LEARNING OBJECTIVES

After reading this article you will be able to:

• • summarise the history of telepsychiatry

• • summarise research evidence regarding the efficacy of telepsychiatry

• • describe the limitations of telepsychiatry.

The theoretical foundation of a psychiatric interview is the mental state examination. Incorporating elements of descriptive phenomenology and elements of empathic observation, the clinician seeks to describe in a semi-structured way processes and behaviours that the patient themselves may or may not be aware of (Taylor 1967; Bell Reference Bell and Hall1977). Since the mid-1990s the generally accepted format of this examination has divided an individual's mental state into six categories – appearance and behaviour, mood and affect, speech and language, thought process and content, cognition, and insight (Trzepacz 1993) – of which three rely on prolonged physical observation.

Conducting a thorough psychiatric examination remotely therefore appears challenging. Despite this, remote psychiatry clinics have existed for at least 50 years (Urness 2004; Shore 2015). More recently the COVID-19 pandemic has made adoption of telepsychiatry near universal – a survey by the American Psychiatric Association found that 98% of psychiatrists were seeing patients remotely in 2021 (American Psychiatric Association 2021). Against this background an understanding of the research evidence comparing the efficacy of telepsychiatry with in-person care, and the limitations of that evidence, is vital for practising clinicians.

Defining telepsychiatry

Telepsychiatry is usually defined as the use of electronic communication and information technologies to provide or support clinical psychiatric care at a distance. It is occasionally subdivided into ‘synchronous’ and ‘asynchronous’ types, the former referring to systems in which the patient and clinician communicate directly and simultaneously (such as a video call) and the latter referring to systems in which they may communicate indirectly or at different times (Drago Reference Drago, Winding and Antypa2016).

A brief history of telepsychiatry, 1970–2019

Initial attempts at establishing remotely delivered psychiatric care universally took place in the USA – a country with a large population living in rural areas. The first system for clinical use was established in the early 1970s when a two-way video system was installed between a teaching hospital and a smaller rural clinic in Nebraska, USA. Patients still went to the clinic, sat in a waiting-room and were shown into their consultation by a member of staff, who remained present during the session (Wittson 1972). This programme was a success and appreciated by the local population, and a number of similar programmes were subsequently developed, with staff usually present in a remote clinic to help the patient use the video equipment (Dwyer Reference Dwyer1973; Murphy 1974; Dongier Reference Dongier, Tempier and Lalinec-Michaud1986).

Although psychiatry seemed a specialty naturally suited to telehealth (owing to the low requirements for physical examination of the patient) technological problems with early systems, including extremely slow data transfer times, made natural conversation impossible. By the mid-1980s the only telepsychiatry programmes active were funded by research grants and were not delivering routine clinical care (Preston 1992). There remained low interest in telepsychiatry for the next two decades – a 1997 review article found 18 studies on telepsychiatry. Of these, five were economic or feasibility evaluations and eight were cohort studies with no control group. The five studies available at that time comparing in-person psychiatry with telepsychiatry all had small samples, and three had no group randomisation. The authors concluded that ‘evidence currently available is insufficient to suggest its widespread implementation’ (Baer Reference Baer, Elford and Cukor1997).

The next decade saw a rapid rise in interest in telepsychiatry, driven in part by the commercial availability of videoconferencing systems (Monnier 2003). The feasibility of these systems for multiple aspects of mental healthcare was demonstrated, including child and adolescent services (Alessi Reference Alessi2002), geriatric services (Johnston 2001), cognitive–behavioural therapy (CBT) (Bouchard Reference Bouchard, Payeur and Rivard2000; Cowain Reference Cowain2001) and neuropsychology (Schopp 2000). Although interest was growing, the efficacy of these methods compared with typical care was less clear: a 2005 meta-analysis found 14 studies which used standardised tools to directly compare diagnostic accuracy of in-person and remote psychiatry, with insufficient evidence to draw any firm conclusions (Hyler 2005). The chief limitation identified was a lack of homogeneity in patient groups studied – among the 14 studies are a cohort of adults with obsessive–compulsive disorder, two cohorts of geriatric nursing home residents, an adolescent cohort and a cohort of forensic in-patients. The authors did not, however, find any evidence that telepsychiatry was inferior to in-person care, a finding that was replicated in another review in 2010, which included 10 randomised controlled trials (RCTs) comparing outcomes following in-person and remote psychiatric interventions (García-Lizana Reference Guinart, Marcy and Hauser2010), albeit with the same limitation of very differing patient groups and interventions.

