Research Articles
Associations Between the Department of Veterans Affairs’ Suicide Prevention Campaign and Calls to Related Crisis Lines
Robert M. Bossarte, PhDa,b Elizabeth Karras, PhDb,c Naiji Lu, PhDc,d Xin Tu, PhDb,c,d Brady Stephens, MSc John Draper, PhDe Janet E. Kemp, RN, PhDf
ABSTRACT Objective. The Transit Authority Suicide Prevention (TASP) campaign was launched by the Department of Veterans Affairs (VA) in a limited number of U.S. cities to promote the use of crisis lines among veterans of military service. Methods. We obtained the daily number of calls to the VCL and National Suicide Prevention Lifeline (NSPL) for six implementation cities (where the campaign was active) and four control cities (where there was no TASP campaign messaging) for a 14-month period. To identify changes in call volume associated with campaign implementation, VCL and NSPL daily call counts for three time periods of equal length (pre-campaign, during campaign, and postcampaign) were modeled using a Poisson log-linear regression with inference based on the generalized estimating equations. Results. Statistically significant increases in calls to both the VCL and the NSPL were reported during the TASP campaign in implementation cities, but were not reported in control cities during or following the campaign. Secondary outcome measures were also reported for the VCL and included the percentage of callers who are veterans, and calls resulting in a rescue during the study period. Conclusions. Results from this study reveal some promise for suicide prevention messaging to promote the use of telephone crisis services and contribute to an emerging area of research examining the effects of campaigns on help seeking.
Department of Veterans Affairs, Office of Public Health, Epidemiology Program, Post-Deployment Health, Washington, DC
a
University of Rochester, Department of Psychiatry, Rochester, NY
b
Department of Veterans Affairs, VISN 2 Center of Excellence for Suicide Prevention, Canandaigua, NY
c
University of Rochester, Department of Biostatistics and Computational Biology, Rochester, NY
d
National Suicide Prevention Lifeline, Link2Health Solutions, Inc., New York, NY
e
Department of Veterans Affairs, Suicide Prevention and Community Engagement, Washington, DC
f
Address correspondence to: Elizabeth Karras, PhD, Department of Veterans Affairs, VISN 2 Center of Excellence for Suicide Prevention, 400 Fort Hill Ave., Canandaigua, NY 14424; tel. 585-393-7236; fax 585-393-7985; e-mail .
516
Public Health Reports / November–December 2014 / Volume 129
Promoting Help Seeking 517
Public communication campaigns have demonstrated efficacy for disseminating strategic messages that address risk and encourage healthy behaviors.1,2 Such communication efforts are regularly employed to target a range of health issues (e.g., exercise/diet, contraception use, and cancer screening),3–5 and have recently drawn attention for their potential in the 2012 National Strategy for Suicide Prevention.6 Although this type of messaging is prevalent in public health approaches to suicide prevention,7 empirical evaluation of it is rare. Limited research suggests that suicide prevention campaigns may be associated with modest, short-term improvements in mental health-related knowledge and attitudes toward treatment;8 however, their impact on help-seeking behaviors is unclear. To date, several studies have attempted to identify associations between communication campaigns and help-seeking behavior, such as calls to crisis lines, an important suicide prevention intervention tool that provides immediate, quality care. These studies have yielded conflicting results in that some have reported statistically significant increases in overall call volume during campaign implementation,9,10 while others have found no change.11 Given that timely help seeking may mitigate the risk for developing suicidal behaviors,12 further investigation of the role campaigns play in promoting the use of crisis lines, a readily accessible service, is needed. Crisis line promotion may be beneficial to U.S. veterans—who may be at increased risk for suicide13,14— by providing a tangible resource and instruction for accessing help. Several reports have underscored the importance of using public messaging to support viable help seeking among this population;15,16 however, little research has been published on such efforts. We addressed this gap by assessing the Transit Authority Suicide Prevention (TASP) campaign, which was launched by the Department of Veterans Affairs (VA) to promote crisis line use. Specifically, we aimed to identify changes in call volume to the Veterans Crisis Line (VCL) and the National Suicide Prevention Lifeline (NSPL) associated with the TASP campaign. THE TASP CAMPAIGN The VCL is a toll-free, confidential service that provides specialized care to veterans and active duty service members in crisis, and resources for their families and friends. The VCL is accessed by calling the NSPL toll-free number (1-800-273-8255) and then pressing “1” to reach the VCL. In 2009, the VA launched the TASP campaign, a series of mass transit-based suicide prevention messages, in a limited number of cities. The
campaign’s primary objective was to increase awareness and promote use of the VCL among U.S. veterans. Suicide prevention placards promoting the 1-800-273-8255 (“press 1”) number were tailored to appear compelling and personally relevant to veterans (e.g., by depicting a soldier and an American flag). They were displayed in each participating city for approximately 12 weeks (mid-August through early November 2009). The same message was used across campaign materials: “It takes the courage and strength of a warrior to ask for help.... If you are in emotional crisis call 1-800-273-8255, ‘press 1’ for veterans.” While it was expected that this campaign would primarily impact calls to the VCL, because both the VCL and NSPL are accessed through a common phone number, we explored call volume patterns for both crisis lines. Cities selected for TASP campaign implementation were chosen based on concerns about elevated suicide rates among veterans in these areas, opportunities for promoting the VCL, demographic characteristics (e.g., estimated veteran population and racial/ethnic composition) of the surrounding population, and the mass-transit infrastructure and use. In total, more than 6,500 advertisements were placed on local buses (interior and exterior), bus shelters, rail cars, and transit stations. METHODS For the purpose of our study, calls to the VCL were identified by record of use of the “press 1” option at the beginning of the call. Callers who did not “press 1” were categorized as having made calls to the NSPL. Data We obtained the number of calls to the VCL and NSPL for area codes within each of the implementation cities as well as for four control cities for a 14-month period beginning on January 5, 2009. Implementation cities included Las Vegas, Nevada; Dallas, Texas; Miami, Florida; Phoenix, Arizona; San Francisco, California; and Los Angeles, California. Control cities, which included Denver, Colorado; New Orleans, Louisiana; San Antonio, Texas; and San Diego, California, were selected based on similarities in region, demographic characteristics, and prevalence of veterans in the general population. Measurement and analysis Several steps were taken to identify changes in call volume associated with TASP campaign implementation. First, we calculated and compared call rates to the VCL and NSPL for the year preceding the study
Public Health Reports / November–December 2014 / Volume 129
518 Research Articles
