National Study of Telepsychiatry Use in U.S. Emergency Departments

In recent decades, psychiatric services have become increasingly difficult to access (1) because of a shortage in mental health professionals (2) and a decrease in dedicated beds (3). With no other services to turn to, many experiencing mental health crises have no other option than to receive care in the emergency department (ED) (4). As hospital and community psychiatric services have been reduced, EDs have seen increasing numbers of psychiatric visits and increased ED boarding of psychiatric patients awaiting bed placement (4). Overall, about 13% of U.S. ED visits involved a diagnosis related to a mental health or substance abuse condition, or both, in 2007 (5). Between 2007 and 2011, ED visits due to mental health conditions increased by 15% (6). Frequent ED use for any reason has also been linked with poor mental health (7). Psychiatric patients often wait much longer in the ED than do nonpsychiatric patients, and during this time, patients receive limited care as they wait for placement (8, 9). EDs must balance the needs of all patients with limited resources, and an increase in visits can result in ED crowding, which can negatively affect the outcomes of all patients (1, 10).

A recent innovation to help mitigate the dearth of psychiatric services available in the emergency setting is the implementation of telepsychiatry in EDs (11–14). In this study, we define ED telepsychiatry as the receipt of telemedicine services, using various telecommunications methods, for the evaluation of ED patients who may require mental or behavioral health care. With telepsychiatry, providers can evaluate patients remotely, thus making psychiatric services more accessible, even to patients located in rural, underserved areas (15, 16). Despite the potential of telepsychiatry, there have not been any nationally representative studies regarding trends in telepsychiatry use and application in EDs. The objective of this study was to investigate the prevalence and application of telepsychiatry in general EDs in the United States.

Methods

Study Design

This study was based on two institutional-level surveys of U.S. EDs. The study was approved by the Human Research Committee of Partners HealthCare.

Survey and Administration

We first conducted a national survey of all U.S. EDs in 2017 to characterize emergency care in 2016. Using the National Emergency Department Inventory (NEDI)-USA database (17), we identified 5,375 EDs open in 2016. EDs were included in NEDI-USA if they were open 24 hours per day, 7 days per week, year-round (24/7/365) and were available for use by the general public; this includes freestanding EDs (i.e., EDs that are not physically or geographically connected to a hospital) (18). We mailed a one-page survey to all ED directors up to three times and then contacted staff from nonresponding EDs to complete the survey by telephone interview.

To learn more about the applications of telepsychiatry, we randomly sampled 15% (130) of the 885 EDs that reported receiving telepsychiatry in 2016 and contacted the EDs primarily by telephone for our 2017 ED Telepsychiatry Survey. Respondents from three of the 130 EDs answered via an online version of the second survey.

Measurements

The 2016 NEDI-USA survey asked ED directors to report on basic characteristics of their ED (e.g., annual visit volumes, presence of a pediatric emergency care coordinator [PECC]) and telemedicine use (e.g., providing versus receiving telemedicine in the ED and clinical applications for which they received telemedicine). A PECC is someone who manages pediatric care in the ED and who helps educate other ED staff on pediatric emergency care (19). Receipt of ED telepsychiatry was assessed with the question: “Does your ED receive telemedicine for patient evaluation?” (20, 21). Those who responded yes were then asked to report the clinical applications for which they used telemedicine. ED directors who reported use of telemedicine for psychiatry or wrote in that they used telemedicine for “behavioral health” or “mental health” were classified as receiving telepsychiatry. (The survey is available for viewing in an online supplement). ED type was classified as hospital based or freestanding (autonomous or satellite). EDs were categorized as satellite freestanding EDs (FSEDs) if they were connected to a flagship hospital, whereas autonomous FSEDs operated independently (18). We also collected additional ED data: rural location, defined as ED location outside of a core-based statistical area (CBSA) (22); Critical Access Hospital (CAH) designation (23); and location in a geographical health professional shortage area (HPSA) for mental health providers (24).

In the 2017 ED Telepsychiatry Survey, after confirming telepsychiatry receipt in 2017, we asked ED leadership about the overall availability of psychiatric services of any form to EDs (25) and ED psychiatric boarding times, using questions similar to those used in previous national surveys (26, 27). Emergency psychiatric services were categorized (25) as follows: general medical ED with mental health professionals who actually staff the ED; general medical ED with psychiatric consultation as needed; psychiatric emergency service (PES) or acute psychiatric service located within or adjacent to general medical ED; stand-alone PES not affiliated with an ED; consultation with an external service that is not on the hospital’s campus (a representative is sent in to the ED to complete the evaluation); and telemedicine services for the evaluation of patients requiring mental health care, which are the only psychiatric services available for the ED.

