Implementation and Clinical Outcomes of Facilitated Telemedicine in Opioid Treatment Programs: A Hybrid Effectiveness-Implementation Analysis

doi:10.21203/rs.3.rs-4486952/v1

Download PDF

Research Article

Implementation and Clinical Outcomes of Facilitated Telemedicine in Opioid Treatment Programs: A Hybrid Effectiveness-Implementation Analysis

https://doi.org/10.21203/rs.3.rs-4486952/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Hybrid effectiveness-implementation designs evaluate the effectiveness and implementation of interventions. We retrospectively evaluated the implementation of a stepped-wedge cluster randomized controlled trial (RCT) of a facilitated telemedicine model (experimental) integrated into opioid treatment programs (OTPs) compared to offsite referral (control) for hepatitis C virus (HCV) treatment. We compared organizational and implementation characteristics associated with an HCV cure and with high healthcare delivery satisfaction.

Methods

We used the Reach, Effectiveness, Adoption, Implementation, Maintenance (RE-AIM)-QuEST framework to guide data collection and evaluation. We evaluated the clinical effectiveness outcome (HCV cure) and patient centered outcomes (changes between in-person and telemedicine patient satisfaction questionnaire subscales: time spent with doctor, interpersonal manner, accessibility and convenience, and general satisfaction). We combined patient and organization-level information to conduct a configurational comparative method of coincidence analysis (CNA) to identify strategy configurations associated with each outcome. We also utilized the random forest method to identify the most important variables and reran the CNA analysis. For contextual interpretation, we discussed stakeholders’ perceptions of the findings through focus group discussions.

Results

Using RE-AIM-QuEST, in terms of reach, the RCT enrolled 96.5% of the original projected sample. We found that low patient load per provider or counselor, site liaison presence, and approaches that promoted case manager integration into site workflows, were associated with increased clinical effectiveness. When evaluating organization factors associated with high satisfaction with healthcare delivery, we found that approaches that promote communication, patient and staff education, and site liaison presence were important in the facilitated telemedicine arm. In the referral arm, additional important factors identified included the ability of the case manager to explain the referral process and OTP affiliation with a university. In terms of telemedicine implementation, free-standing sites and low patient volume were associated with high satisfaction. In the referral arm, university affiliation was an important factor.

Conclusions

Factors associated with communication, addressing participants’ concerns, and case manager integration within OTP workflows are important for successful implementation of facilitated telemedicine. Using these combinations may be important for expanding and evaluating HCV treatment uptake and for understanding successful implementation of telemedicine-based interventions in underserved populations.

Trial registration

Clintrials.gov registration number NCT02933970; Comparison of Telemedicine to Usual Care for HCV Management for Methadone-maintained Individuals - Full Text View - ClinicalTrials.gov

hepatitis C virus (HCV)

telemedicine

substance users

people with opioid use disorders

stepped wedge design

practice-based research

Although pragmatic clinical trials seek to conduct research in settings that are routine providers of healthcare, successful implementation approaches are not well defined, especially for underserved populations.
We evaluated implementation factors that promoted intervention effectiveness and high levels of satisfaction with healthcare delivery through facilitated telemedicine integrated into opioid treatment programs.
The most effective combinations promote full intervention integration into program workflows, facilitate interdisciplinary communication, and address staff and patient concerns largely through education.
Our results highlight approaches that 1) facilitate integrative care through telemedicine and 2) promote successful conduct of pragmatic trials in clinical care settings for underserved populations.

Telemedicine can increase healthcare access by transcending temporal and geographic challenges to healthcare delivery. Implementation of telemedicine to underserved populations necessitates consideration of broadband connectivity, digital health literacy, and patient-provider trust, issues that may be more pronounced in this population subset [1, 2]. Addressing these issues, while simultaneously incorporating health equity principles, promotes telemedicine as an approach to address health disparities [3, 4]. Since data are limited on the approaches to implement telemedicine for underserved populations, we conducted a randomized controlled trial (RCT) evaluating a facilitated telemedicine model (FTM) for the treatment of hepatitis C virus (HCV) infection in opioid treatment programs (OTPs). Because of shared transmission routes, people with opioid use disorder (OUD) have the highest HCV incidence and prevalence [5]. At the same time, recent HCV therapeutic advances through direct acting antivirals (DAAs) have near 100% efficacy after a 2–3 month treatment course [6]. Based upon DAAs’ therapeutic successes, HCV elimination by 2030 has been promoted by several organizations [7, 8, 9]. Simultaneously, however, DAA uptake has been low in the United States as only one-third of treatment-eligible patients received DAAs from 2013–2022 [10]. Without increasing HCV treatment access, HCV elimination by 2030 will not be achieved.

We designed the RCT as a comparative effectiveness study of HCV treatment outcomes utilizing the stepped wedge design. We found that 90.3% of FTM participants (experimental condition) achieved a sustained virologic response (SVR-HCV cure) compared to 39.4% of those in offsite referral (usual care, control condition) odds ratio = 2.9 (p-value < 0.001; 95%confidence interval [2.0, 3.5]) [11]. Additionally, participants were highly satisfied with FTM encounters, rated equivalent to in-person healthcare interactions [12]. To implement the RCT, we developed strategies to conduct research in venues with limited research experience and to engage an underserved population [13]. Furthermore, the RCT incorporated many attributes of well-designed clinical trials, such as involving frontline staff in routine clinical care settings [14]. Since the rigorous RCT was embedded in routine clinical care settings that provide substance use treatment, we sought to evaluate the organizational and implementation characteristics that led to the successful clinical and patient-centered outcomes [12]. These results may benefit research conduct for underserved populations and engaging underserved populations in telemedicine.

We used a mixed-methods approach to evaluate organizational and implementation factors that led to successful implementation of a stepped wedge cluster RCT, clinical effectiveness, and high satisfaction with healthcare delivery [15]. We employed configurational comparative methods (CCM) to identify combinations of characteristics that led to successful implementation and high satisfaction. We then returned the results to a multidisciplinary team of stakeholders to provide further context for the identified factors and combinations associated with the best outcomes. In addition to the factors that promoted an HCV cure, we evaluated characteristics associated with high satisfaction with healthcare delivery, either through FTM or offsite referral. We assessed these characteristics through changes in the subscales from the Patient Satisfaction Questionnaire (PSQ) that was collected at two time points, the initial FTM encounter and at the last treatment timepoint [16].