period (2008) and demographic characteristics of populations in the implementation and control cities to identify differences in pre-campaign call volume and population characteristics that might have contributed to differences in call volume during this study. As a second step, we plotted the four-week moving average of calls to the VCL during a continuous 14-month period (January 5, 2009–February 28, 2010) to identify trends in call volume before, during, and after TASP campaign implementation. We modeled VCL and NSPL daily call counts for three time periods of equal length (pre-campaign: January 5–March 29, 2009; during TASP campaign: August 17–November 8, 2009; and post-campaign: December 7, 2009, to February 28, 2010) using a Poisson log-linear regression model with inference based on the generalized estimating equation (GEE).17,18 These time periods are hereinafter referred to as “campaign periods.” We selected the GEE-based Poisson regression approach because of its ability to accommodate dependence among the callers nested within each city. We used the working independence correlation model for all analyses. Analyses were conducted using SAS® version 9.2.19 Additionally, to glean an understanding of the types of calls received by the VCL before, during, and after the TASP campaign, we calculated secondary outcome measures for both implementation and control cities and plotted the four-week moving average. The VCL primarily receives calls from members of veteran or military communities that include veterans or active duty service members themselves, as well as their families and friends. Secondary outcomes assessed in this study included weekly percentages of (1) calls placed by veterans and (2) calls resulting in a “rescue” or when there is a determination of imminent risk and local emergency services are dispatched. Each category was plotted to identify trends before, during, and after TASP campaign implementation. Calls placed by veterans. We calculated the proportion of weekly calls made by veterans as the number of calls made within the week by individuals who identified themselves as veterans divided by the number of total calls made within the week (i.e., percentage of veterans callers 5 number of veterans calls in week divided by total calls to the VCL in one week). Because this outcome measure was based on the total number of calls, a veteran who called multiple times during the week would be counted multiple times in the weekly figure. Rescues. We calculated the proportion of rescue calls by dividing the number of calls resulting in the dispatch/ coordination of emergency services by the weekly total
number of calls to the VCL (i.e., percentage of rescues 5 calls resulting in dispatch of emergency service in week divided by weekly total number of calls). Because we based this outcome measure on the total number of calls, those who called multiple times during the week would be counted multiple times in the weekly figure. RESULTS Pre-study assessment In 2008, the total rate of calls to the shared 1-800 phone number (both VCL and NSPL) was 4.0 per 1,000 residents in the implementation cities and 5.4 per 1,000 residents in the control cities. Despite the lower overall rate of calls in the TASP campaign implementation cities, the rate of calls to the VCL (i.e., those who “pressed 1”) during the 2008 calendar year was higher in implementation cities than in control cities (7.5 vs. 7.0) (Table 1). When the characteristics of those living in implementation cities were compared with those of residents in control cities, members of the first group were significantly less likely to be veterans (5.8% vs. 10.7%, p,0.001) or white (58.5% vs. 67.3%, p,0.001). Implementation cities were comparable with control cities in other respects, such as population aged $65 years (10.4% vs. 10.6%, p,0.001), proportion married (41.5% vs. 42.8%, p,0.001), and proportion unemployed (4.2% vs. 4.1%, p,0.05) (data not shown). Change in call volume Throughout the study period, January 5, 2009–February 28, 2010, the volume of calls to the VCL was higher in the implementation cities than in the control cities. In the implementation cities, but not in the control cities, marked increases in volume occurred before and after TASP campaign implementation (Figure 1). Additional analyses revealed statistically significant increases in calls to both the VCL and the NSPL during the TASP campaign period that were not present in the control cities or following the campaign (Table 2). Interactions between (1) time and (2) calls during the different campaign periods (pre-campaign, during campaign, and post-campaign) were also examined to assess changes in calls received by each crisis line (Table 2). Analyses identified a statistically significant increase in NSPL call volume over time in implementation cities. Additionally, significant interactions between time and “calls during the TASP campaign period” for both the VCL and NSPL were found in implementation cities. Implementation/control group comparisons. Implementation/control group comparisons of calls to the VCL
Public Health Reports / November–December 2014 / Volume 129
Promoting Help Seeking 519
Table 1. Call rates to the Veterans Crisis Line and National Suicide Prevention Lifeline, in implementation and control cities, before the Transit Authority Suicide Prevention campaign, U.S., 2008 Call rate
Cities Overall Implementation cities Las Vegas, Nevada Dallas, Texas Miami, Florida Phoenix, Arizona San Francisco, California Los Angeles, California Control cities Denver, Colorado New Orleans, Louisiana San Antonio, Texas San Diego, California
Totala (all calls) 4.5 4.0 11.6 3.8 1.5 1.7 11.7 5.3 5.4 8.2 5.2 4.0 3.1
Veterans Crisis Lineb (“press 1”) 7.3 7.5 9.1 6.6 5.6 2.9 21.6 17.0 7.0 12.0 4.8 4.2 6.3
National Suicide Prevention Lifelinec (did not “press 1”) 4.2 3.8 11.9 3.6 1.4 1.5 11.2 4.8 5.3 7.9 5.3 3.9 2.7
a Call rates are per 1,000 overall population (both veterans and non-veterans) during the calendar year. Population estimates are from the U.S. Census Bureau, 2008 American Community Survey.
Call rates are per 1,000 veteran population during the calendar year. Population estimates are from the U.S. Census Bureau, 2008 American Community Survey.
b
Call rates are per 1,000 non-veteran population during the calendar year. Population estimates are from the U.S. Census Bureau, 2008 American Community Survey. c
identified statistically significant increases in calls precampaign and during the campaign in implementation cities. Statistically significant increases in calls to the NSPL were identified in the implementation cities during the pre-campaign, campaign, and post-campaign periods. Statistically significant interactions between (1) time and (2) total call volume were also identified for calls to the NSPL during the pre- and post-campaign periods. No statistically significant interactions were identified for calls to the VCL during any of the three time periods included in implementation/control group comparisons (Table 3). Secondary outcome measures from VCL Trends in the percentage of calls to the VCL placed by veterans in TASP campaign implementation and control cities are shown in Figure 2a. Overall, there was a clear, sustained decrease in the proportion of calls placed by veterans during and following the TASP campaign in implementation cities. There was no change in the proportion of calls to the VCL placed by veterans during the campaign period in control cities. Trends in the percentage of calls to the VCL resulting in a rescue in implementation and control cities are also shown in Figure 2b. There was a general decrease in calls resulting in a rescue during the first half of the TASP campaign implementation cities, with a marked
increase near the end and immediately following the campaign. In control cities, there was no consistent pattern in the proportion of rescues during the campaign. DISCUSSION The primary aim of this research was to examine call volume to crisis lines (VCL and NSPL) and identify changes associated with implementation of the VA’s TASP campaign. Comparisons of call volume were made between implementation cities (where the campaign was active) and control cities (which had no TASP campaign messaging) for a 14-month period before, during, and after the campaign. Results reveal some promise for suicide prevention messaging to promote the use of telephone crisis services. Despite the low intensity of this campaign (i.e., a single form of media with messages passively consumed), a small yet statistically significant increase in calls to both the VCL and NSPL was reported during the campaign in implementation cities. However, the durability of this change is uncertain. For example, calls to the VCL oscillated during the campaign, leading to a distinct drop toward its end. These decreases could be the result of seasonal fluctuations or increased awareness and changes in patterns of care during campaign periods in implementation cities. It is also
Public Health Reports / November–December 2014 / Volume 129
520 Research Articles
Figure 1. Four-week moving average of calls to the Veterans Crisis Line before, during, and after the Transit Authority Suicide Prevention campaign, in implementation and control citiesa: U.S., January 5, 2009–February 28, 2010b
a Study implementation cities included Las Vegas, Nevada; Dallas, Texas; Miami, Florida; Phoenix, Arizona; San Francisco, California; and Los Angeles, California. Control cities were San Diego, California; San Antonio, Texas; Denver, Colorado; and New Orleans, Louisiana.
