Miranda A. Moore, PhD, Megan Coffman, MS, Anuradha Jetty, MPH, Kathleen Klink, MD, Stephen Petterson, PhD and Andrew Bazemore, MD, MPH From the Department of Family and Preventive Medicine, Emory University, Atlanta, GA (MAM); the Robert Graham Center, Washington, DC (MC, AJ, SP, AB); and the Office of Academic Affiliations, Department of Veterans Affairs (KK).
- Miranda A. Moore, PhD,
- Megan Coffman, MS,
- Anuradha Jetty, MPH,
- Kathleen Klink, MD,
- Stephen Petterson, PhD and
- Andrew Bazemore, MD, MPH
From the Department of Family and Preventive Medicine, Emory University, Atlanta, GA (MAM); the Robert Graham Center, Washington,
DC (MC, AJ, SP, AB); and the Office of Academic Affiliations, Department of Veterans Affairs (KK).
- Corresponding author: Miranda A. Moore, PhD, Department of Family and Preventive Medicine, Emory University School of Medicine, Emory Family Medicine, 4500 N Shallowford
Rd, Suite B, Atlanta, GA 30338 (E-mail: ).
Prior presentation: This study was presented in part at the 43rd North American Primary Care Research Group Annual Meeting, Cancun, Mexico (October
2015); at the American Academy of Family Physicians State Legislative Conference, Minneapolis, MN (November 2015); and the
Eastern Economic Association Conference, Washington, DC (February 2016).
Purpose: Little is known about the attitudes toward and adoption of telehealth services among family physicians (FPs), the largest
primary care physician group. We conducted a national survey of FPs, randomly sampled from membership organization files,
to investigate use of and barriers to using telehealth services.
Methods: Using bivariate analyses, we examined how telehealth usage affected FPs’ identified barriers to using telehealth services.
Logistic regressions show the factors associated both with using telehealth services and with barriers to using telehealth
Results: Surveys reached 4980 FPs; 1557 surveys were eligible for analysis (31% response rate). Among FPs, 15% reported using telehealth
services during 2014. After controlling for the characteristics of the physicians and their practice, FPs who were based in
a rural setting, worked in a practice owned by an integrated health system or other ownership structure, and provided hospital/urgent/emergency
care were more likely to use telehealth. Physician and practice characteristics by telehealth use status, sex of the physician,
practice location, years in practice, care provided, and practice ownership were associated with the barriers identified.
Conclusions: Telehealth use was limited among FPs. Many of the barriers to using telehealth services cited by FPs are amenable to policy
- Logistic Regression
- Family Physicians
- Primary Care Physicians
- Surveys and Questionnaires
As health care delivery in the United States transitions to a patient-centered, value-based system with improved access to
services, physician availability is a challenge. Telehealth could help address this problem. Although the term telehealth has been widely applied and well recognized for more than 4 decades, it lacks a singular definition. Broadly, telehealth
is the use of technology to deliver health care services and information from a distance. Telehealth usage has evolved from
static “store-and-forward” applications in which information, such as radiologic images, is stored and then forwarded for
diagnostic review or a second opinion. Today’s clinicians are providing virtual visits in real time through secure, interactive
video exchange. These telehealth visits address a wide range of issues, from urgent to chronic, from primary care to subspecialty
consultation, and from initial diagnosis to follow-up and management.
As the largest health care delivery platform in the United States, the primary care setting1 offers great potential for expanding telehealth use. Telehealth is increasingly being demonstrated as a means of expanding
access to primary care2,3; enhancing core primary care features, including continuity of care3,4 and coordination across different care settings4⇓–6; and reducing health care costs7 and improving health outcomes.8,9 A systematic review investigating the merits of telehealth interventions in primary care found that telehealth was generally
more acceptable to patients than providers, outcomes were at least as effective as in-office care, and costs were no higher
than for in-office care.10
Improvements, however, are still needed. A recent study by Teladoc, Inc., an independent company offering mobile device–based
and Internet-based medical care delivered by a physician who does not have an established relationship with the patient, found
that users were not preferentially located in underserved communities, and providers had poorer performance than physician
office visits on measures such as ordering diagnostic testing and prescribing appropriate antibiotics for bronchitis.11
Many of the benefits of telehealth services are realized by patients, and patient acceptability and use are increasing. Specifically,
remote blood pressure monitoring devices have been shown to be easy for patients to use.12 In addition, parents of children seen for acute pediatric telehealth services in urban neighborhoods, using the Health-e-Access
model, indicated they were highly satisfied with the care their child received and noted advantages over alternatives in terms
of convenience, location, and service.13 In another study, primary care patients with depression indicated that participating in telepsychotherapy was acceptable.14 However, few patients had initiated or engaged in a telepsychotherapy visit.
