We sought to identify whether rates of face-to-face attendance at DBT group skills training and individual therapy were comparable to rates of attendance via telehealth. In addition, we sought to compare whether there were differences in drop-out rates from DBT group skills training and individual therapy between treatment delivered face-to-face and treatment delivered via telehealth. Lastly, we were interested in whether First Nation clients’ attendance rates and drop-out rates were disproportionately impacted by telehealth service delivery compared to face-to-face service delivery.
There were no significant differences between attendance for either DBT skills training groups or individual therapy sessions between face-to-face and delivery via telehealth. The drop-out data numbers were too small to accommodate statistical analyses. There were no significant differences between attendance rates for First Nation clients between face-to-face and telehealth delivery.
The study focused on testing a common clinician assumption and concern that clients with severe mental health symptoms will not engage as much on telehealth as they will face-to-face and hence, that it may be inappropriate to offer DBT over telehealth because of concern regarding poor attendance. This study’s findings provide important preliminary data on the potential for telehealth therapy options to increase access and equity to mental health treatment for this vulnerable group whose treatment options are already limited.
Given the number of programs that are continuing to deliver DBT via telehealth (out of choice or necessity), it is pleasing to know that clients across Australia and New Zealand in this study received a comparable amount of treatment regardless of attending face-to-face or via telehealth. This indicates that on the whole, clients are willing to engage with treatment via telehealth.
While the literature is still emerging in this area, our findings reflect positive attendance rates previously found for clients engaging in telehealth therapy (14, 32). While our study found no difference in attendance rates between face-to-face and telehealth therapy sessions, Silver and colleagues reported a statistically significant improvement in attendance at a primary health care facility when the service transitioned to exclusively telehealth sessions because of COVID-19 risk. While interesting, the findings are limited as the sample was of 8 clinicians who worked at the one hospital who tracked their clients’ rates of missed therapy appointments. Some of Silver and colleagues’ hypotheses for this were improved convenience, decrease in certain barriers such as transport issues, telehealth provided some distance between client and therapist which may be comforting for clients who struggle with intimacy, and therapists taking initiative to start the session as they were the ones to make the phone call and send the video link and clients may have enjoyed this “active pursuit”. Lopez and colleagues (32) found significantly better attendance for a telehealth DBT group compared to a face-to-face group. They also hypothesized that telehealth may have reduced access barriers.
Our study only captured attendance data, and not the lived experiences of the clients. However, available data from qualitative studies of both clients with severe emotion dysregulation (36) and other clinical populations (32, 41, 42, 43) complement the finding of comparable attendance rates between telehealth and face-to-face DBT. Taken together, these qualitative studies indicate that even though clients would prefer a face-to-face group, if possible, telehealth is preferable to nothing. As such, it is possible that the convenience of a telehealth group outweighs both the negatives of telehealth and the positive aspects of in-person groups that are missed over telehealth, explaining a net equivalence in attendance rates between these two methods of delivery.
Within our sample, people who identified as First Nations’ peoples of Australia and New Zealand did not appear to be disproportionately impacted by telehealth service delivery. These groups were just as likely to attend telehealth sessions as face-to-face sessions compared with individuals who didn’t identify as First Nations people. Even though the research is limited, this finding appears to be consistent with some of the findings about digital access in this group reported by Rennie et al. (38) who compared data from the Australian Digital Inclusion Index (ADII), the National Aboriginal Torres Strait Islander Social Survey (NATSISS) and the Census of Population and Housing. They reported that the digital access gap between those who identify as Aboriginal and Torres Strait Islander in Australia and those who don’t identify who live in urban areas is shrinking and that individuals who identify as First Nations’ peoples reported having more positive attitudes towards technology than non-identifying. However, they are also more likely to be mobile-only users and may have disadvantages in regards to this. Our study did not assess whether clients were attending their appointments via a laptop, computer or mobile phone. Therefore, while our results are promising for people living in urban settings, future research should consider the methods in which clients’ attend appointments and whether this affects outcomes.
Certain contextual factors may have enhanced telehealth attendance, independent of the desirability of this delivery method. The telehealth time period occurred when geographical areas were in lock-down. Many of the clients would have been at home without other occupational or social demands hence it may have been easier to attend in this time period. Attending group and individual therapy may have given them something to do as many people reported being bored during this time (44) as well as helped them maintain their connection with others in a time of crisis (14). In addition, clients’ options were typically binary (telehealth or nothing) and may not reflect telehealth uptake during a context when it is optional.
Conversely some factors may have had a negative impact on attendance via telehealth. Teams had not previously delivered DBT skills training via telehealth. Many clinicians were apprehensive about delivering DBT via telehealth and were resistant to change, which may have impacted on their enthusiasm and the quality of the treatment being provided (27, 45). Some research has shown that clinicians have expressed a preference for face-to-face treatment delivery (3) and a reluctance to use telehealth for a number of reasons (8). These reasons include: lack of training, and concerns about efficacy, safety, privacy, and ability to navigate technology, which may all impact on attendance. Many clinicians and teams were learning lessons about how to deliver therapy via telehealth as they were delivering it and delivery was often beleaguered by technological challenges, and stressful (13, 25, 26, 27) and therefore unlikely to be a polished product. There are likely to be a range of factors that influence clinician attitude towards telehealth including familiarity and experience with it, treatment model, diagnostic group, organisation support and available resources. For example, in our experience, doing groups via telehealth is much easier with some teleconference platforms that have greater functionality than it is with other teleconference platforms with lesser functionality.
