A Systematic Evidence Review on Suicide in Kosovo

The after-war Kosovo society was characterized by an increasing number of suicides. The aims of the study were the identification, appraisal, selection, and synthesis of studies on suicide in Kosovo. A classical theoretical approach has been used to analyze studies in known databases as well as dissertations and presentations at national conferences. Fifteen studies that met our selection criteria were analyzed. These studies were psychiatric/psychological, forensic, sociological, epidemiological, and legal type. We have not found any study of qualitative, physiological, or anthropological character. None of these studies noticed the focus on specific cultural elements. The investigation of social elements did not provide valuable data. Most of the studies focused on adolescents. Neither research has convinced us for it comprehensible view of suicide phenomenon. Only two studies analyzed predictors of suicide. Identification of the protective/risk factors is difficult because the reviewed studies have many methodological shortcomings. Although the studies analyzed include many valuable data of epidemiological nature, they do not allow us to develop a model for suicide.


Introducing Kosovo
Kosovo is the youngest country in Europe as 45% of the population is below the age of 25. Kosovo was an autonomous unit in the Yugoslavian Federation until the collapse of communism and the disintegration of the Federation. Efforts toward Kosovo independence escalated the conflict between Kosovo Albanians and Serbs during the years 1998-1999. During the conflict, roughly a million ethnic Albanians fled or were expelled from Kosovo (BBC, 1999). Altogether, more than 11,000 deaths have been reported and around 3,000 people are still missing (BBC, 1999). After the Kosovo War and the 1999 NATO bombing of Yugoslavia, the territory came under the interim administration of the United Nations Mission in Kosovo (UNMIK) and proclaimed its independence in 2008. Post-war Kosovo has around 1.8 million inhabitants and is characterized by an annual 0.9% rate of population growth (World Bank, 2014). Although secularity is prevailing in the country, a large fraction of Albanians are Muslims (95.6%; Kosovo Agency of Statistics, 2014).
In terms of economic growth, Kosovo's citizens are the poorest in Europe with an average annual per capita income (PPP) of US$7,400 (Central Intelligence Agency [CIA], 2014). Unemployment rates are quite high, especially among young people; in fact, Kosovo is considered as the 528717S GOXXX10.1177/2158244014528717SAGE OpenFanaj and Melonashi research-article2014 1 Child and Adolescent Mental Health Unit, Prizren, Kosovo 2 European University of Tirana, Albania country with the highest level of unemployment worldwide among youth below the age of 25 (73%; CIA, 2014). In this context, informal economy has flourished and migration rates are increasing (U.S. Agency for International Development [USAID], 2008). Therefore, the country is characterized by a low rate of economic development and is experiencing the typical transition problems of developing countries (CIA, 2014;USAID, 2008). Moreover, the Kosovo socio-cultural context is one in which traditional cultural elements are still important and generation gaps or conflicts are quite evident. Beside this, war traumas, economic poverty, political instability, democracy in transition have a big influence on social development and quality of life. In this context, the increasing number of suicides in post-war Kosovo became very soon an issue of concern for researchers and policy makers (UNMIK, Ministry of Health of Kosovo, 2007).

Importance of Research on Suicide
According to epidemiological data, Kosovo had the lowest coefficient of suicides in Europe (1.3/100,000 in 1981 and 2.7/100,000 in 1985; Birinxhiku, 1990), until the post-war phase. Several studies conducted after the war have suggested an increase in the number of suicides (Drevinja, Berisha, Serreqi, Statovci, & Haxhibeqiri, 2013;UNMIK, Ministry of Health of Kosovo, 2007;Zhjeqi et al., 2009). Nonetheless, it should be mentioned that both pre-and postwar data recording processes have important methodological limitations. For instance during the pre-war phase, suicide data were based exclusively on the forensic doctor reports to the Statistical Office of Kosovo. Conversely, in the after-war period, the registration of cases was done by the police or prosecution. However, during the last years the Statistical Office of Kosovo has been primarily engaged in this process by using the Demographic-2 statistical form (DEM-2 form)-a statistical demographic official form which also includes information on cause of death. In this context, important discrepancies have been found between the police reports (KP Press and Public Relation Office, personal communication, July 23, 2013) and Statistical Office reports (the police reported twice as many cases of suicide as the Statistical Office of Kosovo; Kosovo Agency of Statistics, 2011). Also, changes of population in the preconflict, during and post-conflict period and the lack of an official population register, have made even more difficult the achievement of a reliable rate of suicides. Also few studies have focused on understanding the phenomenon in terms of several predictors or risk/protective factors.

Aims
The aims of the present article were to identify, select, and synthesize studies on the topic of committed suicide in Kosovo. The present study 1. provides knowledge on the actual condition of suicide research in Kosovo, 2. contributes to the development of one descriptive model of suicide in Kosovo, and 3. supports/suggests future directions for research.

