Journal of Traumatic Stress Disorders & Treatment ISSN: 2324-8947

All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.

Research Article, J Trauma Stress Disor Treat Vol: 3 Issue: 1

Recruiting Partners of Veterans with Posttraumatic Stress Disorder for Survey Research: Feasibility, Response Rate, and Representativeness

Alyssa J Mansfield1*, Kim M Schaper1* and Craig S Rosen2,3
1National Center for PTSD Pacific Islands Division, VA Pacific Islands Health Care System, Honolulu, HI, USA
2National Center for PTSD Dissemination & Training Division, VA Palo Alto Health Care System, Menlo Park, CA, USA
3Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Palo Alto, CA, USA
Corresponding author : Alyssa J Mansfield
National Center for PTSD, VA Pacific Islands Health Care System, 3375 Koapaka Street, Suite I-560, Honolulu, HI 96819-5202, USA
Tel: 808-566-1889; Fax: 808-566-1885
E-mail: [email protected]
Received: September 17, 2013 Accepted: November 13, 2013 Published: January 03, 2014
Citation: Mansfield AJ, Schaper KM, Rosen CS (2014) Recruiting Partners of Veterans with Posttraumatic Stress Disorder for Survey Research: Feasibility, Response Rate, and Representativeness. J Trauma Stress Disor Treat 3:1. doi:10.4172/2324-8947.100018


Recruiting Partners of Veterans with Posttraumatic Stress Disorder for Survey Research: Feasibility, Response Rate, and Representativeness

Intimate partners of veterans diagnosed with posttraumatic stress disorder (PTSD) can be instrumental in the veterans’ PTSD recovery. Despite interest from both veterans and their significant others for greater partner involvement in Veterans Affairs (VA) treatment and services, significant challenges exist to engage partners directly for research, education or therapeutic activities. Furthermore, initiatives aimed at reaching a large, representative group of veterans’ intimate partners are unlikely to achieve sufficient sample size through direct contact with veterans or their significant others alone.

Keywords: Posttraumatic stress disorder; Veterans; Intimate partners; Survey methods; Research subject recruitment


Posttraumatic stress disorder; Veterans; Intimate partners; Survey methods; Research subject recruitment


Engaging partners of individuals diagnosed with posttraumatic stress disorder (PTSD) in education, support, and treatment efforts has far-reaching potential benefits for both individuals and couples. Partners who are knowledgeable about PTSD may recognize symptom patterns more readily, facilitate medication and treatment adherence, and can be instrumental in educating other family members about PTSD and its effects [1,2]. Increasing partner knowledge about PTSD symptoms and treatment, and greater support and involvement in treatment processes foster improved communication and relationship quality, as well as meaningful symptom reduction among individuals with PTSD [3,4]. In contrast, relationship dissatisfaction and marital stress have been linked to increased PTSD symptoms and negative PTSD treatment outcomes among patients [5-7], as well as psychological distress and caregiver burden among partners [1,6,8].
Veterans with PTSD and their partners have expressed interest in partner-focused activities such as educational programs, couples’ therapy, and support groups. Many veterans view their PTSD as a source of significant marital conflict and desire greater partner and family involvement in their treatment [9]. Partners—often the primary caretakers of individuals with PTSD—seek opportunities to learn effective coping skills and reduce caregiver burden through participation in individual-, family-, and partner-based PTSD treatment and education programs [1,10,11]. To date, both cognitivebehavioral couple therapy for PTSD and Veterans Affairs (VA) piloted family programing for veterans with PTSD have been well received among veterans, family members, and providers, and demonstrate promise for VA-wide implementation in the future [4,11]. Still, research exploring partners’ understanding of PTSD, involvement in care, and quality of life remains quite limited even asthe prevalence of PTSD increases among veterans.
Despite interest from both veterans and their partners in greater involvement in VA treatment and services, engaging significant others directly to participate in research, education or therapeutic activities can often be challenging. Unlike the Department of Defense, the VA does not traditionally provide beneficiary care for family members. Furthermore, detailed information on a veteran’s spouse or partner is not generally included within the medical record. Communication with partners must, therefore, typically occur through intermittent contact during veterans’ medical appointments. For clinical and educational efforts targeting a relatively small number of participants, such methods may not be prohibitive. For example, past attempts to assess spousal attitudes and behaviors have relied upon flyers placed in clinics serving veterans [1,8,10], recruitment through VA providers during scheduled appointments [11], or have examined marital support and involvement by sampling only veterans [2,9]. Such clinic-based and convenience samples may not be representative of the population of veterans receiving treatment for PTSD within the VA and their significant others, and may disproportionately include veterans receiving regularly scheduled treatment, those living in close proximity to VA facilities, or individuals—partners and veterans— with fewer barriers to regular care (e.g., individuals without job, transportation, or child/family care issues).
For studies requiring a large, representative set of veterans’ intimate partners (e.g., survey research), achieving a sufficient sample size through direct contact alone is unlikely. Using the veteran as an intermediary is one potential workaround and resembles what has been used with other populations. Specifically, recruitment of minor children or vulnerable persons often occurs through a secondary contact where direct contact or informed consent is constrained. For minor children, consent typically requires a parent or guardian as the primary contact. Similarly, research with vulnerable persons whose cognitive state precludes providing informed consent often relies upon a caregiver’s involvement. Though comparable, both examples differ from our desire to recruit partners of veterans with PTSD in that we are aiming to study a group able to consent on their own, but for whom we are unlikely to have interaction or for whom specific contact information may be lacking. Even successful recruitment of partners through veterans in prior work has been limited by the use of clinic-based and convenience methods often yielding small (N< 100), relatively homogeneous samples [1,4,8,10,11].
We utilized the veteran as an intermediary to examine the effects of PTSD on a variety of outcomes among spouses and partners of veterans receiving care for the disorder within the VA system. Our approach forewent in-person interaction in favor of access to a broader clinical and geographic population. Here, we describe our strategy for recruiting a large number of veteran partners for PTSD survey research. Specifically, we discuss methodological approaches, successes and challenges, and describe opportunities to expand on these methods.