Implementation Science Framework:

We utilized the Reach, Effectiveness, Adoption, Implementation, and Maintenance (RE-AIM) Qualitative Evaluation for Systematic Translation (QuEST) expanded framework [17, 18] to guide collection of organizational and implementation characteristics that led to successful RCT implementation [19].

We reviewed the RE-AIM [20, 21] framework as a strategy to identify potential candidate measures [22], which had been collected or were readily accessible and could be acquired retrospectively [23] (Supplementary Table S1). Based on iterative discussions, we outlined patient, practice, and implementation measures against the RE-AIM QuEST dimensions of reach, effectiveness, adoption, implementation, and maintenance as described below and consistent with prior publications [15]. We next examined available data sources and reduced the list of candidate measures for further analysis and generated a list of available patient characteristics and organizational level (Table 1) and implementation characteristics (Table 2). We considered two alternative emerging approaches to conduct quantitative analyses: (1) CCM and (2) random forest (RF) analysis initially [24] followed by CCM. The RF procedure enabled us to select the potentially most influential variables prior to inclusion in the CCM analysis.

Reach: Reach is defined as “the intervention reaching the target population” [15]. The consent rate, defined as the percentage of approached eligible participants who consented to participate in the study, could be used as an estimate of reach at the patient level. In our case, the study was conducted entirely in OTPs that dispense methadone. (See supplementary Section 1.1.1 for information on calculation of the number of potentially eligible individuals).

Effectiveness: Clinical effectiveness of FTM is measured by the percentage of participants with undetectable HCV RNA assessed 12 weeks post-treatment completion. Supplementary Section 1.1.2 contains information on SVR rates. SVR rates per site are illustrated (Supplementary Table S2) [11].

Adoption: While a conventional definition of adoption is “to what extent are those to whom the intervention is targeted participating”, we have noticed a very high correlation between the number of people who initiate HCV treatment and the number who achieve SVR [25]. Given the unique characteristics of the population and the considerations of the OTP in which the FTM was integrated, we decided to rephrase the question to “what are the factors that affect participation in the intervention by the target population?”

From our prior analysis, we recognized that empathy was the most important aspect for people with OUD when evaluating the satisfaction with a healthcare encounter [12], based upon the weighting of individual PSQ questions (Supplementary Table S3). In the RCT, we assessed patient satisfaction with healthcare delivery on two occasions, at the completion of the first and final encounters. For this analysis, we evaluated the ratio of responses from the second (i.e., the SVR timepoint) measurement over the initial (i.e., the initial participant visit) among the 344 participants who completed the PSQ at both timepoints. Means of the ratios computed per site according to PSQ subscales stratified by study arm are illustrated (Supplementary Table S4). From the PSQ data, we evaluated three subscales: TSWD, IM, and AC. Supplementary material (Section 1.1.3) presents the approach to calculating the ratios of the PSQ. A list of PSQ outcome variables is illustrated (Supplementary Table S5).

Implementation: Implementation has been defined as “to what extent was the intervention continuously implemented?” Since we had PSQ subscale measurement of GS with healthcare delivery, we evaluated organizational and implementation characteristics associated with GS.

Maintenance: At the organizational level, maintenance can be defined as “to what extent did onsite HCV care, either through FTM or some component of the intervention, become routine within the OTPs, consistent with the definition used by prior investigators [15]. ”At the individual level, maintenance can be measured by the durability of SVR as indicated by the number of reinfections that occurred during the two-year follow-up period [11].The TIDieR (Template for Intervention Description and Replication) Checklist is included as an online supplement.

Quantitative Analysis:

The specific CCM approach we chose was Coincidence Analysis (CNA) [24]. CNA ascertains which combinations of conditions are most likely to be associated with a specific outcome, in this instance, high rates of SVR and patient satisfaction with healthcare delivery. CNA is a method of causal inference that identifies combinations of conditions that are minimally sufficient for an outcome, as well as multiple causal pathways (i.e., solution pathways) to achieve the best outcome. After exclusion of specific variables (see Supplementary material sections 1.2 and 1.3), the data included 23 features and 10 outcome variables. The features were divided into 7 organizational (Supplementary Table S6) and 16 implementation (Supplementary Table S7) variables and assessed consistency (i.e., proportion of cases with an outcome condition covered by a solution compared to all cases covered by a configuration), coverage (i.e., proportion of cases with the outcome covered by the solution compared to all cases with the outcome), and complexity (i.e., the number of conditions in a specific configuration). We examined these parameters separately for the FTM intervention versus offsite referral to determine which factors are associated with improved outcomes and those that differed between the two study arms.

To select the configurations, we ran the CNA using the “minimally sufficient condition” function within the R package “cna”, version 3.5.1 [26, 27]. This function provides a condition table that lists all the configurations that meet the consistency threshold, with columns listing the outcome, condition, consistency, and coverage. We start with consistency 1 and coverage 0.25 and obtain the configurations. We then repeat this process each time by reducing the consistency by 0.05. Eventually, we collect all the configurations that meet consistency of greater or equal to 0.75 and coverage of greater or equal to 0.25. These conditions range in complexity from 1 to 4. Beginning with complexity 1, we select the first reasonable condition with highest coverage in each complexity.

We performed analyses using the implementation and organizational variables (i.e., marginal analyses) separately as well as using the two sets of variables simultaneously (i.e., joint analyses). The total number of solutions for SVR over all pairs (consistency, coverage) (Supplementary Table S8) and with the highest consistency and coverage per level of complexity (Supplementary Table S9) by both analytical methods is illustrated. Similarly, the total number of solutions for satisfaction with healthcare delivery over all pairs (consistency, coverage) (Supplementary Table S10) and with the highest consistency and coverage per level of complexity (Supplementary Table S11) is illustrated.

Random Forest (RF) Analysis: An alternative approach to identify the variable combinations most likely to be associated with specific outcomes is to conduct a two-step analysis. In the first step, the RF algorithm is used to identify variables that are most important for the given outcome. The score of the covariate (or feature) measures how important the covariate is in predicting the given outcome. We use the model-independent permutation feature importance evaluation method to obtain the importance score of each variable, and then we select the variables starting from the one with the highest score until we obtain all variables that account for at least 95% of the importance. These selected variables form a subset of all the variables. In the second step, we apply the CNA to the subset of variables obtained from the first step using the same method and thresholds as described in the CNA section above.