From 2010 onwards the rationale for telepsychiatry use began to evolve from the need to provide care to geographically remote areas to the need to provide care that patients find convenient (Pakyurek 2010). Several studies published between 2005 and 2013 demonstrated subjective patient preference for video interviews over face to face, especially among adolescents (Hilty Reference Hubley2013). Proliferation of services was widespread, but evidence of their efficacy compared with in-person clinics remained limited. A review in 2016 found eight studies directly comparing the efficacy of telepsychiatry with in-person care – seven found no difference in outcome and one (focusing on CBT in bulimia nervosa) found in-person care to be superior (Hubley Reference Hubley2016).

Models of telepsychiatry care evolved rapidly from 2010. Prior to this time telepsychiatry typically referred to either the use of a fixed videoconferencing system connecting a satellite clinic to a larger site, or the use of telephone calls to engage a patient in their own home. With the rise of internet and mobile phone availability, psychiatric care could now be delivered via text message, online chat rooms and video calls (Mermelstein 2017).

Until the end of 2019 telepsychiatry services were still, however, largely delivered as short-term projects funded by research grants rather than substantial clinical services (Fig. 1).

FIG 1 Number of published manuscripts containing the word ‘telepsychiatry’, 1960 to the present. Data retrieved from Google Scholar, February 2022.

Telepsychiatry today

COVID-19 has transformed the practice of out-patient psychiatry in the UK and worldwide (Unützer 2020; Öngür 2020; Shore 2020). Substantial periods of social contact minimisation in many countries necessitated the rapid development of remotely delivered community care, and even as pandemic restrictions have eased, in many countries telepsychiatry remains the default mode of out-patient care (Mehrotra 2021; American Psychiatric Association 2021). Some professional bodies have released guidelines concerning the optimal use of telepsychiatry, but these are currently brief outline documents (the American Psychiatric Association's recommendations are just over two A4 sides in length for example, and the Royal College of Psychiatrists’ advice is a similar length) (RCPsych 2020; von Hafften 2021). Most are generally in agreement that telepsychiatry is viewed as safe and effective, and that clinicians are encouraged to offer appointments remotely if possible. In the UK the view that remote assessments are as robust as in-person ones was echoed by NHS England and the Department of Health and Social Care, who in 2020 endorsed remotely delivered Mental Health Act assessments. During the pandemic it was stated that ‘developments in digital technology are now such that staff may be satisfied, on the basis of video assessments, that they have personally seen or examined a person in a “suitable manner”’ (NHS England 2020), although notably this decision has since been overturned (Devon Partnership NHS Trust v Secretary of State for Health and Social Care [2021] EWHC 101). These changes were not unique to the UK – several other countries adjusted their legal frameworks to accommodate a rapid adoption of remote delivery of psychiatric care (Kinoshita 2020).

Acceptability to patients

There is a large body of evidence studying patients’ perceptions, which in general demonstrates that patients are happy with remotely delivered care (Rohland 2000; Polinski 2016; Cowan Reference Cowan, Johnson and Close2020; Guinart Reference Hassan and Sharif2020; Lal 2020). There is some emerging evidence to suggest that certain patient groups, such as female older adults, may favour in-person care (Christensen Reference Christensen, Moller and Hansen2020) but more research is needed to draw firm conclusions in this regard. For a summary of current evidence of intergroup patterns of acceptability to patients see Barnett et al (Reference Barnett, Goulding and Casetta2021).

