Psychology of consulting and coaching

Implementing a Home-Based Suicide Prevention Program With Rural Veterans

Mira Brancu Department of Veterans Affairs Mid-Atlantic

Mental Illness Research, Education and Clinical Center, Durham, North Carolina, and Duke

University

Bethann Cleary Department of Veterans Affairs Mid-Atlantic

Mental Illness Research, Education and Clinical Center, Durham, North Carolina

Katherine C. Cunningham Salem Veterans Affairs Medical Center, Salem, Virginia, and Virginia Polytechnic Institute and

State University

Summer R. Anderson Department of Veterans Affairs Mid-Atlantic

Mental Illness Research, Education and Clinical Center, Durham, North Carolina

Georgia R. Gerard Department of Veteran Affairs Rocky Mountain Mental Illness Research, Education and Clinical

Center, Aurora, Colorado

Paola E. Fernandez Department of Veterans Affairs Mid-Atlantic

Mental Illness Research, Education and Clinical Center, Durham, North Carolina

Bridget B. Matarazzo Department of Veteran Affairs Rocky Mountain Mental Illness Research, Education and Clinical Center,

Aurora, Colorado, and University of Colorado

Veterans living in rural areas are at elevated risk for suicide compared with the veteran population as a whole. The Home-Based Mental Health Evaluation (HOME) program is an evidence-based suicide prevention intervention for veteran patients being dis- charged from a psychiatric hospitalization. The initial program was developed and tested with patients returning to nonrural home environments. The current article aims to describe the feasibility of adapting the HOME program when implemented with rural

Mira Brancu, Department of Veterans Affairs Mid- Atlantic Mental Illness Research, Education and Clinical Center, Durham, North Carolina, and Department of Psychiatry and Behavioral Medicine, Duke University; Bethann Cleary, Department of Veterans Affairs Mid- Atlantic Mental Illness Research, Education and Clinical Center; Katherine C. Cunningham, Salem Veterans Af- fairs Medical Center, Salem, Virginia, and Fralin Bio- medical Research Institute, Virginia Polytechnic Insti- tute and State University; Summer R. Anderson, Department of Veterans Affairs Mid-Atlantic Mental Illness Research, Education and Clinical Center; Geor- gia R. Gerard, Department of Veterans Affairs Rocky Mountain Mental Illness Research, Education and Clin- ical Center, Aurora, Colorado; X Paola E. Fernandez, Department of Veterans Affairs Mid-Atlantic Mental Illness Research, Education and Clinical Center; Bridget B. Matarazzo, Department of Veterans Affairs Rocky Mountain Mental Illness Research, Education and Clin-

ical Center, and Department of Psychiatry, University of Colorado.

Bethann Cleary is now at the Durham Veterans Affairs Medical Center, Durham, North Carolina.

This work was supported by funding from the Depart- ment of Veterans Affairs Office of Rural Health, Veterans Affairs Mid-Atlantic Mental Illness Research, Education and Clinical Center, and Rocky Mountain Mental Illness Research, Education and Clinical Center. The authors thank Madrianne Wong, Meredith Berger, Joy Close, and Juanita Hill. Their contributions helped to shape this pro- gram into one that could benefit at-risk rural veteran pa- tients.

Correspondence concerning this article should be ad- dressed to Mira Brancu, Department of Veterans Affairs Mid-Atlantic Mental Illness Research, Education and Clin- ical Center, 3022 Croasdaile Drive, Suite 300, Durham, NC 27705. E-mail: mira.brancu@va.gov

Journal of Rural Mental Health In the public domain 2020, Vol. 44, No. 3, 146 –155 ISSN: 1935-942X http://dx.doi.org/10.1037/rmh0000138

146

patients at a new facility in a new region of the country. Implementing and adapting the clinical HOME program for rural veteran patients did not require a significantly different approach to achieve similar levels of program engagement as the original research trial. Despite the ease with which these adaptations were made, the barrier of distance remained an obstacle to delivering home-based interventions to rural patients. Methods of addressing barriers to program and treatment engagement are offered for future adaptations of this and other in-home mental health-care programs for rural patients.

Public Health Significance This article aims to facilitate the adaptation and implementation of an evidence- based suicide prevention intervention program by describing how to overcome barriers to in-home mental health care for rural patients in order to achieve levels of program engagement similar to those of the original program.

