Telehealth In Health Centers: Key Adoption Factors, Barriers, And Opportunities

Telehealth is “the use of telecommunications and information technologies to share information and provide clinical care, education, public health, and administrative services at a distance.”¹ It is used through different technologies, such as videoconferencing, the internet, or store and forward of data. Telehealth encompasses telemedicine, which is focused on the remote provision of clinical services that have historically been delivered face-to-face. Telehealth services have the potential to improve access to care, especially in rural or urban areas with scarce health care resources.^2,3 Telehealth has demonstrated benefits for patients by improving the continuity of care, access to specialty care, and health outcomes.^4,5 Yet despite its benefits, this modality has not been fully implemented across different health care organizations. For example, only 42 percent of hospitals and 15 percent of family physicians in the United States have adopted telehealth.^6,7

The Health Resources and Services Administration (HRSA) funds community-based federally qualified health centers under section 330 of the Public Health Service Act to deliver affordable, accessible, comprehensive, and high-quality primary health care services in medically underserved areas. In 2016 there were nearly 1,400 HRSA-funded health centers, which collectively delivered services to nearly twenty-six million patients at more than 11,000 sites across the United States. Recognizing the value of telehealth for improving access to care, HRSA has provided direct grant funding and technical assistance to support health centers, particularly in rural areas, with telehealth implementation. While telehealth has been an allowable care delivery method and expenditure under HRSA’s Health Center Program grant funding, Medicaid—a major payer for health center services—also sees telemedicine as a cost-effective modality for providing services and allows state programs to cover it.⁸ A 2010–11 survey of federally qualified health centers (both HRSA funded and not) found that 38 percent of respondents provided at least one clinical telemedicine service, while over half expected to adopt telehealth in the next one to two years.⁹ However, this survey was not a comprehensive analysis of telehealth use by every HRSA-funded health center.

Understanding why some health centers choose to adopt telehealth while others do not is critical, as improving access to care is central to health centers’ mission. Telehealth adoption among health centers could be affected by key center attributes, including location, patients, operation, and reimbursement policies.¹⁰ If telehealth adoption is found to be associated with certain patient groups, such as homeless or uninsured patients, policies promoting telehealth could be targeted at the specific barriers health centers face when serving those patients. Meanwhile, although the majority of health center patients are Medicaid beneficiaries for whom telehealth services are considered cost-effective,⁸ state Medicaid policies are largely heterogeneous. Thus, it is important to investigate whether telehealth adoption is affected by specific Medicaid reimbursement policies. Our study provides a unique overview of telehealth adoption among health centers—an important safety-net setting that serves vulnerable and underserved populations in the US.

Study Data And Methods

This study examined factors associated with and potential barriers to the adoption of telehealth by HRSA-funded health centers through a mixed-methods approach. We make this distinction since not all federally qualified health centers are funded by HRSA. Specifically, we were interested in how rural versus urban location, quality infrastructure, and Medicaid reimbursement policies affect telehealth use. We conducted statistical modeling to identify factors associated with telehealth adoption among health centers, as well as a qualitative analysis to examine reasons why health centers were not using telehealth. This study was exempt from Institutional Review Board review, as the data did not include any personally identifiable information.

Data

We analyzed data for 2016 from the Uniform Data System, which collects annual, standardized data reported by health centers that include information about patient demographics, services, performance measures, and organizational characteristics. All HRSA-funded health centers are required to report data to the system. HRSA takes multiple steps to help ensure the accuracy and validity of the data reported. The 2016 data are the first comprehensive national data about the adoption and use of telehealth in health centers as well as about perceived barriers to implementing telehealth.

Outcome Measures

The primary outcome of our study was whether a health center was using telehealth, which was constructed from the answer to the question, “Are you using telehealth?” A definition of telehealth was provided with this question. Health centers that answered “yes” also answered another closed-ended question, “How are you using telehealth?” Respondents could choose from six optional answers. Accordingly, we constructed binary outcomes to identify each type of service provided through telehealth, including mental health care, primary care, specialty care, managing patients with chronic conditions, oral health care, and other services. Since a health center could use telehealth capacity for multiple functions, we also constructed a binary outcome for whether the health center was using telehealth to provide two or more types of services.