In the RF application, we used Python 3.9.12, the machine learning library “scikit-learn”, function “RandomForestRegressor” for fitting RF models [28, 29]. The permutation feature importance is obtained using the function “permutation_importance”. We used 500 trees, and each feature was permuted 100 times. Definitions of frequent terms are listed in Supplementary material (Section 1.4).

Qualitative Analysis

Over a series of three 90-minute stakeholder consultation discussions, recorded and transcribed through Zoom, we presented the CNA findings to a range of RCT participant stakeholders including OTP leadership, administrators, clinical staff, case managers (CMs) and research staff. We sought to assess whether the factors identified through our analyses aligned with the stakeholders’ lived experiences. Session transcriptions were verified for accuracy, and we subsequently performed a thematic analysis [30]. Two investigators (AT, AD) coded the transcriptions and merged the higher-level codes into themes.

Organizational Characteristics: All participating OTPs were located in New York State, six downstate and six upstate. Organizational characteristics of 12 participating OTPs are illustrated (Tables 1 and S1). Four OTPs were university-affiliated, four were health system-affiliated, and four were community-based or free-standing. Between 9% and 38% of upstate site patients came from rural areas. The upstate sites also tended to have a younger age and fewer minority patients compared with downstate sites.

RE-AIM Evaluation: We used the RE-AIM QuEST framework (Table S1) to guide the identification and evaluation of the factors important for implementation of FTM into OTPs (Table 3). In our description below, we initially describe the organizational variables as these are relatively unmodifiable. Secondly, we describe the implementation variables, which were modifiable between sites. The description includes those solutions identified by CNA followed by the 2-step procedure. Thirdly, we interpret the findings based upon qualitative evaluation provided by stakeholders.

Reach: In the RCT, 761 participants were assessed for eligibility and 602 were enrolled, 96.5% of the original sample size of 624 participants. A lack of an indication for HCV treatment was the major reason for ineligibility among 144 individuals. The six downstate sites assessed HCV RNA annually, so very few downstate patients were ineligible. From the upstate sites, we estimated that approximately 1414 HCV-infected patients could have been approached (Supplemental section 1.1.1), and we enrolled 311 participants (22.0%).

Recruitment approaches differed across the 12 sites. At the downstate sites, CMs approached HCV RNA positive individuals. At upstate sites, CMs used many approaches to publicize the RCT to patients. An accurate number of people approached is infeasible from some methods, such as publicizing the RCT amongst OTP staff or placing posters with recruitment information within the OTP.

Effectiveness: In the telemedicine arm, we initially evaluated organizational and implementation variables separately. For the organizational variables, both CNA and the two-step procedure identified ‘upstate OTP’ and ‘low nurse practitioner (NP)/physician assistant (PA) patient volume’ as important individually and in combination (Table 3A1). In terms of implementation variables, both CNA and the two-step procedure identified ‘site liaison’ as an important single solution. In terms of combinations, CNA identified the combination of ‘site liaison’ + ‘CM attended group sessions’ as important. The two-step procedure identified the combination of ‘site liaison’ + ‘high CM onsite presence’ as importantly associated with increased SVR.

When we combined the organizational and implementation variables evaluating SVR among telemedicine arm participants with only CNA, we achieved the following solution for two organizational variables: ‘site affiliation (free-standing)’ and ‘low NP/PA patient volume’. In addition, we obtained five implementation variables: ‘site liaison’ ‘low CM turnover’, ‘CM attended group sessions’, ‘CM gave informal presentations’, and ‘CM attended staff meetings’. When we applied the two-step procedure, we obtained a more parsimonious list of solutions consisting of three organizational variables: ‘upstate location’, ‘low counselor patient volume’, and ‘site affiliation (free-standing)’ and two implementation variables: ‘site liaison’ and ‘high CM onsite presence’.

Among referral participants, both CNA and the two-step procedure identified the same organizational single factor, ‘site affiliation (free-standing)’ and the same combination factors, ‘site affiliation (free-standing)’ + ’small practice size’ as important solutions (Table 3A2). In terms of implementation variables, both the CNA and the two-step procedure identified the same single factor, ‘CM gave informal presentations’ as being important. In terms of implementation, the CNA and two-step procedure both identified the same three-two pair combinations as being important: ‘site liaison’ + ‘high CM onsite presence’, “high CM onsite presence’ + ‘poster with contact information’, and ‘CM attended group sessions’ + ‘poster with contact information’. The CNA method identified the additional combination of ‘poster with contact information’ + ‘CM attended staff meetings’.

When combining organizational and implementation variables together evaluating SVR among referral arm participants, the CNA procedure identified four organizational and eleven implementation variables. The organizational variables are ‘small practice size’, ‘site affiliation (free-standing)’, ‘upstate location’, and ‘low NP/PA patient volume’. The implementation variables we identified are: ‘site liaison’, ‘high CM presence onsite’, and ‘high CM education level’, plus an additional 8 variables. When we applied the two-step procedure, we obtained the same two organizational variables: ‘small practice size’ and ‘site affiliation (free-standing)’ and three implementation variables: ‘site liaison’, ‘high CM onsite presence’, and ‘high CM education level’. In summary, the two-step procedure provides more parsimonious solutions than CNA alone.

OTP staff in practices with fewer patients had the potential advantage of delivering greater patient-centered care. “The positive side of having smaller practice sizes [is] being able to support the patient every step of the way, counselors have more space to meet with them, [and] physician assistants can see them as well.” In terms of staffing ratios, upstate practices typically had lower patient volumes. Some upstate OTPs had multiple providers with provider-to-patient ratios ranging from 146–780 patients for medical doctors (MDs), 219–671 for NPs or PAs, and 20–75 patients per counselor. At the downstate sites, each MD/PA/NP covered 410–539 patients and 46–90 per counselor.

Obtaining buy-in and support from OTP leadership and staff was vital for RCT conduct. The site liaison served as a point-of-contact for implementation of study procedures. Sites that lacked a site liaison sometimes suffered from disjointed teams.