Recruitment Methods

The VA Pacific Islands Health Care System (VAPIHCS) spans 4.8 million square miles, serving nearly 130,000 veterans in the Pacific basin, and stretching from the Big Island of Hawaii in the east to Guam in the west [12]. The expansiveness of the VAPIHCS catchment area often restricts in-person study recruitment and participation by limiting non-clinical research staff interaction with potential participants. Reaching family members for PTSD research opportunities presents an even greater challenge given they are not the primary recipient of VA services, and may not accompany the enrolled veteran to VA appointments. A larger, parent study surveyed the intimate partners of VAPIHCS veterans to assess their 1) knowledge, attitudes, and beliefs about PTSD symptoms and treatment; 2) level of participation in VA and non-VA PTSD treatment with their veteran partner; and 3) quality of life domains impacted by their veteran partner’s PTSD. This article will focus on the recruitment strategy used to recruit these partners.

Recruitment Strategy

In the absence of names and other direct contact information for partners, we elected to recruit participants using the veteran as an intermediary. We mailed letters explaining the study to veterans whose electronic medical record data reflected the following inclusion criteria: 1) a marital status of “married” as of May 2011; 2) a PTSD diagnosis code (309.81) from the International Classification of Diseases, 9th Revision (ICD-9) for at least one outpatient medical visit between June 2010 and May 2011, and 3) a home address of record in the VAPIHCS catchment area. The introductory letter briefly described the study and asked each veteran to pass the research information and a postage-paid response card along to their current spouse—or intimate partner, if they were no longer married but were in a long-term, committed relationship—to be mailed back either opting into or out of the study. Along with this letter, we mailed a fact sheet which provided answers to questions such as privacy of responses, restriction of information to the study team, and the minimal risks associated with study participation. We mailed one reminder letter to each veteran for whom a partner response card was not received within 4-6 weeks. For partners opting into the study, we mailed a packet including the survey and instructions, a $20 gift card, and a list of resources for individual and family support. Partners who did not return the survey received up to two reminder letters before data collection ceased. The Institutional Review Board of the VA Pacific Islands Health Care System approved all study procedures and waived the requirements for HIPAA authorization and documentation of informed consent.


We created a brief (4-page) paper-and-pencil instrument in an attempt to better understand how partners of care-seeking veterans diagnosed with PTSD perceived the condition. We assessed the following: demographic information; each partner’s knowledge, beliefs, and attitudes about PTSD, including symptoms based upon clinical diagnosis [13] and related sequelae; beliefs and attitudes about PTSD etiology, treatment, and prognosis; and familiarity with different PTSD treatment modalities. We asked partners about the extent of their involvement with the veteran’s PTSD care and associated barriers, and if they believed having an intimate relationship with someone diagnosed with PTSD had affected a number of quality of life domains (e.g., physical, emotional, financial, interpersonal relationships)—whether positively, negatively, or not at all. Participants were asked to checkmark items with which they agreed or believed to represent their experiences. We also included a space where respondents could submit written comments about their experiences. Surveys were devoid of identifying information for both the partner and veteran, with responses instead linked to the veteran’s non-identifying demographic and clinical datausing a random ID number. It was expected that no more than 15 minutes were needed to complete the survey.