Keywords: veterans, suicide prevention, implementation, adaptation

The risk of death by suicide is 1.5 times greater for veterans than members of the general population (Department of Veterans Affairs, 2019). Compared with the general veteran pop- ulation, rural veterans are at further elevated risk (McCarthy et al., 2012). Rural veterans comprise more than a quarter of all veterans, and over a third of those served in the most recent conflicts in Afghanistan and Iraq in sup- port of Operation Iraqi Freedom (OIF) and Op- eration Enduring Freedom (OEF; Department of Veterans Affairs, 2012).

One factor that may contribute to elevated suicide risk among rural veterans is limited ac- cess to mental health care. According to the Department of Veterans Affairs (VA) Office of Rural Health (ORH), rural veterans are less likely than nonrural veterans to access mental health services, either through the Veterans Health Administration (VHA) or the private sector (VA ORH, 2015). This disparity is par- ticularly concerning because a failure to engage in mental health treatment is associated with a higher risk of suicide (Desai, Dausey, & Rosen- heck, 2005).

In addition to rurality, recent discharge from a psychiatric hospitalization is another risk fac- tor for both veterans and civilians. In a popula- tion-based study of veterans with a history of depression, the highest-risk period for suicide was the first 12 weeks following psychiatric hospitalization, during which the rate of suicide was 5 times higher than the base rate (Valen- stein, Kim, Ganoczy, et al., 2009).

Some evidence suggests that engaging with patients during the brief but vulnerable transi- tion between inpatient and outpatient care can significantly reduce suicide risk (e.g., Linehan et al., 1999; Motto & Bostrom, 2001). Although some interventions developed for this transition period focus on connecting patients to care, they do not necessarily help patients address barriers to initiating and maintaining that care, nor do they provide ongoing risk assessment and guid- ance on how to effectively cope with stressors and/or crises.

Alternatively, the Home-Based Mental Health Evaluation (HOME) program, devel- oped and evaluated by Matarazzo and col- leagues (2017, 2019), focuses on (a) providing a connection to care during the transition from inpatient to outpatient care, (b) collaboratively addressing barriers to initiating and maintaining outpatient care engagement, and (c) assessing suicide risk and ensuring the patient knows how to cope and seek help during a crisis. The pro- gram involves meeting the veteran patient while hospitalized, conducting a telephone call the first day after discharge, making a home visit during the first week after discharge, and weekly telephone contact until the patient is engaged in outpatient mental health care. Each of these contacts includes conducting a suicide risk assessment, reviewing and updating the pa- tient’s suicide prevention safety plan (Stanley & Brown, 2008), reviewing hospital discharge plans, and collaboratively problem-solving bar- riers to accessing and engaging in outpatient

147SUICIDE PREVENTION PROGRAM FOR RURAL VETERANS

care. In this collaborative, patient-centered in- tervention, patients and the HOME provider brainstorm ways to overcome potential barriers and are supported in their efforts to address these barriers.

The HOME program was initially developed and tested at a large western VA medical center as a clinical demonstration project (Matarazzo et al., 2017) and most recently tested through a nonrandomized multisite clinical trial (Mat- arazzo et al., 2019). The HOME program has demonstrated efficacy in engaging veteran pa- tients in care after discharge (Matarazzo et al., 2017, 2019), but until recently, it was only implemented with veterans living near nonrural VA medical centers. Given the additional risk incurred by living in a rural setting, the research team implemented the HOME program with rural veteran patients receiving care at a VA medical center in the mid-Atlantic region.

Mental health interventions may be more ef- fective when they are adapted for a specific cultural group (Griner & Smith, 2006). Culture refers to the shared values, beliefs, and norms of a community or group (as cited by Hartley, 2004). Rural culture has been suggested as a health determinant; however, rural culture is variable, complex, and influenced by factors that may be unique to individual towns or dif- ferent regions of the country. Rural places are geographically, culturally, and ethnically di- verse. They are also economically, racially, re- ligiously, and topographically different from one another and can vary greatly in population size and population density. Therefore, knowl- edge—through either experience, education, or ideally, both— of rural-specific health-care fac- tors is important when working with rural vet- erans. These include issues of availability and accessibility of mental health care in the local community and the differences in how rural patients utilize VA versus community health- care services. They also include understanding barriers to mental health treatment engagement that may exist in rural communities. For exam- ple, rural patients’ perceptions of access to mental health care are more important to de- termining access to care than the actual num- ber of providers or clinics in a community (Bischoff et al., 2014). In addition, in small and rural communities, unlike in larger com- munities where multiple provider options are available, patients may face additional chal-

lenges related to dual relationships and boundaries. Providers may need to discuss with rural patients the various local supports that can be accessed in a way that minimizes dual-relationship concerns.

