In this country and worldwide, despite a changing distribution in sex, race and ethnicity, and age, the number of suicides and suicide attempts each year has remained stable or has increased. Why, given greater public awareness, heightened prevention efforts, and increased efficacy of treatment interventions, has success in reducing suicidal behavior been so limited? The answer may lie partially in the need to improve the engagement of suicidal individuals in treatment and to provide better continuity of care—dimensions that have been understudied in the suicide prevention process.

If investigators view suicide prevention as a multistage process, we can identify several points of intervention (Figure 1). The first point of intervention, the individualized risk assessment, begins when a suicidal individual presents for treatment. If the individual goes to a hospital and receives an evaluation in the emergency department, this represents the next level of intervention, treatment planning and disposition. During this point of intervention, clinicians decide whether to admit the individual. If a patient is not hospitalized, a treatment plan is formulated, the appropriate type of treatment is identified, and the patient is discharged with a referral to individual treatment, which may entail psychosocial strategies, pharmacological strategies, or a combination. Stanley and Brown (1) have recently proposed a brief intervention to be administered at this point to reduce further suicidal crises. However, a vital intervention point before discharge and the start of treatment is almost always unaddressed, that of helping patients move from the assessment and referral stages to being motivated to engage in and complete a successful course of treatment. This level of intervention, treatment engagement, is the subject of this review article and is crucial to address if patients most at risk are to avail themselves of the efficacious treatments that are being tested and developed.

To find relevant articles, searches of the literature published between the years 1950 and 2008 were conducted by using MEDLINE, PsycINFO, and PubMed databases with combinations of the following terms: treatment, intervention, engagement, adherence, compliance, utilization, participation, and suicide attempt. Mapped terms were also included. Articles were selected that focused on interventions that targeted—or factors that related to—the treatment engagement of suicide attempters. Additional searches used terms from identified articles, and reference lists of these articles were used to locate other articles. Studies were included if they addressed a relevant aspect of the material covered under the scope of this review.

What is treatment engagement?

No universal definition of treatment engagement is used consistently across studies. However, we propose that the term refers to being committed to the therapeutic process and being an active participant in a collaborative relationship with a therapist to work to improve one's condition.

Studies examining engagement in psychiatric treatment typically operationalize engagement in one of two ways: first, whether or not the patient attended his or her first scheduled outpatient appointment, and second, the total number of sessions attended by the patient (2,3,4,5,6,7).

Thirteen studies focusing on brief interventions targeting suicide attempters were identified in this review. Seven of the studies focused on interventions specifically designed to enhance treatment engagement of suicidal adults and adolescents, and the other six focused on reducing risk of repeat suicidal behavior, with implications for engaging suicide attempters in treatment.

Interventions designed to improve continuity of care

In an effort to enhance treatment attendance, three studies have explored interventions aimed at improving treatment engagement of suicidal adults (2,3,4). Cedereke and colleagues (4) examined whether follow-up telephone contact with patients after their discharge from the emergency department had an effect on rates of treatment adherence. The intervention consisted of follow-up telephone calls four months and eight months after discharge from the emergency department. During the calls, semistructured interviews were conducted to assess the individual's suicidal behavior, psychosocial functioning, physical health, and level of satisfaction with treatment received. Patients who were in treatment were encouraged to continue attending treatment, and patients who had not attended or had dropped out of treatment were encouraged to attend treatment. When necessary, interviewers provided advice and assisted patients in seeking treatment. The goal of the phone contact was to increase and maintain treatment engagement throughout the follow-up period, thus reducing suicidal behavior. The intervention was not found to be effective, suggesting, perhaps, that contact needs to be initiated immediately on presentation to the emergency department or that the level of contact was not sufficient.

Van Heeringen and colleagues (2) tested the effectiveness of sending a visiting nurse to former patients' homes as a means of improving treatment adherence within a one-year period. After discharge from the emergency department, patients who were nonadherent with treatment received home visits from a community nurse, who assessed reasons for nonadherence and specific treatment needs. A significant improvement in adherence was found among the nonadherent patients in the intervention group once they received home visits. It is unclear which components of the home visit were effective because the study did not use standardized procedures for the visits, and data regarding the process of the visits were not gathered. Although the specific factors of the intervention that led to improvement in treatment adherence could not be identified, it seems likely that problem solving with regard to obstacles to treatment adherence may have been helpful in increasing treatment attendance. In addition, because of its burdensome nature—in terms of the economic and time commitments required—this intervention may not be practical and could not easily be incorporated into standard treatment procedures.