We also obtained information about telepsychiatry providers, hours of operation, frequency of use, clinical functions, and populations served (see online supplement). The 2017 ED Telepsychiatry Survey was piloted in a randomly selected sample of 20 EDs that reported receiving telepsychiatry. Then it was administered to a separate, random sample of 15% of the EDs (N=130) for this study. We chose to study 130 EDs (rather than all 885 eligible EDs from the national survey) because that sample size provided sufficient precision for the study objectives (see online supplement).

Data Analysis

Descriptive statistics are presented as proportions and medians with interquartile ranges. In unadjusted analyses, national receipt of telepsychiatry was examined by ED characteristic by using chi-square and Wilcoxon rank-sum tests as appropriate. An alpha level of p<0.05 was considered statistically significant (two tailed).

Multivariable logistic regression was performed to examine the independent associations between telepsychiatry receipt and multiple ED characteristics. Covariates considered relevant to telepsychiatry (annual total ED visits, hospital-based ED versus autonomous or satellite FSED, ED region, rural location [21, 27], CAH designation [15], and HPSA status [16]) were selected a priori for inclusion in the model, while other factors were evaluated for possible inclusion in the model if found to be suggestively associated with the outcome in unadjusted analyses (p<0.20; presence of a PECC) (15, 16, 21, 27). Results are reported as odds ratios (ORs) with 95% confidence intervals (95% CIs). All analyses were performed with Stata software, version 14.2.

Results

Telepsychiatry Receipt in U.S. EDs, 2016

The 2016 NEDI-USA Survey received responses from 4,507 (84%) of 5,375 EDs (see online supplement). Of the 4,410 responders completing the telepsychiatry question, 885 (20%) EDs indicated that they received telepsychiatry. In unadjusted analyses, EDs that received telepsychiatry, compared with those that did not, were more often hospital based and satellite FSEDs and less often autonomous FSEDs (Table 1). The proportion of rural EDs was also higher among EDs receiving telepsychiatry (26% receivers versus 21% nonreceivers), as was the proportion of EDs with a CAH designation (31% receivers versus 25% nonreceivers).

In a multivariable analysis (Table 2), the ED characteristics associated with a higher likelihood of telepsychiatry receipt were an annual total ED visit volume of ≥10,000, presence of a PECC, satellite FSED (relative to hospital based), location in the South (relative to the Northeast), rural location, and CAH designation. Characteristics associated with lower likelihood of EDs’ receipt of telepsychiatry were a larger annual total visit volume by children, autonomous FSED (relative to hospital based), and location in a HPSA for mental health providers.

ED Telepsychiatry Study

In our 2017 ED Telepsychiatry Survey, 105 (81%) of 130 EDs responded (see online supplement). More than 90% (95 of 105) confirmed the earlier report that their ED received telepsychiatry. Of the 10 that did not confirm receiving telepsychiatry services, one ED clarified that it provided telepsychiatry (for other EDs), and the other nine received some other form of telemedicine in 2017 but not specifically for patients who require mental health care.

ED Psychiatric Services

When asked about overall availability of psychiatric services in the ED, 59% (95% CI=49%–69%) of the 95 EDs that confirmed telepsychiatry receipt in 2017 reported that telepsychiatry was the only form of emergency psychiatric services available (Figure 1). Among the 95 EDs, 15% reported receiving emergency psychiatric services from an off-campus external service, which then sent a representative to the ED to complete evaluations. Additionally, 13% of EDs receiving telepsychiatry had no dedicated space devoted to psychiatric emergency patients but did have mental health professionals (e.g., licensed mental health counselors, licensed clinical social workers) who staffed the ED; among these EDs, none had any board-certified or board-eligible psychiatrists on duty in the ED.

ED Wait Time for Psychiatric Inpatient Beds

To gain more understanding of the context in which telepsychiatry is applied, we also investigated ED psychiatric wait times (see online supplement). When asked about the average time that elapsed between request for adult patient transfer and departure from ED to a psychiatric inpatient bed, 23 (24%) EDs reported it as 6 to 11.9 hours, and 45 (47%) EDs reported it as 12 or more hours. Of the 45 EDs that reported wait times of 12 hours or more for adult patient transfer, 22 (49%) reported telepsychiatry as their only form of psychiatric services. Most EDs (68%) reported that the maximum elapsed wait time for a psychiatric inpatient bed was more than 1 day in 2017, with those wait times ranging from just over 1 day to 30 days.