CM onsite presence and involvement in OTP activities were critical to engaging and supporting patients in HCV care. These findings are supported by the solutions of ‘high CM onsite presence’, ‘CM attending group sessions and staff meetings’, and ‘posting contact information’. “If the CM is onsite frequently, they can stay on top of the patient and make sure they follow through with seeking care and help them navigate the process.” An additional staff member remarked, “[I cannot] underscore the importance of the CM … A dedicated staff [member] who would keep track of patients who had HCV and follow [them] throughout the process.” Furthermore, CM integration into OTP workflows fostered several attributes. “The CM attending the staff meetings was very valuable because you would get high-level insights… [It] ensured that the CM is part of the staff, the team, and that they're all in communication.” Being in frequent communication with staff and patients assisted CMs in identifying and mitigating study-related issues and supporting patients.

Adoption: In prior work, we found that TSWD and IM were two of the three most highly weighted PSQ subscales [12]. Therefore, these constructs, along with AC, are important adoption measures to identify solutions associated with high satisfaction with healthcare delivery.

TSWD and IM: In the telemedicine arm, when assessing TSWD, the important organizational variables identified by both CNA and the two-step procedure were ‘low MD patient volume’ (Table 3B1). Additionally, CNA identified the combination of ‘low MD patient volume’ + ‘upstate location’. In terms of implementation variables, CNA identified ‘poster with contact information’ as well as the combination of ‘flag alert’ + ’high CM onsite presence’ were associated with high patient satisfaction. The two-step procedure identified 3 combinations as being important: ‘site liaison’ + ‘high CM onsite presence’, ‘high CM onsite presence’ + ‘low CM turnover’, ‘low CM turnover’ + ‘site MD/PA involved in recruitment’.

In referral, when assessing TSWD, important single organizational variables identified by both CNA and the two-step procedure were ‘OTP affiliation (university/hospital)’ (Table 3B2). The CNA procedure identified the combination of ‘upstate location’ + ‘low counselor patient volume’ while the two-step procedure identified ‘high percentage of rural patients’ + ‘low counselor patient volume’ as important. In terms of implementation variables, CNA identified the combination of ‘high CM educational level’ + ‘CM gave informal presentations to site staff”. The two-step procedure identified the combination of ‘patients educated using liver model’ + ‘CM attended site staff meetings’ as being associated with high patient satisfaction.

In terms of IM in telemedicine, both procedures identified ‘OTP affiliation (university/hospital)’ as being important (Table 3C1). The CNA procedure identified the triple combination of ‘flag alert’ + ‘site liaison’ + ‘CM verbally educated patients’ as associated with higher levels of patient satisfaction. The two-step procedure identified the combination of ‘high CM onsite presence’ + “low CM turnover’.

When assessing IM in referral, CNA identified the single factor of ‘OTP affiliation (university/hospital)’ as being important (Table 3C2). Both procedures identified the combinations of ‘small practice size’ + ‘low MD patient volume’ as being important. Additionally, the two-step procedure identified the combination of ‘onsite phlebotomy’ + ‘high CM onsite presence’ as important. Furthermore, the triple combination of ‘onsite phlebotomy’ + ‘CM attended group sessions’ + ‘CM verbally educated patients’ were important.

Increasing knowledge about addiction, HCV, and health systems was foundational to achieving high patient satisfaction. “Offsite liver specialists typically had no understanding of addiction, not showing as much empathy to our patients as we would see in our own clinic. As a result, patients may not be trusting, and may not follow through with their care.” Understanding opioid addiction was a prerequisite to building trust and respect with people with OUD. “Having the knowledge and understanding of addiction… Because once you have that understanding and empathy, you’re able to deliver personal care.” OTP patients and staff also needed to be educated about HCV. “All clinic staff need to have basic understanding of HCV and DAAs.” Providing education facilitated OTP staff’s ability to guide their patients through HCV care. Similarly, ‘high CM education level’ was explained as, “Understanding how to navigate systems… Through my education, I understand Medicaid and referrals… [I am] able to understand and reiterate information to patients.” When speaking to patients, CMs refrained from using medical jargon and utilized relatable, everyday language to ensure that patients truly understood the health information and treatment plan.

Patient-centered care, specifically time to address patient concerns and clear communication channels, increased patient satisfaction. “Satisfaction comes from the quality of the conversation [and] how you’re interacting with the patient. If you have a low amount of patients in the clinic, there’s more familiarity. You have more time to [have] those conversations.” Another stakeholder commented, “Having actions that show that everyone is on board and working as a team.” Similarly, “Flag alert is another reminder to go and see the PA because they’re going to help you complete whatever study activity that’s needed… coordination of care, everyone’s working together to help [them] become cured.” Onsite phlebotomy removed patients’ transportation barriers to accessing care and provided “a lot of touch points… check-ins and reminders, which is useful”. Opportunities for frequent interactions, clear communication channels, and expressing genuine interest in patients’ well-being promoted high satisfaction with healthcare delivery.

In telemedicine, the same single organizational variable identified by both CNA and the two-step procedure was ‘low MD patient volume’ (Table 3D1). The CNA procedure identified that the combination of ‘upstate location’ + ‘low NP/PA patient volume’ were associated with high satisfaction. In terms of implementation variables, both procedures identified ‘site MD/PA involved in recruitment’ as a single important factor. The CNA and two-step procedures both identified ‘high onsite CM presence’ as an important component of a two-factor combination, in CNA, the other factor was ‘CM attended site staff meetings’ and the two-step procedure identified ‘site liaison’ as the other important component.

In referral, CNA identified ‘OTP affiliation (health system)’ as a single important factor. The combination of ‘upstate location’ + ‘low counselor patient volume’ were identified as important organizational variables associated with high patient satisfaction (Table 3D2). Meanwhile, the two-step procedure identified the combination of ‘low PA/NP patient volume’ + ‘low counselor patient volume’ as associated with high AC. Similarly, the combination of ‘site liaison’ + ‘CM attended staff meetings’ were important variables identified by the CNA procedure. The two-step procedure identified the combination of ‘low CM turnover’ + ‘CM attended site staff meetings’ as important.