Although studies agree that patients find telepsychiatry acceptable there is a key limitation to be aware of. The vast majority of patient opinion studies in telepsychiatry suffer extremely poor response rates: 22% in Guinart et al (Reference Hassan and Sharif2020), 33% in Rohland et al (2000) and 54% in Polinski et al (2016). There are clearly cohorts of individuals for whom telepsychiatry is not practical – people who do not own the necessary technology or are unfamiliar with it, people who struggle to adjust routines rapidly and people who lack a private space in which to talk freely, for example. Current qualitative methods that rely on voluntary participation are possibly failing to capture their perspectives.

Acceptability to clinicians

There is limited evidence on the acceptability of telepsychiatry to clinicians, with most studies having small samples and the notable methodological flaw of only including clinicians who have already adopted telepsychiatry into their practice. Those that do exist largely demonstrate that clinicians find telepsychiatry systems easy to use and feel that their clinical judgement is not affected by seeing patients via a video screen (Elford Reference Elford, White and St John2001; Hensel Reference Holmqvist2020), although a minority report a clinician preference for in-person care (Hassan Reference Hilty, Ferrer and Parish2019). As ever in telepsychiatry, comparison between studies is limited by different patient groups, telepsychiatry systems and analysis methods. Elford et al studied a cohort of adolescent patients using a dedicated out-patient clinic video system, Hensel et al studied a system for remote emergency assessments in a general adult population and Hassan & Sharif focused on assessments of refugees. There is currently insufficient evidence to determine whether acceptability to clinicians is higher in certain circumstances (seeing familiar patients, for example) – for a review article discussing these questions in more detail see Hubley et al (Reference Hubley2016).

Diagnostic reliability

A small number of randomised trials have assessed the diagnostic reliability of a telepsychiatry assessment compared with an in-person assessment, all of which view an in-person diagnosis as their gold standard. Results of percentage agreement are reassuring: 69–82% in a child and adolescent sample (Brøndbo Reference Brøndbo, Mathiassen and Martinussen2012), 75% in a random in-patient sample (Mazhari 2019), 76% in a general out-patient sample (Shore 2007), 92% in autism (Schutte 2015) and 96% in a child psychiatry sample (Elford Reference Elford, White and Bowering2000). There is a larger body of evidence of the reliability of structured tools being delivered remotely which shows that most commonly used tools are as effective via video call as they are in person – for a review see Drago et al (Reference Drago, Winding and Antypa2016). A review focusing specifically on telephone assessments versus in-person assessments found an insufficient quality of literature to draw any conclusions (Muskens 2014).

Therapeutic alliance

Therapeutic alliance, usually defined as the collaboration between patient and therapist to achieve a shared goal (typically recovery), is an important aspect of any psychiatric consultation. Body posture, eye contact and non-verbal communication (factors in building a positive alliance) are naturally affected by a remote care delivery method (Wootton 2003). A 2018 meta-analysis explicitly comparing measures of alliance in remote and in-person care (which focused exclusively on the delivery of formal therapy rather than general psychiatric assessments) included five studies and found that therapeutic alliance was marginally less strong in remote care, but that this did not translate to worse outcomes (Norwood 2018).