Similarly, for pediatric patients, 28 (29%) EDs reported the average time elapsed waiting for a psychiatric inpatient bed was 6 to 11.9 hours, and 38 (40%) EDs reported a wait of at least 12 hours. Of the 38 EDs that reported wait times of 12 hours or more for pediatric patient transfer, 19 (50%) reported telepsychiatry as their only form of psychiatric services. More than half of EDs reported that the maximum wait time for pediatric patients was 1 day or less. For EDs that reported maximum wait times of more than 1 day (42%), the maximum wait times ranged from just over 1 day to 90 days.

Telepsychiatry Operations and Applications

As shown in Table 3, telepsychiatry providers tended to be other hospitals in an ED’s own hospital system (40%) or external, private organizations or companies (39%). Most EDs had access to telepsychiatry services 24/7/365. Additionally, most (62%) EDs reported receiving telepsychiatry services at least six times per week, with 25% of EDs receiving telepsychiatry services at least once per day (Table 3). Telepsychiatry often assisted with the diagnosis and treatment of psychiatric conditions. However, we found that telepsychiatry most frequently assisted in admission or discharge decisions and with transfer coordination (Table 3). Telepsychiatry services were used for staff education by only 11% of EDs.

Nearly all EDs received telepsychiatry for adults, and many received telepsychiatry for all age demographics, including children and adolescents, geriatric patients, and nongeriatric adults (Table 3). Two (2%) EDs reported that they received telepsychiatry only for children and adolescents.

Discussion

Our national investigation into the understudied topic of ED telepsychiatry yielded new information regarding the prevalence and clinical applications of telepsychiatry in emergency settings. Through our national survey, 2016 NEDI-USA, we found that 20% of U.S. EDs received some form of telepsychiatry, and these respondents were more likely to be EDs with higher annual visit volumes, rural location, and a CAH designation (Table 2). Furthermore, in the 2017 ED Telepsychiatry Survey, we found that 59% of 95 EDs receive telepsychiatry as their only form of emergency psychiatric services. Telepsychiatry was most frequently used to assist EDs in admission or discharge decisions (80%) and transfer coordination (76%).

Nationally representative research on characteristics of ED telepsychiatry is very limited. The California Health Care Foundation examined seven ED telepsychiatry service programs as case studies (28). Four of the programs provided services through a hospital system or from large academic hospitals to surrounding sites (both affiliated and unaffiliated). Two of the programs were provided through private, for-profit companies, and one program was provided through the state department of mental health. These findings parallel the range of providers reported in this study. Additionally, two programs noted increased psychiatric knowledge among ED physicians after implementation of telepsychiatry and thus reduced requests for consultation later on. This parallels the 11% of EDs in this study that reported using ED telepsychiatry for staff education.

Although research on the prevalence and characteristics of ED telepsychiatry is limited, there have been investigations into the effectiveness of telepsychiatry with favorable ED outcomes. Such research is vital in understanding how ED telepsychiatry may or may not work as an alternative to traditional psychiatric emergency services, considering that telepsychiatry was the only option for 59% of EDs. Two studies examining the usefulness of ED telepsychiatry in pediatric populations found that satisfaction among patients, parents, and providers as well as clinical outcomes after telepsychiatry assessment were similar to those for patients who received face-to-face psychiatric care (29, 30). Pediatric telepsychiatry also reduced costs and ED lengths of stay (30). Similar outcomes were seen in a statewide ED telepsychiatry program in which telepsychiatry patients were more likely to receive follow up, were less likely to be admitted to an inpatient setting at their ED visit, had a shorter ED length of stay, and had lower overall 30-day inpatient costs (31).

Another study that examined the agreement between telepsychiatry assessment and face-to-face assessment of 73 ED psychiatric patients found there were no significant differences between evaluation methods (32). Since more than half of the EDs surveyed in this study use telepsychiatry for diagnosis of psychiatric conditions, consistency in telepsychiatry evaluations is essential.

As the United States continues to face a deficit of available psychiatric services, EDs will likely continue to see high volumes of psychiatric patients in need of such services (1, 2). General EDs in the United States are innovating to address this challenge. This is particularly true for EDs with PECCs and EDs in rural or CAH settings, as the NEDI-USA survey found that sites receiving telepsychiatry were more likely to have these characteristics. We hypothesize that, since a PECC’s role includes facilitating improvements in pediatric emergency care, this quality improvement initiative may encourage the installation of telemedicine services (19)—or vice versa. The results showing that EDs receiving telepsychiatry were more likely to be rural are consistent with regional data showing that rural New England EDs were more likely to receive telemedicine services (in any form) (27). Previous research has also shown a rapid growth in telepsychiatry use among rural Medicare beneficiaries (33). For rural EDs that have not adopted telepsychiatry but lack the availability of psychiatric consultations, many providers are comfortable with the concept of telepsychiatry (34). ED telepsychiatry is frequently implemented as a solution to expand psychiatric services in the midst of major workforce constraints, especially in rural and CAH settings. As telepsychiatry receipt becomes more common, nationally representative research is integral to the future investigation of patient outcomes, quality of care, and overall cost-effectiveness.