The Transit Authority Suicide Prevention campaign was launched by the Department of Veterans Affairs from mid-August through early November 2009 to promote the use of crisis lines among veterans of military service.
b
possible that the observed increases in call volume in these areas were associated with related efforts to promote help seeking among veterans living there (i.e., state, local, or regionally initiated suicide prevention programs). However, unlike the TASP campaign, these related efforts were not uniformly distributed across implementation cities. This variability is not unusual when examining relationships between public messaging (i.e., campaigns) and health behavior change. Communication programs often take time to find their place with their audience, and it can take several months or even years to achieve their full impact.20 There may be some evidence of this phenomena in our study data, as a striking increase in calls to the VCL occurred a few months after the
completion of the TASP campaign that continued through the end of our study. Individuals within an audience may also adopt the recommended behavior (i.e., call the crisis line) at different paces. As such, persistent messaging efforts may be required during long periods of time to motivate and reinforce the desired behavior. Although the TASP campaign messaging was intended to promote the VCL, statistically significant associations were also identified during the TASP campaign for the NSPL in implementation cities. The largest increase in calls following TASP campaign implementation was among those who accessed the NSPL (rather than VCL) during this period. While the TASP campaign targeted veterans, the locations
Public Health Reports / November–December 2014 / Volume 129
Public Health Reports / November–December 2014 / Volume 129
2.589 0.772 0.278 0.002 20.004 20.001
Intercept Calls during campaignd Calls post-campaignd Time Time*during campaignd Time*post-campaignd ,0.001 ,0.001 0.170 0.117 0.013 0.853
P-valuec 4.138 0.525 20.163 0.003 20.004 0.001
Coefficient 0.027 0.066 0.099 0.001 0.001 0.001
SE ,0.001 ,0.001 0.098 ,0.001 ,0.001 0.333
P-valuec 2.166 0.369 0.316 0.002 20.004 20.003
Coefficient 0.073 0.196 0.318 0.002 0.002 0.002
SE
,0.001 0.060 0.321 0.257 0.083 0.178
P-valuec
Calls to the Veterans Crisis Line
Study control cities included Denver, Colorado; New Orleans, Louisiana; San Antonio, Texas; and San Diego, California.
SE 5 standard error
d
The reference category for all analyses was calls during the pre-campaign period.
Statistically significant at p,0.05
b
a
c
0.059 0.134 0.202 0.001 0.002 0.002
SE
Calls to the National Suicide Prevention Lifeline
3.790 20.480 20.172 20.001 0.001 0.001
Coefficient
0.034 0.094 0.156 0.001 0.001 0.001
SE
,0.001 0.604 0.268 0.134 0.143 0.189
P-valuec
Calls to the National Suicide Prevention Lifeline
Control citiesb
Study implementation cities included Las Vegas, Nevada; Dallas, Texas; Miami, Florida; Phoenix, Arizona; San Francisco, California; and Los Angeles, California.
Coefficient
Variable
Calls to the Veterans Crisis Line
Implementation citiesa
Table 2. Regression analyses of daily call volume to the Veterans Crisis Line and National Suicide Prevention Lifeline before, during, and after the Transit Authority Suicide Prevention campaign, for implementation and control cities: U.S., January 5, 2009–February 28, 2010
Public Health Reports / November–December 2014 / Volume 129 0.034 0.043 0.001 0.001
3.790 0.349 20.001 0.004 ,0.001
20.001
3.741 0.922 ,0.001
0.000
0.826 20.002
2.534
Coefficient
0.001
0.087 0.106 0.001
0.002
0.219 0.001
0.182
SE
Study control cities included Denver, Colorado; New Orleans, Louisiana; San Antonio, Texas; and San Diego, California.
SE 5 standard error
TASP 5 Transit Authority Suicide Prevention
VCL 5 Veterans Crisis Line
d
The reference category for all analyses was calls in the control cities.
Statistically significant at p,0.05
b
a
c
0.920
0.002
,0.001 ,0.001 ,0.001 0.134
,0.001 0.257
0.094 0.002
0.423 0.002
,0.001
P-valuec
0.073
SE
2.166
Coefficient
0.085
,0.001 ,0.001 0.573
0.998
,0.001 0.188
,0.001
P-valuec
August 17–November 8, 2009 During TASP campaignc
0.003
3.617 0.358 ,0.001
0.003
0.385 20.001
2.482
Coefficient
0.001
0.152 0.179 0.001
0.002
0.365 0.002
0.310
SE
0.001
,0.001 0.046 0.705
0.116
0.291 0.440
,0.001
P-valuec
December 7, 2009–February 28, 2010 Post-campaign
Study implementation cities included Las Vegas, Nevada; Dallas, Texas; Miami, Florida; Phoenix, Arizona; San Francisco, California; and Los Angeles, California.
VCL Intercept Calls to the VCL (implementation cities)d Time Calls to VCL*time (implementation cities)d National Suicide Prevention Lifeline Intercept Total calls (implementation cities)d Time Total calls*time (implementation cities)d
Variable
January 5–March 29, 2009 Pre-campaign
Implementationa/controlb cities comparisons
Table 3. Implementation and control cities comparisons of daily call volume to the VCL and National Suicide Prevention Lifeline before, during, and after the TASP campaign: U.S., January 5, 2009–February 28, 2010
Promoting Help Seeking 523
and mechanisms selected for the display of messages were not specific to veteran populations. An increase in calls associated with message exposure may also have been extended to non-veterans in these areas. Not all veterans may have felt comfortable seeking assistance through the VA (i.e., the “press 1” option) and may have instead chosen to use the NSPL. Therefore, it is also possible that the observed post-campaign increase in calls to the NSPL was, in part, a function of increased call volume from veterans who did not use the “press 1” option. Assessment of secondary outcomes provides additional evidence of the TASP campaign’s impact on use of the VCL. A decrease in the proportion of calls
placed by veterans in TASP campaign implementation cities suggests use by a broader population of callers. A marked increase in rescues during the end stages and immediately following the TASP campaign may also be indicative of increased awareness of the crisis line. Limitations This study was subject to several limitations. Due to the ecological nature of the study, it was not possible to estimate direct associations between campaign exposure and crisis line use. The observed increase in calls to the VCL prior to TASP campaign implementation may be the result of some other unobserved factor. It should be noted that the TASP campaign was part
Figure 2a. Four-week moving average of percentage of calls to the Veterans Crisis Line placed by veteransa before, during, and after the Transit Authority Suicide Prevention campaign, in implementation and control citiesb: U.S., June 7, 2009–February 27, 2010
a The percentage of weekly calls to the Veterans Crisis Line made by veterans was calculated as the number of calls made within the week by individuals who identified themselves as veterans divided by the number of total calls made within the week. Veterans who called multiple times during the week were counted multiple times in the weekly figure.