The results are aggregated across all clients, and therefore don’t illuminate at an individual level if some clients are more likely to attend on face-to-face and some more likely to attend on telehealth. Dunn et al (36) asked clients in their DBT clinic once Covid-19 was no longer a factor what mode of delivery they would prefer. For individual therapy, 34% chose face-to-face, 30% chose a combination of both, 24% chose telehealth and 12% were unsure. For group skills training, 33% chose telehealth, 30% chose face-to-face, 26% chose a combination of both and 10% were unsure (1% did not respond). Based on the responses from that study, it seems likely that some clients may have been more likely to attend therapy on telehealth than face-to-face and vice versa for other clients. Averaging attendance data obscures this important consideration.
Unfortunately, we did not have sufficient information to look at significant differences in drop-out rates comparing face-to-face delivery and telehealth or transitioning between the modalities. This would require a much larger sample to understand whether there are significant differences between DBT delivered via telehealth and via face to face for clients graduating from therapy or dropping out of therapy. The DBT program that the first and second author work in has been moving between periods of treatment delivery face-to-face and via telehealth for nearly three years since early 2020 in line with increased risk of transmission associated with the Covid-19 pandemic. Anecdotally, we have noticed greater rates of drop-out when clients commenced on telehealth and the program then transitioned back to face-to-face. This requires further examination with larger samples.
Another limitation of this study is that the data were collected over a brief time period (approximately 2 months for each time period). As such, we don’t know how generalizable these results are to treatment after the first peak of the pandemic. This study relied on the accuracy of recording of attendance and clinicians may have had different approaches to how they calculated missed and scheduled appointments per week. However, the research team dedicated time to corresponding with team leaders to get as much clarification as possible on this to ensure the data were as clean and representative as possible.
We also do not know about the fidelity to the model of the treatment programs that were involved, nor what the standard clinical outcomes were of the programs involved in this study. Hence, we were comparing whether overall there was a reduction in attendance when clients were participating via telehealth in contrast to when they had attended face-to-face. Future research would be strengthened by investigating attendance and drop-out with programs that were certified allowing the assumption of a minimum level of fidelity.
As only one adolescent team was included, we could not examine whether there are differences between adolescent programs compared with adult programs. A recent review by Reis et al (46) did not find any studies of online intervention for adolescents with personality disorders, hence, more research is needed in this area. Given the well-documented facility with online solutions experienced by digital natives, it may well be that patterns of telehealth attendance for adolescents may be very different from their adult counterparts.
This study compares attendance levels in telehealth and face-to-face DBT for those clients who were able to engage remotely. Barnett et al (3) in their umbrella discussion of systematic reviews note that researchers did not generally report what percentage of clients were excluded because they did not have the resources to engage in telehealth. This was not assessed in our study, but in Cooney et al.’s (27) qualitative study practical technological issues and resource deficits for clients and clinicians were one of the most-cited barriers to providing DBT over telehealth. Without attending to the lack of resources to engage in telehealth, excluding people may result in the exacerbation of existing inequalities.
This study has a number of strengths. It was sufficiently powered and risks of selection bias were low given that we collected service level data for all clients of attendance and drop-out rates. In addition, the programs contributing data are relatively homogenous in terms of the structure and represent a range of programs across Australia and New Zealand rather than from a single service.
To our knowledge, this is one of the first studies to provide quantitative data for people with severe emotion dysregulation who were enrolled in a DBT program and participated via telehealth. Ultimately, we need data comparing clinical outcomes between those who attend face-to-face and those who attend on telehealth, however, this is beyond the scope of what we could measure given that this was an opportunistic study conducted during the Covid-19 pandemic. This is an important area of future research. To build on the findings of this paper, it would also be valuable to track attendance and drop-out rates of clients enrolled in face-to-face and telehealth DBT programs over a longer timeframe with a larger sample. It would also be helpful to examine telehealth vs face-to-face attendance for adolescents specifically.
In the absence of empirical research regarding the use of DBT delivered via telehealth, when the Covid-19 pandemic began and lockdowns prevented delivery of face-to-face treatment, DBT programs that transitioned to telehealth needed to make a number of assumptions. One of them was that delivery of the treatment via telehealth would be better than no treatment at all. Whilst there are no trials comparing those two options, anecdotally from the qualitative research that has been conducted (13, 25, 26, 27), this study contributes to the evidence for utility in delivering DBT via telehealth.
Comer (2) suggests that with the widespread uptake and acceptability among both clinicians and clients that the use of telehealth will outlast the Covid-19 pandemic and become the dominant mode of mental health delivery. This is consistent with the data from a survey in the US of mental health clinicians and organisation who reported a high likelihood that telehealth would continue post-pandemic (47). However, as we move out of the Covid-19 pandemic, ethically, we need to try and determine whether delivery of treatment via telehealth for people with a diagnosis of BPD or severe emotion dysregulation is equivalent to outcomes achieved where the therapy is delivered face-to-face or in what situations it is a preferred or less preferred option in order to be able to be transparent with clients regarding the state of the evidence. It is not ethical to fall into a method of delivery that yields poorer outcomes, through convenience. In their comprehensive review of the availability, efficacy and clinical utility of DBT delivered via telehealth Van Leeuwen et al (37), suggest a return to face-to-face contact as soon as is possible, with a shift to DBT delivered via telehealth only being justified if it is the only way to get an evidence-based treatment like DBT to patients that need it. To build on the findings of this paper and our current knowledge about DBT delivered via telehealth, it would be valuable to track attendance and drop-out rates of clients enrolled in face-to-face and telehealth DBT programs over a longer timeframe with a larger sample.