Although we found no studies specifically investigating the cost-effectiveness of telehealth services in primary care settings,
telehealth services in other settings have shown mixed results with respect to improving health outcomes and reducing costs.
Acute care telehealth under the Health-e-Access program was found to be safe and effective.15 Residents of senior-living communities with access to acute care telehealth services reduced their emergency department visits
without increasing other health care service utilization or mortality16; frail older adults who received telemonitoring (monitoring of a patient’s vital signs in a remote setting such as their
home) services were found to have the same rates of hospitalizations and ED visits, and the same total days spent in a hospital,
as hospital receiving usual care.17
In addition, Medicare Advantage members with heart failure and a recent hospitalization randomized to telemonitoring services
provided by case managers experienced fewer inpatient days compared with the previous year.18 Postoperative telehealth visits for patients who had undergone parathyroidectomy were found to be cost-effective and efficient.19 In Europe, costs for implantable cardiac defibrillator remote follow-up visits were found to be essentially the same as those
for in-office follow-up visits.20
A few studies provide information about the knowledge, attitudes, and adoption of telehealth services among health care providers.
A systematic review of articles published through February 2013 investigating the perceptions of primary care clinicians,
administrators, and clinical staff regarding the acceptability and feasibility of remote monitoring technology in routine
adult primary care found only 15 studies meeting inclusion criteria.21 These studies revealed many negatives: barriers to implementation; the clinical relevance of the data collected; fewer patient
visits and the potential for overtreatment; insufficient time to monitor and discuss the data collected with the patient;
electronic health record incompatibilities; and uncertain legal liability regarding response protocols.
In a separate study, primary care physicians agreed that teledermatology increases access to dermatologic care, improves patient
care, and is acceptable to patients.22 Teledermatologists, however, indicated barriers to providing services that included delays in reimbursement, no reimbursement
for services, lack of communication with referring providers, and costly and inefficient software platforms.23,24
In the primary care setting, 1 study found that clinicians were satisfied with making clinical decisions regarding follow-up
care for 10 common chronic conditions when using information gathered from an online patient questionnaire instead of during
an in-office visit.4 Another study of physicians and allied health professionals, which used focus group interviews to investigate the use of
patient E-mails in the clinical care of patients with diabetes, found that most physicians did not E-mail patients directly
and were uncertain about its potential to improve care.25 By contrast, the allied health professionals used E-mail frequently and perceived value in the service.
When focusing more narrowly on live interactive video visits, few studies have investigated use and provider perceptions in
primary care settings. The Health-e-Access providers indicated they were comfortable collaborating with the telehealth assistants
and confident that their telehealth communications met parents’ needs; however, they felt slightly less confident in their
remote diagnosis compared with in-person visits.26 In Spain, a binary logistic regression analysis of survey data investigating the determinants of telemedicine use in clinical
practice revealed that primary care physicians placed the greatest importance on telemedicine’s potential to reduce costs
and on its usefulness to the medical profession. Secondary explanatory factors were the perception of the security of medical
information and confidentiality, and the patients’ predisposition toward telemedicine. The physician’s own opinion formed
a third set of factors influencing the use of telemedicine.27
In summary, while evidence does show that advances have been made in the use of telehealth in primary care settings, little
is known about the penetration of the use of various telehealth methods, that is, store and forward versus live interactive
video; the characteristics of the users versus the nonusers; or the factors associated with identified barriers to use.