Study implementation cities included Las Vegas, Nevada; Dallas, Texas; Miami, Florida; Phoenix, Arizona; San Francisco, California; and Los Angeles, California. The control cities were San Diego, California; San Antonio, Texas; Denver, Colorado; and New Orleans, Louisiana.
b
Public Health Reports / November–December 2014 / Volume 129
524 Research Articles
Figure 2b. Four-week moving averagea of percentage of calls to the Veterans Crisis Line resulting in the dispatch of emergency services (rescue) before, during, and after the Transit Authority Suicide Prevention campaign, in implementation and control citiesb: U.S., June 7, 2009–February 27, 2010
a The proportion of rescue calls was calculated by dividing the number of calls resulting in the dispatch of emergency services in the week by the weekly total number of calls to the Veterans Crisis Line. Because the outcome measure is based on the total number of calls, those who called multiple times during the week would be counted multiple times in the weekly figure.
Study implementation cities included Las Vegas, Nevada; Dallas, Texas; Miami, Florida; Phoenix, Arizona; San Francisco, California; and Los Angeles, California. The control cities were San Diego, California; San Antonio, Texas; Denver, Colorado; and New Orleans, Louisiana.
b
of a multifaceted effort to address suicide risk among veterans, and it is possible that complementary suicide prevention activities contributed to an increase in calls to crisis lines in these areas. Additionally, national attention and media reports about the VA’s Suicide Prevention Program and VCL may have increased awareness. We did not measure these related efforts in this study. However, complementary activities that may have contributed to an increase in calls to crisis lines would also have been available in control cities and would be expected to have a similar influence on call volume in these areas. Implementation cities were selected, in part, because of reports of increased suicide among veterans in these
areas. The high baseline prevalence of suicide risk may have contributed to a greater number of calls to crisis lines. The TASP campaign used mass-transit space to advertise the VCL, and message exposure may have been limited to those who used mass transportation. The proportions of all calls placed by veterans and calls resulting in a rescue were considered independently. However, it is possible that increases in calls to the VCL by family members or friends did not directly result in the identification of veterans at imminent risk for suicide. Finally, it is possible that some TASP campaign advertisements were displayed prior to the early August target date or remained longer than the scheduled 12-week period.
Public Health Reports / November–December 2014 / Volume 129
Promoting Help Seeking 525
Further research is needed to test for the presence of lagged effects, dose response, or increased health service use during periods of media campaign implementation. Analyses of reasons for calling, exposure to media messages, and willingness to access the VCL are needed to fully consider the impact of suicide prevention media campaigns among veterans. Next steps should include reporting evaluation data, when available, from assessments of ongoing VA suicide prevention campaigns.
5. 6.
7. 8. 9.
CONCLUSION This article underscores the important role communication campaigns can play in promoting positive prevention messages, specifically those endorsing the VCL, and contributes to an emerging body of research examining the effects of strategic communications on help seeking.
10.
This research was approved by the Institutional Review Board at the VISN 2 Syracuse Veterans Affairs Medical Center in Syracuse, New York.
13.
REFERENCES
15.
1. Synder LB, Hamilton MA, Mitchell EW, Kiwanuka-Tondo J, FlemingMilici F, Proctor D. A meta-analysis of the effect of mediated health communication campaigns on behavior change in the United States. J Health Commun 2004;9 Suppl 1:71-96. 2. Noar SM. A 10-year retrospective of research in health mass media campaigns: where do we go from here? J Health Commun 2006;11:21-42. 3. Huhman ME, Potter LD, Duke JC, Judkins DR, Heitzler CD, Wong FL. Evaluation of a national physical activity intervention for children: VERB campaign, 2002–2004. Am J Prev Med 2007;32:38-43. 4. Valente TW, Saba WP. Campaign exposure and interpersonal
11. 12.
14.
16. 17. 18. 19. 20.
communication as factors in contraceptive use in Bolivia. J Health Commun 2001;6:303-22. Anderson JO, Mullins RM, Siahpush M, Spittal MJ, Wakefield M. Mass media campaign improves cervical screening across all socioeconomic groups. Health Educ Res 2009;24:867-75. Department of Health and Human Services (US), Office of the Surgeon General. 2012 national strategy for suicide prevention: goals and objectives for action: a report of the U.S. Surgeon General of the National Action Alliance for Suicide Prevention. Washington: HHS; 2012. Chambers DA, Pearson JL, Lubell K, Brandon S, O’Brien K, Zinn J. The science of public messages for suicide prevention: a workshop summary. Suicide Life Threat Behav 2005;35:134-45. Dumesnil H, Verger P. Public awareness campaigns about depression and suicide: a review. Psychiatr Serv 2009;60:1203-13. Jenner E, Jenner LW, Matthews-Sterling M, Butts JK, Williams TE. Awareness effects of a youth suicide prevention media campaign in Louisiana. Suicide Life Threat Behav 2010;40:394-406. Oliver RJ, Spilsbury JC, Osiecki SS, Denihan WM, Zureick JL, Friedman S. Brief report: preliminary results of a suicide awareness mass media campaign in Cuyahoga County, Ohio. Suicide Life Threat Behav 2008;38:245-9. Daigle M, Beausoleil L, Brisoux J, Raymond S, Charbonneau L, Desaulniers J. Reaching suicidal people with media campaigns: new challenges for a new century. Crisis 2006;27:172-80. Wilson CJ, Bushnell JA, Caputi P. Early access and help seeking: practice implications and new initiatives. Early Interv Psychiatry 2011;5 Suppl 1:34-9. Kang HK, Bullman TA. Risk of suicide among US veterans after returning from the Iraq or Afghanistan war zones. JAMA 2008;300: 652-3. McCarthy JF, Valenstein M, Kim HM, Ilgen M, Zivin K, Blow FC. Suicide mortality among patients receiving care in the Veterans Health Administration health system. Am J Epidemiol 2009;169:1033-8. Langford L, Litts D, Pearson JL. Using science to improve communications about suicide among military and veteran populations: looking for a few good messages. Am J Public Health 2013;103:31-8. Kemp J, Bossarte R. Suicide data report, 2012. Washington: Department of Veterans Affairs (US), Mental Health Services, Suicide Prevention Program; 2013. Kowalski J, Tu XM. Modern applied U-statistics. New York: Wiley; 2007. Lian KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986;73:13-22. SAS Institute, Inc. SAS®: Version 9.2 for Windows. Cary (NC): SAS Institute, Inc.; 2010. Hornik RC. Public health communication: evidence for behavior change. Mahwah (NJ): Lawrence Erlbaum Associates; 2002.
