Supporting Veterans, Caregivers, and Providers in Rural Regions with Tele-Geriatric Psychiatry Consultation: A Mixed Methods Pilot Study

Abstract

Objectives:

Cross-facility tele-geriatric psychiatry consultation is a promising model for providing specialty services to regions lacking sufficient geriatric psychiatry expertise. This evaluation focused on assessing the feasibility and acceptability of a consultation program developed by a geriatric psychiatrist in a Veterans Health Administration regional telehealth hub.

Design:

Concurrent, mixed methods program evaluation

Setting:

A region served by a VA health care system telehealth hub

Participants:

Patients with at least 1 geriatric mental health encounter with a geriatric psychiatrist consultant during a 1 year-period; referring providers.

Intervention:

Virtual psychiatric evaluation of Veterans with time-limited follow-up and e-consultation with providers.

Measurements:

Interviews with consultant, medical record data, and referring provider surveys

Results:

315 Veterans (M = 76.0 ± 9.64 years; 40% rural-dwelling) had 666 encounters (M = 2.11 ± 1.78) with most occurring via clinical video telehealth (n = 443; 67.6%), e-consultation (n = 99; 15.1%), or video to home (n = 95; 14.5%). Most encounters were related to neurocognitive disorders, depressive disorders, trauma-related disorders, or serious mental illness. Referring providers (N = 58) highly recommended the program, reported high satisfaction, followed through with recommendations, and believed that this program increased access to geriatric psychiatry.

Conclusions:

This single program was shown to be feasible, acceptable, and valued by the referring providers. The findings highlighted the complex presentations of Veterans referred, and the current unmet need of providers of such Veterans, providing impetus for wider implementation.

Introduction

More than 9 million United States (U.S.) military Veterans are aged 65 years or older.¹ Aging Veterans present with medical and behavioral comorbidities that include depression, dementia with behavioral disturbances, and Posttraumatic Stress Disorder (PTSD).² The Veterans Health Administration (VHA) and our nation face significant workforce shortages of geriatric mental health (MH) professionals needed to support the complex care needs of aging Veterans.^3-5 MH professionals have increasingly been integrated in care settings where older Veterans are served, such as primary care,⁶ home-based care,⁷ and nursing home care.⁷ Despite this progress, many providers do not have the specialized training necessary to comprehensively address the co-occurring psychiatric, medical, and neurocognitive disorders that older patients may experience.

Geriatric psychiatrists specialize in assessing and treating patients with complex comorbidities, which include new or longstanding psychiatric disorders, acute and chronic medical conditions, and neurocognitive disorders. Geriatric psychiatrists’ expertise is essential when patients present with either a rapid or gradual increase in psychiatric symptoms or distress behaviors that could lead to placement in a more restrictive care setting (e.g., moving to long-term care or a locked unit). Nationwide, there are 64 geriatric psychiatry fellowship programs with 45 fellows trained during academic year 2021-2022.⁸ This number is insufficient to address the needs of aging Americans as the national ratio of geriatric psychiatrists to adults aged 65 years or older is projected to be 1 geriatric psychiatrist for every 27,000 older adults in 2030.⁹ Further, the availability of board-certified geriatric psychiatrists does not mirror where older adults reside; these specialists tend to be concentrated in larger, urban, academically-affiliated areas.⁵ Similarly, within the VHA, these specialists are concentrated in larger facilities primarily in non-rural areas.¹⁰ Thus, rural older Veterans, their caregivers, and health care teams have limited access to these specialists. Providers caring for patients with complex needs living in rural areas must do so without specialty training or support.^11,12

One potential solution, telehealth, increases access to MH care and has been successfully implemented in caring for older patients.^13-15 Within VHA, a hub-to-spoke model that focuses on increasing rural patients’ access to geriatric medicine and dementia care has been developed and implemented across 12 sites.¹⁶ Another study demonstrated high patient satisfaction with geriatric psychiatry continuity of care services delivered via telehealth.¹⁴ Other models trialed include telehealth services delivered by geriatric psychiatrists to Program of All-Inclusive Care for the Elderly (PACE) patients¹⁷ and provider-to-provider consultation via the Extension for Community Healthcare Outcomes (ECHO) medical education model.^18,19 A literature review¹³ found evidence supporting the use of telehealth for geriatric psychiatry services, specifically, for assessment, diagnosis, and ongoing treatment. Thus, tele-geriatric psychiatry telehealth consultation is feasible and acceptable, but the existing models may not allow for the timely consultative care needed in settings such as residential and/or nursing home care. Specialties such as consultation-liaison psychiatry provide this prompt care in acute medical settings.