Method
The inclusive criteria for studies included the following in hierarchical order: • • Papers that were published in international journals (three articles); • • Doctoral dissertations (two) and magisterial theses (four); • • Presentations in international conferences, published in online proceedings (five); • • Presentations in international conferences in Kosovo (three).
Published studies and published abstracts were found in the known scientific databases, PubMed (n.d.) and ScienceDirect (n.d.). From Internet search in PubMed (n.d.) database with the words "suicide, Kosovo," 208 articles were retrieved. Nonetheless, in most of these articles, Kosovo was mentioned in the context of other studies (e.g., former Yugoslavia). In fact, only three fully published studies were conducted in Kosovo.
A total of 206 articles were retrieved from the ScienceDirect (n.d.) database. In addition, there were only five abstracts from international congresses (online proceedings), which were considered for review. A single study was found in Google search (a presentation PowerPoint format). Research in National and University Library of Kosovo resulted in two doctoral dissertations and four magisterial theses. Only one doctoral thesis belonged to the pre-war period (the 1990s).
As regards presentations, there were three from the First Scientific Conference of the Albanian Psychiatric League (2006), two from the International Conference on "Prevention of Suicide Behaviour Among Youth of Kosovo" (2007), and four from the International Scientific Conference-"Towards Effective Prevention of Suicide and Self-Harm Among the Youth in Kosovo (2009)." Only three of these were available for review; the rest was impossible to review, because their authors did not respond to our inquiry to have access to them. In total, the studies reviewed for the purpose of the research (see Table 1) can be classified according to the specific fields as follows: 1. Psychology/psychiatry-11 studies, 2. Epidemiology-1 study, 3. Forensic studies-2 studies, 4. Law studies-1 study, and 5. Sociology-1 study.  [2000][2001][2002][2003][2004][2005], the number of cases increased from 29 to 55 in this region, hard to find a coefficient in the absence of an official population record (Fanaj & Berzani, 2009) conference presentation Media analysis Monitored and analyzed writings on mental health topics of eight most popular daily newspapers in Kosovo and Albania for 6 months.

• •
In articles for suicide have had numerous weaknesses.

• •
The issue of suicide encouragement Note. PTSD = post-traumatic stress disorder.

Results
A classical theoretical approach was used to analyze these studies (Montero & Leon, 2007). Sixteen studies (three journal articles, two dissertation theses, four magisterial theses, and seven conference presentations/online abstracts), which meet the selection criteria, were analyzed (see Table 1). Of these, nine studies have examined suicide in the general population, two in the forensic facilities, one in media, and four among adolescents (community sample). Six studies analyzed suicidal ideation/behavior and nine studies analyzed suicide in general. Studies examining suicide in general did not focus on a specific age group, whereas studies on suicidal ideation/behavior focused both on adults (in two cases) and adolescents (four cases). Finally, no studies mention or describe intervention programs for suicide in Kosovo, suggesting that these might be nonexistent.

Main Findings
Rate. Kosovo had the lowest percentage of suicide rates in Europe, which was 1.3/100,000 (1981), whereas in 1985, it was 2.7/100,000 (Birinxhiku, 1990). In 2008 to 2012, the calculated rate based on police service data was 3.17/100,000 . Data are indicative of an increase in suicide rates in Kosovo (Arënliu, 2010), close to double (Fanaj, Dana, Shala, & Poniku, 2006). An increase in cases of 122% has been reported for year 2000 as compared with 1983 (Zhjeqi et al., 2009).
Age group. Birinxhiku (1990) found that 49% of reported suicide cases belonged to the age group 15 to 34 years. Along the same lines, Fanaj et al. (2006) reported rates of 45.5% belonging to the same age group (Fanaj et al., 2006). Finally, another study (Haliti, 2007) reported 40.8% of cases in the age group 30 to 39 years.
Gender. Studies converged in suggesting important gender differences in suicide. Indeed, significantly more men than women commit suicide, the ratio being 2 to 1 for men (Birinxhiku, 1990;Drevinja et al., 2013). Percentages of individuals committing suicide and belonging to the male gender varied in the range from 69.4% (Haliti, 2007) to 76.7% (Zhjeqi et al., 2009).
Residence. Data on this particular aspect were inconclusive; one study found that 60% of the sample was from rural areas (Fanaj et al., 2006), another study found no significant differences based on residence (Zhjeqi et al., 2009). Finally, Musa (2010) found the opposite result as slightly more suicide cases were from urban rather than rural areas (57%).

Place of suicide.
A study examining the place where suicide was committed found that in 34.7% of the cases the place was far from home, in 32.7% of cases near home, and in 14.3% of cases at home (Haliti, 2007).