Ultimately, there is a burden– benefit trade- off faced by rural veterans who decide to en- gage in mental health care: the perceived bal- ance between the effort and challenges required to engage in care (i.e., availability, accessibility, confidentiality, and acceptability) and the per- ceived potential for benefit (e.g., symptom alle- viation or reduction). Veterans in rural areas may be more likely to delay seeking care, less likely to report needing mental health or sub- stance use treatment (Davis et al., 2011), and less likely to receive care from a mental health specialist (Spoont et al., 2011).

Based on these factors, and in response to the real-time experiences of the providers and patients, several of these considerations were incorporated into the implementation of the HOME program with rural patients. This ar- ticle (a) describes adaptations made to the HOME program when implemented with ru- ral patients and (b) reports descriptive data regarding the feasibility of implementing the HOME program with rural versus nonrural patients.

Implementation Method

Patients admitted to the inpatient psychiatric unit at a mid-Atlantic VA medical center were screened for HOME program eligibility. Screening methods were essentially similar to those of the original HOME program clinical trial protocol. Specifically, screening data were collected via review of the electronic medical record (EMR) and criminal history (to assess for potential safety concerns; see the eligibility cri- teria) and through collaboration with the inpa- tient treatment team. Patients who met the fol- lowing criteria were offered the HOME program: (a) had been hospitalized on the inpa- tient psychiatric unit at the participating VA medical center; (b) were willing to engage in outpatient mental health treatment following the hospital discharge; (c) were willing and able to provide a working phone number and physical address of the residence to which they would be discharged; (d) were being discharged to a lo-

148 BRANCU ET AL.

cation within a 2-hr drive from the participating medical center.

Patients were not eligible for the HOME pro- gram if they (a) were being directly discharged to a residential treatment facility, jail, or prison; (b) were enrolled in an intensive case manage- ment program that included one or more in- home visits (e.g., VA Mental Health Intensive Case Management [MHICM] Program); or (c) could not provide a safe home environment or location for the HOME program provider to visit after discharge.

During enrollment in the intervention, pa- tients completed an information sheet, which included demographic information (including race, ethnicity, age, and military service history) and contact information (including phone num- ber and discharge location address). One minor adaptation made to the protocol was to deter- mine rurality based on the home address pro- vided by the patient and the distance to the nearest VA medical center (see the Adaptations to the HOME Program Protocol section).

Similar to the original clinical trial, HOME providers were generally full-time clinical so- cial workers or community-focused clinical psychologists. The total number of staff needed to meet the demands of the program averaged approximately 1.5 full-time staff, which dif- fered based on the number of inpatient beds and turnaround at any given hospital facility. A typ- ical workload was to serve approximately 10 –12 patients at a time.

The HOME provider recorded data regarding the patient’s engagement in the HOME program in order to compare this information to the original protocol. This included the number of phone calls made, whether a home visit or other face-to-face contact occurred (e.g., coffee shop, public library, place of employment), contact duration, and if the patient withdrew from the intervention. The provider also recorded data regarding the patient’s posthospitalization treat- ment engagement. This included whether the patient engaged in at least one outpatient mental health appointment and how long it took to engage in that care. These data were obtained through patient report and confirmed by review of the electronic medical record (EMR). This work was part of a larger project approved by the Institutional Review Boards at the Durham VA Medical Center and the Rocky Mountain Regional Medical Center. Informed consent

was waived for this work, given the archival and retrospective nature of the program descrip- tion.

