Novel clinical demonstrations
Many published reports on telepsychiatry are anecdotal accounts or case reports of novel applications. This kind of report is appropriate given how recently the technologies involved have been developed.
Case reports describe use of telepsychiatry for conducting interviews in community mental health clinics with patients who have schizophrenia (
8,
9,
10,
11,
12); multiple-session evaluations for a range of disorders (
13); ten sessions of psychological treatment of a female-male transsexual (
14); 16 sessions of cognitive-behavioral treatment of a child with a disruptive behavior problem (
15); a session of family therapy (
16); treatment of a depressed Hispanic patient (
17); clinical supervision and trauma debriefing of a remote mental health treatment team (
18); psychiatric sessions using low-cost videophones (
19); and a group treatment session for veterans with posttraumatic stress disorder (
20). These reports provide good early evidence of the wide range of potential applications for telepsychiatry.
Program descriptions
Several larger-scale telepsychiatry programs have also been described in the literature. The largest of these programs, and the most thoroughly described, is an integral part of South Australia's Rural and Remote Mental Health Service (
21,
22,
23,
24,
25,
26). Since 1994 the program has used telecommunications for delivery of a wide range of clinical psychiatric services to community patients dispersed over nearly one million square kilometers. More than 2,000 clinical consultations—emergency services, inpatient liaison, postdischarge follow-up, and the like—were conducted in the first four years of the program's operation. More than 80 percent of the system's usage is for direct clinical purposes; clinical supervision is also provided.
Other general telepsychiatry programs have been described, including Oregon's RodeoNet and a program at the University of Kansas Medical Center (
27,
28,
29); a program at the University College Hospital in Galway, Ireland, as a link to the island of Inishmore (
13); the South Carolina Department of Mental Health's Deaf Services Program (
30); programs in rural Appalachia (
31), the Highlands of Scotland (
32), and the U.S. Federal Bureau of Prisons (
33); and programs at the University of Nebraska Medical Center (
34), the University of Kentucky (
35), Texas Tech University (
7), the University of Oulu, Finland (
36), and Cairns Base Hospital in Queensland, Australia (
37). These programs have provided a wide range of services, including general adult and child psychiatric consultations and treatment of incarcerated inmates. Finally, brief surveys of active telepsychiatry programs have been published (
4,
38).
Prospective empirical evaluations
Reliability of clinical assessments. Several studies have furnished evidence of the reliability of psychiatric evaluations conducted by telepsychiatry. Baigent and colleagues (
39) examined the interrater reliability of two psychiatrists who conducted semistructured interviews, including the Brief Psychiatric Rating Scale (BPRS), with 63 randomly assigned subjects in an observer-interviewer split configuration in remote and same-room settings. Although several differences between the telepsychiatry and same-room settings emerged— "degree of concern" for the patient and the frequency of some mental status findings were lower in the telepsychiatry settings—generally the diagnoses made were equally reliable in both settings.
Baer and associates (
40) examined rating scales for obsessive-compulsive, depressive, and anxiety symptoms administered to 16 patients in a same-room setting and ten remote patients and found that reliability was excellent in both settings. Elford and coworkers (
41) found an interrater agreement of 96 percent among same-room and remote diagnostic interviews conducted with 23 children ranging in age from four to 16 years.
Ruskin and associates (
42) examined the interrater reliability of psychiatric diagnoses made by telepsychiatry and in same-room settings of 30 psychiatric inpatients using the Structured Clinical Interview for DSM-III-R. Fifteen patients had two same-room interviews, and 15 patients had one same-room and one remote interview by telecommunication. Interrater agreement was calculated for diagnoses of major depression, bipolar disorder, panic disorder, and alcohol dependence. The resulting reliability coefficients were nearly identical for the two groups.
Zarate and associates (
43) used the BPRS, the Scale for the Assessment of Positive Symptoms (SAPS), and the Scale for the Assessment of Negative Symptoms (SANS) with 45 patients with schizophrenia assigned to one of three interview conditions: same-room setting, telepsychiatry at high bandwidth (384 kilobits per second [kbps]), and telepsychiatry at low bandwidth (128 kbps). Results showed that the global severity of schizophrenia and the overall severity of positive symptoms were reliably evaluated by both telepsychiatry conditions, although negative symptoms were less reliably assessed in the low-bandwidth condition. In addition, telepsychiatry was well accepted by patients, although they generally preferred the high-bandwidth transmission. Similar findings were obtained in another study with a small sample (N=7) (
44).