Spooren and colleagues (3) attempted to improve treatment engagement by providing a comprehensive referral, with a fixed appointment and a flexible treatment approach. Treatment could include family involvement, meeting with the outpatient treatment provider, incorporating the patient's perspective, and motivating the patient to attend treatment. Results showed that adherence rates were significantly higher in the intervention group than in the control group. However, the results held only for the two hospitals in which most of the aftercare referrals were made within the institution. Thus the results suggest that initial reaching out to patients must be followed with relatively easy access to outpatient treatment.

Four studies focused on treatment engagement of suicidal adolescents (5,6,7,8). Zimmerman and colleagues (7) examined the effectiveness of a contact intervention that consisted of a two-hour multifamily orientation group before intake at an outpatient clinic. The group provided an overview of the clinic and evaluation processes, psychoeducation regarding suicidality, and alternate adaptive behaviors. The goal of the intervention was to improve treatment adherence by reducing the period between initial telephone contact with the clinic and the first clinic appointment. Adherence was measured both by attendance at the initial clinic appointment and by continuation in treatment beyond two appointments. Those in the orientation group had a significantly better attendance rate than others who had received usual treatment at the first clinic appointment, but the intervention did not have an impact on treatment adherence beyond the first appointment.

Rotherham-Borus and colleagues (9) tested the effectiveness of a contact intervention designed to address structural factors believed to impede treatment adherence, such as negative attitudes, expectations, and behaviors of both emergency department staff and adolescent suicide attempters. The intervention comprised three components: staff training workshops, a video that was shown to attempters and their families and aimed at improving the understanding of adolescent suicide and outpatient treatment, and a brief family session held in the emergency department. After emergency department treatment, all attempters were referred to a specialized outpatient clinic, where they received a six-week treatment program designed for the study. Three levels of treatment adherence were assessed: adherence to one session, completion of the six-week program, and the total number of sessions attended. Adolescents in the specialized emergency department program were significantly more likely than those receiving treatment as usual to attend any treatment postdischarge from the emergency department. However, the program led to only a marginally significant increase in the total number of sessions attended and the likelihood of completing the six-week treatment program.

Donaldson and colleagues (5) examined the effectiveness of an intervention involving a verbal contract to attend treatment, which was followed up with phone contact. The intervention focused on psychoeducation regarding outpatient treatment (including a review of expectations for treatment and resistance to treatment) and a verbal contract to attend at least four psychoeducation sessions. Over a period of eight weeks postdischarge, patients received three structured telephone contacts that used a problem-solving approach targeting suicidal ideation and adherence to treatment. Results indicated that although there was a trend toward differences between the two groups in rates of adherence, the differences were not significant.

In an attempt to address patients' low engagement with aftercare treatment after first-time episodes of nonfatal deliberate self-harm, Spirito and colleagues (8) offered the patients rapid, easy access to on-call trainee psychiatrists in the event of further difficulties. Patients were also encouraged to seek help at an early stage should such problems arise. The follow-up data obtained after one year showed a significant reduction of actual or seriously threatened deliberate self-harm in the experimental group, who also utilized fewer general medical and psychiatric services compared with a control group.

These studies reinforce the importance of follow-up contact and suggest that the level of contact has to be substantial to be effective. They also highlight the value of immediate contact postdischarge from the index hospitalization as well as the need for a comprehensive, multifaceted intervention. In addition, the findings demonstrate how little is actually known about what factors improve treatment engagement.