Rural/Urban Location, Quality Infrastructure, And Other Operational Characteristics

We hypothesized that a decision to adopt telehealth was associated with a health center’s rural versus urban location and its quality infrastructure. We included a rural/urban location indicator and two health center quality infrastructure variables: recognition as a patient-centered medical home (PCMH) and participation in a Health Center Controlled Network (HCCN). The PCMH is a model of comprehensive primary care delivery that is believed to improve the quality and efficiency of, and access to, the health care system.¹¹ HCCNs are groups of health centers that work together to address operational and clinical challenges related to the use of health information technology (IT).¹² HCCNs support the Federal Health IT Strategic Plan by helping health centers improve access to care, enhance the quality of care, and achieve cost efficiencies through practice redesigns.¹² We hypothesized that both PCMHs and HCCNs support improvements in the quality of care that are associated with a center’s decision to use telehealth. We also controlled for other operational measures, including two variables on behavioral staffing capacity: having any mental health full-time-equivalent (FTE) providers and any substance abuse care FTE providers. (For a comprehensive list of other operational variables, see online appendix exhibit A1.)¹³

Patient Characteristics

The patient population served at each health center varies and can affect the decision to implement telehealth. We included patients’ insurance status (Medicaid, Medicare, no health insurance, and other insurance) and several patient demographic characteristics measured as the percentage of the patient population with such characteristics served at a center: sex, age, member of a racial/ethnic minority group, homeless, migrant or seasonal agricultural worker, income below 100 percent of the federal poverty level, and not having English as the first language.

State-Level Medicaid Policy Characteristics

A health center’s investment in telehealth could be influenced by heterogeneous state Medicaid policies. We constructed four indicators to identify whether Medicaid reimbursed the following services: live video (services that occur in real time), store and forward (the transmission of videos and digital images such as x-rays and photos through a secure electronic communications system), remote patient monitoring, and transmission/facility fee.^14,15 Two additional variables identified whether Medicaid imposed the following specific requirements: a “location requirement,” including restricting telehealth reimbursement to care provided only in rural or underserved areas or limiting the type of facility that could be an originating site, and reimbursement of out-of-state providers for rendering services via telehealth.¹⁴

Regression

We estimated a multivariate probit model on the probability of telehealth use among all 1,367 health centers. Among the 523 centers that were using telehealth, we further estimated seven probit models on the probabilities of using telehealth for each of the six service types and for using telehealth for multiple services. After each probit estimation, the marginal effects of each factor on the probability of each outcome of interest were calculated with delta-methods standard errors.

Robustness Check

Our specification assumed a two-part decision process. However, it is possible that the decision to invest in telehealth is not a linear two-part process. Health centers might decide whether to implement telehealth and what type of service to provide through telehealth at the same time. Therefore, we assessed whether our findings persisted by estimating probit models for each type of service among all 1,367 health centers instead of only among those currently using telehealth. The results were fairly similar and therefore are not reported separately.

Qualitative Analysis

All health centers that reported no telehealth use described their reasons. Two coders generated a coding framework based on an initial review of the responses. These codes encompassed health centers’ orientations toward telehealth (six codes) and their reported barriers to telehealth (eight codes). Intercoder reliability for the first 100 responses indicated substantial agreement (Cohen’s kappa: 0.61; mean agreement: 94.72 percent; range: 88.00–98.97 percent).¹⁶ After resolving coding discrepancies, each coder coded 422 responses.

Limitations

This study had several limitations. First, the cross-sectional nature of this analysis with only one year of data made it difficult to infer causality.

Second, because of the limitations of the data, we were not able to study the intensity of telehealth use or any specific modalities or types of use such as store and forward versus live video or clinical care versus patient education.

Third, we were not able to investigate site-specific factors, as the data contained only center-level information. However, since decisions on expanding services and care delivery methods are usually made at the center level, we believe that this limitation is not critical.

Finally, the reasons for not using telehealth were based on self-reports by health center administrators and might not have captured all of the barriers to its adoption. Considering these limitations, we believe that our findings are still instructive for policy development.