Adaptations to the HOME Program Protocol

Defining Rurality

As noted previously, a small adaption to the original protocol required establishing a defini- tion of “rurality” in advance of enrolling pa- tients. Rurality can be defined and categorized in a wide variety of ways. The VA ORH utilizes Rural-Urban Commuting Area (RUCA) codes for rural classification (WWAMI Rural Health Research Center, 2005). On a scale from 1 to 10, an area is classified as Metropolitan (1–3), Micropolitan (4 – 6), Small Town (7–9), or Ru- ral (10). To align with the VA ORH use of RUCA classifications, the team defined a RUCA code of 1 as “urban” and RUCA codes of 2–10 as “rural” (WWAMI Rural Health Re- search Center, 2005).

Providers visiting the various rural environ- ments noted important differences between the RUCA descriptions and the local topography and resources available. These factors could help inform rural classifications in addition to RUCA codes and included the following: prox- imity to and accessibility of mental health treat- ment services, access to transportation, and dis- tance to other resources/areas.

Distance

After defining rurality, we had to adapt our eligibility criteria with respect to the distance of the patient’s home from the VA medical center where staff worked. Distance and/or driving time to a rural home location can be quite vari- able depending on the geographic area. For ex- ample, 20 miles in one area of the country might take twice as long to drive compared with other areas because of topographical features, geo- graphical features, road infrastructure, or traffic. When establishing driving-distance parameters, we considered the cost of provider hours spent traveling to one patient’s home compared with seeing more patients closer to the medical fa- cility, car availability, individual provider con- cerns about certain driving durations, and cel- lular service connection in rural areas. The

149SUICIDE PREVENTION PROGRAM FOR RURAL VETERANS

original HOME program protocol excluded pa- tients who lived more than a 45-min drive from the medical center. To accommodate rural pa- tients while mitigating provider burden, our team adjusted this to 2 hr.

Phone Service Access

A potential barrier to implementing the HOME program with rural veterans is limited, intermittent, or unreliable cell service in some rural areas. As such, an additional adaptation was for the providers to have more in-depth conversations regarding phone access as part of the enrollment process. When available, pa- tients were encouraged to provide their landline phone number to ensure a more stable connec- tion. If a landline was not accessible, but the patient had a cell phone with inconsistent cel- lular service, the patient would commute to an area of better cell service reception for the scheduled phone contacts.

Transportation

Our providers quickly learned that difficulty finding reliable transportation was a frequent barrier for rural patients attempting to engage in outpatient mental health care after an inpatient hospitalization. Specifically, patients’ travel time might be significantly longer than their appointment length, and they might have mul- tiple appointments in a given week. This dis- tance burden is compounded when there are additional transportation barriers, such as few public transportation options, the cost of travel, or a lack of access to a personal vehicle.

In response to these common transportation barriers among rural patients, the HOME program providers adapted the protocol to explicitly ask about and problem-solve potential transportation barriers to care. Problem-solving discussions in- cluded considering asking a friend or family mem- ber (particularly one listed on the safety plan) for a ride, using public transportation or transportation available through county social services, using ride-sharing, or linking with providers through phone or videoconferencing when possible (see the Discussion section). For qualifying veterans, the VA may also offer support for the cost of certain transportation options and may have shut- tle service available in some areas through volun- teer service mechanisms (e.g., Disabled American Veterans).

Appointment Management

HOME providers learned over time that many of their rural patients tried to bundle multiple health-care appointments on the same day to re- duce appointment-related travel time. In larger medical facilities, however, appointments for dif- ferent types of medical or mental health needs are not always available on the same day or around the same time periods. Thus, additional adapta- tions included serving in coordination and advo- cacy roles to assist rural veteran patients in sched- uling these additional services when appropriate.

Engagement Outcomes of the Adapted HOME Program

Patient Characteristics

Table 1 presents participant characteristics, including rural and nonrural subsamples. Out of 242 eligible patients who were approached and offered the HOME program at this site, 42 (17.35%) declined, and 200 (83.64%) enrolled in the program. Two were excluded from anal- yses because of missing rurality data. Of the 198 remaining patients, 71 (35.5%) were con- sidered to live in rural areas. The majority of the sample identified as male (n � 168, 85%). De- mographics were similar for rural and nonrural patients, although there was a relatively higher percentage of White/Caucasian patients living in rural areas. Compared with nonrural veterans, rural veterans drove, on average, 20 more miles and 27.4 more minutes to get to the nearest medical center for their appointments.

