Nearly 1 million people worldwide die by suicide annually [1]. In the United States (U.S.), about 80% of suicide deaths occur in men, and rates have increased by 31.6% among 35–64 year-old men (hereafter, middle-aged men) since 1999. Suicide is now the fourth leading cause of death in this group in the U.S., ahead of diabetes, pneumonia, and other common conditions, at staggering human and economic costs [2– 5]. Similarly sobering findings have been observed globally [6]. Thus, reducing suicide among middle-aged men is a high priority worldwide [1, 2].

Nearly half of all adults who die by suicide see a primary care clinician (PCC) within the month prior to death; about 20% see a mental health specialist [7, 8]. Yet most clinical efforts to reduce suicide target specialty mental health settings [9–14]. Such approaches have value, but do not reach most at-risk individuals, suggesting the critical role of primary care initiatives [15, 16].

An impediment to primary care suicide prevention is that in current practice the topic is rarely broached by individuals with thoughts of suicide. Prior research indicates that patients generally are open to clinician inquiry about suicide thoughts [17], yet clinicians inquire rarely and inconsistently [18, 19], even when risk factors are present [20–27]. Indeed, currently PCCs may detect only 10–20% of suicidal patients [19, 28]. Factors limiting middle-aged men’s disclosure of suicide thoughts to PCCs include gender-linked norms [29, 30]; stigma [31, 32]; the belief that PCCs are not equipped to deal with mental health issues [33]; and competing health concerns in brief visits [34]. PCCs also struggle with competing demands in brief visits, and many lack up to date training and resources to engage optimally with suicidal patients [35–40].

While clinician-targeted risk assessment skills training interventions have increased detection of suicidal patients, still many visiting trained clinicians go undetected, indicating room for improvement [19]. No primary care interventions have sought to address reticence to disclose suicide thoughts by activating affected patients to disclose or signal their receptiveness to discussing the thoughts with a clinician, thereby prompting clinician inquiry and intervention.

We previously developed a personally-tailored interactive multimedia program aimed at encouraging adults with depression symptoms to disclose their symptoms to a PCC and increase treatment receptivity. Used in primary care offices before visits, the program went beyond the typical depression screener in providing motivational video and text messages tailored to patients’ symptom levels, visit agenda, perceptions about depression (e.g., causes, stigma), and comfort discussing mental health. Patients controlled how much material to view, a motivation-enhancing feature [41]. In a randomized controlled trial (RCT), beyond other salutary effects the program led to greater discussion of suicide thoughts, despite containing only one screen addressing suicidality [42, 43]. Given these findings, we envisioned a new tailored multimedia program - Men and Providers Preventing Suicide (MAPS) – that would activate middle-aged men with recent thoughts of suicide (regardless of depression status) to disclose the thoughts to their PCC. We envisioned elements of the program that might increase receptivity to participating in three months of integrated follow- up collaborative care, delivered telephonically by a therapist working with the PCC and a psychiatrist. Together these approaches were theorized to remove barriers in addressing suicide thoughts among middle-aged men [44, 45].

Here, we describe a qualitative study using semi-structured interviews to solicit the views of stakeholders in primary care-base suicide prevention. The study was conducted to guide development of MAPS before finalizing the program and examining its effects in a RCT [46]. We had two aims: (1) to verify (or not) that factors previously reported to influence patient disclosure of, and PCC inquiry about, suicide thoughts remained relevant to address; and (2) to discover additional themes relevant to program design not elaborated in prior studies. The latter aim was important since few studies had solicited stakeholders’ views related to suicide prevention, and none had focused on elucidating factors influencing disclosure of and engagement in treatment for suicide thoughts among middle-aged men in primary care [30, 33, 47–51].

Our analysis revealed two groupings of themes relevant to MAPS, broadly supported by comments across stakeholder categories, addressing its structure, delivery, and content. Some of the structure and delivery-based themes have been reported previously in the context of other educational and motivational interventions. These included the desire to receive relevant information in plain language, and the preference for use of media (e.g., video) over on- screen text alone [67]. By contrast, the theme affirming the resonance and perceived efficacy of MAPS was a novel finding. Stakeholders across groups believed that a personally-tailored multimedia program like MAPS could be a useful tool for activating men with suicide thoughts. They felt the program, if deployed in primary care offices immediately before visits, would effectively encourage men to discuss suicide thoughts with their PCC, and would enhance their openness to treatment aimed at reducing suicide thoughts.