Public Health Reports / November–December 2014 / Volume 129
Associations Between the Department of Veterans Affairs’ Suicide Prevention Campaign and Calls to Related Crisis Lines
Robert M. Bossarte, PhDa,b Elizabeth Karras, PhDb,c Naiji Lu, PhDc,d Xin Tu, PhDb,c,d Brady Stephens, MSc John Draper, PhDe Janet E. Kemp, RN, PhDf
ABSTRACT Objective. The Transit Authority Suicide Prevention (TASP) campaign was launched by the Department of Veterans Affairs (VA) in a limited number of U.S. cities to promote the use of crisis lines among veterans of military service. Methods. We obtained the daily number of calls to the VCL and National Suicide Prevention Lifeline (NSPL) for six implementation cities (where the campaign was active) and four control cities (where there was no TASP campaign messaging) for a 14-month period. To identify changes in call volume associated with campaign implementation, VCL and NSPL daily call counts for three time periods of equal length (pre-campaign, during campaign, and postcampaign) were modeled using a Poisson log-linear regression with inference based on the generalized estimating equations. Results. Statistically significant increases in calls to both the VCL and the NSPL were reported during the TASP campaign in implementation cities, but were not reported in control cities during or following the campaign. Secondary outcome measures were also reported for the VCL and included the percentage of callers who are veterans, and calls resulting in a rescue during the study period. Conclusions. Results from this study reveal some promise for suicide prevention messaging to promote the use of telephone crisis services and contribute to an emerging area of research examining the effects of campaigns on help seeking.
Department of Veterans Affairs, Office of Public Health, Epidemiology Program, Post-Deployment Health, Washington, DC
a
University of Rochester, Department of Psychiatry, Rochester, NY
b
Department of Veterans Affairs, VISN 2 Center of Excellence for Suicide Prevention, Canandaigua, NY
c
University of Rochester, Department of Biostatistics and Computational Biology, Rochester, NY
d
National Suicide Prevention Lifeline, Link2Health Solutions, Inc., New York, NY
e
Department of Veterans Affairs, Suicide Prevention and Community Engagement, Washington, DC
f
Address correspondence to: Elizabeth Karras, PhD, Department of Veterans Affairs, VISN 2 Center of Excellence for Suicide Prevention, 400 Fort Hill Ave., Canandaigua, NY 14424; tel. 585-393-7236; fax 585-393-7985; e-mail .
516
Public Health Reports / November–December 2014 / Volume 129
Promoting Help Seeking 517
Public communication campaigns have demonstrated efficacy for disseminating strategic messages that address risk and encourage healthy behaviors.1,2 Such communication efforts are regularly employed to target a range of health issues (e.g., exercise/diet, contraception use, and cancer screening),3–5 and have recently drawn attention for their potential in the 2012 National Strategy for Suicide Prevention.6 Although this type of messaging is prevalent in public health approaches to suicide prevention,7 empirical evaluation of it is rare. Limited research suggests that suicide prevention campaigns may be associated with modest, short-term improvements in mental health-related knowledge and attitudes toward treatment;8 however, their impact on help-seeking behaviors is unclear. To date, several studies have attempted to identify associations between communication campaigns and help-seeking behavior, such as calls to crisis lines, an important suicide prevention intervention tool that provides immediate, quality care. These studies have yielded conflicting results in that some have reported statistically significant increases in overall call volume during campaign implementation,9,10 while others have found no change.11 Given that timely help seeking may mitigate the risk for developing suicidal behaviors,12 further investigation of the role campaigns play in promoting the use of crisis lines, a readily accessible service, is needed. Crisis line promotion may be beneficial to U.S. veterans—who may be at increased risk for suicide13,14— by providing a tangible resource and instruction for accessing help. Several reports have underscored the importance of using public messaging to support viable help seeking among this population;15,16 however, little research has been published on such efforts. We addressed this gap by assessing the Transit Authority Suicide Prevention (TASP) campaign, which was launched by the Department of Veterans Affairs (VA) to promote crisis line use. Specifically, we aimed to identify changes in call volume to the Veterans Crisis Line (VCL) and the National Suicide Prevention Lifeline (NSPL) associated with the TASP campaign. THE TASP CAMPAIGN The VCL is a toll-free, confidential service that provides specialized care to veterans and active duty service members in crisis, and resources for their families and friends. The VCL is accessed by calling the NSPL toll-free number (1-800-273-8255) and then pressing “1” to reach the VCL. In 2009, the VA launched the TASP campaign, a series of mass transit-based suicide prevention messages, in a limited number of cities. The
campaign’s primary objective was to increase awareness and promote use of the VCL among U.S. veterans. Suicide prevention placards promoting the 1-800-273-8255 (“press 1”) number were tailored to appear compelling and personally relevant to veterans (e.g., by depicting a soldier and an American flag). They were displayed in each participating city for approximately 12 weeks (mid-August through early November 2009). The same message was used across campaign materials: “It takes the courage and strength of a warrior to ask for help.... If you are in emotional crisis call 1-800-273-8255, ‘press 1’ for veterans.” While it was expected that this campaign would primarily impact calls to the VCL, because both the VCL and NSPL are accessed through a common phone number, we explored call volume patterns for both crisis lines. Cities selected for TASP campaign implementation were chosen based on concerns about elevated suicide rates among veterans in these areas, opportunities for promoting the VCL, demographic characteristics (e.g., estimated veteran population and racial/ethnic composition) of the surrounding population, and the mass-transit infrastructure and use. In total, more than 6,500 advertisements were placed on local buses (interior and exterior), bus shelters, rail cars, and transit stations. METHODS For the purpose of our study, calls to the VCL were identified by record of use of the “press 1” option at the beginning of the call. Callers who did not “press 1” were categorized as having made calls to the NSPL. Data We obtained the number of calls to the VCL and NSPL for area codes within each of the implementation cities as well as for four control cities for a 14-month period beginning on January 5, 2009. Implementation cities included Las Vegas, Nevada; Dallas, Texas; Miami, Florida; Phoenix, Arizona; San Francisco, California; and Los Angeles, California. Control cities, which included Denver, Colorado; New Orleans, Louisiana; San Antonio, Texas; and San Diego, California, were selected based on similarities in region, demographic characteristics, and prevalence of veterans in the general population. Measurement and analysis Several steps were taken to identify changes in call volume associated with TASP campaign implementation. First, we calculated and compared call rates to the VCL and NSPL for the year preceding the study
Public Health Reports / November–December 2014 / Volume 129
518 Research Articles