Cross-facility telehealth consultation using a blended model of geriatric MH expertise and a rapid consultation-liaison approach, is a promising approach that increases access in regions lacking geriatric MH expertise. This model was developed and led by a board-certified geriatric psychiatrist (J.F.) in 2015; and in 2018, it was incorporated into a VHA regional telehealth hub. Consultation occurs within a VHA region spanning multiple states. Tele-consultation on dementia-related behavioral issues, depression, anxiety, PTSD, and other geriatric MH conditions supports rural providers, rural older Veterans, and their families. The geriatric psychiatrist delivers consultation via video telehealth, telephone, and e-consults across the VHA continuum of care spanning outpatient settings, nursing homes, and home care.

An evaluation team conducted a mixed methods evaluation of this geriatric psychiatry consultation service with two objectives. The first objective was to characterize the Veterans served, consultative and follow-up services delivered, and encounter diagnoses during a 1-year-period. Variations in services delivered were examined based on rurality. The second objective was to examine provider perceptions of the program through surveys. Survey responses among providers with prescription privileges and those without (i.e., prescribers and non-prescribers) also were examined.

Methods

This program evaluation was reviewed by the Stanford University Institutional Review Board and determined to be program evaluation and not human subjects research.

Design

We employed a concurrent mixed-methods design.²⁰

Data Sources and Study Sample

Our cohort consisted of patients with at least 1 geriatric MH encounter (consult, follow-up or e-consult) with the geriatric psychiatrist consultant in fiscal year 2021 (FY2021; October 2020 – September 2021). Three data sources were used to characterize the program and gather feedback: (1) interviews with the consultant; (2) VHA Corporate Data Warehouse (CDW); and (3) surveys from referring providers.

Interviews with Consultant

To characterize the service, the evaluation team conducted four semi-structured interviews with the geriatric psychiatrist. The interview guide included questions regarding the history and development of the program, services delivered, and common referral questions.

VA Corporate Data Warehouse (CDW)

The VA CDW is a national enterprise data warehouse comprised of clinical and administrative systems data. At point of care, data are entered into the VA’s electronic health record (EHR) by way of manual entry, barcode scanning, or through automated instrumentation. The EHR data are uploaded into the CDW nightly. Access to the CDW for operational activities not constituted as research is granted by the National Data Systems group, a division under the VHA Office of Informatics and Information Governance.

The cohort was defined as patients who had an encounter with the consultant during FY2021. These encounters were reviewed and then limited to patients who were seen in one of the consultant’s psychogeriatric clinics based on VA Managerial Cost Accounting Stop Codes. A stop code consists of 6 digits and characterizes encounters for workload credit. The first 3 digits (primary stop code), represents the main clinical group responsible for care. The last 3 digits (secondary stop code) serves as a modifier to define the provider type or modality. Secondary stop codes were used to characterize modality by noting whether each visit was delivered via e-consultation, telephone, clinical video telehealth, telehealth to home, or care team meetings. Patient setting was ascertained via the referring provider’s setting; however, this information was missing for 125 encounters (18.8%).

Patient characteristics included age at time of first encounter, gender, rurality, race, ethnicity and the VA Care Need Assessment (CAN) score. Rurality (Urban, Rural, Highly Rural) was extracted from the CDW and is defined using Rural Urban Commuting Area (RUCA) codes. The RUCA classification system was developed by the Department of Agriculture and the Department of Health and Human Services.²¹ RUCA codes take into consideration both population density and how closely a community is linked socioeconomically to urban areas. The CAN score²² estimates the probability of a patient being hospitalized or dying within 90 days and ranges from 0 (lowest risk) to 99 (highest risk). These scores are estimated each week, but are not generated for patients who are currently hospitalized or in residential care facilities. The CAN score that was calculated during the week that the encounter occurred was extracted.

Each encounter was characterized in terms of modality of service (stop code pairings), services delivered characterized by the Current Procedural Terminology (CPT) codes, and diagnoses using the International Classification of Diseases Codes (ICD-10). For this we used the Clinical Classifications Software Refined (CCSR),^23,24 an aggregation of ICD classifications for analytical purposes, and performed slight modifications to the groupings based on analytic needs. We combined infrequently occurring mental and behavioral disorders into an “other psychiatric” category and infrequently occurring medical diagnoses into an “other medical” category.