We emphasized patient convenience by integrating telemedicine encounters into the OTP, a venue frequented by patients due to methadone dispensing requirements for frequent in-person appearance. We coupled methadone administration with dispensing DAAs. Simultaneously, we sought to maximize accessibility through flexible scheduling. “If a patient needed to be seen at 6 am, [the telemedicine provider] would do that… There was much more flexibility with telehealth.”

Implementation: We used the PSQ subscale of GS to assess implementation. We measured GS at the initial and final healthcare encounters. In telemedicine, both procedures identified the single factor ‘free-standing OTP’ (Table 3E1). CNA identified that the combination of ‘upstate location’ + ‘low MD patient volume’ were associated with high patient satisfaction. The two-step procedure identified the combination of ‘low NP/PA patient volume’ + ‘low MD patient volume’. Important implementation variables, identified by both procedures, included the single factor ‘poster with contract information’ and the two-factor combination of ‘flag alert’ + ‘low CM turnover’. Finally, the CNA procedure identified a triple combination of ‘flag alert’ + ‘site liaison’ + ‘CM attended group sessions’ as associated with high patient satisfaction. Similarly, the two-step procedure identified ‘flag alert’ + ‘site liaison’ + ‘high CM onsite presence’ as being important.

In referral, both procedures identified the single factor ‘OTP affiliation (university/hospital)’ (Table 3E2). CNA identified the triple combination of ‘upstate location’ + ‘low NP/PA patient volume’ + ‘low counselor patient volume’. The two-step procedure identified the combination of “low NP/PA patient volume” + “low counselor patient volume’ as important organizational variables. In terms of implementation variables, the four following combinations were identified by both procedures: ‘site liaison’ + ‘high CM onsite presence’, ‘high CM onsite presence’ + ‘low CM turnover’, ‘low CM turnover’ + ‘high CM education level’, ‘low CM turnover’ + ‘CM attended staff meetings’.

FTM for HCV promoted general satisfaction with healthcare delivery in the OTP. “Knowing that [patients] have nurses there to speak to about adverse events, and can do the blood draw [onsite]…You feel this is my hub to improve my health, whether it’s with opioids or HCV. [It] is seen as you’re looking out for me. You understand me.”

Maintenance: The RCT offered hands-on experience with telemedicine to participating OTPs. At some sites, FTM served as a conduit over many years to potentiate direct onsite HCV management. Direct participation in FTM by onsite clinicians provided requisite expertise to manage HCV. Since RCT completion, 10 of 12 study sites (83%) report that they have continued to offer HCV treatment, either through telemedicine or onsite clinicians directly prescribing DAAs. These sites have continued to co-dispense DAAs with methadone. In recent years, New York State has lifted HCV treatment restrictions to permit PAs, NPs, and non-specialists to prescribe DAAs. Three sites have continued to utilize FTM for complex patients, such as cirrhotic individuals who require specialty care, or when limited staffing or HCV treatment knowledge does not support onsite treatment. Ten sites continued with onsite phlebotomy, a FTM model component.

The telemedicine experience was extremely useful when the COVID-19 pandemic lockdown required rapid conversion to telehealth-only encounters [31]. OTP administrators commented that COVID-19 and associated changes to reimbursement policies accelerated the adoption and implementation of telemedicine for underserved populations. Many sites have continued to use telehealth encounters for counseling and medication monitoring [32].

Qualitative Analysis: We performed a thematic analysis, depicting three themes under the umbrella of “Trust and Respect in the OTP” (Fig. 1). Specific organizational and implementation characteristics are illustrated (Table 4).

Theme 1: Endorsing facilitated telemedicine.

OTP staff were enthusiastic about FTM for HCV. However, they initially required training on HCV pathogenesis, diagnosis, and treatment, which enabled them to guide their patients in HCV care and answer DAA-related questions. Education also ensured that staff understood HCV-related roles and responsibilities.

Theme 2: Integrating the case manager.

The study-supported CM was integrated into OTP workflows and followed patients throughout the HCV treatment course. CM integration was imperative because it allowed CMs to attend staff meetings and group sessions to learn more about their patients. High CM onsite presence was critical to engaging and supporting patients in HCV care. High CM education levels enabled clear communication with patients and facilitated explanations related to navigating the healthcare system.

Theme 3: Meeting patients where they are.

FTM circumvented referral to offsite hepatitis specialists, which is ineffective at linking people with OUD to HCV care. Understanding opioid addiction was a prerequisite to building trust and respect with patients. Another prerequisite was flexibility in FTM scheduling and onsite phlebotomy. These examples also promoted FTM accessibility and convenience.

Additional stakeholder quotes are illustrated (Table 5).

Hybrid effectiveness-implementation analysis can elucidate strategies that lead to successful interventions. In this investigation, we examined organizational and implementation characteristics that led to successful HCV treatment and satisfaction with healthcare delivery. We observed that identified strategies in FTM were much more parsimonious compared to offsite referral. Important to both treatment approaches was the presence of a site liaison, a dedicated staff person acting as a point-of-contact for implementation of study procedures. Two common themes, patient-provider trust and integration into OTP workloads, underlie many of the other important strategies. For example, ‘CM attended group sessions’ indicates that CMs sought to engage with OTP patients in a safe space with the goal of health improvement [33]. ‘Low provider and counselor patient volumes’ and ‘high CM onsite presence’ both indicate the availability of the site staff and CMs, respectively, to address enrollment, retention, and study procedures. Similarly, variables such as ‘flag alert’, ‘CM attended staff meetings’, and ‘site MD/PA involved in recruitment’ foster integration of CMs into OTP workflows and promote cross-discipline collaboration. Taken together, these solutions underscore the importance of clear communication and close collaboration between OTP patients and staff to achieve both an HCV cure and high satisfaction with healthcare delivery.

FTM implementation aligned with the missions of the OTPs and the New York State Office of Addiction Services and Supports, the state agency that oversees a network of OTPs. A recent guidance has promoted patient-centered healthcare within OTPs [34]. Many of the identified solutions promote patient-centered care, including augmenting staff and patient HCV knowledge. Improved knowledge enabled OTP staff to advocate for patients through HCV treatment. One approach to increase staff education is Project Echo [35]. From the patients’ perspective, improved HCV-related knowledge, specifically related to the treatment course, assuaged anxiety. CMs explained that research staff with high levels of education facilitated explanation of the offsite referral process. Additionally, OTPs with a university affiliation benefited from a straightforward referral process to university-affiliated hepatitis specialists and typically shared a common electronic health record. These features improved patient accessibility and convenience. Other investigators have shown that people with OUD seek to evade negative perceptions of OUD shame [36, 37], which is a major reason why referral arm participants referred offsite from OTPs typically avoid or delay HCV care. Flexibility in scheduling and “meeting patients where they are” were additional attributes identified as important for high healthcare delivery satisfaction [12, 13].