Brief interventions to reduce suicidal behavior

Although very brief interventions designed to reduce suicidal behavior are appealing because of their simplicity and cost-effectiveness, their efficacy has been limited. To date, none of these tested interventions have been administered in an emergency setting. All have occurred postdischarge. Carter and colleagues (10), Motto and Bostrom (11), and Morgan and colleagues (12) examined the effectiveness of written communication (a postcard, a shot letter, and a card with the phone number of a 24-hour hotline, respectively) on reducing future suicidality. Vaiva and colleagues (13) and Fleischmann and colleagues (14) tested the effectiveness of phone contact on reducing risk of suicide. Greenfield and colleagues (15) examined the effect of rapid follow-up on reducing rehospitalization rates and levels of suicidality of suicidal adolescents.

Morgan and colleagues (12) assessed patients who presented to the emergency department for a first episode of nonfatal deliberate self-harm, defined as suicidal and nonsuicidal self-injurious behavior. In an effort to decrease the rate of repetition of deliberate self-harm, the investigators randomly assigned patients to receive either standard emergency department care or an intervention consisting of a green index card that provided a number for rapid, easy access to on-call psychiatry trainees and encouragement to seek help should they begin to experience difficulties again in the future. Patients in the intervention group received the green card in the emergency department before discharge and were then mailed an additional copy of the card to their homes three weeks later. General practitioners were also sent copies of the green card and were encouraged to remind their patients about it as needed. Intervention and control group patients were compared on number of inpatient days, day-treatment days, community nurse visits, and kept outpatient appointments over a one-year period. Results indicated that although patients in the intervention group harmed themselves less frequently than those in the control group, the reduction was not statistically significant.

Motto and Bostrom (11) explored a long-term contact intervention that consisted of written communication with suicidal patients. The investigators hypothesized that reducing demands placed on patients (by approaching them rather than expecting them to seek out services) and reducing their sense of isolation would decrease future suicidality. One month postdischarge from the emergency department, patients were contacted to determine whether they had accepted and adhered to their postdischarge treatment plan for that 30-day period. The contacts were primarily intended to show concern for patients and their well-being. Those who were in treatment were designated as the treatment group. Those who were not in treatment were then randomly assigned to either the intervention group, in which they received periodic contact in the form of a letter (24 letters over five years), or the control group, which received no further contact. Results indicate that although the intervention was initially successful at reducing suicidality, as the frequency of the contact was reduced its protective influence ceased.

Greenfield and colleagues (15) examined suicidal adolescents presenting to the emergency department of a pediatric hospital. Patients were assigned to either rapid-response outpatient follow-up (the experimental group) or to a control group that involved a ten-day wait for outpatient follow-up postdischarge from the emergency department. At the six-month follow-up assessment, no between-group differences were observed in changes in levels of suicidality or in overall functioning over the follow-up period. Overall, suicidal adolescents who received rapid-response outpatient follow-up had a lower hospitalization rate than those who did not. The two groups achieved similar increases in levels of functioning and decreased levels of suicidality, suggesting that suicidal adolescents can be treated within a rapid-response outpatient model and thus avoid hospitalization.

Carter and colleagues (10,16) sent one postcard a month for the first four months postdischarge from the emergency department for self-poisoning and then one postcard every two months for the next eight months (for a total of eight postcards over one year). The postcard consisted of two sentences, the first expressing hope that the patient was doing well and the second indicating that if the patient wished to make contact, it would be welcomed. The intervention reduced repetitions of deliberate self-harm, although it did not significantly reduce the proportion of individual repeaters.

Vaiva and colleagues (13) tested the effectiveness of phone contact on reducing repeat attempts. The intervention consisted of one phone call at either one month or three months postdischarge from the emergency department. The goal of the telephone call was to review the reasons that outpatient treatment had been advised and to provide support, advice, and, if necessary, crisis intervention. Individuals contacted one month after their discharge from the emergency department experienced a significantly lower rate of repeat attempts than a control group that received no contact. Those contacted at three months postdischarge from the emergency department did not differ on rate of repeat attempts from those who did not receive contact.

Fleischmann and colleagues (14) conducted a one-hour informational session in the emergency department before the patient was discharged. The session covered correlates of suicidal behavior and alternative coping measures. Nine follow-up phone calls or visits were conducted over the 18 months postdischarge. Compared with treatment as usual, the intervention was found to significantly reduce the rate of completed suicides over the course of the 18-month follow-up period. However, the study did not report on the effectiveness of the intervention at reducing the number of repeat suicide attempts that did not result in death.