Our second grouping of themes, MAPS informational and motivational content, shed light on nuanced aspects of the suicide experience for men, including self-expectations and perceptions about PCC responses to men with suicide thoughts. Here, another previously unreported theme emerged: the fear that disclosing suicide thoughts to a PCC must always result in hospitalization. Concerns about mandatory hospitalization after disclosure of suicide thoughts have been reported by patients cared for in other settings (e.g., emergency departments). However, to our knowledge this is the first time such concerns have been identified around disclosure in primary care. Some men in our study who had previously attempted suicide specifically cited the fear of mandatory hospitalization as a likely barrier to their disclosure of suicide thoughts to a PCC, should such thoughts recur. In response to this theme, we added an item to the MAPS tailoring questionnaire asking whether the user agreed or disagreed with the statement: “A man who tells his doctor about thoughts of suicide must always be admitted to a psychiatric hospital.” We also added a brief video patient vignette to the program (presented only to users agreeing with the statement about hospital admission) addressing this concern and clarifying that hospitalization usually is not necessary.

Many of the remaining content-related themes emerging from our analysis resonated broadly with the findings of prior (mostly non-qualitative) studies, and provided general support for planned MAPS program elements and approaches. The influence of masculinity and male identity in interpreting and care-seeking for suicide thoughts was a strong and pervasive theme, across stakeholder categories. Beyond being consistent with prior work on mental health care-seeking in men [29, 30, 68, 69], this theme also resonates with the concept of perceived burdensomeness to others, a key construct in the Interpersonal Theory of Suicide that is predictive of suicide behaviors [70]. Stakeholders’ observations that men are generally taught not to express emotion and that some men with suicide thoughts may lack the words to express what they are experiencing prompted further adjustments to MAPS. This included modifying a patient video vignette to include a segment showing a man during his primary care visit telling his clinician about recent suicide thoughts, normalizing and role-modeling the disclosure process for men who might otherwise struggle with this task.

The theme of concerns about PCC preparedness to engage effectively with activated men who have suicide thoughts also echoed the findings of prior studies [18, 19, 33, 71]. Here again, we felt that most of the comments supporting this theme underscored the appropriateness of planned MAPS elements (e.g., a video clarifying that all PCCs receive training in and frequently care for mental health concerns including suicide thoughts). However, responding to comments about lack of appropriate treatment and follow-up resources in primary care, we developed one additional video, presented to all men using the program. It depicted a man who previously had suicide thoughts describing his favorable experience with collaborative mental health care (offered to all men participating in the subsequent RCT of MAPS as noted previously), summarizing the team members, their roles, the process, and how the approach contributes to resolution of suicide thoughts [72–74].

There was uncertainty and disagreement among stakeholders regarding whether MAPS should include questions assessing suicide intent. Ultimately, we included such items in the program, for two reasons. First, while some stakeholders worried that asking about suicide intent could reduce a man’s likelihood of disclosure of suicide thoughts and increase suicide risk, prior research indicates this is unlikely [35]. Second, we felt deploying MAPS in primary care offices immediately before visits and offering follow-up collaborative mental health care to all men using the program would address stakeholders’ concerns regarding access to appropriate treatment and support for men with high intent.

Stakeholders’ comments also prompted us to change another planned aspect of our RCT: the inclusion of complementary training for all PCCs participating in the trial, to avoid activating men with MAPS only to have them experience unprepared clinicians. Prior to stakeholder engagement, we had planned for the clinicians to be briefed on a popular suicide risk assessment and triage tool. However, as noted in Table 3 (Theme 1), some stakeholders felt clinicians already tend to fall into a de-empowering risk management mode once a patient endorses suicide thoughts, and in doing so may fail to convey empathy and solicit the circumstances and emotions contributing to thoughts of suicide. Thus, we substituted a set of brief video modules that encourage clinicians to conduct risk assessment within a broadly empathetic, empowering, patient-centered framework [75].

Strengths of our study were the elicitation of feedback from multiple categories of stakeholders in suicide prevention, and the use of a rigorous qualitative analytic approach. Our study also had limitations. It involved a relatively small, highly selected convenience sample of mostly well-educated, white participants. Whether the findings may generalize to other types of individuals is uncertain, requiring exploration with more diverse samples. For our category of interviewees who had previously attempted suicide, we had hoped to enroll only middle aged men, given our focus. However, despite multiple recruitment strategies, we struggled to do so and therefore also enrolled women. Still, the women we interviewed (including those who had previously attempted suicide) all had a male family member or loved one who had attempted or died by suicide. Future studies limited to men who previously attempted suicide would be useful, though may be difficult to achieve.

In deliberately focusing the interviews on gaining information to inform the development of MAPS, we may not have solicited information relevant to other potential approaches to primary care-based suicide prevention. Likewise, some themes we identified might be broadly applicable across settings, but others might be fairly circumscribed, a product of local norms, health care delivery, or health policy.