Veterans Crisis Line Callers With and Without Prior VHA Service Use

Abstract

This study examines differences between Veterans with and without prior Veterans Health Administration service use who received a clinical referral from the Veterans' Crisis Line. Approximately 91% of Veterans had a history of service use and 9% did not. Callers with prior service use were older, had more mental health disorders, made in-person contact more quickly, and used more outpatient mental healthcare. Those without prior service use were younger, had more mental health problems, and presented for care later. Callers with suicide-related diagnoses had high rates of service contact. These groups represent different subpopulations with unique healthcare needs and practices.

Research indicates that subgroups of Veterans, such as those diagnosed with a mental health disorder and those who receive healthcare from the Veterans Health Administration (VHA), are at greater risk for suicide than the general population (Blow et al., 2012; Bullman & Kang, 1996; Kang & Bullman, 2008; Kang & Bullman, 2009). Reviews of the relevant prevalence, risk factor, and intervention literatures suggest that a public health approach is needed to address risk in this population (Bossarte, Claassen, & Knox, 2010; Knox & Bossarte, 2012; Rice & Sher, 2012). Recognizing this, VHA has implemented a comprehensive suicide prevention strategy designed to increase access to health services, promote help-seeking among those in distress, and provide intensive care to Veterans identified as being at high risk for suicidal behavior (i.e. attempts or suicide) (Department of Veterans Affairs, 2008a). The National Veterans Crisis Line (VCL), previously known as the National Veterans Suicide Hotline, provides Veterans with 24-hour access to trained responders with the ability to link them to care in their local community, making it an important component of the VA's comprehensive suicide prevention program (Knox et al., 2012).

The VCL is a telephone crisis intervention and support service available to Veterans, active duty/reserve service members and National Guard personnel, and family and friends who are concerned about them. When a call is received by the VCL, responders identify the crisis, take steps to reduce distress, assess the caller's level of risk, and link them with local VHA or community resources, when appropriate (Department of Veterans Affairs, 2011). Critical differences between the VCL and crisis lines in the civilian sector indicate a need for research on the VCL (Britton et al., 2013). The VCL serves a predominantly male population with military experience with evidence of increased risk for suicide when compared to the general population served by civilian lines (Blow et al., 2012). VCL responders are trained clinicians who can view the medical charts of VHA patients when given permission, while civilian line responders are mostly trained volunteers who rarely have access to medical records. VCL responders also refer high-risk callers not in need of immediate care to a network of Suicide Prevention Coordinators (SPC) located at VHA medical centers across the country, a service unavailable to most civilian lines. Once a referral is initiated, SPCs are responsible for contacting the caller within one 24-hour business day and coordinating appropriate care. SPCs have access to a wide range of VHA services including primary care, mental health, and homeless services, among others (Department of Veterans Affairs, 2008b).

Although a comprehensive description of Veteran callers is available in the 2012 Suicide Data Report, Veterans are a heterogeneous population and different groups of Veterans may call the VCL for different reasons (Kemp & Bossarte, 2013). One potentially informative distinction is whether or not Veteran callers have previously used VHA services, as only an estimated 34% of all Veterans are currently enrolled in VHA services (Department of Veterans Affairs, 2012). This distinction could be important because the VCL may serve a different function for each group, complementing or improving existing healthcare for those with prior VHA use and acting as a gateway into VHA care for those without prior use. Understanding the characteristics and service utilization patterns of these two groups of callers may provide critical information concerning the characteristics of Veterans who use the VCL, why they use it, and how VCL services can be tailored to better meet their needs.

Methods

Data

Data for this study were obtained from VCL call records and VHA medical files. Information for each call is registered in a database using a standardized call management system. Depending on the nature of the call and information provided by the caller, call records may contain information that is needed to link VCL records with clinical data. These linked records are used for multiple purposes, including tracking caller characteristics and trends, identifying changes in service needs, and assessing call outcomes (Department of Veterans Affairs, 2011). This study analyzed calls received during the 2010 calendar year, from callers who (1) provided a valid security number, (2) accepted a VCL referral, (3) presented for VHA services, and (4) were identified in VHA clinical records. It is important to note that referrals are made for all callers with an indication of need and a willingness to accept a referral for additional care. Only the first referral was used for callers who accepted more than one referral. Key identifiers in the call log were used to find a corresponding medical record and extract the diagnostic and service utilization data used in the analyses. Prior VHA use was defined as receiving any VHA service since October of 2005 (which was approximately five years before the referral), with the remainder being defined as having no prior use. VHA service utilization was obtained for the 180 day period following the index referral. The analytic database was compiled and analyzed in 2012. This study was granted IRB approval by the Veterans Affairs Medical Center in Syracuse, New York.