HOME Program Engagement

Table 2 displays data regarding how rural and nonrural patients engaged in the HOME pro- gram (e.g., number of phone calls, length of home visits). The team was able to attain similar program engagement outcomes to the standard HOME protocol as described in the original Matarazzo et al. (2019) publication despite the previously described adaptions. For example, in the original protocol, 87% of patients had at least one HOME program call after hospitaliza- tion; in the current adapted version, 85% of rural patients, 90% of nonrural patients, and 88.9% of the combined group received at least one call after hospitalization. In the original

150 BRANCU ET AL.

protocol, the median number of phone calls was 2, similar to the current adaptation. In the orig- inal protocol, 66% of patients received an in- home visit; in the current adapted version, 64.7% of rural patients and 63.1% of nonrural patients received an in-home visit (and when adding other in-person visits, 73% of the total sample had a face-to-face visit after hospitaliza-

tion). Of the patients who did not receive any posthospitalization face-to-face visits (n � 54; 27.3%), the following reasons were noted: pa- tient scheduled and then later canceled a home visit (n � 14), patient declined home visit/ declined to schedule a home visit (n � 13), patient confirmed visit but subsequently was a no-show (n � 2), clinician was unable to reach

Table 1 Patient Characteristics

Demographic Rural

(n � 68) Nonrural

(n � 130) Total

(N � 198)

Gender Male 58 (85%) 110 (85%) 168 (85%) Female 10 (15%) 18 (14%) 28 (14%) Transgender 0 (0%) 2 (2%) 2 (1%)

Agea

Years 51 (14) 50 (15) 50 (15) Race

Black/African American 29 (43%) 67 (52%) 96 (48%) White/Caucasian 38 (56%) 57 (44%) 95 (48%) Otherb 0 (0%) 5 (3.8%) 5 (2.5%)

Discharge location Own home 57 (83.8%) 91 (70.0%) 148 (74.7%) Family/friend’s home 7 (10.3%) 19 (14.6%) 26 (13.1%) Transitionalc 2 (3.0%) 16 (12.3%) 18 (9.1%)

Nearest VAMCd

Distance in miles 53.7 (8.8–120) 25.8 (1.2–112) 32.4 (1.2–120) Time in minutes 61.0 (13.0–115) 32.0 (4.0–135) 40.0 (4.0–135)

Note. VAMC � Department of Veterans Affairs (VA) medical center; SD � standard deviation. Percentages do not sum to 100% for race or discharge location because counts for missing data for race and discharge location are not listed. a Age � mean age in years (SD). b Other includes collapsed cell counts for Asian, Native American, and mixed race. c Transitional includes shelters, group homes, halfway houses, extended-stay motels, and so forth. d Nearest VAMC � median time and distance (range).

Table 2 HOME Program Engagement

Program engagement type Rural

(n � 68) Nonrural

(n � 130) Total

(N � 198)

Phone calls Patients with �1 intervention call 58 (85.3%) 117 (90.0%) 176 (88.9%) Median calls per patienta 2 (1–18) 2 (1–9) 2 (1–18) Median call duration in minutes 15 (3–60) 15 (4–45) 15 (3–60)

Home visits No in-person contact 18 (26.5%) 36 (27.7%) 54 (27.3%) Received an in-home visit 44 (64.7%) 82 (63.1%) 126 (63%) Received other in-person contact 6 (8.8%) 12 (9.2%) 18 (9.0%) Median in-person contact duration in minutes 47 (15–60) 45 (15–80) 45 (15–80)

Disengagement Withdrew from HOME program 4 (5.9%) 3 (2.3%) 7 (3.5%) Withdrawn due to nonresponse 13 (19.1%) 23 (17.7%) 36 (18.2%)

Note. HOME � Home-Based Mental Health Evaluation. a Median calls per patient indicates per patient with at least one completed intervention call.

151SUICIDE PREVENTION PROGRAM FOR RURAL VETERANS

patient to schedule a home visit (n � 22), and other (n � 3). Overall, these descriptive data indicate that the team achieved a level of pro- gram engagement for this adapted version that was similar to the original trial, regardless of whether the patients were rural or nonrural.