Surveys of Referring Providers

The survey was designed to assess providers’ perceptions of the tele-geriatric psychiatry consultation program. Referring providers who placed a consult in FY2021 were invited to participate in an anonymous, confidential, and voluntary survey to provide feedback on the program. Regardless of the number of consults requested, respondents only completed the survey once. Qualitative interviews were conducted with three referring providers for the purpose of piloting and refining the questions used in the survey (findings not reported).

The survey assessed providers’ discipline, type and rurality of work setting, whether the provider received training to work with older adults, and type of training received if relevant. Referring providers rated their level of comfort in working with geriatric mental health-related presenting problems on a 5-point scale ranging from “very comfortable” to “very uncomfortable” and including a sixth option “comfort varies widely by case.” The remaining questions asked about the tele-geriatric psychiatry consult service. Providers were asked about how they learned about the service, what type of consult service they requested (i.e., video/telephone or e-consultation), the number of consults requested within FY2021, and reasons for requesting the consult(s). Using a 5-point Likert-type scale ranging from “strongly agree” to “strongly disagree,” six questions asked about satisfaction with the results of the consult, follow through with recommendations from consultant, perceived benefits to the care team or patient/caregiver, follow, whether the consult increased access to geriatric psychiatry expertise, and whether they would recommend the consult to other providers with similar patients. Open-ended questions asked about barriers to implementing the recommendations and about any additional information about the service.

Analysis

Quantitative data were summarized using frequencies, percentages, and descriptive statistics. Chi-square analyses examined exploratory questions about variations in survey responses by setting (rural/nonrural) and referring provider discipline (prescriber/non-prescriber).

Qualitative analysis summarized the findings from the semi-structured interviews with the consultant and open-ended survey questions. Content analysis with inductively developed codes was used to categorize the survey responses.²⁵ Two primary coders (AJA, CC) coded the data and one co-author (CEG) reviewed the code applications. Discrepancies were discussed; a fourth co-author (RT) addressed unresolved disagreements. Thematic analysis was employed to identify overarching patterns across the code categories in the open-ended survey questions.^26,27

Results

Consultant Interviews

Common referral questions were related to: dementia-related behaviors that are distressing and/or dangerous to Veterans and others; psychiatric conditions in context of cognitive impairment; depression; anxiety; PTSD; psychosis; and sleep concerns. Initial telehealth services typically included psychiatric diagnostic evaluations with extensive chart review, and a patient interview with caregiver or VHA staff present. Recommendations were behavioral (i.e., modifying antecedents for behaviors) or medication-related with some patients receiving time-limited follow-up encounters for evaluation and management. E-consultations included detailed recommendations regarding psychotropic medication(s) and dose/titration to address presenting concerns, suggestions for deprescribing medications, and suggestions of laboratory tests and imaging for further evaluation. Other activities included attending behavioral rounds at VHA nursing homes, providing informal consultation to staff to adjust recommendations, and delivering didactic presentations to direct care staff.

CDW Data

During the 1-year period, 315 Veterans (see Table 1) were seen with an average of 2.11 encounters (SD = 1.78, range 1 to 18). The clinical settings varied with 42.2% from VHA nursing homes followed by 28.6% from outpatient settings. Forty percent of Veterans were rural-dwelling. Half of the sample (n = 166; 52.7%) had only one visit. There were 666 completed encounters; however, 11 were missing CPT codes, 5 were missing diagnoses (ICD-10 codes), and 7 were missing secondary stop codes needed to categorize visit modality.

Table 1.

Characteristics of Patients Served during October 2020 to September 2021 (N = 315)