To address this overall lack of information about telehealth, we designed a survey to investigate whether and how family physicians
(FPs) used telehealth services. The survey was fielded to a randomly selected, representative sample of FPs and collected
information on the individual and practice characteristics of FPs, their use of telehealth services, and the barriers to using
The survey instrument was developed with guidance from (1) a review of the literature on telehealth and primary care conducted
in 2013,28 (2) an expert panel discussion convened at the Robert Graham Center on January 23, 2014, and (3) an expert survey methodologist
from RTI International. The instrument and study protocol were approved by the institutional review board of the American
Academy of Family Physicians (AAFP), and all human participants consented when they returned the survey. After field testing
the instrument using a small group of FPs, the final survey instrument was reduced to 30 questions separated into 5 sections
(See Online Appendix). These sections focused on (1) physician characteristics, (2) practice characteristics, (3) attitudes
toward and barriers to using telehealth, (4) use of telehealth among telehealth users, and (5) beliefs about telehealth, with
separate sections for users and nonusers. The survey provided a specific definition of telehealth that focused on primary care services, primary care and specialist referral services, and the sharing of electronic health
data between a primary care provider and specialist (Figure 1).
Definition of telehealth.
We drew a random sample of 9000 FPs in direct patient care from the 2014 American Medical Association (AMA) Physician Masterfile.
To ensure sufficient responses from rural FPs, we oversampled rural FPs at a rate of 2 to 1. Because the AMA Physician Masterfile
has been known to have outdated addresses, this sample was then linked with AAFP membership data in an effort to obtain current
mailing addresses. The linkage produced a list of 5119 FPs who are members of the AAFP.
From this sample, 5000 FPs were mailed in January 2015 an introductory letter along with the survey questionnaire and a $2
bill as an incentive to complete the survey. Approximately 10 business days later, an E-mail reminder was sent to the nonrespondents
containing a link to an online version of the questionnaire. Two additional follow-up communications were sent to nonrespondents
over the next 6 weeks. Data collection closed after 8 weeks.
To determine whether an FP provided telehealth services, the survey asked whether they had used telehealth services in the
past 12 months (ie, calendar year 2014) and if so, “Approximately how many times in the past 12 months did you engage or refer
your patients for a telehealth consultation?” Users were also asked to indicate their (1) method of use (non–mutually exclusive
categories of real-time interactive video, shared computer screen images with audio, and stored or forwarded image or text
transmission), (2) clinical purpose of use (non–mutually exclusive categories of diagnosis or treatment, second opinion, follow-up,
chronic disease management, emergency care, and administrative purposes), and (3) the type of clinicians referred via telehealth
(non–mutually exclusive categories of specialists, other FPs, general internal medicine physicians, mental health treatment
providers, physical therapists, and dieticians). In addition, both users and nonusers were asked to identify barriers to telehealth
use from a non–mutually exclusive list including cost of equipment, lack of training on how to use telehealth, reimbursement
by insurers, and liability issues with telehealth.
Sampling weights were used in all analyses so the estimates were representative of FPs in direct patient care across the United
States, not just AAFP members. Specifically, the sample was weighted to account for the oversampling of rural physicians.
We also adjusted the weights for the underrepresentation of international medical school graduates, younger physicians, and
osteopathic physicians among respondents compared with all FPs in the AMA Physician Masterfile. Descriptive statistics were
computed for select items on the survey instrument. The total number of responses and percentages are reported for categorical
variables, with means and standard deviations reported for continuous variables (Table 1). The results are presented separately for the nonrespondents, the total sample, telehealth users, and nonusers. The nonrespondent
sample is distributed similar to the respondent sample; thus, the respondents should be a good representation of the population
of US family physicians. Bivariate analysis was performed to examine statistically significant differences between the users
and nonusers of telehealth, using χ2 tests for categorical variables and analysis of variance for continuous variables. Logistic regression was used to investigate
the factors associated with using telehealth services (Table 2) and with identifying each barrier to using telehealth (Table 3). The data analysis was conducted using Stata 14.0 (StataCorp, College Station, TX).