Although some positive findings are reported in these studies, none of these interventions was effective in reducing the number of repeat attempters over the long term. Although such interventions are attractive because of their cost-effectiveness and transportability, their limited efficacy is not surprising given the extent and severity of suicide attempters' problems. Treatment engagement was not tested in these studies. It is possible that these interventions might not be effective in reducing suicidal behavior in the long term but may be helpful in encouraging patients to seek and remain in treatment. They should be explored further for their effectiveness as contact interventions focused on improving treatment engagement.

The significance of early intervention

Early intervention after a suicide attempt is vital because the three-month period after an initial attempt is the period at which an individual is at highest risk of additional suicidal behavior (17). Yet suicide attempters have been found to be very difficult to engage in treatment (18). Between 11% and 50% of attempters refuse outpatient treatment or drop out of outpatient therapy very quickly (19). Up to 60% of suicide attempters do not attend more than one week of treatment postdischarge from the emergency department (20,21). Of attempters who do attend treatment, 38% are no longer in outpatient treatment after three months (17), and 73% are not in treatment after one year (22).

Furthermore, the poor initial treatment attendance and high dropout rate after an attempt are particularly troublesome because history of suicide attempt is the most significant risk factor for repeat attempts and completed suicide, with prior suicide attempt found to increase the likelihood of a future attempt as much as four times (23,24). About 15%—25% of suicide attempters make another attempt, and 5%—10% die by completed suicide (25,26,27,28,29).

Research on early intervention after a suicide attempt that directly focuses on developing strategies to improve the rate of treatment engagement of suicide attempters is critical. No such strategies have yet been developed specifically for suicide attempters.

Help-seeking, gender, and suicidal behavior

In developing strategies for the successful treatment engagement of suicide attempters, it is important to consider gender differences related to both help seeking and suicidal behavior. Although females attempt suicide more frequently than males, males complete suicide at a much higher rate than females. Furthermore, males with less social support (those who are single, recently separated, and divorced) have higher suicide rates (30,31). Males more often than females have an alcohol use disorder (32,33), which is associated with increased suicide risk (34,35). Males are also more likely than females to have access to (36,37,38) and to use violent methods of suicide, such as hanging, shooting, or jumping (39,40), and, therefore, the likelihood of survival is very small.

Although the more lethal methods chosen by males may explain their higher suicide rate, another factor at play may be gender-related differences in help seeking. Females tend to be more willing than males to seek help (41). Suicide attempters and completers often suffer from depression (42,43), which is more common among females than males (44,45). There is also evidence that females and males seek help for problems that are "gender typical"; for example, females are more likely to seek help for depression (46,47), whereas males are more likely to seek help for alcohol problems (48). Differences in help-seeking behavior are explained not by the fact that males need help less often (that is, have better health than females) but by differences in how help is sought and in the perception of need for treatment. Females more often seek consultation for psychosocial distress and emotional difficulties (44,49). Males have more difficulty expressing emotions (50,51) and more often seek treatment when physical symptoms are present (44,49). As a result, it has been suggested that the higher rate of depression among females (nearly twice the rate among males) is not accurate and instead reflects underreporting of depressive symptoms by males (44,52).

Sharpe and Arnold (53) and Sanden and colleagues (54) found that males do not view mental health treatment as a primary solution to physical problems and tend to rationalize any pain experienced. They often have a "wait it out" attitude, feeling that problems will improve on their own. This attitude has been attributed to gender role norms and stereotypes, such as males fearing that they will appear weak, dependent, and lacking masculinity (55,56,57).

Studies suggest that males may need help recognizing their symptoms and understanding the impact of their illness in relation to gender role expectations and socialization (58,59,60,61). This may be particularly important to consider for individuals suffering from depression and at risk of suicide, given the impact that depression and the resulting stereotype of weakness can have on psychosocial functioning (44,62).

Furthermore, when considering effective strategies for males, interventions in which the power differential between patient and clinician is minimized may be particularly useful. For example, Alcoholics Anonymous, which has many more male members than female, uses a model of collaborative participation (63,64) as opposed to a model that emphasizes the clinician as the expert and the patient as treatment recipient. A collaborative approach generally has greater acceptability for males. An approach that incorporates this stance may be more effective for suicidal males given gender role expectations and concerns about appearing weak (58,59,60,61,62). Suicidal males may be more likely to seek help in a forum where their expertise and strengths are supported rather than one that emphasizes a role of the "sick patient."

