Group Problem-Solving Training as a Low-Intensity Psychosocial Intervention for Suicidal Outpatients: An Implementation Study

Background : Current research on suicide prevention revealed that deficits in problem solving skills belong to the most important cognitive risk factors of suicide vulnerability. Poor problem solving skills are associated with depression, hopelessness, maladaptive coping and an increased risk of suicidality. Evidence suggests that good problem solving protects against depression or hopelessness and improvements in problem solving skills can contribute to the prevention of suicide attempts. Problem solving training (PST) as a low intensity brief cognitive-behavioural intervention can be effective in suicide prevention also. Aims : our main aim was to accomplish the cultural adaptation of the PST in group format to improve access to evidence based low intensity psychosocial interventions for patients with attempted suicide in order to prevent further attempts. The second aim was to assess changes in problem solving skills, depression, hopelessness and negative thinking style after the intervention. Methods: Subjects were outpatients with major depression (N=90) with attempted suicide in their psychiatric history. The procedure included a manual-based 12-week outpatient problem solving training in group format (G-PST). Measure administered before and after the intervention: Beck Depression Inventory, Beck Hopelessness Scale, Paykel Life Events Scale, Ways of Coping Questionnaire and Mean-Ends Problem Solving Test. Results: patients showed very significant improvements in problem-solving skills, including planful problem solving, problem analysis and seeking social support. The level of depression and hopelessness as well as the negative evaluation of stressful life events were found to decrease significantly after the G-PST. Conclusions: the findings indicate that G-PST is an effective low-intensity brief intervention which contributes significantly to the improvements of modifiable risk factors of suicide. Group Problem-Solving Training as a Low-Intensity Psychosocial Intervention for Suicidal Outpatients: An Implementation Study.


Introduction
Suicide is a very complex phenomenon with psychological, social, biological, cultural and environmental factors involved.
Mental disorders and previous suicide attempts are the main risk factors for a completed suicide act [1,2]. Clinical studies show that untreated major depression and bipolar disorder are the most important risk factors, 90% of all suicide-related death occur among patients with current major psychiatric disorders [3,4]. Several unmodifiable factors such as traumas, negative life events, gender, social insecurity and poverty may increase suicide risk. Hungary had the highest avarage sucide rate in the world between 1960 and 2000 [5]. In 1980-1985 the WHO registered 44.4 death per 100.000 inhabitants in Hungary, which is the triple of the global suicide rate of the world (14 suicides per 100.000 inhabitants. Since then there has been a huge decline in suicide rates and for now suicide rate is 17/100.000 due to a much better recognition of depression and to more effective and comprehensive treatment approaches [1]. Effective suicide prevention strategies should focus on the modifiable risk factors such as low mood, hopelessness, dysfunctional thinking style, lack of social support and poor problem-solving abilities [1,6]. Several studies confirmed the predictive value of hopelessness in suicidal behaviour and identified it as the final step before attempting suicide [2,[6][7][8][9].
Hopelessness can be understood as the loss of perspective, an extremely negative view of the future. According to Beck's cognitive model of depression the negative view of the future is part of the negative cognitive triad, namely a negative view of the self, the world and the future [10]. The starting point of the hopelessness theory of suicide is that not the low mood or depression per se but the loss of perspective of the future, namely hopelessness can be viewed as the final stage before suicide. It has also been a wellestablished finding that hopelessness is a key factor in increasing the risk of suicidal behaviour and it has a predictive value for the repetition of suicide attempts as well as for completed suicides up to 10 years later [7].
Studies concerning the psychological processes underlying suicidal behaviour has also highlighted deficits in social problemsolving abilities. Several researchers including Walker, Chang and Hirsch [11] suggested that individuals with neurotic tendencies tended to consider problems as threats and employed impulsive, passive and/or avoidance type coping. These maladaptive ways of coping present greater risk for suicide vulnerability as mediators towards hopelessness. Several investigations highlight that emotion-oriented coping is strongly associated with poorer mental health and self-harm thoughts as well as with suicide attempts whilst problem-oriented coping is associated with better mental health [12]. Collectively, the results of these studies suggest that suicidal people are less active in their problem-solving efforts, they tend to give avoidant and impulsive responses, and also tend to have difficulty in generating alternative solutions to their problems. In a clinical study over 60% of patients with at least one suicide attempt reported that they had had at least one problem that they believed to be unsolvable and they had experienced high level of hopelessness at that time as well as strong suicidal ideation [13]. Strong research evidences have confirmed the view that meaningful association exists between suicidal ideation and poor social problem-solving skills. Pollock and Williams [14] found that suicidal patients with at least one attempt subsequently demonstrated maladaptive problem-solving skills such as increased passivity and avoidance in problem situations. A more recent study by Collinson et al. [15] also confirmed that suicidal patients presenting to hospitals with self-harm for the second or subsequent time, show poor problemsolving skills. However, some studies failed to find differences in problem-solving skills between suicidal and non-suicidal psychiatric patients. Pollock & Williams [14] and Poós, Annus & Perczel Forintos [16] compared patients who attempted suicide with non-suicidal psychiatric controls concluded that passive and avoidant coping were not unique to suicidal patients; however, a passive coping style in combination with the smaller number and less effective alternatives generated may increase suicide vulnerability. If this is true, major improvements in problem solving and coping skills should be an important target for interventions which aim for the effective prevention of suicidality.