Variables

Demographics

Age and gender were obtained from the Veteran's medical record. Age was categorized into 18-29, 30-39, 40-49, 50-59, 60-69, 70-79, and 80 and older age groups.

Primary Diagnoses

Primary diagnoses were made by healthcare providers at first service utilization following the referral, and entered into the Veteran's medical record as International Classification of Diseases, Ninth Edition (ICD-9) codes. Diagnoses were divided into four categories. The first category was suicide-related diagnoses and V codes (i.e. suicidal ideation, suicide attempts, and preparatory behavior). The second category included mental health and substance use disorder diagnoses (e.g., depression, alcohol dependence). ICD-9 V codes representing other factors influencing health status and contact with health services were a third category that included diagnoses associated with persons encountering health services in other circumstances such as other persons seeking consultation (e.g., persons consulting on behalf of another, counseling not otherwise specified), housing, household and economic circumstances, other psychosocial diagnoses (e.g., unspecified psychosocial conditions, other administrative purpose), and other factors influencing health status and contact with health services (e.g., suspected mental conditions, aftercare, depression screening). The fourth category included non-mental health, substance use or psychosocial health-related diagnoses (e.g., non-specific health diagnoses, diabetes, and back pain). The specific coding algorithm is available upon request.

Service Utilization

Service utilization was obtained for all mental and non-mental health inpatient and outpatient services provided by VHA providers following the referral. These analyses focused on the first in-person contact, as SPCs often contact Veterans by telephone within one 24-hour business day of their call to schedule these appointments. The number of days until the caller presented for care was calculated and categorized as same day, 1-7 days, 8-14, 15-21, 22-30, and 30 or more days.

Mental health and inpatient VHA services accessed on or within the first 180 days following the referral were also assessed and grouped into (1) outpatient mental health encounters, (2) mental health inpatient stays, (2) mental health residential/domicillary (VHA housing facilities) stays, and (3) all other non-mental health in-patient stays. Analyses were limited to these groups to increase the probability that the visit was related to the VCL call. Outpatient encounters covered all mental health services including medication management, mental health, substance abuse and mental health-related homeless, case management, vocational, and justice-related in-person treatments that occurred during outpatient, inpatient, residential, day-treatment, or during home visits. Contact with each service was examined in thirty-day increments up to 180 days. The amount of treatment received was also assessed. Length of inpatient stays was measured in bed days, and the mean number of in-person outpatient encounters for each time period was also calculated, allowing for multiple encounters per day.

Statistical Analysis

Statistical analysis included chi-square (χ²) tests of difference for categorical variables and Student's t-test (t) for continuous variables. Because of the large number of comparisons, a p-value of <0.01 was considered a significant difference. SAS Version 9.3 was used for all analyses.

Results

Subject Population

In calendar year 2010, the VCL received 102,720 calls. Referrals were made for 21,130 (20.6%) calls. Calls from 13,444 callers (64% of referrals) met inclusion criteria. Callers with prior VHA use made up 91% (n=12,265) of the subject population, with 9% (n=1,179) having no prior use.

Demographic Characteristics

Males comprised 88% of callers with prior VHA use and 89% of callers without prior use, suggesting no difference for gender between groups (Table 1). The overall mean (SD) age for those with prior VHA use was 48 (13.45) years, compared to 42 (14.31) years for those without prior use. Categorically, callers with prior VHA use were more likely to be middle aged and callers without prior use were more likely to be younger.

Table 1. Caller demographics for callers with and without prior VHA use ^a.