Comparison of Demographic Characteristics and Barriers to Telehealth Use in 2014 for Family Physicians Responding to the January
2015 Robert Graham Center Survey,* by Telehealth Use
Factors Associated With Telehealth Use in 2014 for Family Physicians Responding to the January 2015 Robert Graham Center Survey*
Factors Associated With Indicating That the Item Is a Barrier to Using Telehealth in 2014 for Family Physicians Responding
to the January 2015 Robert Graham Center Survey*
The survey obtained basic demographic and practice characteristics from respondents: years in practice, practice location,
use of electronic health record, and practice size. Based on the number of years in practice, respondents were categorized
into 4 groups: (1) 0 to 10 years, (2) 11 to 20 years, (3) 21 to 30 years, and (4) >31 years. Most respondents indicated they
provided general primary care. Respondents who provided hospital, urgent, or emergency care and “other” care were combined
into a single “other” category. Ownership of the primary practice location was categorized as “privately owned practice,”
a “hospital- or health system–owned practice,” and “integrated health system” (eg, Kaiser Permanente), or “other.” From the
AMA Masterfile, we identified the FPs’ medical school location (international or in the United States) and their type of medical
degree: osteopathic (DO) or allopathic (MD). Addresses were geocoded, and census tract information was used to determine the
urban or rural location of each practice setting.
Characteristics of Telehealth Users and Nonusers
From the 5000 FPs targeted, 1630 responded. The final analysis sample consists of 1557 respondents (response rate, 31.1%)
who had a valid survey identifier (38 did not) and reported engaging in direct primary care (35 were either not in direct
patient care or chose not to answer this question and were excluded).
Approximately 15% of the sample (n = 225) indicated they had used telehealth services in calendar year 2014. Telehealth users
differed from nonusers in many ways. They were more likely to be located in a rural setting (26% vs 15%; P < .001), to use an electronic health record (97% vs 92%; P = .006), and to work in a practice with ≥6 FPs (40% vs 29%; P = .0047). In addition, telehealth users were less likely to work in a privately owned practice (22% vs 39%; P < .001) and to provide general primary care to their patients (76% vs 86%; P = .004).
Telehealth users were less likely to report at least 1 barrier to providing telehealth services in their office than nonusers
(84% of users vs 90% of nonusers; P = .008). Lack of training and reimbursement were the most common barriers identified by both users and nonusers.
Methods and Clinical Purpose of Telehealth Use
FPs who used telehealth did so infrequently; with 22% using it 1 to 2 times in calendar year 2014, and another 26% using it
3 to 5 times (Table 4). Almost half of telehealth users indicated their telehealth usage involved real-time interactive video, 31% used store-and-forward
image or text transmission, and 11% used shared computer screen images with audio. Over half (55%) of the FPs who used telehealth
services in calendar year 2014 indicated they used it for diagnosis and/or treatment purposes. Other common purposes included
chronic disease management, follow-up, second opinions, and emergency care.
Characteristics of Telehealth Users in 2014 (n = 225) among Family Physicians Responding to the January 2015 Robert Graham
The majority of the FPs who reported using telehealth services used them to refer their patients to specialists (68%). Approximately
28% of telehealth users referred their patients to mental health treatment providers.
Factors Associated with Telehealth Use
Logistic regression results showed that the use of telehealth was significantly associated with practice location, practice
size, the type of care provided, and the ownership of the physician’s practice (Table 2). FPs practicing in a rural setting had 3 times higher odds of using telehealth in the past year than FPs in urban settings
(odds ratio [OR], 3.05; 95% confidence interval [CI], 2.19–4.25). FPs who provide general primary care were less likely to
use telehealth (OR, 0.51; 95% CI, 0.33–0.80). FPs working in practices owned by an integrated health system were more likely
to use telehealth than those in privately owned practices (OR, 3.53; 95% CI, 1.79–6.98).