Patient outcomes

A number of RCTs with samples over 100 have compared patient outcomes after telepsychiatry and in-person psychiatry (Table 1). Non-inferiority of telepsychiatry has been demonstrated in treating depression using medication, psychotherapy and psychoeducation (Ruskin 2004; Chong Reference Chong and Moreno2012; Moreno 2012; Egede Reference Egede, Acierno and Knapp2015), although in two of these studies (which use the same patient population) individuals using telemedicine were prescribed antidepressants at a significantly higher rate (Chong Reference Chong and Moreno2012; Moreno 2012). One study demonstrated that structured therapy for post-traumatic stress disorder (PTSD) was more effective if delivered via video call than in person, although the telepsychiatry group received more overall care (Fortney Reference Fortney, Pyne and Kimbrell2015), and a second RCT of telepsychiatry versus in-person care in PTSD demonstrated no significant difference in outcome between the groups (Morland 2010). Two RCTs did not stratify by diagnosis but demonstrated non-inferiority of telepsychiatry (De Las Cuevas Reference De Las Cuevas, Arredondo and Cabrera2006; O'Reilly 2007). In one of these studies neither group of patients improved during the follow-up period, making the effectiveness of either intervention hard to judge (O'Reilly 2007). To our knowledge only one RCT (focusing on therapy for bulimia nervosa) to date has shown superior outcomes for in-person care compared with remote care, although differences were small and of dubious clinical significance (Mitchell 2008).

TABLE 1 Summary of randomised controlled trials (n > 100) directly comparing in-clinic or at-home telepsychiatry and in-person care

PTSD, post-traumatic stress disorder; CBT, cognitive–behavioural therapy.

Three further RCTs use adjunctive care designs, comparing a group receiving treatment as usual with a group additionally receiving a telepsychiatry intervention (Fortney Reference Freeman, Boggs and Zachrison2013; Myers 2015; Hulsbosch 2016). All show that the group receiving adjunctive care had better outcomes. One used a complex design in which one group's direct care providers received expert advice remotely (Fortney Reference Freeman, Boggs and Zachrison2013). Previous reviewers have expressed concern as to the methodological quality of adjunctive telepsychiatry research. A 2018 review included eight studies, of which five were judged to be of low quality and three of adequate quality. All five low-quality studies found that the technological adjunct reduced remission rates, whereas none of the adequate-quality studies did (Koblauch 2018).

An extremely important limitation of the RCTs above is that they all use a model of care in which the ‘telepsychiatry’ intervention was delivered in a clinic – the patient experience is identical to a typical out-patient appointment other than the psychiatrist is replaced by a video screen. This is obviously quite different from the manner in which telepsychiatry has been delivered during the COVID-19 pandemic. There are a small number of studies comparing in-person care with care delivered via video with the patient at home. Two studies focusing on PTSD found generally similar results between groups, with slightly improved 3-month follow-up scores in the in-person group (Acierno Reference Acierno, Gros and Ruggiero2016, Reference Acierno, Knapp and Tuerk2017). This slight preference for in-person therapy is also reflected in one study comparing at-home versus in-person treatment for depression which showed better outcomes in the in-person care group (Luxton 2016) and one showing better outcomes for in-person therapy for PTSD (Glassman Reference Harerimana, Forchuk and O'regan2019). Conversely, however, one RCT comparing at-home telepsychiatry with in-person clinic appointments found no difference between groups (Choi Reference Choi, Marti and Bruce2014). Finally, one study compared depression care via telephone with in-person clinic care, finding in-person care slightly superior (Mohr 2012). One limitation in comparing in-clinic and at-home studies is that the available at-home RCTs exclusively examine formal therapies, whereas in-clinic studies examine a broader range of care modalities, including psychiatric assessments of new patients, medication management and psychoeducation sessions. RCTs of in-clinic therapies universally show no outcome difference, so it is reasonable to hypothesise that setting, not just the type of care delivered, is important in outcome prediction.

Research discussed so far concerns live appointments delivered remotely. There is also a significant body of literature examining ‘asynchronous telepsychiatry’ – i.e. care delivered by messaging systems, chat rooms or other modalities in which the clinician and patient are not present at the same time. Discussing this in detail is outside the scope of this article – for a recent review see O'Keefe et al (2019).

Limitations of telepsychiatry

Three chief barriers to adoption of telepsychiatry are frequently discussed in the literature.