Treatment Engagement

Table 3 displays the percentage of rural and nonrural patients who engaged in follow-up outpatient mental health care. The median number of days between inpatient discharge and the first outpatient appointment was 5 days for both rural and nonrural patients. Similarly, 83.8% of rural patients and 81.5% of nonrural patients engaged in at least one outpatient mental health-care appointment within a time frame of 5 days.

Discussion

The HOME program provides a critical bridge service as patients transition between a structured inpatient unit and outpatient care. Suicide risk assessment, safety planning, ongo- ing assessment of treatment engagement barri- ers, and collaborative brainstorming for over- coming barriers to treatment engagement assist patients in making this transition safely and connecting with outpatient care. Our observa- tions suggest that although several adaptations should be considered, the HOME program is feasible to implement with rural patients.

Health-care system administrators and pro- viders who are interested in implementing the HOME program with patients living in rural areas first must define rural. Utilizing RUCA codes, as we did in the current study, is one strategy. Although RUCA codes facilitate clas- sification, there are many varying characteris- tics of rural areas with the same RUCA code that should be observed and potentially incor-

porated into implementation, in addition to the factors noted by our team (e.g., proximity to mental health care; see also Fox, Blank, Rovn- yak, & Barnett, 2001). For example, RUCA codes of 2 or 3 in some areas, such as the geographic area of the program in this article, often contained locations considered to be more similar to rural environments, whereas in other areas of the country, these same RUCA codes would point to large/urban areas. The use of different RUCA codes may have yielded differ- ent outcomes. Thus, as described earlier, clini- cians must assess the true nature of their area to determine rurality classifications and adjust- ments.

Distance, access to phone services, transpor- tation, and appointment management are also important considerations for the potential adap- tation of the HOME program. Our team found that no major modifications to the program’s most critical elements had to be made to attend to these factors; rather, the adaptations were reframed as important elements of clinical prob- lem-solving interactions with rural patients. These conversations were anchored in the cli- nicians’ knowledge regarding this cultural group, the specific barriers identified by a vet- eran patient, and the clinician’s collaborative problem-solving skills. For example, our clini- cians are aware of the common value of inde- pendence among individuals living in rural ar- eas, as well as the reality of limited public transportation. They were sensitive to both of these issues as they problem-solved with pa- tients around options to get to appointments (e.g., asking a friend or family member for a ride).

Despite the ease with which these adaptations were made, it must be acknowledged that the barrier of distance remains an obstacle to deliv- ering home-based interventions to rural pa- tients. There were some highly rural patients

Table 3 Posthospitalization Treatment Engagement

Outpatient engagement Rural

(n � 68) Nonrural

(n � 130) Total

(N � 198)

Attended �1 MH appointment 57 (83.8%) 106 (81.5%) 163 (82.3%) Days until first MH appointment 5 (1–57) 5 (0–65) 5 (0–65)

Note. MH appointment � outpatient mental health appointment. Days until first MH appointment indicates the median number of days from discharge until the first MH appointment.

152 BRANCU ET AL.

who were not eligible for the HOME program because the distance to their homes was beyond clinically reasonable for a home visit (�2 hr of driving each way). Having more highly rural veteran patients might have affected the out- comes to an extent that is currently unknown. It is possible that further adaptations would have to be made for highly rural veterans. Currently, our team is evaluating whether one additional adaptation to the HOME program might be fea- sible and yield the same level of program en- gagement: offering patients who cannot be seen for the initial in-home visit the option of a telephone or videoconferencing appointment. In the VA health-care system, the videoconferenc- ing option is often referred to as telehealth or video on demand, and many facilities have tele- health coordinators who provide video-on- demand technology tablets to create a “tele- health provider appointment and visit.” Telehealth and phone-based care are increas- ingly becoming available and offered as options for certain health-care needs. In fact, the VA recently attained “Anywhere to Anywhere” au- thority, allowing health-care providers to utilize telehealth to treat patients across state lines, including directly in patients’ homes (Authority of Health Care Providers to Practice Telehealth, 2018).

When there are barriers to traditional imple- mentation with respect to distance, provider safety in the home environment, or patients’ expressed discomfort with a home visit, these options may help ensure that patients can still receive a nearly similar level of contact. Our team’s experience with a few patients using this alternative option suggests that there may be an equivalent benefit, although anecdotally, many patients reported still preferring an in-person visit. Based on these preliminary observations, further research trials and evaluation of adapta- tions using these technologies appear well war- ranted.