Problem-Solving Training
Problem-solving training (PST) as a brief structured cognitivebehavioural and psycho-educative intervention has been adapted for several goals, context and target populations. However, it was found to be promising in decreasing the modifiable risk factors of suicide. Primary prevention of suicide should focus on the screening of groups at risk in the general population. Secondary suicide prevention aims to decrease the likelihood of a suicide attempts in high-risk individuals such as borderline or bipolar patients, while the goal of tertiary prevention should be to diminish suicide contagion after completed suicides [17]. According to some meta-analysis the following methods proved to be effective in suicide prevention: pharmacotherapy, cognitive-behavioural therapy, reduced access to lethal means as well as responsible media reporting of suicide (e.a. [18][19][20]. In terms of psychosocial interventions suicidal patients often benefit from therapies that directly address intrusive suicidal thoughts, motivation and attempts. Cognitive therapy for suicide prevention, cognitive-behavioural therapy, dialectical behaviour therapy, problem-solving therapy, mentalization-based treatment and interpersonal therapy have all some positive effects for preventing suicide by decreasing both suicidal ideation and the reattempt rate of suicide attempts [21]. Systematic reviews have indicated that problem-solving therapy can be a promising cognitive-behavioural intervention following self-harm [19,22].
For example, in an early study Salkovskis, Atha, and Storer [23] reported that PST was effective when was delivered in individual setting. Other studies did not find any evidence for the effectiveness of PST when applied in group setting [24], while in a meta-analysis Hawton et al. [18] argued that PST has the quality of evidence range between only moderate and low. Almost all studies reported that relatively short training was used consisting of 4-6 sessions.
However, guidelines by the National Institute for Clinical Excellence (NICE) recommend the use of problem-solving therapy as a wellestablished intervention to prevent repeated self-harm [25].
Problem-solving training (PST) is designed to improve problemsolving skills, initially developed by D'Zurilla and Godfried in 1971 [26]. Since then, it has been extended and used in many research trials which proved its effectiveness in several mental disorders [27], especially in suicide prevention [22,23,28,29]. Meta-analysis of randomized controlled trials revealed that suicidal patients receiving PST had significantly greater improvement in their level of depression and hopelessness as well as in their problemsolving skills than suicidal patients in the control treatment groups [22]. In a one-year follow-up study conducted by Hatcher et al. [30] the proportion of people at high risk for further suicide attempt (based on the Beck Hopelessness Scale) was significantly lower in the group receiving problem-solving training than in the group receiving usual care. However, regarding long-term suicide prevention, current literature and the NICE guidelines strongly emphasize the importance of further research on the effectiveness of interventions carried out in clinical settings [31]. Finally, there is a large gap in the demand and the supply of the evidence based, high intensity, individually addressed psychotherapies, since the resources for providing psychotherapy in mental health care are very limited. Evidence based psychotherapies have always been hardly available due to the limited number of trained therapists in the funded healthcare. Low-intensity psychosocial interventions (LIPI) can fill this gap in clinical practice by increasing the access to psychological treatments [32]. Group interventions are widely used in clinical setting as well as in health psychology, since they allow intervening on low level / low intensity. As the group format PST is based on the principles of cognitive behaviour therapy it can enable participants to foster effective social problem-solving, overcome negative thinking and to treat underlying psychological problems. The implementation of group PST can facilitate the social support between the group members while they are able to serve as sources of feedback to each other. Furthermore, group members can facilitate brainstorming and generate alternative solutions for problem-solving together. Up to now research on the effectiveness of PST, especially of G-PST gave mixed results [33].

Aims of the Study
Suicide prevention is a serious concern world-wide and the availability of effective, evidence based psychological prevention methods is of utmost importance. PST as a short, structured lowintensity cognitive-behavioural intervention which was found to be promising in decreasing the modifiable risk factors of suicide.
Although suicide rates show a steady decline in our country, effective ways of suicide prevention are much needed therefore our aim was to adopt the method of problem-solving training. In this implementation study our main goal was to use a culturally adapted problem-solving training in group format (G-PST) and investigate its effect on depression, hopelessness, problem-solving as well as coping skills as they belong to the modifiable risk factors of suicide.
We hypothesized that after a 12-week G-PST participants would show improvement in problem-solving abilities, in coping skills as well as decreased level of depression and hopelessness. In addition, we investigated the number of stressful the life events and their emotional impact on the patients before and after the intervention.