period (2008) and demographic characteristics of populations in the implementation and control cities to identify differences in pre-campaign call volume and population characteristics that might have contributed to differences in call volume during this study. As a second step, we plotted the four-week moving average of calls to the VCL during a continuous 14-month period (January 5, 2009–February 28, 2010) to identify trends in call volume before, during, and after TASP campaign implementation. We modeled VCL and NSPL daily call counts for three time periods of equal length (pre-campaign: January 5–March 29, 2009; during TASP campaign: August 17–November 8, 2009; and post-campaign: December 7, 2009, to February 28, 2010) using a Poisson log-linear regression model with inference based on the generalized estimating equation (GEE).17,18 These time periods are hereinafter referred to as “campaign periods.” We selected the GEE-based Poisson regression approach because of its ability to accommodate dependence among the callers nested within each city. We used the working independence correlation model for all analyses. Analyses were conducted using SAS® version 9.2.19 Additionally, to glean an understanding of the types of calls received by the VCL before, during, and after the TASP campaign, we calculated secondary outcome measures for both implementation and control cities and plotted the four-week moving average. The VCL primarily receives calls from members of veteran or military communities that include veterans or active duty service members themselves, as well as their families and friends. Secondary outcomes assessed in this study included weekly percentages of (1) calls placed by veterans and (2) calls resulting in a “rescue” or when there is a determination of imminent risk and local emergency services are dispatched. Each category was plotted to identify trends before, during, and after TASP campaign implementation. Calls placed by veterans. We calculated the proportion of weekly calls made by veterans as the number of calls made within the week by individuals who identified themselves as veterans divided by the number of total calls made within the week (i.e., percentage of veterans callers 5 number of veterans calls in week divided by total calls to the VCL in one week). Because this outcome measure was based on the total number of calls, a veteran who called multiple times during the week would be counted multiple times in the weekly figure. Rescues. We calculated the proportion of rescue calls by dividing the number of calls resulting in the dispatch/ coordination of emergency services by the weekly total
number of calls to the VCL (i.e., percentage of rescues 5 calls resulting in dispatch of emergency service in week divided by weekly total number of calls). Because we based this outcome measure on the total number of calls, those who called multiple times during the week would be counted multiple times in the weekly figure. RESULTS Pre-study assessment In 2008, the total rate of calls to the shared 1-800 phone number (both VCL and NSPL) was 4.0 per 1,000 residents in the implementation cities and 5.4 per 1,000 residents in the control cities. Despite the lower overall rate of calls in the TASP campaign implementation cities, the rate of calls to the VCL (i.e., those who “pressed 1”) during the 2008 calendar year was higher in implementation cities than in control cities (7.5 vs. 7.0) (Table 1). When the characteristics of those living in implementation cities were compared with those of residents in control cities, members of the first group were significantly less likely to be veterans (5.8% vs. 10.7%, p,0.001) or white (58.5% vs. 67.3%, p,0.001). Implementation cities were comparable with control cities in other respects, such as population aged $65 years (10.4% vs. 10.6%, p,0.001), proportion married (41.5% vs. 42.8%, p,0.001), and proportion unemployed (4.2% vs. 4.1%, p,0.05) (data not shown). Change in call volume Throughout the study period, January 5, 2009–February 28, 2010, the volume of calls to the VCL was higher in the implementation cities than in the control cities. In the implementation cities, but not in the control cities, marked increases in volume occurred before and after TASP campaign implementation (Figure 1). Additional analyses revealed statistically significant increases in calls to both the VCL and the NSPL during the TASP campaign period that were not present in the control cities or following the campaign (Table 2). Interactions between (1) time and (2) calls during the different campaign periods (pre-campaign, during campaign, and post-campaign) were also examined to assess changes in calls received by each crisis line (Table 2). Analyses identified a statistically significant increase in NSPL call volume over time in implementation cities. Additionally, significant interactions between time and “calls during the TASP campaign period” for both the VCL and NSPL were found in implementation cities. Implementation/control group comparisons. Implementation/control group comparisons of calls to the VCL
Public Health Reports / November–December 2014 / Volume 129
Promoting Help Seeking 519
Table 1. Call rates to the Veterans Crisis Line and National Suicide Prevention Lifeline, in implementation and control cities, before the Transit Authority Suicide Prevention campaign, U.S., 2008 Call rate
Cities Overall Implementation cities Las Vegas, Nevada Dallas, Texas Miami, Florida Phoenix, Arizona San Francisco, California Los Angeles, California Control cities Denver, Colorado New Orleans, Louisiana San Antonio, Texas San Diego, California
Totala (all calls) 4.5 4.0 11.6 3.8 1.5 1.7 11.7 5.3 5.4 8.2 5.2 4.0 3.1
Veterans Crisis Lineb (“press 1”) 7.3 7.5 9.1 6.6 5.6 2.9 21.6 17.0 7.0 12.0 4.8 4.2 6.3
National Suicide Prevention Lifelinec (did not “press 1”) 4.2 3.8 11.9 3.6 1.4 1.5 11.2 4.8 5.3 7.9 5.3 3.9 2.7
a Call rates are per 1,000 overall population (both veterans and non-veterans) during the calendar year. Population estimates are from the U.S. Census Bureau, 2008 American Community Survey.
Call rates are per 1,000 veteran population during the calendar year. Population estimates are from the U.S. Census Bureau, 2008 American Community Survey.
b
Call rates are per 1,000 non-veteran population during the calendar year. Population estimates are from the U.S. Census Bureau, 2008 American Community Survey. c
identified statistically significant increases in calls precampaign and during the campaign in implementation cities. Statistically significant increases in calls to the NSPL were identified in the implementation cities during the pre-campaign, campaign, and post-campaign periods. Statistically significant interactions between (1) time and (2) total call volume were also identified for calls to the NSPL during the pre- and post-campaign periods. No statistically significant interactions were identified for calls to the VCL during any of the three time periods included in implementation/control group comparisons (Table 3). Secondary outcome measures from VCL Trends in the percentage of calls to the VCL placed by veterans in TASP campaign implementation and control cities are shown in Figure 2a. Overall, there was a clear, sustained decrease in the proportion of calls placed by veterans during and following the TASP campaign in implementation cities. There was no change in the proportion of calls to the VCL placed by veterans during the campaign period in control cities. Trends in the percentage of calls to the VCL resulting in a rescue in implementation and control cities are also shown in Figure 2b. There was a general decrease in calls resulting in a rescue during the first half of the TASP campaign implementation cities, with a marked
increase near the end and immediately following the campaign. In control cities, there was no consistent pattern in the proportion of rescues during the campaign. DISCUSSION The primary aim of this research was to examine call volume to crisis lines (VCL and NSPL) and identify changes associated with implementation of the VA’s TASP campaign. Comparisons of call volume were made between implementation cities (where the campaign was active) and control cities (which had no TASP campaign messaging) for a 14-month period before, during, and after the campaign. Results reveal some promise for suicide prevention messaging to promote the use of telephone crisis services. Despite the low intensity of this campaign (i.e., a single form of media with messages passively consumed), a small yet statistically significant increase in calls to both the VCL and NSPL was reported during the campaign in implementation cities. However, the durability of this change is uncertain. For example, calls to the VCL oscillated during the campaign, leading to a distinct drop toward its end. These decreases could be the result of seasonal fluctuations or increased awareness and changes in patterns of care during campaign periods in implementation cities. It is also
Public Health Reports / November–December 2014 / Volume 129
520 Research Articles
Figure 1. Four-week moving average of calls to the Veterans Crisis Line before, during, and after the Transit Authority Suicide Prevention campaign, in implementation and control citiesa: U.S., January 5, 2009–February 28, 2010b
a Study implementation cities included Las Vegas, Nevada; Dallas, Texas; Miami, Florida; Phoenix, Arizona; San Francisco, California; and Los Angeles, California. Control cities were San Diego, California; San Antonio, Texas; Denver, Colorado; and New Orleans, Louisiana.