	With Prior Use (N=12,147) N (%)	Without Prior Use (N=1,170) N (%)	χ2 or t	p
Male	10,651 (87.68)	1,042 (89.06)	1.89	0.1696
18-29	1,540 (12.68)	316 (27.01)	182.72	<0.0001*
30-39	1,684 (13.86)	214 (18.29)	17.11	<0.0001*
40-49	2,520 (20.75)	237 (20.26)	0.16	0.6931
50-59	3,807 (31.34)	260 (22.22)	41.83	<0.0001*
60-69	2,148 (17.68)	114 (9.74)	47.71	<0.0001*
70-79	315 (2.59)	20 (1.71)	3.40	0.0652
80+	133 (1.09)	9 (0.77)	1.073	0.3003
Mean Age (SD)	48 (13.45)	42 (14.31)	14.82	<0.0001*

^aN=143 cases of either missing or unavailable data

^*significant at p<0.01

Primary Diagnoses

For primary diagnoses at first service contact, callers with prior VHA use were more likely to receive mental health or substance use disorder diagnoses (Table 2). Specifically, they were more likely to have diagnoses of post-traumatic stress disorder (PTSD), major depression, bipolar disorder, schizophrenia, and substance use disorders.

Table 2. Primary diagnoses at first service use following VCL referral.

	With Prior Use (N=12,147) N (%)	Without Prior Use (N=1,170) N (%)	χ2	p
Suicide and Suicidal Ideation	476 (3.92)	72 (6.15)	13.51	0.0002*
Any Mental Health/Substance Use Disorder	7,343 (60.45)	555 (47.44)	74.91	<0.0001*
Any Mental Health	6,410 (52.77)	505 (43.16)	39.46	<0.0001*
Depression	2,832 (23.24)	264 (22.56)	0.27	0.6006
Major Depression	945 (7.78)	66 (5.64)	6.96	0.0083*
Other Depression	1,878 (15.46)	798 (16.92)	1.71	0.1878
Bipolar	566 (4.66)	12 (1.79)	20.79	<0.0001*
Post-Traumatic Stress Disorder	1,394 (11.48)	54 (4.62)	51.83	<0.0001*
Other Anxiety	353 (2.91)	29 (2.48)	0.70	0.4029
Schizophrenia	363 (2.99)	6 (0.51)	24.28	<0.0001*
Other Mental Health-Related	911 (7.50)	131 (11.20)	20.22	<0.0001*
Any Substance Use Disorder	933 (7.68)	50 (4.27)	18.12	<0.0001*
Alcohol Use Disorder	669 (5.51)	38 (3.25)	10.84	0.0010*
Other Substance Use Disorder	264 (2.17)	12 (1.03)	6.93	0.0085*
Factors Influencing Health Status and Contact with Health Services	2,897 (23.85)	382 (32.65)	44.53	<0.0001*
Persons Encountering Health Services in Other Circumstances	2,471 (20.34)	317 (27.09)	29.39	<0.0001*
Other Persons Seeking Consultation	1,448 (11.92)	124 (10.60)	1.79	0.1806
Housing, Household & Economic Circumstances	448 (3.69)	94 (8.03)	51.63	<0.0001*
Other Psychosocial, excluding Suicidal Ideation	575 (4.73)	99 (8.46)	30.87	<0.0001*
Other Factors Influencing Health Status and Contact with Health Services	426 (3.51)	65 (5.56)	12.61	0.0004*
Non-Mental Health/Substance Use/Psychosocial	1,431 (11.78)	161 (13.76)	3.98	0.0462

^*significant at p<0.01

The group without prior VHA use was more likely to receive suicide-related diagnoses and more likely to receive some “other” mental health-related diagnoses (e.g., adjustment disorder, unspecified non-psychotic disorders). They were also more likely to receive diagnoses associated with other factors influencing health status and contact with health services, including housing, household, and economic circumstances, other psychosocial and administrative reasons, and other factors such as suspected mental conditions, aftercare, and depression screening.

Service Utilization

Within seven days of the referral, 91% of callers with prior VHA use and 71% without prior use presented for in-person VHA care (Table 3). Callers with prior VHA use were more likely to present for care the same day as their index referral (44% with prior use vs. 25% without prior use), and those without prior use were more likely to present for care after 15 days.