Participants
A total of 90 psychiatric patients were recruited from our university outpatient clinical psychology service. Patients suitable for inclusion were those who had one or more suicide attempts within the last 5 years, fulfilled the DSM-5 criteria of major depressive disorder and aged between 18-60 years [34]. All of them were previously treated in hospital after their suicide attempt and were on antidepressant medication. In terms of the method of suicide attempt, 73% of the patients used drug overdose, 27% cut him/herself. Patients were excluded if they had psychosis, hypomanic or manic episode or addictive disorders in their psychiatric history or they were currently at acute risk for suicide.
Eligible patients were consecutively assigned to G-PST in the order of their appearance at the outpatient clinical psychology service.
All of the participants provided written informed consent, as A sociodemographic and clinical data sheet was used to obtain information on the demographic and clinical history. Demographic characteristics of the participants (gender, age, marital status, education, occupation and psychiatric symptoms) can be seen in Table 1.

Measures
The following questionnaires were carried out at the baseline and after the intervention.

Beck Depression Inventory (BDI):
To measure the severity of depression, the widely used BDI was administered [7,34,35]. The BDI consists of 21 questions on a four-point Likert-scale where a higher score correlates with more severe depression. Participants are asked to assess their level of their symptoms on each item. The questionnaire shows a very strong internal consistency of 0.911.

Beck Hopelessness Scale (BHS): For measuring hopelessness
we used the widely highly recognized BHS [36,37]. BHS is a selfreport inventory designed to measure three major aspects of hopelessness: feelings about the future, loss of motivation, and expectations. The test consists of 20 yes or no items; score above 9 indicates severe suicide risk. The scale showed acceptable reliability, Cronbach's alpha 0.71.

Ways of Coping Questionnaire (WAYS)
: Is a 22-item selfreport measure in order to identify methods people use to cope with stressful situations [38,39]. Participants are asked to make ratings on a 4-point Likert-scale. Factors of the WAYS include problem analyses, planful problem-solving, emotion focused coping, seeking social support, escape-avoidance, emotion balancing coping. The questionnaire indicated an acceptable internal consistency as the Cronbach's alpha was 0.7.

Mean-Ends Problem-Solving Test (MEPS):
Is one of the most commonly used methods of assessing interpersonal problemsolving deficits [40,41], adapted by Poós, Annus & Perczel Forintos [16]. The MEPS provides 10 vignettes that describe the beginning of a problem situation (e.g., conflict with the boss or partner) and its desired end-point (good relationship with boss or partner).
Participants requested to describe how they would solve the situation for each scenario. Regarding the most common scoring method MEPS is marked for 'relevant means' that is evaluated by a trained clinical psychologist on a Likert-type scale from 0 (not effective at all) to 7 (very effective). The "relevant means" stand for the number of discrete steps that enable the protagonist to reach were recruited from the outpatient clinical psychology service of our University according to the inclusion criteria described above.
Following initial assessment, they were consecutively assigned to This was followed by practice in pairs using worksheets and discussion in "large" group format. Feedback was provided at each step. Homework was also assigned after the sessions in order to practice the actual skills.

Descriptive Statistics and Baseline Data
The mean age of participants was 39.5 years; 13.3% of them were male and 86.7% female. The marital status showed a normal distribution between married (36.7%), divorced/widowed (23.3%) and single (40%). 43% of the participants were employed, 20.3% was unemployed, 11.4% seasonal employed, 8.8% was on maternity leave and 16.5% was retired/disabled. Most of them (90%) were diagnosed with major depression while the rate of comorbid personality disorders was 10% (Table 1). Drop out: 83% of the participants (74 persons) completed the G-PST, only 17% (16 persons) left the program. This level of dropout rate was consistent with the literatures' findings [43]. Missing data was handled by multiple imputation using the Fully Conditional Specification method (FCS). In this approach, missing data are imputed by a chain of conditional distributions based on regression models using available baseline data. To assess for sampling bias, a comparative analysis was conducted between participants who completed the final measures and those whose data were missing at the posttreatment measurement. Results revealed no significant betweengroup differences on any baseline measure (e.g., demographics, diagnostic variables, depression, hopelessness and problemsolving abilities).