Patient Characteristics	N (%)	M (SD)
Age (years)		76.0 (9.64)
Sex
Female	29 (9.2%)
Male	286 (90.8%)
Rurality
Highly Rural	28 (8.9%)
Rural	98 (31.1%)
Urban	184 (58.4%)
Missing	5 (1.6%)
Race
American Indian, Alaska Native	2 (0.6%)
Black or African American	24 (7.6%)
White	262 (83.2%)
Declined to answer	6 (1.9%)
Other	21 (6.7%)
Ethnicity
Hispanic or Latino	13 (4.1%)
Not Hispanic or Latino	205 (65.1%)
Declined to answer	3 (1.0%)
Marital Status
Single	22 (7.0%)
Married	141 (44.8%)
Separated/Divorced	118 (35.8%)
Widowed	30 (9.5%)
Unknown	2 (0.6%)
Missing	2 (0.6%)
Setting
VHA Nursing Homes (CLC)	133 (42.2%)
HBPC	12 (3.8%)
Inpatient	40 (12.7%)
Outpatient	90 (28.6%)
Unknown	40 (12.7%)
CAN Scores
Available	111 (35.2%)	84.1 (17.8)
Unavailable	204 (64.8%)

Note. CAN = Care Need Assessment Score; CLC = Community Living Center; HBPC = Home-Based Primary Care. These scores are not generated for patients who are currently in a residential setting (e.g., VHA Community Living Center)

Of the 661 encounters with diagnostic information, on average there were 1.45 (SD = .54) diagnoses per encounter. Most encounters included a neurocognitive disorder diagnosis (n = 504; 76.2%), which encompassed dementias with behavioral disturbance (n = 314; 62.3%), dementia without behavioral disturbance (n = 127; 25.2%), mild cognitive impairment (n = 62; 12.3%), and delirium (n = 15; 3.0%) (see Supplemental Table 3). Within encounters including neurocognitive diagnoses, the most frequent co-occurring diagnoses were depressive and trauma and stressor disorders (see Table 2). Of encounters without neurocognitive disorders, the most frequent diagnoses were depressive disorders, schizophrenia spectrum and psychotic, and trauma and stressor disorders. One hundred fifteen encounters were related to serious mental illness (see Supplemental Table 4).

Table 2.

Neurocognitive Disorders and Co-Occurring Diagnoses (N = 661)

	Neurocognitive disorder (n = 504)		No Neurocognitive disorder (n = 157)
CCSR	n	%	n	%
Alcohol-related disorders	2	0.4%	10	6.4%
Anxiety and fear-related disorders	11	2.2%	9	5.7%
Bipolar and related disorders	14	2.8%	10	6.4%
Depressive disorders	77	15.3%	64	40.8%
Other psychiatric disorders	1	0.2%	9	5.7%
Other specified and unspecified mood disorders	4	0.8%	2	1.3%
Other specified substance-related disorders	0	0.0%	2	1.3%
Personality disorders	2	0.4%	1	0.6%
Schizophrenia spectrum and other psychotic disorders	45	8.9%	45	28.7%
Sleep wake disorders	3	0.6%	2	1.3%
Trauma- and stressor-related disorders	74	14.7%	44	28.0%
Mental and substance use disorders in remission	3	0.6%	5	3.2%
Nervous system signs and symptoms	1	0.2%	3	1.9%
Other medical	1	0.2%	8	5.1%
Other nervous system disorders (neither hereditary nor degenerative)	0	0.0%	1	0.6%
Neurocognitive disorder only	266	52.8%	N/A	N/A

Note. n = 5 encounters were missing diagnostic information.

Of the 655 encounters with complete data, each encounter had an average of 2.05 CPT codes used (SD = .73). Two hundred seventy encounters were initial consults (n = 132; 48.8% rural), and 385 consults (n = 150; 39.0% rural) were follow-up visits. Most encounters were clinical video telehealth (CVT) (n = 443; 67.6%), followed by e-consults (n = 99; 15.1%), video to home (n = 95; 14.5%), telephone (n = 16; 2.4%), and clinical conferences (n = 2; 0.3%). Table 3 displays the services provided across all modalities. A comparison of four visit types (video to home, phone, CVT, e-consults) by rural versus non-rural Veteran status revealed that the distribution significantly varied, χ²(3) = 22.52, p < .001 (see Figure 1a). Non-rural Veterans were more likely to receive CVT than rural Veterans. Rural Veterans were slightly more likely to receive video to home or e-consultation.

Table 3.