The Transit Authority Suicide Prevention campaign was launched by the Department of Veterans Affairs from mid-August through early November 2009 to promote the use of crisis lines among veterans of military service.
b
possible that the observed increases in call volume in these areas were associated with related efforts to promote help seeking among veterans living there (i.e., state, local, or regionally initiated suicide prevention programs). However, unlike the TASP campaign, these related efforts were not uniformly distributed across implementation cities. This variability is not unusual when examining relationships between public messaging (i.e., campaigns) and health behavior change. Communication programs often take time to find their place with their audience, and it can take several months or even years to achieve their full impact.20 There may be some evidence of this phenomena in our study data, as a striking increase in calls to the VCL occurred a few months after the
completion of the TASP campaign that continued through the end of our study. Individuals within an audience may also adopt the recommended behavior (i.e., call the crisis line) at different paces. As such, persistent messaging efforts may be required during long periods of time to motivate and reinforce the desired behavior. Although the TASP campaign messaging was intended to promote the VCL, statistically significant associations were also identified during the TASP campaign for the NSPL in implementation cities. The largest increase in calls following TASP campaign implementation was among those who accessed the NSPL (rather than VCL) during this period. While the TASP campaign targeted veterans, the locations
Public Health Reports / November–December 2014 / Volume 129
Public Health Reports / November–December 2014 / Volume 129
2.589 0.772 0.278 0.002 20.004 20.001
Intercept Calls during campaignd Calls post-campaignd Time Time*during campaignd Time*post-campaignd ,0.001 ,0.001 0.170 0.117 0.013 0.853
P-valuec 4.138 0.525 20.163 0.003 20.004 0.001
Coefficient 0.027 0.066 0.099 0.001 0.001 0.001
SE ,0.001 ,0.001 0.098 ,0.001 ,0.001 0.333
P-valuec 2.166 0.369 0.316 0.002 20.004 20.003
Coefficient 0.073 0.196 0.318 0.002 0.002 0.002
SE
,0.001 0.060 0.321 0.257 0.083 0.178
P-valuec
Calls to the Veterans Crisis Line
Study control cities included Denver, Colorado; New Orleans, Louisiana; San Antonio, Texas; and San Diego, California.
SE 5 standard error
d
The reference category for all analyses was calls during the pre-campaign period.
Statistically significant at p,0.05
b
a
c
0.059 0.134 0.202 0.001 0.002 0.002
SE
Calls to the National Suicide Prevention Lifeline
3.790 20.480 20.172 20.001 0.001 0.001
Coefficient
0.034 0.094 0.156 0.001 0.001 0.001
SE
,0.001 0.604 0.268 0.134 0.143 0.189
P-valuec
Calls to the National Suicide Prevention Lifeline
Control citiesb
Study implementation cities included Las Vegas, Nevada; Dallas, Texas; Miami, Florida; Phoenix, Arizona; San Francisco, California; and Los Angeles, California.
Coefficient
Variable
Calls to the Veterans Crisis Line
Implementation citiesa
Table 2. Regression analyses of daily call volume to the Veterans Crisis Line and National Suicide Prevention Lifeline before, during, and after the Transit Authority Suicide Prevention campaign, for implementation and control cities: U.S., January 5, 2009–February 28, 2010
Public Health Reports / November–December 2014 / Volume 129 0.034 0.043 0.001 0.001
3.790 0.349 20.001 0.004 ,0.001
20.001
3.741 0.922 ,0.001
0.000
0.826 20.002
2.534
Coefficient
0.001
0.087 0.106 0.001
0.002
0.219 0.001
0.182
SE
Study control cities included Denver, Colorado; New Orleans, Louisiana; San Antonio, Texas; and San Diego, California.
SE 5 standard error
TASP 5 Transit Authority Suicide Prevention
VCL 5 Veterans Crisis Line
d
The reference category for all analyses was calls in the control cities.
Statistically significant at p,0.05
b
a
c
0.920
0.002
,0.001 ,0.001 ,0.001 0.134
,0.001 0.257
0.094 0.002
0.423 0.002
,0.001
P-valuec
0.073
SE
2.166
Coefficient
0.085
,0.001 ,0.001 0.573
0.998
,0.001 0.188
,0.001
P-valuec
August 17–November 8, 2009 During TASP campaignc
0.003
3.617 0.358 ,0.001
0.003
0.385 20.001
2.482
Coefficient
0.001
0.152 0.179 0.001
0.002
0.365 0.002
0.310
SE
0.001
,0.001 0.046 0.705
0.116
0.291 0.440
,0.001
P-valuec
December 7, 2009–February 28, 2010 Post-campaign
Study implementation cities included Las Vegas, Nevada; Dallas, Texas; Miami, Florida; Phoenix, Arizona; San Francisco, California; and Los Angeles, California.
VCL Intercept Calls to the VCL (implementation cities)d Time Calls to VCL*time (implementation cities)d National Suicide Prevention Lifeline Intercept Total calls (implementation cities)d Time Total calls*time (implementation cities)d
Variable
January 5–March 29, 2009 Pre-campaign
Implementationa/controlb cities comparisons
Table 3. Implementation and control cities comparisons of daily call volume to the VCL and National Suicide Prevention Lifeline before, during, and after the TASP campaign: U.S., January 5, 2009–February 28, 2010
Promoting Help Seeking 523
and mechanisms selected for the display of messages were not specific to veteran populations. An increase in calls associated with message exposure may also have been extended to non-veterans in these areas. Not all veterans may have felt comfortable seeking assistance through the VA (i.e., the “press 1” option) and may have instead chosen to use the NSPL. Therefore, it is also possible that the observed post-campaign increase in calls to the NSPL was, in part, a function of increased call volume from veterans who did not use the “press 1” option. Assessment of secondary outcomes provides additional evidence of the TASP campaign’s impact on use of the VCL. A decrease in the proportion of calls
placed by veterans in TASP campaign implementation cities suggests use by a broader population of callers. A marked increase in rescues during the end stages and immediately following the TASP campaign may also be indicative of increased awareness of the crisis line. Limitations This study was subject to several limitations. Due to the ecological nature of the study, it was not possible to estimate direct associations between campaign exposure and crisis line use. The observed increase in calls to the VCL prior to TASP campaign implementation may be the result of some other unobserved factor. It should be noted that the TASP campaign was part
Figure 2a. Four-week moving average of percentage of calls to the Veterans Crisis Line placed by veteransa before, during, and after the Transit Authority Suicide Prevention campaign, in implementation and control citiesb: U.S., June 7, 2009–February 27, 2010
a The percentage of weekly calls to the Veterans Crisis Line made by veterans was calculated as the number of calls made within the week by individuals who identified themselves as veterans divided by the number of total calls made within the week. Veterans who called multiple times during the week were counted multiple times in the weekly figure.