Finally, remote cognitive examinations have been conducted with teleconferencing technology (
45). Montani and associates (
46,
47) evaluated samples of ten and 15 elderly patients for cognitive deficits using the Mini Mental Status Examination (MMSE) and the clock face test. Each patient participated in two evaluations, conducted a week apart, one by telepsychiatry and the other in a same-room setting. Results showed that scores on both tests were lower in the telepsychiatry condition; the difference was small but statistically significant. The authors attributed the lower scores to patients' difficulties hearing and maintaining attention in telepsychiatry evaluations and concluded that telepsychiatry can be used reliably in the examination of elderly patients (
46).
Ball and colleagues (
48) demonstrated with a sample of 12 patients that the MMSE could be reliably conducted via remote technology. A study by Craig and associates (
49) conducted with 23 patients produced evidence that even neurological examinations can be conducted reliably via remote technology.
Satisfaction and acceptance. Studies have demonstrated generally high rates of patients' and clinicians' satisfaction with telepsychiatry. Satisfaction has been high among incarcerated patients (
28,
32), hospitalized elderly patients (
47), rural outpatients (
21,
36,
50), clinician referrers (
21), caregiver support groups (
51), psychiatric inpatients (
5,
52), patients with schizophrenia (
43), and rural children and their parents (
30,
31,
41). Telepsychiatry interviewing has been well accepted even among patients with paranoid schizophrenia who had ideas of reference, with no exacerbation of delusional symptoms (
52). In Australia, 90 percent of persons referred for mental health services reported that they would use telepsychiatry if it were offered (
53), and two-thirds of rural respondents in the United States were willing to use it (
54).
Clinical outcome. Only two clinical outcome studies have been completed to date. Brown and associates (
51) examined the use of telepsychiatry compared with a traditional caregivers support group. In a quasi-experimental design comparing the two groups, similar outcome results were found on the Profile of Mood States and other relevant measures.
Zaylor (
55) retrospectively compared Global Assessment of Functioning (GAF) scores for 49 patients with depression or schizoaffective disorder who were treated by telepsychiatry or same-room sessions for at least six months. No differences were found in the percentage change in GAF scores from initial visit to six-month visit between the two groups, suggesting the clinical efficacy of telepsychiatry. Furthermore, the patients treated by telepsychiatry had greater attendance rates and required dramatically shorter sessions compared with those in the same-room group.
Cost-effectiveness. Only one cost study of a telepsychiatry application has been reported (
37). In a comparison of telepsychiatry and conventional mental health services in rural Queensland, Australia, the cost savings produced by telepsychiatry were estimated to be $181,716 in the first year, primarily from reduced travel by patients and clinicians; annual savings in subsequent years rose to $209,126. However, cost estimates failed to incorporate costs for maintenance and upgrading of equipment.
Limitations of the literature
Key limitations of the current literature are the paucity of reports of any kind on the use of telepsychiatry and the lack of rigorous empirical study of telepsychiatry applications. Most reports published to date are case reports or program descriptions. In empirical studies the focus has been primarily on the reliability of telepsychiatry assessments and general user satisfaction rather than on broader biopsychosocial research. Critically lacking are reliable baseline data before implementation of telepsychiatry programs, evaluation of clinical process and outcome, randomized assignment and control groups, and efforts to determine the efficacy of telepsychiatry for specific patient populations.
Furthermore, most reports of telepsychiatry programs indicate that little effort has been made to monitor program changes and adaptations as newer technology has been introduced, making systematic comparisons within and across programs problematic. Perhaps most significant is the lack of cost analyses. Given that the primary reason for implementing telepsychiatry applications is to save money in the course of expanding access to care, the degree to which telepsychiatry fulfills its promise of providing an affordable means of solving workforce problems remains largely unknown.
These limitations may be ascribed to the novelty of telepsychiatry applications; thus far the process of implementation has taken precedence over evaluation of outcomes. Below we offer recommendations about the procedural and methodological issues that should shape future empirical research.