Encounters CPT Codes by Service Modality (N = 655)

Type of encounter	CPT Code	Brief Description	n	%
E-consultation			99	15.1
	99446	Medical consultation and review 5-10 min	2	(2.0)
	99447	Medical consultation and review 11-20 min	8	(8.1)
	99448	Medical consultation and review 21-30 min	11	(11.1)
	99449	Medical consultation and review 31+ min	78	(78.8)
Video to Home *			95	14.5
Consult
	90792	Psychiatric diagnostic evaluation with medical services	36	(37.9)
Follow-up
	99213	Evaluation and Management 20-29 min	2	(2.1)
	99214	Evaluation and Management 30-39 min	51	(53.7)
	99215	Evaluation and Management 40-54 min	5	(5.3)
	99442	Telephone Evaluation and Management 11-20 min	1	(1.1)
Additional services	90833	Psychotherapy 30 min	43	(45.3)
	90885	Record evaluation	34	(35.8)
	90785	Interactive Complexity	68	(71.6)
Clinical team conference			2	0.3
	99367	Medical team conference	2	(100)
Additional services	99337	Evaluation and management/domiciliary	1	(50)
Clinical Video Telehealth			443	67.6
Consult	90791	Psychiatric diagnostic evaluation	2	(0.5)
	90792	Psychiatric diagnostic evaluation with medical services	132	(29.8)
Converted to e-consult	99449	Interprofessional health record assessment	1	(0.2)
Follow-up	99213	Evaluation and management 20-29 min	17	(3.8)
	99214	Evaluation and management 30-39 min	261	(58.9)
	99215	Evaluation and management 40-54 min	28	(6.3)
	90833	Psychotherapy 30 min	1	(0.2)
Additional services	90833	Psychotherapy 30 min	89	(20.2)
	90785	Interactive complexity	324	(73.1)
	90882	Environmental intervention	1	(0.2)
	90885	Record evaluation	125	(28.2)
Telephone			16	2.4%
Converted to e-consult	99447	Medical consultation and review 11-20 min	1	(6.3)
Follow-up	99214	Evaluation and management 30-39 min	1	(6.3)
	99442	Telephone evaluation and management 11-20 min	9	(56.3)
	99443	Telephone evaluation and management 21-30 min	5	(31.3)
Additional services	90785	Interactive complexity (add on)	1	(6.3)
	90885	Record evaluation	2	(12.5)
	99358	Prolonged service	2	(12.5)

Note.

^*Video to Home = VA Video Connect (VVC). CPT = Current Procedural Terminology. n = 11 encounters missing CPT codes.

Figures 1a & 1b.

Visit Modality and Clinical Setting Varies by Rurality

Note. Visit types significantly varied among rural and non-rural Veterans, χ²(3) = 22.52, p < .001. *p < .05

The setting of the consultation varied by rurality (see Figure 1b). Rural patient encounters were more likely in HBPC (χ²(1) = 14.24, p < .001), and outpatient settings (χ²(1) = 12.89, p < .001), but were less likely in VHA nursing homes (χ²(1) = 6.12, p = .013). Rural and non-rural Veterans did not differ on any other variables.

Provider Survey Data

Fifty-eight of 98 referring providers responded to the survey, yielding a 59% response rate. As shown in Table 4, most respondents were providers with prescription privileges (i.e., prescribers; 71%), specifically physicians, physician assistants, pharmacists, or advance practice nurses. Most respondents were from either primary care (37.9%) or VHA nursing home (35.4%). Most were comfortable to very comfortable (67.2%) caring for older patients with geriatric mental health-related problems. The majority learned of the consult through a colleague/word of mouth (n = 33; 56.9%) followed by personal interaction with the consultant (n = 25; 43.1%). Forty-two (72.4%) respondents used the consult service two or more times. Respondents most often used both regular consults and e-consults (n = 30; 51.7%), followed by regular consults only (n = 17; 29.3%) and e-consults only (n = 11; 19.0%).

Table 4.

Provider Survey Characteristics (N = 58)