In terms of feasibility, the HOME program was successfully implemented with rural pa- tients. Compared with HOME program delivery with nonrural patients, delivery with rural pa- tients was comparable with respect to the num- ber of phone calls, the length of the calls, the duration of the home visit, and the ability to conduct the home visit. The impact of the HOME program on treatment engagement for rural and nonrural patients was similar as well.

Despite rural patients living, on average, ap- proximately 28 miles further away from the closest VA medical center, the proportion of patients who attended an initial outpatient men- tal health appointment and the number of days until initiating outpatient care was similar for both groups.

A limitation of this study is that we did not formally examine the acceptability of the HOME program for rural patients, although the acceptability of the HOME program more gen- erally has been evaluated and reported else- where (e.g., Matarazzo et al., 2017). We did, however, record quotes of feedback spontane- ously offered by patients. For example, many patients voiced appreciation for the program, making comments such as, “This is an excellent program; you didn’t dump me out the door and say ‘fly free’—you said, ‘we’re with you; we’re here,’ so I don’t get panicked or overwhelmed. You’re a transition between the hospital and real life.” Outpatient providers also expressed appreciation for the HOME program, noting that it was helpful to have a thorough suicide risk assessment and updated safety plan com- pleted prior to their session because it allowed them more time to focus on other aspects of care. One outpatient provider shared, “Having HOME is a relief . . . I know all the suicide assessment and related stuff will be done, so I can focus on the other main goals!”

Another limitation of this study relates to the generalizability of our observations. We tested the HOME program with rural patients in one mid-Atlantic portion of the United States. Al- though many of the adaptations made would likely be relevant to other rural areas, additional considerations are also indicated. Implementing the HOME program in mountainous areas, for example, will likely introduce unique barriers and facilitators to intervention delivery. It is recommended that the HOME program always be tailored to the patients being served while maintaining fidelity to the core model of the program.

In addition to exploring phone-only and tele- health-only options of HOME program deliv- ery, our team is also conducting a formal cost analysis. The preliminary data are encouraging and suggest that the cost of the HOME program may be offset by reductions in emergency visits and rehospitalization rates. These ongoing anal- yses, in conjunction with the findings reported

153SUICIDE PREVENTION PROGRAM FOR RURAL VETERANS

in the current study and those reported by Mat- arazzo et al. (2017, 2019), suggest that the HOME program is effective and is feasible to implement with both rural and nonrural pa- tients.

Overall, our team learned that with a few adaptations, most of which are clinically rele- vant problem-solving opportunities, it is feasi- ble to implement HOME with rural veteran patients in a manner that follows the original protocol and achieves the same outcomes. In certain areas, telehealth for HOME may poten- tially further reduce cost and improve workload efficiencies. Furthermore, both outpatient pro- viders and patients viewed HOME as extremely valuable during the transition from psychiatric inpatient to outpatient mental health care. The HOME program is an important intervention to address a gap in treatment and reduce suicide risk among high-risk rural veterans.

References

Authority of Health Care Providers to Practice Tele- health, 38 C.F.R. § 17 (2018).

Bischoff, R. J., Reisbig, A. M. J., Springer, P. R., Schultz, S., Robinson, W. D., & Olson, M. (2014). Succeeding in rural mental health practice: Being sensitive to culture by fitting in and collaborating. Contemporary Family Therapy: An International Journal, 36, 1–16. http://dx.doi.org/10.1007/ s10591-013-9287-x

Davis, L., Mahaney-Price, A. F., Tabb, K. D., Mc- Neal, S. F., Hamner, K., Hilgeman, M. M., . . . Hawn, M. T. (2011). Alabama Veterans Rural Health Initiative: A preliminary evaluation of un- met health care needs. Journal of Rural Social Sciences, 26, 14–31.