Study Results

Health Center Characteristics

Almost half of the health centers were in rural areas (exhibit 1). Two-thirds of them were recognized as patient-centered medical homes, while 70 percent participated in a Health Center Controlled Network. Ninety-one percent of health centers had mental health FTE providers, but fewer than a third had substance abuse care FTE providers. The average patient population of a health center was 44 percent male, 52 percent racial/ethnic minority, and 64 percent ages 18–64. Health centers served higher proportions of Medicaid and uninsured patients and lower proportions of patients with other insurance and Medicare.

Almost all of the centers were in a state where Medicaid reimbursed some form of live video. Thirty-two percent had Medicaid reimbursement for store-and-forward services; 35 percent had it for remote patient monitoring; and 54 percent had it for the facility fee. Fourteen percent faced a location requirement by Medicaid, and 23 percent were in a state where out-of-state providers could be reimbursed by Medicaid.

Overview Of Telehealth Use

Thirty-eight percent of health centers reported using telehealth, with almost half of them using it for mental health care. In addition, 24 percent used telehealth for specialty care, 25 percent for primary care, 22 percent for managing patients with chronic conditions, 4 percent for oral health care, and 22 percent for other types of services. Slightly more than a quarter of the health centers with telehealth used it to provide more than one type of service. A map depicting the variation in telehealth use among health centers nationwide shows that health centers in similar locations could differ in their adoption of telehealth (appendix exhibit A2).¹³

Key Factors For Overall Telehealth Use

Health centers located in rural areas were associated with a 10.0-percentage-point increase in the probability of telehealth use, compared to those in urban areas (exhibit 2). Health centers in HCCNs were associated with a 6.3-percentage-point decrease in that probability, compared to those not in HCCNs. Health centers with mental health or substance abuse care FTE providers were associated with a 15.1-percentage-point and a 5.3 percentage-point increase, respectively, in the probability. In addition, health centers in states where Medicaid reimbursed some form of live video or store-and-forward services were associated with a 16.1-percentage-point or an 11.0-percentage-point increase, respectively. Finally, having a higher proportion of patients ages 18–64 and a higher proportion of Medicaid patients were each associated with a lower likelihood of telehealth use, although the marginal effects were relatively small.

Key Factors For Mental Health Services And Multiple Types Of Services

Health centers in rural areas were 12.2 percentage points more likely to use telehealth for mental health care, compared to those in urban areas. Health centers with PCMH recognition, any mental health care FTE providers, and any substance abuse care FTE providers had a higher probability of using telehealth for mental health services, compared to their counterparts. Medicaid reimbursement for live video and facility fees was also associated with a higher likelihood of providing mental health care through telehealth.

Several factors were associated with whether a health center used its telehealth capacity to provide multiple types of services. Centers with higher proportions of patients ages sixty-five and older or homeless patients were all associated with lower probability of using telehealth for multiple types of services, although the marginal effects were relatively small. Centers in states where Medicaid reimbursed live video or store-and-forward services had a higher probability of using telehealth for multiple types of services.

How Do Health Centers Not Currently Using Telehealth View It?

The reasons that health centers gave for not using telehealth indicated vastly different levels of readiness for, or understanding of, telehealth—generalizable as “soon,” “not yet,” and “not likely.” The “soon” health centers were exploring the use of telehealth or were in the process of implementing it (exhibit 3). The “not yet” centers were focused on other priorities, such as building patient portals. In contrast, the “not likely” centers did not perceive a need for telehealth or were not interested in or lacked information about telehealth. Some believed that telehealth was not as effective as face-to-face care. Finally, a handful were not using telehealth but had tried it before.

SOURCE Authors’ analysis of data for 2016 from the Uniform Data System. NOTES $N=844$. Health centers are those funded by the Health Resources and Services Administration (see the text).

The levels of readiness in rural and urban health centers were roughly the same. Notably, 14.2 percent of rural centers did not perceive a need for telehealth, despite the limited access to primary and specialty care in rural areas, because they felt that they were able to provide care on site, were fully staffed, or had partnerships with visiting or local specialists.

What Are Health Centers’ Barriers To Using Telehealth?