Referring Provider Characteristics	N (%)
Respondent’s Discipline (N = 50)1
Physician	18 (31.0%)
Nurse Practitioner or Advanced Practice Nurse	11 (19.0%)
Psychologist	9 (15.5%)
Social Worker	4 (6.9%)
Pharmacist	3 (5.2%)
Physician Assistant	3 (5.2%)
Nurse	1 (1.7%)
Unknown	1 (1.7%)
Prescribing Providers	35 (71%)
Description of work setting2
Primary Care Clinic	22 (37.9%)
Community Living Center (VHA nursing home)	20 (35.4%)
Mental Health Clinic	8 (13.8%)
Inpatient	6 (10.3%)
Community Based Outpatient Clinic3	5 (8.6%)
Inpatient Mental Health	2 (3.4%)
Caregiver Support Program	2 (3.4%)
Primary Care Mental Health Integration	2 (3.4%)
Other setting (Emergency department, Outpatient clinic, Outpatient clinic with inpatient consults)	3 (5.2%)
Where services are provided
Rural only	25 (43.1%)
Rural and Nonrural	24 (41.4%)
Nonrural only	9 (15.5%)
Received specialized training for working with older adults
No	37 (63.8%)
Yes, On the job experience	19 (32.8%)
Yes, Continuing education	17 (29.3%)
Yes, Geriatric coursework	15 (25.9%)
Yes, Rotation experience	14 (24.1%)
Yes, Fellowship focus on Geriatrics	3 (5.2%)
Yes, Other	3 (5.2%)
Comfort working with geriatric mental health-related presenting problems
Very comfortable	14 (24.1%)
Comfortable	25 (43.1%)
Neutral	13 (22.4%)
Uncomfortable	4 (6.9%)
Very uncomfortable	1 (1.7%)
Comfort varies widely by case	1 (1.7%)
Use of consult in past year4
None, but used in previous years	1 (1.8%)
Once	13 (22.4%)
Two or more times	42 (72.4%)
How learned of consult5
Colleague/word of mouth	33 (56.9%)
Personal interaction with the consultant	25 (43.1%)
Meeting/presentation	11 (19.0%)
E-mail	8 (13.8%)
Electronic medical record	8 (13.8%)
Unknown	2 (3.4%)

Note.

¹N=50; eight individuals left this blank.

²Total doesn’t equal 100% due to multiple respondents marking multiple options.

³Three of 5 respondents also selected other settings.

⁴N=56; two individuals left this question blank.

⁵Some indicated that they learned of the consult through multiple sources; hence totals exceed 100%.

Most respondents (82.8%) strongly agreed that they would recommend the consults and that the consults increased access to geriatric psychiatry care (see Figure 2). Most (72.4%) strongly agreed that they were satisfied with the results of the consult and that the care team (70.7%) and patient/caregivers (72.4%) benefitted. Moreover, 74.1% strongly agreed that they followed through with the recommendations. An open-ended question probed about potential barriers. Fifteen respondents (26%) reported barriers (see Supplemental Table 2) and of those reported, the most prominent was the need to put in orders (n = 9). No differences in either the frequency or type of barriers emerged for prescribing and non-prescribing providers (p < .05).

Figure 2.

Referring Provider Survey Responses (N = 56)

Qualitative analysis of the question regarding reasons for referral yielded six reasons (see Supplemental Table 1): diagnostic clarification (n = 8; 14%), medication recommendations (n = 18; 31%), distress or challenging behaviors (n = 11; 19%), patients with complex needs (n = 9; 16%), requests for treatment and/or treatment planning (n = 6; 10%), and general requests for geriatric psychiatry assistance (n = 15; 26%). Many respondents discussed asking for specific advice related to prescribing including dose adjustments, medication review, and consideration of lab tests. The consultation reasons somewhat varied among non-prescribers and prescribers, findings suggested that non-prescribers more frequently reported a general need for geropsychiatry expertise as a reason for consultation than prescribers (χ²(1) = 5.53, p = .025).

Thematic analysis helped show how diagnostic evaluations and treatment plans were connected. Identifying a diagnosis for patients who have complex and sometimes perplexing presentations serves as more than a label. Rather, the diagnosis may lead to a path forward in terms of treatment planning, and especially in the case of patients who may exhibit behaviors that may be challenging to manage and/or patients with co-occurring neurocognitive and psychiatric conditions.

Thematic analysis of the feedback emphasized that respondents perceived that the service increases access to a specialist, particularly in rural locations. The essential components of this service extend beyond access to the specialist and include the timely, prompt, and collaborative approach that the consultant uses. One respondent indicated “[this service] has greatly improved quality of life for our Veterans, improved access to timely psychiatric care, and has provided our teams a consultation liaison who can help us more successfully manage this specialty population.” Another respondent noted that the consultant “contacted [the provider] directly prior to sending her recommendations to ensure she understood the case and I agreed with her recommendations. It was nice to have a discussion with a specialist and develop a final plan we both agreed on.” These comments demonstrate the collaboration created through the consultant’s approach to working with the patients, their families, referring providers, and care teams.