Veteran PTSD Data

Veterans with more severe PTSD or experiencing a greater number of PTSD symptoms may have greater dysfunction in their intimate relationships, and may be less likely to pass study information on to their partners. To test whether the partners of veterans with more severe cases of PTSD were less likely to participate in the survey, we examined each veteran’s self-reported PTSD symptoms using results from the civilian and military versions of the PTSD Checklist [14,15], and the Primary Care PTSD Screen [16] on file. Specifically, in addition to demographic data for all veterans meeting initial study inclusion criteria, we queried veterans’ electronic medical records to obtain PTSD symptom assessment scores and the time interval during which these assessments were completed relative to the timing of the first study mailing to the veteran (i.e., < 30 days prior, 30 days to 6 months prior, 6 to12 months prior, >1 year prior). Assessment score and time interval for all veterans who were invited into the study was included to determine whether the final sample overrepresented veterans with less severe PTSD symptomatology as assessed by the VA. The Institutional Review Board of the VA Pacific Islands Health Care System approved all study procedures and waived the requirement for documentation of informed consent.

Results of Recruitment

Using an initial roster of 1,958 veterans diagnosed with and seeking care for PTSD in the VA Pacific Islands Health Care System as a proxy contact for their significant others resulted in 877 returned response cards: 573 spouses/partners who opted into the study, and 304 who opted out of the study. Of the 571 surveys mailed (two opt-in response cards were rec eived after data collection ceased; Figure 1), 385 were returned without further prompting. Initial reminder letters to partners with outstanding surveys (n=186) yielded an additional 70 returned surveys, and final reminder letters sent to the remaining 116 participants with outstanding surveys yielded 12 additional completed surveys. The 82 surveys returned after mailing any reminder letter— initial or final—represent 44% of the 186 initial reminder letters sent and 14% of the 571 surveys.
Figure 1: Participant Flow.
A total of 471 surveys were returned and contained usable data with two categories of exceptions. Specifically, cases were excluded from further analysis if responses appeared to originate from someone other than the veteran’s intimate partner or if a veteran match could not be found. This yielded a final sample of 455 valid surveys matched to veteran data (Figure 1), 252 of which included qualitative comments. Corresponding study response rates ranged from 23.2 percent to 82.6 percent when considering all phases of sample recruitment and assemblage among veterans and their partners (Table 1).
Table 1: Response rates for all phases of study recruitment.
Table 2 presents characteristics of all veterans sent an introductory letter (hereafter “initial sample”), the subset comprising those matched to the final sample of partner respondents (hereafter “final sample”), and those whose partners did not participate (hereafter “nonresponders”). The final sample differed in statistically significant ways from nonresponders on several factors. On average, the final sample was slightly older, and included lower proportions of partners whose veterans served in more recent (i.e., Persian Gulf, which includes post-9/11 operations in Afghanistan and Iraq; Post- Vietnam) and distant (i.e., Korean War era and earlier) service periods, but a higher proportion of Vietnam era veterans’ partners compared to nonresponders. The final sample also included a greater number of dyads residing in the Hawaiian Islands (82.4% vs. 74.1% among nonresponders). Our concern that veterans reporting less severe PTSD symptoms would be overrepresented in the final sample was not realized. The proportion of veterans with a positive PTSD screen during their VA last assessment was comparable between nonresponders and the final sample (48.9% vs. 50.1%, respectively; not statistically significant), with slightly more veterans from the nonresponder group completing a PTSD symptom assessment in the six months prior to the study compared to the final sample (33.0% vs. 24.4%, respectively).
Table 2: Characteristics of veterans meeting study inclusion criteria and by partner response.