The barriers to using telehealth reported by health centers could be assigned to one of eight categories (exhibit 4). The issue of broadband, generally understood as insufficient bandwidth or lack of any high-speed internet access, was considered separately from other technical issues such as equipment and software because it is commonly perceived as a barrier.⁹ The most cited reasons by both urban and rural health centers for not using telehealth were cost and reimbursement, followed by technical issues aside from broadband. The first category included issues with billing insurers and paying providers, and the second included lacking equipment and having incompatible electronic health record (EHR) systems.

SOURCE Authors’ analysis of data for 2016 from the Uniform Data System. NOTES $N=844$. Health centers are those funded by the Health Resources and Services Administration (see the text).

Rural health centers were substantially more likely than urban ones to name miscellaneous technical issues (such as inadequate space or time), partners and providers, and broadband as barriers. Partners and providers proved to be a critical—and complex—barrier to telehealth, including such issues as internal staffing shortages, no clear workflow process, and the unavailability of specialty providers.

Some urban and rural health centers reported concerns about using telehealth for their patient populations, stating that some patients were not a good fit for telehealth because they were homeless, lacked access to technology, or required complex care. Regulations, policies, and scopes of work combined to create paperwork barriers that delayed or prevented both urban and rural health centers from adopting telehealth.

Discussion

Overall Telehealth Adoption And Rural-Urban Differences

In 2016 nearly 40 percent of health centers were using telehealth to provide services. The adoption rate in health centers was much higher than that found in a 2014 survey of family physicians (15 percent).⁷ The difference might imply that health centers have a greater need for using telehealth or are in a better position to adopt it, compared to other primary care settings. Given that many health centers reported that they were in the process of implementing telehealth or exploring the possibility of doing so, an increase in the adoption rate is likely in the future.

Overall, health centers’ decisions to implement telehealth appear to be influenced by their location and a combination of operational, patient, and state-level Medicaid policy factors. Our findings suggest that telehealth adoption is driven in part by the need to improve access, with health centers in rural areas more likely to invest in telehealth and use it for mental health care, compared to those in urban areas. This finding is consistent with our hypothesis, as telehealth could reduce the resource differential between urban and rural areas by enhancing access to care. Additionally, among health centers using telehealth, those in rural areas were more likely to provide mental health care, reflecting both the shortage of mental health providers in many rural areas and the need to use telehealth for care coordination. Mental health providers might not be physically available in some rural areas, and multiple providers could be separated by large distances—as is the case with many health centers in Alaska, Hawaii, and other Pacific islands. Despite their uptake of telehealth, rural health centers were more likely to report barriers with logistics, such as inadequate space, a lack of partners or providers, and broadband. Future policies to promote telehealth adoption in rural areas will need to address those issues.

Health Center Quality Infrastructure

Although being a patient-centered medical home was not significantly associated with overall telehealth adoption, it might encourage telehealth use for mental health care among centers using telehealth. As PCMHs often require integration of behavioral health and primary care, telehealth is a potential solution for those with limited local access to mental health specialists. A tenet of the PCMH is seamless care coordination across settings and providers, and telehealth could be an essential tool for facilitating this process.

Somewhat surprisingly, health centers that participated in Health Center Controlled Networks were less likely to use telehealth than nonparticipants. HCCNs provide specialized training and technical assistance to participants, such as shared resources and training, data analytics expertise to support quality measurement and improvement, and the ability to share and apply lessons learned across providers—while providing economies of scale for health centers through their strong group purchasing power. One explanation for our results is that grant funding from HCCNs focuses on the adoption and meaningful use of EHRs, including using EHR data to improve quality and support health information exchange across care settings. Health centers tend to have precious few human and capital resources to allocate to both technical and nontechnical aspects of using new technologies, such as process redesign or workflow. Centers in HCCNs might prioritize EHR implementation over telehealth. Moving forward, since most health centers have adopted EHRs, those in HCCNs might focus their technical capacity building more on telehealth adoption.

Medicaid Reimbursement

As both our statistical and qualitative analyses indicated, reimbursement was a key determinant of telehealth adoption among health centers. Specifically, state-level Medicaid policies were significantly associated with health centers’ adoption of telehealth. Medicaid policy barriers also potentially explained the negative association between telehealth adoption and the proportion of Medicaid patients at health centers.