Table 3. Days to first VHA service utilization for callers with and without prior VHA use ^a.

Timing	With Prior Use (N=12,147) N (%)	Without Prior Use (N=1,170) N (%)	χ2	p
Same Day b	5,372 (44.22)	296 (25.3)	156.36	<0.0001*
1-7 Days	5,750 (47.34)	537 (45.9)	0.89	0.3463
8-14 Days	391 (3.22)	50 (4.27)	3.71	0.0542
15-21 Days	154 (1.27)	36 (3.08)	24.84	<0.0001*
22-30 Days	106 (0.87)	29 (2.48)	27.43	<0.0001*
30+ Days	374 (3.08)	222 (18.97)	630.72	<0.0001*

^aN=127 cases of either missing or unavailable data

^bSame day could include use that occurred before the index call

^*significant at p<0.01

Because callers without VHA prior use were more likely to receive suicide-related diagnoses at first service contact and more likely to receive care after 15 days, post hoc analyses compared time to contact between callers in callers who received suicide-related diagnoses (Table 4). There was a non-significant trend such that a greater percentage of those with prior VHA use received care on the same day as their call than those without prior use. However, it is not known whether this treatment contact occurred before or after the call. No other differences were observed as 96% of callers without VHA prior use made contact within a week of their call compared to 97% with prior use.

Table 4. Days to first VHA service utilization for callers with a suicide-related diagnosis.

Timing	With Prior Use (N=476) N (%)	Without Prior Use (N=72) N (%)	X2	p
Same Day a	219 (46.01)	25 (34.72)	3.23	0.0725
1-7 Days	245 (51.47)	44 (61.11)	2.33	0.1267
8-14 Days	5 (1.05)	2 (2.78)	1.48	0.2238
15-21 Days	1 (0.21)	0 (0.00)	0.13	>0.9999
22-30 Days	3 (0.63)	0 (0.00)	0.35	>0.9999
More than 30 Days	3 (0.63)	1 (1.39)	0.50	0.4813

^aSame day could include use that occurred before the index call

Few differences were observed between groups for out- and in-patient service use. For outpatient services, callers with prior VHA use had a greater mean number of in-person outpatient mental health encounters in the first 30 days following the referral (0-30 days, mean [SD] = 4.4 [5.4] with prior use vs. 3.6 [4.7] without prior use), t(13.19) = 4.10, p<0.01).

Discussion

This study compared two groups of VCL callers who received an SPC referral for VHA care; those with and without prior VHA service use in the five years preceding their referral. Approximately 91% of callers had a history of prior VHA service use, indicating that the VCL is most frequently used by Veterans engaged in VHA care, potentially as a complement to other VHA services. This finding is not surprising as Veteran callers already receiving VHA care would be expected to be more likely to informed of an utilize additional VHA care than those who were previously unaware or unwilling to receive VHA services. This finding does, however, highlight the importance of the VCL as an active component of VHA care, providing an additional source of support, access point for new services, or immediate resource when other services are unavailable. The remaining 9% had no prior VHA use, and their call to the VCL initiated their first in-person encounter with a VHA provider within at minimum five years. To increase use of the VCL among those with and especially without prior VHA use, efforts are being made to increase awareness of this resource and study the effectiveness of related media campaigns implemented in support of this effort (Karras et al., 2013).

Significant demographic and diagnostic differences were observed between the two groups. Callers with prior VHA use tended to be older than those without prior use indicating that there are important demographic differences between the two groups (Department of Veterans Affairs, 2009). Those with prior VHA use were also more likely to have a mental health diagnosis at first contact and may use the VCL to address emerging problems when regular mental healthcare providers are unavailable or to initiate new treatment. For this group, having 24-hour access to VCL responders may meet mental healthcare needs that are not met by other VHA services, such as crisis resolution. Callers without prior VHA use were younger than those with prior use. Although this group was comparatively small, the number of service members returning from combat zones and increasing rates of suicide in the military suggest that many of them may be at increased suicide risk and may benefit from the crisis resolution and linkage to VHA services (Kuehn, 2009). Callers without prior VHA use were more likely to receive suicide-related primary diagnoses at first contact than those with prior use. This finding may be a result of diagnostic practices at first contact as suicidal ideation and behavior in callers with prior VHA use may be considered symptoms of already documented psychiatric disorders such as depression. Callers without prior VHA use were more likely to report non-specific mental-health or stress-related problems (e.g., adjustment disorder or unspecified mental health problems), and psychosocial problems (e.g., housing, economic). They may be less likely than callers with prior VHA use to conceptualize their problems as psychological and may require a wider range of healthcare and psychosocial referrals.

