We developed an online survey (Multimedia Appendix 1) to measure the use of telephone and video technologies for delivering a specific set of SUD services [7], gauge the intent to use telephonic and video services after COVID-19, and explore the perceived readiness to use telephonic and video services, using TAMs previously developed by Gustafson et al [21] and Davis et al [31]. Designated regional Addiction Technology Transfer Centers (ATTCs) distributed the survey in their respective regions. Substance Abuse and Mental Health Services Administration (SAMHSA)-funded ATTCs support the workforce for addiction treatment and recovery. These regional ATTCs correspond to the 10 regional offices of the US Department of Health and Human Services. In total, 7 of 10 regional ATTCs representing 43 states partnered in this survey. The 3 nonparticipant regional ATTCs represented the remaining 7 states; these ATTCs declined participation, citing survey overburden due to other unrelated surveys in the field.

The survey was distributed and data were collected from May 15 to August 31, 2020. Principal investigators at the regional ATTCs distributed survey links to SUD treatment and recovery support administrators or personnel (physicians, counselors, and peer recovery coaches). They used various methods to disseminate the survey and obtain a convenience sample as large and representative as possible, including the use of regional ATTC listservs and partnering with state policymakers to share the invitation widely. The survey questionnaire contained 79 questions. The respondents could change their answers before submitting the survey. Cookies were not used to identify unique users, but no incentive was provided to the respondents for completing the questionnaire, thus limiting the likelihood of duplicate survey submissions. Multiple responses from the same organization were averaged and accounted for in regression analysis. Only submitted surveys were analyzed. The University of Wisconsin’s institutional review board approved the survey distribution and the recruitment of study participants (approval number 2020-0551). All data were collected using REDCap, a secure web application [32].

The survey included the following components and scales: organization type (ie, health system, opioid treatment programs, recovery community organizations, and specialty addiction treatment providers [nonopioid treatment programs]), organization location (ie, rural, small city, suburban, and urban), and organizational role of the respondents (ie, administrators and personnel providing treatment and recovery services, including counselors, physicians, and recovery coaches). Since rural individuals with SUDs are typically underserved [33] and are more markedly impacted by the opioid epidemic [34], and rural providers are more prone to adopt telehealth [20], the rural representativeness of the sample will be assessed using data from the National Treatment Center study [33].

For telehealth use, the survey assessed (1) the use of telephone and video technologies for the following services: screening and assessment, buprenorphine therapy, case management, intensive outpatient treatment, peer recovery support, regular outpatient treatment, and residential counseling sessions with binary yes/no variables; and (2) the projected intent to use telephonic and video services after implementing COVID-19 safety measures for these services, as per the following categories: “less than before,” “about the same,” “little more than before,” or “much more than before.”

The Organizational Readiness for Technology Use predictive tool developed by Gustafson et al [21] was used to assess dimensions of organizational readiness for the use of telephone and video technologies. Each item was evaluated using a 5-point Likert scale with endpoints of 1=“strongly disagree” and 5=“strongly agree.” The inventory assessed the perceived feasibility of reimbursement for the technology during and after COVID-19; access to information technology experts, clinical champions, and billing experts to support the use of these technologies; ease of technology integration into the workflow; staff, facilities, and equipment to promote the technology; leadership, staff, and patient support; technology accessibility and affordability; and staff training.

The technology acceptance survey included two subscales from the TAM [35,36]: ease of use and perceived usefulness. The ease of use scale assesses the ease of learning, customizing, and using the technology. Perceived usefulness assesses the extent to which the technology is perceived to enhance effectiveness, improve performance, increase productivity, and be useful. Items in these subscales were scored on 5-point Likert scales with endpoints ranging from “strongly disagree” to “strongly agree.” These subscales were used in conjunction with the projected intent to use determined by the survey participants.

Frequency distributions were used to describe organizational characteristics (setting and type), participant job descriptions, the use of telephone and video technologies for different SUD services, and the intent to use these technologies to deliver various services post COVID-19. In 3 regions, an overall survey response rate was calculated using a query on unique organizations identified from regional ATTC databases. For Regional ATTCs that lacked the capacity to conduct this query (n=5), SAMHSA’s Treatment Episode Database was used to estimate the available number of total SUD treatment organizations in that region’s states. The rural representativeness of the sample was calculated using data from the National Treatment Center Study [33], in accordance with a robust literature base indicating that rural organizations have been disproportionately affected by the opioid epidemic [34] and are more likely to adopt telehealth [20]. We used linear mixed-effects models (LMMs) to investigate differences in the intent to use telephone and video technologies based on organization location or setting and the job functions of the survey respondents, accounting for multiple respondents within the same organization. These models are expressed using the following equation:

where X denotes either the organization location, setting, or job function of the survey respondent; u_i is the random intercept for organizations, and ε_ij is the within-organization random error. 

Composite scores for the intent to use telephone and video technologies were generated by averaging those of the intent to use services after the implementation of COVID-19 safety measures across the different SUD services considered herein. To compare the odds of using telephone vs video technologies for the different post–COVID-19 services, a generalized LMM was used. These analyses compared the odds of reporting “more use”/“little more use” of these technologies post COVID-19 to those of reporting “about the same”/“little less” for telephone and video technologies post COVID-19. These models are expressed using the following equation: 

where u_i denotes the random intercept for organizations, and ε_ij is the within-organization random error. 