Factors Associated with Identifying Barriers to Telehealth Use
As expected, being a telehealth user was associated with lower odds of identifying any barrier to providing telehealth (OR,
0.51; 95% CI, 0.33–0.77) (Table 3). FPs providing general primary care were significantly more likely to identify cost (OR, 1.53; 95% CI, 1.12–2.09), training
(OR, 1.61; 95% CI, 1.18–2.18), and reimbursement (OR, 1.68; 95% CI, 1.24–2.27) as barriers to using telehealth. Compared with
FPs practicing in a private practice, FPs practicing in an integrated health systems (OR, 0.55; 95% CI, 0.33–0.92) and FPs
in practices with some other type of practice ownership (OR, 0.51; 95% CI, 0.37–0.71) were less likely to identify reimbursement
as a barrier to use. Female FPs were more likely to identify training as a barrier (OR, 1.52; 95% CI, 1.21–1.91).
Compared with FPs who had practiced for ≤10 years, FPs who had practiced longer had lower odds of identifying training as
a barrier to providing telehealth services. Rural FPs had 33% lower odds than their urban counterparts of identifying liability
issues as a barrier (OR, 0.67; 95% CI, 0.56–1.06). By contrast, international medical graduates, compared with US medical
graduates, had 57% higher odds of identifying liability issues as a barrier to providing telehealth services (OR, 1.57; 95%
As telehealth technological capabilities improve and the demand for accessible health care services increases, telehealth
represents an important venue to meet the needs of patients. Few previous studies have examined the use of telehealth in primary
care settings. To our knowledge, our study represents the first in-depth assessment of these issues using a nationally representative
The findings highlight the low adoption rate of telehealth services in the provision of primary care. Fifteen percent of FPs
used telehealth services in the preceding 12 months (calendar year 2014), and many of the users did so infrequently. The greater
use of telehealth services by rural FPs may reflect the greater demand among rural patients as a result of travel impediments
and provider shortages.29
Although AMD Global Telemedicine, a provider of telemedicine equipment and technology, believes providing telehealth services
does not require a physician to be “tech-savvy,”30 FPs who currently use an electronic health record are more likely to use telehealth services. Economies of scale, which reduce
barriers associated with cost, may be the reason FPs in larger practices and integrated health care systems were more likely
to use telehealth. A surprisingly large 84% of telehealth users and 90% of nonusers reported at least 1 barrier to providing
telehealth services in their practice.
Several limitations must be considered when interpreting these findings. With a response rate of about 31% and only 15% of
the respondents indicating they had used telehealth services in the past year, our results on the use of telehealth are based
on the responses of only 225 FPs. In addition, as with all surveys asking respondents to report on past activity, recall bias
is a limitation. Although the survey design was based on a robust review of the literature, and a survey expert was engaged
to validate the survey, survey question order and phrasing could affect the way the FPs responded to the questions.
If telehealth services are to have a major impact in the primary care setting, more physicians will need to become experienced
in the use of these services. Many impediments to wider adoption exist; however, many of these barriers are amenable to policy
modifications. One suggestion for overcoming the training barrier is for family medicine residency programs to ensure that
graduating residents are offered opportunities to use telehealth services. To address issues of reimbursement, governmental
and private payers could engage in outreach efforts to increase awareness of their current allowed payments for telehealth
and either expand the types of telehealth services currently eligible for payment or develop new ways to reimburse telehealth
The authors thank WellPoint/Anthem, Inc., for funding of this survey and the American Academy of Family Physicians for fielding
the survey. Claire Gibbons, PhD, provided valuable work on the background for this article. Douglas Kamerow, MD, MPH, provided
many edits to the text of the article. Murrey Olmsted, PhD, a survey methodologist at RTI International, provided valuable
help with the survey design.
This article was externally peer reviewed.
Funding: This study was funded by WellPoint, Inc., under the AAFP/Graham Center Telehealth Survey contract. The funder had no control
over the survey questions, sample selection, or analysis.
Conflict of interest: none declared.
To see this article online, please go to: http://jabfm.org/content/30/3/320.full.
- Received for publication June 22, 2016.
- Revision received December 12, 2016.
- Accepted for publication January 6, 2017.
All copyrights for this article are reserved to Rural Health