Overall, 91% of callers with history of prior VHA service use engaged in VHA care within a week of the referral, indicating that in-person contact was made quickly. Those with prior VHA use were more likely to present the same day as their referral and may have seen a provider before the call, already had an appointment scheduled, or were seen immediately after the call. In the first month (i.e., 0-30 days) following the referral, callers with prior VHA use had more in-person outpatient mental health encounters, which is consistent with the increased prevalence of mental health diagnoses in this group. In contrast, 71% of callers without prior VHA use had an in-person visit within seven days of the referral, and they were also more likely to utilize VHA services after 15 days. There are a number of possible causes for these delays including enrollment, eligibility, or scheduling issues, difficulty establishing follow-up contact with callers, challenges negotiating caller preferences and schedules, caller ambivalence about receiving care, or having fewer problems requiring immediate attention. Importantly, this delay did not extend to callers without prior VHA use who received suicide-related diagnoses, 96% of whom made contact within a week of the call.

These findings support current VCL policy which instructs responders to attend to callers' current distress, assess their suicidal thoughts and behaviors, and provide SPC referrals to address their mental and physical healthcare needs. Comparisons of diagnoses and patterns of service use suggest that the VCL may serve an important therapeutic function for callers with prior VHA use such as improving callers' ability to cope with crises; an effect that may be independent of receipt of in-person clinical services. Results also suggest that callers without prior VHA use may be more likely to conceptualize their problems as stress-related, physical, or psychosocial, and require a broader range of referrals from SPCs. Referrals to such services may engage these callers in VHA care, provide tangible relief (i.e., housing, food), reduce associated psychiatric symptoms, and reduce their risk for suicide (Kyle & Dunn, 2008). These findings also highlight the crucial role of SPCs in helping callers without prior VHA use navigate enrollment and scheduling issues, and address treatment engagement concerns.

This study was not without limitations. To obtain information on diagnoses and treatment engagement, the analytical sample was limited to callers who provided identifying information and accepted an SPC referral. Calls from Veterans who were not offered or did not accept an SPC referral were excluded from the analyses. Treatment utilization did not assess Vet Centers or the community treatment as they are not recorded in Veterans' medical records. Suicide-related diagnoses were used as an indicator of suicide-related problems, which clearly underestimated the prevalence of such phenomena. Information on time and date of clinical service and VCL referral needed to determine the order of call and same-day visits is available from VA administrative data, but it was not available in the dataset and it is unclear whether clinicians reliably entered the time that their sessions took place. There may also have been differences in the consistency with which responders offered referrals and entered them into VHA medical records, which could limit the generalizability of these findings.

Future research should studies should examine factors associated with the offer and acceptance of a referral in these two groups. Such studies may explain why some callers do not accept referrals and how VCL responders can facilitate their doing so. To better understand differences in engagement between the groups, studies should also examined causes in the delay of face-to-face contact in callers without prior VHA use. Future analyses should evaluate characteristics of same-day service following calls and consider characteristics such as reasons for calling, type of service utilized, and treatment history. Subsequent studies should examine the link between the crisis line calls and follow-up treatment to ensure that treatment addresses related problems. In callers with prior VHA use, assessment of prior mental healthcare utilization may be helpful in determining the effectiveness of the VCL in initiating mental healthcare. Use of silent monitoring may provide an opportunity to better understand the variance between responders in call log documentation, and may also provide information on how the VCL can be used therapeutically with different groups of callers.

Abbreviations

VHA

Veterans Health Administration

SPC

Suicide Prevention Coordinator

VCL

Veterans Crisis Line

VA

U.S. Department of Veterans Affairs

ICD-9

International Classification of Diseases, 9th Revision