The Effects of PST on Problem-Solving Abilities and
Coping Skills Table 2 shows the results of the paired samples t-test which was Cohen's d=.9). To investigate the effectiveness of G-PST on ways of coping (based on WAYS) we also performed paired samples t-test and a significant increase was found on the subscales of WAYS (see (Table 2)

The Effects of G-PST on Depression and Hopelessness
Comparing pre-, and post-treatment data, there was a significant decrease in depression (t (89)= 4.66, p<.001, Cohen's d=0.5, medium effect size) and hopelessness also showed large decrease with a very strong significance (t(89)=7.03, p<.001, Cohen's d=0.8, large effect size, (Table 3). This hypothesis was also confirmed since both depression and hopelessness decreased significantly after the G-PST. Means and standard deviations are shown in Table 3. To examine if problem-solving abilities and coping skills are interrelated either to depression or hopelessness, Pearson's correlation analysis was performed. Table 4 shows that there was a moderate and very strong significant negative correlation between depression (BDI) and WAYS subscales as follows: problem analysis  Paykel's Life Event Scale (PLES); NS = not significant, ** p<0.01, * p<0.5, + p<.1 Table 4: Correlations between coping measures (WAYS, MEPS) with depression and hopelessness.  Table 3.

Discussion
In countries with moderate to high suicide rates the increased availability and easy accessibility of effective, evidence based psychological prevention methods is of utmost importance. Suicide prevention is a priority for health services all over the world especially in countries with high suicide rate. Since Hungary can be characterized by relatively high suicide rates (24.3/100,000 in 2011 and 17/100,000 in 2017), [44] several efforts have been made to reduce it. Rihmer et al. [5] in their review conclude that suicide rates can most effectively be reduced by the identification of major depression and bipolar disorder and by the better care of depressed patients. However, in addition to medication, psychosocial interventions can also be effective ways of suicide prevention. In the last decades, psychosocial interventions and learning adaptive skills to cope with low mood or stressful life events are more preferred by the patients than antidepressant medication alone [48]. Beck's cognitive model of depression and hopelessness had strongly been supported by numerous scientific research where hopelessness considered a significant predictor of suicidality. By measuring hopelessness, it is possible to identify patients at risk of suicide and deliver specific interventions on factors associated with hopelessness such as poor problem-solving skills [14,45] which are in close connection with each other. Problem-solving therapy as a brief cognitive-behavioural intervention has a sound theoretical basis; however, research papers reported mixed results about its effectiveness in suicide prevention [22,24,41].
In this study we presented the implementation of a culturally adapted manual based problem-solving training in our country. We  Table   2). This finding is in accordance with the model of coping described by Connor-Smith, Compas, Wadsworth, Thomsen and Saltzman [46]. It is well-validated that disengagement coping (avoidance, denial, wishful thinking) can be linked to internalizing problems.
Our results corroborate that PST enhances the development of primary control coping (problem-solving, emotional expression, emotional modulation) and secondary control coping (acceptance, cognitive restructuring, positive thinking, distraction).
Coping skills also showed significant improvement on all subscales of the WAYS but the escape/avoidance subscale ( According to the cognitive theory of emotional disorders, this is not the event, but its evaluation has an emotional impact on the mood of a person. Exactly this is what was found in our study: participants evaluated the impact of their same life events significantly less serious after the intervention than before. It might mean that even if G-PST is a low-intensity cognitive intervention it still seems to be effective in restructuring distorted negative thinking and weakening the distressing effect of life events (see Table 2; p<.01). However, the result that the number of stressful life events increased significantly at post-treatment is surprising (see Table 2; p<.01). One explanation might be that during the course of G-PST participants became more aware of their previous distressing life events or these events were somehow activated and recollected.
The main limitation of our study is that no control group was used for comparison therefore the results can't solely be interpreted as an evidence of effectiveness of the G-PST. Spontaneous recovery and placebo effects might have played part in the changes also. Another limitation, that no long-term follow-up was performed after the intervention. Post-treatment assessment was carried out at the end of G-PST; therefore the scores could be proper indicators of the actual mood and hopelessness but did not allow prediction about the decrease of future suicide attempts. Follow-up assessments after a longer period would have provided a more accurate picture about the long-term effects of the interventions.
To sum up, despite of the limitations described above, the results of the study support the view that applying G-PST after suicide attempts can lead to significant positive changes in the modifiable risk factors such as depression, hopelessness and great improvement in problem-solving skills, therefore it seems to be a promising way of preventing further suicide attempts. The 10-session G-PST delivered in 12 weeks was highly supported by the participants who expressed their needs to have even more sessions in order to consolidate more deeply their newly acquired skills. Finally, we can conclude that G-PST can be used as an effective, cost-effective low-intensity, evidence based psychosocial intervention for depressed outpatients with attempted suicide in their psychiatric history. The implementation of a culturally adapted manual based problem-solving training in group format improves access to psychological treatments, allows to reach wider populations at risk. Furthermore, G-PST allows to widen the availability of PST as a low intensity intervention for mild to moderate mental health problems in schools or in general practice.