planning. J Rural Health, 2010; 26:301–309.

22.

Dillman

, Smyth

, Christian

. Internet, mail and mixed-mode surveys: The tailored design method. New York: John Wiley & Sons, 2009 .

23.

Nagin

. Group-based modeling of development. Cambridge, MA: Harvard University Press, 2005 .

24.

Enders

, Bandalos

. The relative performance of full information maximum likelihood estimation for missing data in structural equation models. Structural Equation Modeling, 2001; 8:430–457.

25.

Booz Allen Hamilton, Inc.. Analysis for the Department of Veteran Affairs fee basis program for the Office of Rural Health. Final Report submitted to Office of Rural Health, Contract #W74V8H-04-D-0078. 2008 .

26.

Juretic

, Hill

, Luptak

, et al. Reaching out to older Veterans in need: The Elko Clinic Demonstration Project. J Rural Health, 2010; 26:325–332.

27.

Luptak

, Dailey

, Juretic

, et al. The Care Coordination Home Telehealth (CCHT) Rural Demonstration Project: A symptom-based approach for serving older veterans in remote geographical settings. Rural Remote Health, 2010; 10:1375.

28.

Brooks

, Turvey

, Augusterfer

. Provider barriers to telemental health: Obstacles overcome, obstacles remaining. Telemed J E Health, 2013; 19:433–437.

29.

Czaja

, Lee

. The impact of the internet on older adults. In: Charness

, Schaie

, eds. Impact of technology on successful aging. New York: Springer, 2003:113–133.

30.

Brown

, Schafft

. Rural people and communities in the 21st century: Resilience and transformation. Malden, MA: Polity Press, 2011 .

31.

Spotswood

. Use of VA telehealth soars as focus moves beyond Veterans in rural areas. US Medicine. 2013. Available at www.usmedicine.com/agencies/department-of-veterans-affairs/use-of-va-telehealth-soars-as-focus-moves-beyond-veterans-in-remote-areas/ (last accessed September 30, 2013 ).

32.

Ahern

, Woods

, Lightowler

, Finley

, Houston

. Promise of and potential for patient-facing technologies to enable meaningful use. Am J Prev Med, 2011; 40(5 Suppl 2):S162–S172.

33.

MobiHealthNews. In-depth: A brief history of digital patient engagement tools. 2014. Available at http://mobihealthnews.com/29985/in-depth-a-brief-history-of-digital-patient-engagement-tools/2/ (last accessed May 22, 2014 ).