Effective strategies from related populations

Even though the number of studies on treatment engagement by suicidal individuals is limited, several studies have examined treatment engagement in related populations, which may inform the development of strategies to engage suicidal individuals (65,66,67,68,69,70,71,72,73). In particular, there are several similarities in the areas of substance abuse and suicide attempt. Similar to suicide attempters, substance abusers are considered a hard-to-treat population for a number of reasons. First, as with suicide attempters, most substance abusers (90%—95%) have poor attendance in outpatient treatment (72). Furthermore, substance use recovery is typically characterized by periodic relapse and treatment readmissions, which is similar to the course of major depression, the disorder most frequently associated with suicide attempt (66). Finally, early treatment of substance abusers is essential if recovery is to be successful (66,69), similar to findings regarding the importance of early intervention to reduce risk of future suicide attempts (17).

Several techniques have been found to be effective at improving treatment attendance among substance abusers, which can inform a model of treatment engagement of suicide attempters. Motivational interviewing is a highly effective strategy with this population (67,68,70). Carroll and Nuro (65) incorporated motivational interviewing strategies (such as practicing empathy, providing choice, removing barriers, providing feedback, and clarifying goals) into a standard intake assessment session, using a motivational interviewing style (including asking open-ended questions, listening reflectively, affirming change-related participant statements and efforts, eliciting self-motivational statements with directive methods, and handling resistance without direct confrontation) in a clinical setting. They compared motivational interviewing with treatment as usual and found it to be effective at improving treatment retention when implemented into standard intake assessment procedures in a clinical setting, with rates as high as 64% of patients engaging in treatment (74).

Fiorentine (73) found that patients' perception of the utility of treatment is the greatest predictor of treatment engagement. This may explain why motivational interviewing is effective, given its focus on clarifying goals for treatment, addressing treatment barriers, and providing psychoeducation on the importance of treatment for the particular individual. These findings lend support for the incorporation of a motivational interview into a model of treatment engagement for suicide attempters. Given the relatively short training time required for clinicians to learn motivational interviewing techniques (about two hours) and the fact that it takes half as long to deliver as other common treatments (71), motivational interviewing is a reasonable intervention to draw upon when considering a model of treatment engagement of suicide attempters that can easily be incorporated into practice.

Incorporation of social support systems into treatment procedures has also been an effective method of improving treatment engagement of substance abusers. Several types of involvement have been explored. One method, the Johnson intervention (75), calls for the members of the substance abuser's primary support group to hold an "intervention" in which support group members join together at a specified time to confront the individual about the effect of his or her abuse and need for treatment (75,76). Other methods were developed in response to concerns about the highly confrontational nature of an intervention. The ARISE method (A Relational Intervention Sequence for Engagement) (77) incorporates involvement of concerned family members, who then provide gradually increasing levels of pressure to engage the substance abuser in treatment. This method allows for less confrontation than occurs when initiating a traditional intervention (77).

Involving concerned significant others in the initial stages of the treatment process, even before the first outpatient appointment, has consistently been shown to improve treatment engagement (74,76,77,78). This finding remains across age groups (79,80), race and ethnicity (78), and specific type of substance abuse (74,76).

Social supports act as a source of motivation to engage in treatment, and level of motivation has been linked to level of treatment engagement in several studies (81,82). This may explain the powerful effect that involving concerned others in treatment can have on the engagement process. It also suggests another possible mechanism for improving the treatment engagement of suicide attempters, especially given that suicide attempters typically lack motivation for treatment (83) and often report difficulty garnering support from others (84).

Suicide attempters and engagement in treatment

Specific factors related to both patient and clinician characteristics play a role in engaging and maintaining the suicidal patient in treatment. To our knowledge, no studies have asked suicidal individuals directly why they either refused treatment or left treatment prematurely. Most studies comparing the characteristics of attempters who attend treatment with those who do not attend fail to identify any sociodemographic or clinical characteristics associated with outpatient treatment adherence (4,17,20,85). However, some studies have found that a history of multiple suicide attempts (86), previous psychiatric hospitalizations (86,87), and being married (87,88) are associated in a limited way with increased treatment adherence. Direct feedback from those who decline treatment would facilitate focused prevention and intervention efforts.