Veterans and their significant others desire greater involvement in PTSD-related education, support, and treatment. Such efforts are likely to enhance couple and family relationship quality, medication and treatment adherence, and meaningful symptom reduction among individuals with PTSD [1-4,9-11]. Yet, efforts to date have struggled to reach a large number of partners outside of limited and irregular interaction during veterans’ clinical appointments. The present study illustrates that recruitment of veterans’ intimate partners for survey research can effectively occur through a veteran intermediary. Specifically, we were able to recruit partners from a large, geographically dispersed sample of veterans—one quite similar to the target population (i.e., married VAPIHCS veterans diagnosed with, and recently treated for, PTSD in the VAPIHCS). VAPIHCS is both an ethnoculturally and geographically unique region compared to the mainland U.S. in that there is no true racial/ethnic majority group, and typically ocean rather than land separates many areas. Nonetheless, our results support the potential to engage a large, representative number of intimate partners by first approaching the veteran and are likely to be replicated in similar studies conducted with veterans and their partners nationally.
Our methods included several limitations. Overall response rates were not high, with fewer than one in four partners of eligible veterans enrolling into the study and returning a valid survey. However, we made successful contact with nearly half of the targeted partners from our initial roster (877/1958, 45%) based on returned response cards. Although some individuals opted out of the study, the overall return rate supports our premise that partners can be contacted through the veteran as an intermediary. We also enrolled a high number of respondents compared to previous research among veterans’ intimate partners. It is possible that greater incentives or additional outreach efforts may have also improved participation. A total of 82 surveys were returned after a reminder letter was sent. Although we cannot conclude that surveys were returned in direct response to our letter (e.g., the survey may have already been in the mail), the outreach may have influenced survey return. In addition, our response rates may not reflect the level of interest or likely participation in partneroriented PTSD education and service initiatives our parent study aimed to inform, which could be lower or higher than our observed response. The survey itself may have also influenced our response rate. Although designed to be brief and simple, the overall subject matter or specific item content could have affected the participant’s decision to complete and return the survey. While we do not know why some participants opted into the study but did not return a completed survey, this number was relatively small (n = 98, 17% of those opting in).
Despite general representativeness from the target population, our study included a disproportionate number of partners of Vietnam veterans. These couples are more likely to be of retirement age and without employment and child-rearing responsibilities paralleling those of younger veterans which may have affected participation. Similarly, it is possible that lower participation from partners of veterans in older cohorts (e.g., pre-Vietnam) may be linked to preoccupation with either partner’s physical health. Our sample also only included veterans whose marital status was listed as “married” at the time of recruitment. This strategy excluded veterans who are in long-term relationships but never got married, veterans whose marital status information is incorrect in the VHA database, or veterans involved in same-sex partnerships. Inclusion of these couples is an important consideration for future studies.
Opportunities for participation in research and education efforts may be relatively scarce within the VAPIHCS compared to other VA health care system catchment areas. This possibility may have contributed to greater numbers of participants in the current study. However, previous studies identified substantial interest in partnerrelated programs and services from veterans’ intimate partners [1,10,11]. Although a higher percentage of significant others opting into the study was desirable, this method was still able to recruit a substantial and representative partner subset among veterans contacted based on our preliminary study criteria. Further, the high percentage of partners who opted into the study and returned a valid, matched survey may reflect a broader desire by partners for PTSDrelated education and services.
Several unmeasured variables might have influenced participation in our study. Respondents may reflect partnerships with open communication about mental health issues. Similarly, the couple’s level of closeness or trust may have influenced the veteran’s willingness to pass along the study information. Although de-identified, our survey queried about the partner’s experience living with someone diagnosed with PTSD. Many veterans may feel uncomfortable with their partner providing information about their shared home life and may discourage participation. Our sample may have attracted individuals who are interested in or knowledgeable about mental health. Veterans engaged in PTSD care likely feel less stigma about mental health and may be more willing to discuss these issues with their partner or spouse, and to encourage involvement in PTSD programs and services.
Recruiting through a veteran intermediary is likely to overrepresent couples that have strong bonds, admit that PTSD is a problem, and address mental health issues jointly. This recruitment strategy may be appropriate for research on how family members can best support veterans who are already engaged in mental health care. However, use of a veteran intermediary may not work well in studies that focus on helping the veteran recognize a problem and encouraging the veteran to get help. Moreover, a veteran who is ambivalent or defensive about mental health issues may be less likely to invite his or her partner to participate in a study about mental health and use of care. Use of an intermediary in this study inherently relied upon the veteran to pass the specific information to his or her partner. Because we could not differentiate between those partners who never received information from the veteran and those who did and did not take any action, we cannot know if unreturned response cards truly represent the partner’s choice to opt out of the study. Although these are limitations of our approach, the number of returned valid surveys and diversity in PTSD symptoms and demographic factors suggest that our method remains an effective recruitment strategy. Finally, our success in recruiting partners of veterans treated for PTSD within the VA Pacific Islands Health Care System may not generalize to work with other mental health or medical conditions, other VA health care systems, or people not seeking treatment. Even so, our findings highlight the potential to reach significant others outside of clinical encounters and to expand partner participation in VA programs and services.


This material is the result of work supported by the National Center for PTSD, and with resources and the use of facilities at the VA Pacific Islands Health Care System, Honolulu, Hawaii. The authors would like to thank Sandra Bozik Lyman, Gennifer Alie, and Alana Yanagida for their assistance in conducting this research.

Conflict of Interest

The opinions expressed are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs or the United States government.


Track Your Manuscript