Department of Veterans Affairs. (2012, July). Char- acteristics of rural veterans: 2010. Retrieved from http://www.va.gov/vetdata/

Department of Veterans Affairs. (2019, September). 2019 National Veteran Suicide Prevention Annual Report: Office of Mental Health and Suicide Pre- vention. Retrieved from https://www.mentalhealth .va.gov/docs/data-sheets/2019/2019_National_ Veteran_Suicide_Prevention_Annual_Report_ 508.pdf

Department of Veterans Affairs, Office of Rural Health (VA ORH). (2015). Annual report. Retrieved from https://www.ruralhealth.va.gov/docs/ORH_Annual_ Report_2015_FINAL.pdf

Desai, R. A., Dausey, D. J., & Rosenheck, R. A. (2005). Mental health service delivery and suicide risk: The role of individual patient and facility factors. The American Journal of Psychiatry, 162,

311–318. http://dx.doi.org/10.1176/appi.ajp.162.2 .311

Fox, J. C., Blank, M., Rovnyak, V. G., & Barnett, R. Y. (2001). Barriers to help seeking for mental disorders in a rural impoverished population. Com- munity Mental Health Journal, 37, 421– 436. http://dx.doi.org/10.1023/A:1017580013197

Griner, D., & Smith, T. B. (2006). Culturally adapted mental health intervention: A meta-analytic re- view. Psychotherapy: Theory, Research, Practice, Training, 43, 531–548. http://dx.doi.org/10.1037/ 0033-3204.43.4.531

Hartley, D. (2004). Rural health disparities, popula- tion health, and rural culture. American Journal of Public Health, 94, 1675–1678. http://dx.doi.org/10 .2105/AJPH.94.10.1675

Linehan, M. M., Schmidt, H., III, Dimeff, L. A., Craft, J. C., Kanter, J., & Comtois, K. A. (1999). Dialectical behavior therapy for patients with bor- derline personality disorder and drug-dependence. The American Journal on Addictions, 8, 279–292. http://dx.doi.org/10.1080/105504999305686

Matarazzo, B. B., Farro, S. A., Billera, M., Forster, J. E., Kemp, J. E., & Brenner, L. A. (2017). Con- necting veterans at risk for suicide to care through the HOME program. Suicide and Life-Threatening Behavior, 47, 709–717. http://dx.doi.org/10.1111/ sltb.12334

Matarazzo, B. B., Forster, J. E., Hostetter, T. A., Billera, M., Adler, J., Ganzini, L., . . . Brenner, L. A. (2019). Efficacy of the Home-Based Mental Health Evaluation (HOME) program for engaging patients in care following hospitalization. Psychi- atry Services, 70, 1094–1100. http://dx.doi.org/10 .1176/appi.ps.201900002

McCarthy, J. F., Blow, F. C., Ignacio, R. V., Ilgen, M. A., Austin, K. L., & Valenstein, M. (2012). Suicide among patients in the Veterans Affairs health system: Rural-urban differences in rates, risks, and methods. American Journal of Public Health, 102(Suppl. 1), S111–S117. http://dx.doi .org/10.2105/AJPH.2011.300463

Motto, J. A., & Bostrom, A. G. (2001). A randomized controlled trial of postcrisis suicide prevention. Psychiatric Services, 52, 828– 833. http://dx.doi .org/10.1176/appi.ps.52.6.828

Spoont, M., Greer, N., Su, J., Fitzgerald, P., Rutks, I., & Wilt, T. J. (2011). Rural vs. urban ambulatory health care: A systematic review (VA-ESP Project 09-009). Retrieved from https://www.hsrd.research .va.gov/publications/esp/ambulatory-REPORT.pdf

Stanley, B., & Brown, G. K. (2008). The safety plan treatment manual to reduce suicide risk: Veteran version. Washington, DC: U.S. Department of Veterans Affairs.

Valenstein, M., Kim, H. M., Ganoczy, D., McCarthy, J. F., Zivin, K., Austin, K. L., . . . Olfson, M. (2009). Higher-risk periods for suicide among VA

154 BRANCU ET AL.

patients receiving depression treatment: Prioritiz- ing suicide prevention efforts. Journal of Affective Disorders, 112, 50–58. http://dx.doi.org/10.1016/j .jad.2008.08.020

WWAMI Rural Health Research Center. (2005). Ru- ral-urban commuting area codes (RUCAs). Re-

trieved from http://depts.washington.edu/uwruca/ index.php

Received September 30, 2019 Revision received April 1, 2020

Accepted April 6, 2020 �

155SUICIDE PREVENTION PROGRAM FOR RURAL VETERANS