Study implementation cities included Las Vegas, Nevada; Dallas, Texas; Miami, Florida; Phoenix, Arizona; San Francisco, California; and Los Angeles, California. The control cities were San Diego, California; San Antonio, Texas; Denver, Colorado; and New Orleans, Louisiana.
b
Public Health Reports / November–December 2014 / Volume 129
524 Research Articles
Figure 2b. Four-week moving averagea of percentage of calls to the Veterans Crisis Line resulting in the dispatch of emergency services (rescue) before, during, and after the Transit Authority Suicide Prevention campaign, in implementation and control citiesb: U.S., June 7, 2009–February 27, 2010
a The proportion of rescue calls was calculated by dividing the number of calls resulting in the dispatch of emergency services in the week by the weekly total number of calls to the Veterans Crisis Line. Because the outcome measure is based on the total number of calls, those who called multiple times during the week would be counted multiple times in the weekly figure.
Study implementation cities included Las Vegas, Nevada; Dallas, Texas; Miami, Florida; Phoenix, Arizona; San Francisco, California; and Los Angeles, California. The control cities were San Diego, California; San Antonio, Texas; Denver, Colorado; and New Orleans, Louisiana.
b
of a multifaceted effort to address suicide risk among veterans, and it is possible that complementary suicide prevention activities contributed to an increase in calls to crisis lines in these areas. Additionally, national attention and media reports about the VA’s Suicide Prevention Program and VCL may have increased awareness. We did not measure these related efforts in this study. However, complementary activities that may have contributed to an increase in calls to crisis lines would also have been available in control cities and would be expected to have a similar influence on call volume in these areas. Implementation cities were selected, in part, because of reports of increased suicide among veterans in these
areas. The high baseline prevalence of suicide risk may have contributed to a greater number of calls to crisis lines. The TASP campaign used mass-transit space to advertise the VCL, and message exposure may have been limited to those who used mass transportation. The proportions of all calls placed by veterans and calls resulting in a rescue were considered independently. However, it is possible that increases in calls to the VCL by family members or friends did not directly result in the identification of veterans at imminent risk for suicide. Finally, it is possible that some TASP campaign advertisements were displayed prior to the early August target date or remained longer than the scheduled 12-week period.
Public Health Reports / November–December 2014 / Volume 129
Promoting Help Seeking 525
Further research is needed to test for the presence of lagged effects, dose response, or increased health service use during periods of media campaign implementation. Analyses of reasons for calling, exposure to media messages, and willingness to access the VCL are needed to fully consider the impact of suicide prevention media campaigns among veterans. Next steps should include reporting evaluation data, when available, from assessments of ongoing VA suicide prevention campaigns.
5. 6.
7. 8. 9.
CONCLUSION This article underscores the important role communication campaigns can play in promoting positive prevention messages, specifically those endorsing the VCL, and contributes to an emerging body of research examining the effects of strategic communications on help seeking.
10.
This research was approved by the Institutional Review Board at the VISN 2 Syracuse Veterans Affairs Medical Center in Syracuse, New York.
13.
REFERENCES
15.
1. Synder LB, Hamilton MA, Mitchell EW, Kiwanuka-Tondo J, FlemingMilici F, Proctor D. A meta-analysis of the effect of mediated health communication campaigns on behavior change in the United States. J Health Commun 2004;9 Suppl 1:71-96. 2. Noar SM. A 10-year retrospective of research in health mass media campaigns: where do we go from here? J Health Commun 2006;11:21-42. 3. Huhman ME, Potter LD, Duke JC, Judkins DR, Heitzler CD, Wong FL. Evaluation of a national physical activity intervention for children: VERB campaign, 2002–2004. Am J Prev Med 2007;32:38-43. 4. Valente TW, Saba WP. Campaign exposure and interpersonal
11. 12.
14.
16. 17. 18. 19. 20.
communication as factors in contraceptive use in Bolivia. J Health Commun 2001;6:303-22. Anderson JO, Mullins RM, Siahpush M, Spittal MJ, Wakefield M. Mass media campaign improves cervical screening across all socioeconomic groups. Health Educ Res 2009;24:867-75. Department of Health and Human Services (US), Office of the Surgeon General. 2012 national strategy for suicide prevention: goals and objectives for action: a report of the U.S. Surgeon General of the National Action Alliance for Suicide Prevention. Washington: HHS; 2012. Chambers DA, Pearson JL, Lubell K, Brandon S, O’Brien K, Zinn J. The science of public messages for suicide prevention: a workshop summary. Suicide Life Threat Behav 2005;35:134-45. Dumesnil H, Verger P. Public awareness campaigns about depression and suicide: a review. Psychiatr Serv 2009;60:1203-13. Jenner E, Jenner LW, Matthews-Sterling M, Butts JK, Williams TE. Awareness effects of a youth suicide prevention media campaign in Louisiana. Suicide Life Threat Behav 2010;40:394-406. Oliver RJ, Spilsbury JC, Osiecki SS, Denihan WM, Zureick JL, Friedman S. Brief report: preliminary results of a suicide awareness mass media campaign in Cuyahoga County, Ohio. Suicide Life Threat Behav 2008;38:245-9. Daigle M, Beausoleil L, Brisoux J, Raymond S, Charbonneau L, Desaulniers J. Reaching suicidal people with media campaigns: new challenges for a new century. Crisis 2006;27:172-80. Wilson CJ, Bushnell JA, Caputi P. Early access and help seeking: practice implications and new initiatives. Early Interv Psychiatry 2011;5 Suppl 1:34-9. Kang HK, Bullman TA. Risk of suicide among US veterans after returning from the Iraq or Afghanistan war zones. JAMA 2008;300: 652-3. McCarthy JF, Valenstein M, Kim HM, Ilgen M, Zivin K, Blow FC. Suicide mortality among patients receiving care in the Veterans Health Administration health system. Am J Epidemiol 2009;169:1033-8. Langford L, Litts D, Pearson JL. Using science to improve communications about suicide among military and veteran populations: looking for a few good messages. Am J Public Health 2013;103:31-8. Kemp J, Bossarte R. Suicide data report, 2012. Washington: Department of Veterans Affairs (US), Mental Health Services, Suicide Prevention Program; 2013. Kowalski J, Tu XM. Modern applied U-statistics. New York: Wiley; 2007. Lian KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986;73:13-22. SAS Institute, Inc. SAS®: Version 9.2 for Windows. Cary (NC): SAS Institute, Inc.; 2010. Hornik RC. Public health communication: evidence for behavior change. Mahwah (NJ): Lawrence Erlbaum Associates; 2002.
Public Health Reports / November–December 2014 / Volume 129