EDs receiving telepsychiatry were less likely to be autonomous FSEDs or located in an HPSA for mental health providers. Although satellite FSEDs may receive telepsychiatry services through parent hospitals, many autonomous FSEDs are run by for-profit entities, and the services they receive may be heavily dictated by the company (35). It is unclear why EDs receiving telepsychiatry were less likely to be in an HPSA for mental health providers, but we hypothesize that it is indicative of the overall mental health professional shortage. For these EDs, telepsychiatry services through nearby hospitals may be nonexistent. Additionally, the receipt of telepsychiatry from out-of-state providers may be hindered by state policies that strictly regulate telemedicine services.

As shown in Table 3, a small proportion (16%) of EDs did not receive telepsychiatry 24/7/365. Of these EDs, six reported telepsychiatry as their only form of psychiatric emergency services. More research needs to be conducted to better understand the reasons for these lapses in service and how staffing, reimbursement, and other factors play a role.

In many EDs, there is limited delivery of mental health care for psychiatric patients during the boarding process (1, 5). As patients wait for placement, the delay in receiving care is extended. The use of telepsychiatry to assist with inpatient placement could alleviate ED crowding in several different ways, from helping ED physicians understand the difference between emergent and nonemergent cases to securing inpatient psychiatric beds more effectively (28–32).

In addition to ED patient flow, 56% of EDs reported using telepsychiatry for diagnosis of psychiatric conditions, and 45% used it for treatment of psychiatric conditions, which indicates that, for many, telepsychiatry may allow for earlier delivery of care. This is important for uninsured psychiatric patients who have limited options for inpatient care and, thus, higher boarding times (36).

These uses are somewhat different than that of pediatric telemedicine receivers (20), where 75% used such services for treatment of conditions. Although most literature on ED telepsychiatry focuses on the delivery of psychiatric care via telecommunications, our study suggests that ED psychiatric treatment may be less of a priority, as telepsychiatry is more often used for informed triage of patients through assistance with admission and discharge decisions and transfer coordination. Future inquiries may investigate barriers to treatment via telepsychiatry and ways to combat such barriers.

This study had several potential limitations. For logistical reasons, the NEDI-USA survey asked about telepsychiatry receipt in 2016, whereas the follow-up survey investigated 2017 telepsychiatry receipt. Even so, this enabled us to further confirm ED telepsychiatry receipt for both 2016 and 2017. Although the response rates were more than 80% for both surveys, the smaller sample size for the second survey limited our ability to perform more extensive analyses. Because of the cross-sectional nature of this study, causal inferences were not possible. Furthermore, all data were self-reported, which may have introduced information bias. To minimize this possibility, we targeted the surveys to ED directors. We estimate that over 80% of respondents to the 2016 NEDI-USA survey were ED directors or in ED leadership (i.e., individuals who would definitely know about the status of telepsychiatry in their ED). For the 2017 ED Telepsychiatry Survey, all respondents were ED directors or in ED leadership. Additionally, the NEDI-USA database includes all 24/7/365 EDs available to the general public. As stand-alone PES do not meet these criteria, such entities were not included. We believe that this was appropriate for this study, in which the objective was to understand the prevalence and application of telepsychiatry in general EDs in the United States.

Conclusions

We found that 20% of U.S. EDs receive telepsychiatry services. For most of these EDs (59%), telepsychiatry is the only means by which a patient in crisis can receive psychiatric services in the ED. Telepsychiatry is used quite frequently among receiving EDs, and it commonly assists with admission and discharge decisions and with coordinating transfer and placement, which may affect ED boarding and crowding. Even so, about half of EDs that receive telepsychiatry use it to begin delivering some form of psychiatric care to patients.

Currently, telepsychiatry may be an effective alternative for creating accessible services and streamlining the ED process, with favorable effects on ED boarding and crowding and better utilization of limited resources (28–32). Considering these outcomes, we believe that more EDs should consider the benefits of implementing such services. Additionally, we encourage more research to better understand the barriers (and solutions) to receiving ED telepsychiatry. Future studies might also explore upstream factors, such as lack of psychiatric services for preventive and chronic care and dedicated psychiatric spaces and how such factors may affect the usefulness of ED telepsychiatry.