Many important implementation and organizational characteristics involve high levels of CM availability in the OTP and their integration into OTP workflows. Three identified solutions, ‘low CM turnover’, ‘poster with contact information’ and ‘high CM onsite presence’, are associated with CM accessibility to OTP patients and staff. Other solutions, namely ‘CM attended group sessions’ and ‘CM attended staff meetings’ illustrate specifically how CM integration into OTP workflows was important for successful outcomes. Onsite phlebotomy was similarly viewed by stakeholders as an additional opportunity for patient interactions. Contemporaneously, OTP staff supported their patients through HCV treatment. Factors such as ‘low provider/counselor patient volume’, ‘provider involvement in recruitment process’ and ‘flag alert’ (i.e., as a communication method) are solutions illustrating the benefits of CM integration into OTP workflows. These results are supported by observations from OTP staff interviews where a warm handoff and open communication channels were identified approaches resulting from full FTM integration into OTP workflows [38, 39].

During RCT conduct, we considered the fidelity of the FTM across sites and three study periods due to the stepped wedge design [40]. Additional fidelity challenges resulted from COVID-related lockdowns and disruptions. Furthermore, the RCT was conducted at OTPs with limited clinical research experience. Even from the clinical perspective, HCV is infrequently addressed in OTPs. A recent national OTP survey revealed that 60.9% provided viral hepatitis testing, although only 12.6% provided onsite HCV treatment [41]. People with OUD are considered underserved due to limited healthcare access resulting from substance use stigma, competing priorities, and social vulnerabilities, such as homelessness, unemployment, and lack of social support [42]. These issues underscore the importance of identified solutions for intervention implementation for underserved populations.

Study strengths include use of emerging analytic methods (CNA, RF) and a mixed methods approach to assess variation in clinical and patient-centered outcomes as a function of patient, organization, and implementation variables. We selected candidate variables through the RE-AIM framework using the RF method, which identifies the most important variables contributing to intervention implementation. RF inclusion refines the CNA results. Unlike traditional quantitative methods, CNA does not provide strong statistical inference and relies on researchers’ subjective judgments for variable selection, threshold setting, and configurations. CNA is also sensitive to small changes in the data, which could substantially alter the results by adding or removing variables from the combinations. Similarly, clinical effectiveness (SVR) and patient-centered (satisfaction with healthcare delivery) outcomes permitted evaluation of FTM as a healthcare delivery modality. Furthermore, stakeholder discussions provided context to the CNA findings. Study limitations include a small sample size and the retrospective evaluation of organizational and implementation characteristics. Thus, some RE-AIM measures were acquired late, after RCT completion, or not at all. For example, we could not obtain independent data on in-person communication between potentially eligible patients and study staff as well as FTM fidelity in settings other than OTPs.

Conclusions: Evaluation of organizational and implementation characteristics that lead to successful FTM implementation, while simultaneously improving clinical and patient-centered outcomes, may lead to more effective and sustainable resource utilization. We found that opportunities to establish trusting relationships with patients were important to achieving successful clinical and patient-centered outcomes. Understanding factors of effective FTM targeted to underserved populations can support future research among people with socioecological vulnerabilities, low digital health literacy, and approaches to expand telemedicine as a healthcare delivery approach.

AC accessibility and convenience

CCM configurational comparative methods

CM case manager

CNA coincidence analysis

DAAs direct acting antivirals

FTM facilitated telemedicine model

GS general satisfaction

HCV hepatitis C virus

IM interpersonal manner

MD medical doctor

NP nurse practitioner

OTP opioid treatment program

OUD opioid use disorder

PA physician assistant

PSQ patient satisfaction questionnaire

QuEST qualitative evaluation for systematic translation

RCT randomized controlled trial

RE-AIM reach effectiveness adoption implementation maintenance

RF random forest

SVR sustained virologic response

TIDieR template for intervention description and replication

TSWD time spent with doctor

Ethics approval and consent to participate

The study was approved by the Institutional Review Board at University at Buffalo (parent institution), and by the Institutional Review Boards at each participating site. All participants provided informed consent.

Consent for publication

Not applicable.

Availability of data and materials

The datasets generated and analyzed during the current study are not publicly available due to confidentiality concerns but are available from the corresponding author on reasonable request.

Competing interests

AHT received grant support from the Patient-Centered Outcomes Research Institute (PCORI), the National Institutes of Health, Kaleida Health Foundation, Gilead Sciences, Abbott Laboratories, Novo Nordisk, and AstraZeneca. AHT has also served as an advisor to Gilead Sciences, AbbVie, and Novo Nordisk. MM received grant support from PCORI and Kaleida Health Foundation. JNT received grant support from PCORI, National Institutes of Health/National Heart, Lung, and Blood Institute and the Department of Health and Human Services’ Administration for Community Living. None of the other authors have any competing interests.

Funding

The study was funded the Patient-Centered Outcomes Research Institute (PCORI) award (IHS-1507-31640) and partially funded by the Troup Fund of the Kaleida Health Foundation. The statements in this work are solely the responsibility of the authors and do not necessarily represent the views of PCORI, its Board of Governors or Methodology Committee. Funders were not involved in data collection, analysis, or manuscript preparation.

Authors’ contributions

AHT: Conceptualization, formal analysis, investigation, resources, data curation, writing (original), writing (review/editing), visualization, supervision, project administration, funding acquisition.

MM: Conceptualization, methodology, software, validation, formal analysis, investigation, data curation, writing (original), writing (review/editing), visualization, supervision, project administration, funding acquisition.

MZ: Conceptualization, investigation, writing (original), writing (review/editing), visualization.

AL: Software, data curation, writing (original), writing (review/editing), visualization.

AD: Investigation, data curation, writing (original), writing (review/editing), project administration, visualization.