Medicaid reimbursement for some form of live video (synchronous) and store-and-forward services (asynchronous) was a major policy driver for telehealth adoption by health centers. Live video is the most predominantly reimbursed telehealth modality in all states but Massachusetts and Rhode Island.¹⁴ Although fewer than 4 percent of health centers are in those two states, which lack live video reimbursement, those health centers were significantly less likely to invest in telehealth, especially for mental health (exhibits 1 and 2). On the other hand, store-and-forward services were defined and reimbursed by only eleven state Medicaid programs, as most states require the delivery of telehealth services in real time.¹⁴ Store-and-forward services do not require the presence of the patient with the provider. Medicaid reimbursement for these services could encourage health centers to implement telehealth, as this modality provides greater flexibility in using it.

Location requirement by Medicaid did not have a significant effect on telehealth adoption. Note that this result should be interpreted with caution. It is possible that our analysis did not capture some of the specific limitations on types of facilities serving as distance sites. While Medicare has a clear policy of preventing federally qualified health centers from being distance sites,¹⁷ facility-type regulations are less clear for Medicaid programs. Given that more and more state Medicaid programs have been eliminating the overall location restriction in recent years, however, this issue might become less critical for health centers.

Patient Population

A health center’s decision to invest in telehealth could be influenced by the patient population the center serves. Our analysis showed that patient population could be a barrier. Specifically, patient demographics, including sex and age, influenced how telehealth was used. Notably, health centers with more patients ages sixty-five and older were less likely to use telehealth efficiently by providing multiple types of services. This finding may reflect older patients’ preference for receiving care in person, or it may be that the nature of some types of services provided to older adults is better suited to in-person provision. One might suspect that a higher proportion of older patients could create concerns for health centers in building telehealth capacity because of restrictions from Medicare reimbursement policy.¹⁷ However, our analysis did not find any significant negative association between the proportion of Medicare patients and telehealth use. In other instances, having higher proportions of homeless patients also steered health centers away from using telehealth broadly for different purposes. It is plausible that health centers respond to the transient nature of homeless patients by centralizing most services at the clinic when the patient is present. To reduce population disparities, future policies should support health centers in overcoming those barriers.

Is Broadband An Issue?

Surprisingly, the lack of broadband, or insufficient bandwidth, was not commonly reported as a reason for not using telehealth. Only nineteen health centers reported this as a barrier, although most of them were in rural areas. This is in contrast to previous findings that the cost of broadband or insufficient broadband was a prevalent issue.^9,15 Several factors could contribute to the availability of broadband in health centers. First, nearly all health centers in the US have implemented EHRs certified by the Office of the National Coordinator for Health Information Technology as part of the meaningful-use program of the Centers for Medicare and Medicaid Services. These applications, like telehealth, require broadband. In addition, the availability of broadband funding through the Department of Agriculture, state and regional investments in broadband and programmatic changes, and greater awareness and uptake of the Federal Communications Commission’s Rural Health Care Program have all resulted in decreasing numbers of health centers without internet service. Notably, although relatively few health centers reported inadequate or missing broadband access as a barrier, this does not mean that these centers have sufficient bandwidth to support their increased use of telehealth. For example, multiple, concurrent real-time video sessions and the uploading of ultra-high-density images as part of the store-and-forward telehealth modality might require higher bandwidth. Additionally, whether patients of health centers have access to broadband in their homes or on mobile devices remains an issue.

Future Implications

Looking forward, the rate of adoption of telehealth in health centers is expected to increase, but barriers and challenges remain to be addressed. Furthermore, many health centers did not perceive a need for telehealth or either were not interested in or lacked information about telehealth. Health centers may benefit from more education about what telehealth is, its value, and how it can be used with different population groups and in different settings. Finally, there has been a concerted effort to advance the use of health IT through Health Center Controlled Networks. Although the focus of these networks was not initially on telehealth, there is an opportunity to leverage them to increase telehealth use across all health centers.

ACKNOWLEDGMENTS

The views expressed in this article are those of the authors and do not necessarily reflect the official policies of the Department of Health and Human Services or the Health Resources and Services Administration (HRSA). The authors acknowledge the earlier work by Bill England and Stephanie Begley in the HRSA Federal Office of Rural Health Policy, Office for the Advancement of Telehealth, which helped inform the authors’ qualitative analytic approach. The authors also acknowledge Mei Wa Kwong in the Center for Connected Health Policy for providing the fall 2016 report on state telehealth policy (see note 14).