With respect to clinician characteristics, studies indicate that mental heath professionals are often hesitant to treat suicidal individuals (89,90) because they are at great risk for a repeat attempt and completed suicide and represent potential malpractice claims (91). Fear or hesitancy from the treating clinician has a negative impact on the therapeutic relationship (92,93,94) and is liable to decrease a suicidal individual's likelihood of engaging in treatment.

Recommendations for engaging suicidal individuals

On the basis of our review of extant literature, we found that simply giving patients the name of an outpatient referral or even making an appointment for them is not sufficient. Outreach contact, either by phone or in person, is a key strategy for engaging suicidal individuals in treatment after discharge from an emergency department visit, and it enhances the likelihood that the person will attend treatment. Another strategy is to provide patient psychoeducation regarding the importance of follow-up treatment.

Brief motivational enhancement techniques to increase patients' willingness to pursue treatment and overcome obstacles may also be an effective engagement approach. Motivational interviewing was initially developed to address problem drinking (81) and has been used since in a variety of primary care and hospital settings to address a multitude of problem issues, such as smoking (95), problematic diet (96), medication nonadherence (97), poor exercise habits (98), and obesity (99). Research has demonstrated that motivational interviewing can be effective as both a technique to encourage treatment entry and an intervention in and of itself (100).

Motivational interviewing is focused on helping individuals work through their ambivalence about changing their behavior and explores patients' concerns and beliefs about changing and not changing (101). The interview reviews the importance of change with the patient and how confident the patient is in his or her ability to make the necessary changes (102). According to motivational interviewing, change is most likely to occur when patients view it as highly important and assert greater confidence in their own ability to make change (102). For example, individuals may feel that attending outpatient treatment is important but may lack confidence in their ability to attend treatment regularly. Conversely, they may strongly believe that they are capable of attending outpatient treatment but may feel that such treatment is not necessary or important for them to reduce their risk of repeat suicidality. Change is discussed in terms of how adopting a new behavior, in this case attending outpatient treatment, is consistent with the patients' values and goals. Therefore, the patient plays an active role in the session, and the interviewer is seen as a consultant rather than as an expert or advisor (103).

Most important to motivational interviewing is to engage patients in a discussion that is noncoercive and nonthreatening and to create an atmosphere that is empathic, nonjudgmental, and supportive of the patient's concerns (103). Open-ended questions, affirmations, reflective listening, and summarizing are the cornerstones of this approach (103).

Brief motivational interviewing can easily be incorporated into clinical practice and does not require a lengthier interview or extensive training on the part of the interviewer. Research suggests that there are only minimal time differences between the length of a brief motivational interview and an interview using a traditional approach (104) and training is often conducted in one-day sessions (101).

Emergency department interventions may include a personalized risk assessment that conveys to patients their degree of risk-based factors such as age, gender, and previous psychiatric history. Included, as well, may be a discussion of current attitudes and intentions regarding outpatient treatment, how this may conflict with their goals, and a review of effective strategies for resolving this discrepancy. This setting may also serve as a site for delivering brief interventions to reduce risk of further suicidal crises (1). Although this type of interaction places an additional burden on emergency department staff and may lengthen the amount of time the clinician spends with patients, it has the potential to have an impact on the number of repeat emergency visits and suicidal behavior.

Treatment engagement is an important yet neglected issue in the assessment of treatment of suicidal individuals. Early interventions incorporating varied forms of contact across time as well as motivational interviewing strategies may improve treatment engagement among suicide attempters. In addition, focusing on engaging males in treatment is particularly important because they die from suicide more often than females and are less likely to seek treatment. Future research should develop and test models of treatment engagement for this difficult-to-treat and high-risk population.

This work was supported in part by grants 5P20A A015630-05, 5R01MH62665-05, 2R01 MH 61017-06, and 5R01MH061017-07 from the National Institutes of Health to Dr. Stanley.