SGG: Investigation, data curation, writing (review/editing).

MT: Investigation, data curation, writing (review/editing).

KD: Investigation, data curation, writing (review/editing).

AS: Investigation, data curation, writing (review/editing).

LSB: Investigation, data curation, writing (review/editing).

JNT: Conceptualization, formal analysis, data curation, writing (original), writing (review/editing), visualization, supervision, project administration, funding acquisition.

All authors read and approved the final manuscript.

Acknowledgements

We acknowledge the participants, staff, and administrators at each OTP. We also acknowledge the study case managers.

Talal AH, Sofikitou EM, Jaanimagi U, Zeremski M, Tobin JN, Markatou M. A framework for patient-centered telemedicine: Application and lessons learned from vulnerable populations. J Biomed Inf. 2020;112:103622. 10.1016/j.jbi.2020.103622.
Henderson K, Winkler Y, Wyatt R, American Telemedicine Association Advisory Group on Using Telehealth to Eliminate Disparties and Inequalities. A Framework for Eliminating Health Disparities Using Telehealth. 2021. https://wwwamericantelemedorg/resources/a-framework-for-eliminating-health-disparities-using-telehealth/. Accessed 8 Oct 2023.
Mehrotra A, Bhatia RS, Snoswell CL. Paying for telemedicine after the pandemic. JAMA. 2021;325:431–2. 10.1001/jama.2020.25706.
Henderson K, Winkler Y, Wyatt R, American Telemedicine Association Advisory Group on Using Telehealth to Eliminate Disparities and Inequities. The Disparities Framework in Action: Activating Stakeholders to Eliminate Inequities in Care Through Telehealth. 2022. https://wwwamericantelemedorg/resources/the-disparities-framework-in-action-activating-stakeholders-to-eliminate-inequities-in-care-through-telehealth/ Accessed 8 Oct 2023.
Ryerson AB, Schillie S, Barker LK, Kupronis BA, Wester C. Vital Signs: Newly Reported Acute and Chronic Hepatitis C Cases - United States, 2009–2018. MMWR Morb Mortal Wkly Rep. 2020;69:399–404. 10.15585/mmwr.mm6914a2.
Ghany MG, Morgan TR, Hepatitis AASLD-IDSA. Guidance Panel. Hepatitis C guidance 2019 update: American Association for the Study of Liver Diseases–Infectious Diseases Society of America recommendations for testing, managing, and treating hepatitis C virus infection. Hepatology. 2020;71:686–721. 10.1002/hep.31060.
Fleurence RL, Collins FS. A national hepatitis C elimination program in the United States: a historic opportunity. JAMA. 2023;329:1251–2. 10.1001/jama.2023.3692.
Talal AH, Dharia A, Kapadia SN, Tyson GL, Birch S, Zerzan-Thul J, et al. Hepatitis C Virus Elimination Programs in Louisiana and Washington: Importance of Screening and Surveillance Systems. J Public Health Manag Pract. 2024;30:208–12. 10.1097/PHH.0000000000001808.
U.S. Department of Health and Human Services. Viral hepatitis national strategic plan for the United States: a roadmap to elimination (2021–2025). 2021. https://www.hhs.gov/sites/default/files/Viral-Hepatitis-National-Strategic-Plan-2021-2025.pdf.Accessed 19 Jan 2021.
Wester C, Osinubi A, Kaufman HW, Symum H, Meyer WA 3rd, Huang X, et al. Hepatitis C Virus Clearance Cascade - United States, 2013–2022. MMWR Morb Mortal Wkly Rep. 2023;72:716–20. 10.15585/mmwr.mm7226a3.
Talal AH, Markatou M, Liu A, Perumalswami PV, Dinani AM, Tobin JN, et al. Integrated Hepatitis C-Opioid Use Disorder Care Through Facilitated Telemedicine: A Randomized Trial. JAMA. 2024;331:1369–78. 10.1001/jama.2024.2452.
Talal AH, Sofikitou EM, Wang K, Dickerson S, Jaanimagi U, Markatou M. High satisfaction with patient-centered telemedicine for hepatitis C virus delivered to substance users: a mixed-methods study. Telemed J E Health. 2023;29:395–407. 10.1089/tmj.2022.0189.
Talal AH, George SJ, Talal LA, Dharia A, Ventuneac A, Baciewicz G, et al. Engaging people who use drugs in clinical research: integrating facilitated telemedicine for HCV into substance use treatment. Res Involv Engagem. 2023;9:63. 10.1186/s40900-023-00474-x.
Ford I, Norrie J, Pragmatic Trials. N Engl J Med. 2016;375:454–63. 10.1056/NEJMra1510059.
Forman J, Heisler M, Damschroder LJ, Kaselitz E, Kerr EA. Development and application of the RE-AIM QuEST mixed methods framework for program evaluation. Prev Med Rep. 2017;6:322–8. 10.1016/j.pmedr.2017.04.002.
Marshall GN, Hays RD. The patient satisfaction questionnaire short-form (PSQ-18). Santa Monica: RAND; 1994.
Glasgow RE, Dickinson P, Fisher L, Christiansen S, Toobert DJ, Bender BG, et al. Use of RE-AIM to develop a multi-media facilitation tool for the patient-centered medical home. Implement Sci. 2011;6:118. 10.1186/1748-5908-6-118.
Glasgow RE, McKay HG, Piette JD, Reynolds KD. The RE-AIM framework for evaluating interventions: what can it tell us about approaches to chronic illness management? Patient Educ Couns. 2001;44:119–27. 10.1016/s0738-3991(00)00186-5.
Glasgow RE, Vogt TM, Boles SM. Evaluating the public health impact of health promotion interventions: the RE-AIM framework. Am J Public Health. 1999;89:1322–7. 10.2105/ajph.89.9.1322.
Damschroder LJ, Aron DC, Keith RE, Kirsh SR, Alexander JA, Lowery JC. Fostering implementation of health services research findings into practice: a consolidated framework for advancing implementation science. Implement Sci. 2009;4:50. 10.1186/1748-5908-4-50.
Damschroder LJ, Hagedorn HJ. A guiding framework and approach for implementation research in substance use disorders treatment. Psychol Addict Behav. 2011;25:194–205. 10.1037/a0022284.