NOTES

• 1 Health Resources and Services Administration. Uniform Data System: reporting instructions for 2016 health center data [Internet]. Rockville (MD): HRSA; [cited 2018 Oct 31]. Available from: https://bphc.hrsa.gov/datareporting/reporting/2016udsreportingmanual.pdf Google Scholar
• 2 Committee on Pediatric Workforce, Marcin JP, Rimsza ME, Moskowitz WB. The use of telemedicine to address access and physician workforce shortages. Pediatrics. 2015;136(1):202–9. Crossref, Medline, Google Scholar
• 3 Ronis SD, McConnochie KM, Wang H, Wood NE. Urban telemedicine enables equity in access to acute illness care. Telemed J E Health. 2017;23(2):105–12. Crossref, Medline, Google Scholar
• 4 Kvedar J, Coye MJ, Everett W. Connected health: a review of technologies and strategies to improve patient care with telemedicine and telehealth. Health Aff (Millwood). 2014;33(2):194–9. Go to the article, Google Scholar
• 5 Shane-McWhorter L, McAdam-Marx C, Lenert L, Petersen M, Woolsey S, Coursey JMet al. Pharmacist-provided diabetes management and education via a telemonitoring program. J Am Pharm Assoc (2003). 2015;55(5):516–26. Crossref, Medline, Google Scholar
• 6 Adler-Milstein J, Kvedar J, Bates DW. Telehealth among US hospitals: several factors, including state reimbursement and licensure policies, influence adoption. Health Aff (Millwood). 2014;33(2):207–15. Go to the article, Google Scholar
• 7 Coffman M, Moore M, Jetty A, Klink K, Bazemore A. Who is using telehealth in primary care? Safety net clinics and Health Maintenance Organizations (HMOs). J Am Board Fam Med. 2016;29(4):432–3. Crossref, Medline, Google Scholar
• 8 Medicaid.gov. Telemedicine [Internet]. Baltimore (MD): Centers for Medicare and Medicaid Services; [cited 2018 Oct 31]. Available from: https://www.medicaid.gov/medicaid/benefits/telemed/index.html Google Scholar
• 9 Cunningham M, Lara A, Shin P. Results from the 2010–11 Readiness For Meaningful Use of HIT and Patient Centered Medical Home Recognition Survey [Internet]. Washington (DC): George Washington University, School of Public Health and Health Services, Department of Health Policy; 2011 Nov 3 [cited 2018 Oct 31]. Available from: https://hsrc.himmelfarb.gwu.edu/cgi/viewcontent.cgi?referer=https://www.google.com/&httpsredir=1&article=1031&context=sphhs_policy_ggrchn Google Scholar
• 10 Schwamm LH. Telehealth: seven strategies to successfully implement disruptive technology and transform health care. Health Aff (Millwood). 2014;33(2):200–6. Go to the article, Google Scholar
• 11 Rosenthal TC. The medical home: growing evidence to support a new approach to primary care. J Am Board Fam Med. 2008;21(5):427–40. Crossref, Medline, Google Scholar
• 12 Health Resources and Services Administration. Health Center Controlled Networks [Internet]. Rockville (MD): HRSA; [cited 2018 Oct 31]. Available from: https://bphc.hrsa.gov/qualityimprovement/strategicpartnerships/hccn.html Google Scholar
• 13 To access the appendix, click on the Details tab of the article online.
• 14 Center for Connected Health Policy. State telehealth laws and Medicaid program policies: a comprehensive scan of the 50 states and District of Columbia. Sacramento (CA): The Center; 2016. Google Scholar
• 15 Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation, Office of Health Policy. Report to Congress: e-health and telemedicine [Internet]. Washington (DC): ASPE; 2016 Aug 12 [cited 2018 Oct 31]. Available from: https://aspe.hhs.gov/sites/default/files/pdf/206751/TelemedicineE-HealthReport.pdf Google Scholar
• 16 Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33(1):159–74. Crossref, Medline, Google Scholar
• 17 Center for Connected Health Policy. Community health centers and telehealth: a cost analysis report and recommendations. Sacramento (CA): The Center; 2015. Google Scholar