Variables determining organizational readiness for technology adoption were analyzed by comparing the scores for organizational readiness for the use of telephone and video technologies by using LMMs to investigate differences in factors between these technologies. These models are expressed using the following equation: 

where u_i denotes the random intercept for organizations; and ε_ij is the within-organization random error. 

Lastly, the TAM data were analyzed through mediation analysis, which compared the perceived ease of use and perceived usefulness variables to the composite intent of use variables for telephonic and video services. These mediation analyses were conducted using linear regression models to predict the use of technology [36]. These models are expressed using the following equations: 

where ε is the error in the estimation of the intent and ease of use. All analyses were conducted using the lme4 package (RStudio).

This study surveyed administrators and personnel from SUD treatment and recovery organizations to evaluate their current and projected use of telehealth for different SUD services during the early months of the COVID-19 pandemic. We examined key concepts for telephonic and video services separately, including the current use of technology, intent to use these technologies post COVID-19, and organizational readiness for using these technologies. We hypothesized that organizations would report high usage rates (>50%) of the current telephone and video technologies across a range of SUD services and would report significant odds of intending to increase their usage of these technologies post COVID-19. Furthermore, we conjectured that consistent with the TAM, perceived usefulness of technology would mediate the effects of ease of use on the intent to use these technologies. We did not propose a priori hypotheses about which SUD services would be most conducive to telehealth or whether differences would emerge between telephonic and video services.

Regarding the use of telehealth services during the survey, consistent with our hypotheses, most organizations used various services. Screening was the most common telephonically delivered service, whereas outpatient treatment was the most common video-delivered service.

These results are encouraging, since numerous studies have suggested that drug and alcohol screens can be administered telephonically with high levels of reliability and validity, and that outpatient treatment delivered via videoconferencing has comparable effectiveness to in-person care [37].

Regarding the future intent to use these services, consistent with our hypotheses, organizations reported significant odds of increasing their use of telehealth and video services after COVID-19. The two services for which respondents did not anticipate using telephonic services were buprenorphine therapy and residential counseling. For both of these services, respondents reported their willingness to use video-based services, suggesting their receptiveness to using technology in general, but they expressed specific concerns about using the telephone. The reticence to prescribe buprenorphine telephonically could reflect various factors, including provider mistrust and stigma toward patients with opioid use disorders, as well as concerns about diversion [38]. Use of video services was viewed more favorably, compared to telephonic services, for most SUD services.

Regarding organizational readiness for technology use, systematic differences emerged between telephone and video technologies. Relative to video services, respondents perceived telephonic services as more advantageous in terms of access, ease of use, affordability, and ease of sustainability. By contrast, video services were perceived as more valuable in terms of the likelihood of reimbursement and having the support of a clinical champion. Respondents preferred video services to telephonic services for all but two services. These findings are consistent with those of previous studies reporting that video-based counseling is associated with higher patient satisfaction but is substantially more expensive and not necessarily associated with superior levels of abstinence [39]. With the emergence of new videoconferencing tools, telephonic counseling would likely still have value owing to its simplicity, affordability, and reach, particularly among patients with limited access to video-based technologies.

Finally, on performing TAM analysis, the mediation model supported our hypothesis that the perceived usefulness of technology would mediate the association between the ease of use and the intent to use. This model emphasizes the critical role of the perceived usefulness in the adoption and current use of technologies. The survey outcomes were encouraging, in that SUD program administrators and personnel perceived the use of telephone and video technologies as useful during and after COVID-19.

However, our results should be considered within the context of several limitations. First, participants were recruited through convenience sampling of administrators and leaders contacted by principal investigators of regional ATTCs via email. Hence, the sampling methods might have resulted in a selection bias, such that individuals most comfortable with technology were most likely to complete the electronic survey. Furthermore, this sampling approach limits direct comparisons between participants who completed the survey and those who did not, though we could estimate a response rate and examine the rural representativeness of the sample on the basis of publicly available nationwide data. Second, even though the survey respondents represented 43 states and a range of organizational settings, the present findings based on a limited response rate may not be extrapolated to the general population. Third, patient-level data were not collected during sampling, thus limiting the representativeness of this sample and nationwide SUD organizations on the basis of the characteristics of the patients they serve.

Nonetheless, our results provide a promising outlook toward the use of telephonic and video services after COVID-19. Regarding the future applications of telehealth, this study suggests that the rapid transition to widespread use of telephonic and video services—necessitated almost overnight owing to COVID-19 stay-at-home orders and social distancing guidelines—is associated with high levels of provider receptivity to telephone technology. Clinicians perceived video services more appealing but telephonic services more accessible, suggesting that both channels play a role in the delivery of SUD services. Respondents’ perceptions of the continued use of telehealth post COVID-19 and their general readiness to use it support the positive outlook toward the role of telehealth in SUD services. Future studies are required to review the acceptance of these different service delivery approaches and their impact on care outcomes.