Discussion

Our objective was to conduct an initial evaluation of a regional tele-geriatric psychiatry consultation service. Using multiple methods and data sources, we provide a snapshot of services delivered through a tailored approach that increased perceived access to geriatric psychiatry expertise, particularly for rural areas. Our key findings demonstrated that this regional telehealth approach is feasible, acceptable, and highly valued by referring providers. Examination of the medical record data revealed that the geriatric psychiatrist served patients with complex problems, often including co-occurring psychiatric and neurocognitive disorders. These patients had significant care needs and of those residing in the community, many were at increased risk of hospitalization as demonstrated by the high CAN scores.

In contrast to other studies that have largely examined single modalities of care delivery,²⁸ in this service the geriatric psychiatrist conducts assessments and delivers care across multiple clinical settings using a variety of modalities including e-consultation, video telehealth to home, video telehealth to clinics or nursing homes, or via phone/team conferencing. Feedback from referring providers and the medical record data demonstrated the varied services delivered and questions addressed. Open-ended feedback illustrated that the consultant’s patient-centered approach accommodates for patient and family preferences as well as the needs and preferences of the care team. Moreover, the multiple data sources show that care coordination, education, and team collaboration are hallmarks of the consultant’s approach to caring for these patients with complex needs. Not only are billable services delivered, but informal consultation (i.e., virtual “curbside consultation”) may contribute to the high satisfaction expressed by referring providers. The survey findings demonstrated that this service likely increases access to geriatric psychiatry expertise through both the use of telehealth modalities and the consultative model. Tele-consultation allows for timely access to an expert which would not be as possible if the expert were responsible for continuity care as well. The survey findings suggested that the patient and/or caregiver also benefitted, which should be corroborated in future studies. Many providers reported that they followed through with the consultant’s recommendations. However, in the future, objective data should be gathered to corroborate these subjective findings. Future studies should examine whether certain modalities provide more value to referring providers within specific clinical settings.

System-level barriers to this model of care were identified. The main barrier was related to the consultant not prescribing psychotropic medications and not entering orders. This barrier is consistent with primary care physicians’ overall impressions of e-consults identified in other studies.²⁹ Another barrier identified by the consultant is that time and effort devoted to informal consultation is difficult to capture for workload credit. This systems-related barrier may impede adoption of this model as specialty providers come up against workload and productivity expectations.

Future directions need to be considered in refining and evaluating this model of care. This single provider consultation approach could be adapted for interdisciplinary care teams. A formative evaluation of the adaptation, dissemination, and implementation of regional tele-geriatric MH consultation should be conducted. Other important directions include consideration of how this tele-consultation could be incorporated into the training of future MH providers. However, logistical challenges to including trainees exist due to the delivery of services across multiple health care systems and states.

Several limitations to this investigation need to be considered. First, this was an initial program evaluation that did not have a comparison condition. Second, changes in specific health outcomes were not examined; however, for such a heterogenous sample with complex problems, it may prove difficult to quantify changes. Third, the CDW data focused on encounter diagnoses and does not include the number of medical conditions that patients have or more detailed problem information. Fourth, it is possible some referring providers who were not satisfied with the services chose not to respond to the provider surveys resulting in positive bias.

This concurrent mixed methods evaluation served to characterize the tele-geriatric psychiatry consultation services delivered to a variety of settings through integration of interviews with the consultant, medical record data, and surveys with referring providers. Taken together, these data highlight the complex presentations and referral reasons that often encompass questions related to medications, distress behaviors, diagnoses, and assistance with overall treatment planning for patients with high care needs and co-occurring MH and/or neurocognitive disorders. While this single program was shown to be feasible, acceptable, and valued by the referring providers in one region, it needs to be replicated in other regions. Implementation science frameworks would be helpful to characterize this program and examine its replication and adaptation when implementing it elsewhere. This model shows promise for expanding access to the geriatric psychiatry specialty, particularly in rural areas.

Highlights.

1. What is the primary question addressed by this study? This program evaluation aimed to characterize patients served and referring provider perceptions of a cross-facility tele-geriatric psychiatry consultation program.

2. What is the main finding of this study?— This approach is feasible, acceptable, and highly valued by referring providers. The program served older Veterans with complex psychiatric and cognitive needs across multiple care settings including outpatient and rehabilitation/long-term care.

3. What is the meaning of the finding?—Geriatric psychiatry consultation delivered to multiple VA medical systems via telehealth may increase access to this specialty in rural areas.