Moullin JC, Dickson KS, Stadnick NA, Albers B, Nilsen P, Broder-Fingert S, et al. Ten recommendations for using implementation frameworks in research and practice. Implement Sci Commun. 2020;1:42. 10.1186/s43058-020-00023-7.
Harden SM, Smith ML, Ory MG, Smith-Ray RL, Estabrooks PA, Glasgow RE. RE-AIM in Clinical, Community, and Corporate Settings: Perspectives, Strategies, and Recommendations to Enhance Public Health Impact. Front Public Health. 2018;6:71. 10.3389/fpubh.2018.00071.
Yakovchenko V, Miech EJ, Chinman MJ, Chartier M, Gonzalez R, Kirchner JE, et al. Strategy Configurations Directly Linked to Higher Hepatitis C Virus Treatment Starts: An Applied Use of Configurational Comparative Methods. Med Care. 2020;58:e31–8. 10.1097/MLR.0000000000001319.
Talal AH, Mukhopadhyay R, Veronesi V, Dharia A, Saraceno G, Markatou M. Improvements in Substance Use, and Social Outcomes after Integrated Opioid Use Disorder and HCV Treatment. Submitted. 2024.
Ambuehl M, Baumgartner M, Epple R, Kauffamn A, Thiem A. CNA: An R package for coincidence analysis (CNA). R. package version 3.5.1. http://CRANR-projectorg/package=cna 2015. Accessed 5 Nov 2022.
Whitaker RG, Sperber N, Baumgartner M, Thiem A, Cragun D, Damschroder L, et al. Coincidence analysis: a new method for causal inference in implementation science. Implement Sci. 2020;15:108. 10.1186/s13012-020-01070-3.
Breiman L. Random Forests. Mach Learn. 2001;45:5–32.
Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, et al. Scikit-learn: Machine Learning in Python. J Mach Learn Res. 2011;12:2825–30.
Terry G, Hayfield N, Clarke V, Braun V. Thematic Analysis. In: Stainton-Rogers W, Willig C, editors. SAGE Research Methods: The SAGE Handbook of Qualitative Research in Psychology. London: SAGE; 2017. pp. 1–40.
Ventuneac A, Dickerson SS, Dharia A, George SJ, Talal AH. Scaling and Sustaining Facilitated Telemedicine to Expand Treatment Access Among Underserved Populations: A Qualitative Study. Telemed J E Health. 2023;29:1862–9. 10.1089/tmj.2022.0534.
McDonald J, New York State Department of Health Issues Commissioner. Determination for the Use of Telemedicine to Ensure Continued Access to Life Saving Medication for New Yorkers Struggling with Mental Health and Opioid Use Disorder. Albany, NY: New York State Department of Health. 2023. https://wwwhealthnygov/professionals/narcotic/docs/2023-01-31_telemedicine_initial_exampdf. Accessed 5 Feb 2023.
Talal AH, Jaanimagi U, Dharia A, Dickerson SS. Facilitated telemedicine for hepatitis C virus: Addressing challenges for improving health and life for people with opioid use disorder. Health Expect. 2023;26:2594–607. 10.1111/hex.13854.
New York State Office of Addiction Services and Supports. Person-centered care guidance for OASAS certified programs Albany, NY: New York State Office of Addiction Services and Supports. 2018. https://oasas.ny.gov/system/files/documents/2020/01/oasasperson-centeredcareguidance.pdf. Accessed 9 Sep 2021.
Arora S, Thornton K. Novel Models of Hepatitis C Virus Care Delivery: Telemedicine, Project ECHO, and Integrative Care. Clin Liver Dis. 2020;16:5.
Biancarelli DL, Biello KB, Childs E, Drainoni M, Salhaney P, Edeza A, et al. Strategies used by people who inject drugs to avoid stigma in healthcare settings. Drug Alcohol Depend. 2019;198:80–6. 10.1016/j.drugalcdep.2019.01.037.
Treloar C, Rance J, Backmund M. Understanding barriers to hepatitis C virus care and stigmatization from a social perspective. Clin Infect Dis. 2013;57(Suppl 2):S51–5. 10.1093/cid/cit263.
Dickerson SS, George SJ, Ventuneac A, Dharia A, Talal AH. Care Integration for Hepatitis C Virus Treatment using Facilitated Telemedicine within Opioid Treatment Programs: Qualitative Study. J Med Internet Res. In.
Heath B, Wise Romero P, Reynolds KA. A standard framework for levels of integrated healthcare. Washington, DC: SAMHSA-HRSA Center for Integrated Health Solutions; 2013. https://www.thenationalcouncil.org/wp-content/uploads/2020/01/CIHS_Framework_Final_charts.pdf?daf=375ateTbd56. Accessed 3 Jan 2021.
Talal AH, Markatou M, Sofikitou EM, Brown LS, Perumalswami P, Dinani A, et al. Patient-centered HCV care via telemedicine for individuals on medication for opioid use disorder: Telemedicine for Evaluation, Adherence and Medication for Hepatitis C (TEAM-C). Contemp Clin Trials. 2022;112:106632. 10.1016/j.cct.2021.106632.
Jones CM, Byrd DJ, Clarke TJ, Campbell TB, Ohuoha C, McCance-Katz EF. Characteristics and current clinical practices of opioid treatment programs in the United States. Drug Alcohol Depend. 2019;205:107616. 10.1016/j.drugalcdep.2019.107616.
Talal AH, Jaanimägi U, Davis K, Bailey J, Bauer BM, Dharia A, et al. Facilitating engagement of persons with opioid use disorder in treatment for hepatitis C virus infection via telemedicine: stories of onsite case managers. J Subst Abuse Treat. 2021;108421. 10.1016/j.jsat.2021.108421.

Tables 1 to 5 are available in the Supplementary Files section.

Download PDF

Version 1

posted

You are reading this latest preprint version

Implementation and Clinical Outcomes of Facilitated Telemedicine in Opioid Treatment Programs: A Hybrid Effectiveness-Implementation Analysis

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Trial registration

Figures

Contributions to the Literature

Background

Methodological Overview

Quantitative Analysis:

Qualitative Analysis

Results

Discussion

Abbreviations

Declarations

References

Tables

Supplementary Files

Status:

Version 1