Suicide Attempts Findings
There are not enough data to provide solid conclusions on suicide attempts. For 6 years, 341 cases of suicide attempts were hospitalized in the psychiatric ward (female/male ratio is 2:1). Based on the substances used 53% of the cases used psychotropic drugs and 19% of the cases used chemicals (Fanaj et al., 2006). One study reported 12.8% of suicide attempts within the forensic psychiatric setting (Ibishi, Kola, Ramiqi, & Musliu, 2013) and concluded that attempts were more common in forensic settings than in the general population (Ibishi et al., 2013). For adolescents, the self-report rates of suicide attempts ranged from 3.4% (Fanaj, Fanaj, Drevinja, & Poniku, 2013) to 23.5% (Jemini, 2008). Adolescent girls reported more suicidal attempts (Arënliu, 2009).

Suicide Ideation/Behavior Findings
Studies have provided different rates of suicide plans: 29% (Jemini, 2008) versus 6.4% (Fanaj et al., 2013). In case of ideation, there is no significant difference-43% (Jemini, 2008) versus 37.6% (Fanaj et al., 2013). However, other studies do not provide evidence on suicidal ideation (Arënliu, 2009;Statovci, 2011Statovci, , 2012Wang et al., 2012). As regards gender, all studies reported significantly more suicidal ideation/behavior for girls (Arënliu, 2009;Fanaj et al., 2013;Jemini, 2008;Statovci, 2011Statovci, , 2012. Regarding residence, adolescents from urban (rather than rural) areas reported higher levels of suicidal ideation/behavior (Arënliu, 2009). However, another study found that irrespective of age, gender, or education, subjects in rural areas had higher suicide ideation scores (Wang et al., 2012). Data from forensic settings (sample: 134 individuals) found that 14.9% of the sample reported having a plan to kill themselves, 40% have told someone that they are going to commit suicide, and 23.4% were likely to attempt suicide someday (Ibishi et al., 2013). Suicidal thoughts were more common in forensic settings than in the general population and were significantly associated with personality disorder, psychosis, and post-traumatic stress disorder (PTSD; Ibishi et al., 2013). Finally, the prevalence of suicide ideation was 70% in a group of 125 war victims (Wang et al., 2012).

Predictors
For suicidal ideation/behavior. Only two studies analyzed predictors of suicidal ideation/behavior. The first study (Arënliu, 2009) found that a recent traumatic event was a more significant predictor of suicide ideation/behavior than previous war trauma. Arënliu (2009) andJemini (2008) found prediction variance for suicide ideation/behavior to be 61.6%, by examining the following factors: family dis-cohesion, depression, family mental health history, and suicide in the family.

Protective Factors/Risk Factors Findings
A summary of the protective/risk factors is rather difficult due to the heterogeneity and lack of reviewed studies, different samples, different questionnaires, unclearly defined categories, lack of criteria based on the accepted models or theories, and lack of use of statistical tools.
Regarding risk factors, Birinxhiku (1990) suggested that the most frequent factors of suicides are psychopathological factors (46.3%), socio-pathological factors (22.17%), and somatic-pathological factors (7.66%), and unknown factors are 23.79% (Birinxhiku, 1990). As regards the socio-pathological factors, another study (Fanaj et al., 2006) with a similar design to the previous found similar figures (23.63%). Even more important, in almost half of the cases, the effect of socio-pathological factors was due to economic difficulties (Fanaj et al., 2006). This finding has been supported by other studies Zhjeqi et al., 2009). Other aspects include social isolation (1.8% to 2.8%; Birinxhiku, 1990;Fanaj et al., 2006) and psychiatric history (45.45%; Fanaj et al., 2006). To summarize, Drevinja et al. (2013) state that In over 80% of cases come to the expression of social factors: unemployment (60%), low income per capita, enormous growth of drug users and alcohol, low budget for mental health, dominance in rural settlements, debt, domestic violence, inability to buy drugs, the inability of the education of the children, war trauma/PTSD, large number of people who are missing or dead after the recent war in Kosovo. (p. 1) In a similar fashion, another epidemiologic study analyzed suicide as a consequence of socioeconomic, psychological, and political problems as unemployment and poverty (Zhjeqi et al., 2009).
A range of variables which can be considered as risk or protective factors were identified in the context of adolescents. The following are based on the study of Arënliu (2009): • • Religion has a negative correlation with the scores of suicidal ideation/behavior; • • Those who report separation from partner (romance), move to a city, chronic illness, accident, financial hardship, report higher scores of suicidal ideation/ behavior; • • Adolescent's perception of their own well-being and happiness is related to both suicidal ideation/ behavior; • • Coping mechanisms such as not seeking concrete support, behavioral disengagement, and self-blame seem to be predicting suicidal ideation/behavior; • • Substance use appears to be a risk factor in urban areas; • • History of suicide (attempt or completion) in family increases risk of suicide attempt among adolescents.
Other studies of adolescents have reported that psychological difficulties (emotionality, hyperactivity, behavioral problems, and problems with peers), self-esteem, depression, substance use, and hopelessness were significantly correlated with suicidal ideation/plan (Fanaj et al., 2013;Jemini, 2008;Statovci, 2011Statovci, , 2012. Important risk factors for suicidal behavior in forensic settings were identified in 73% of cases, including abnormal behavior, isolation, hopelessness, depression or other mental disorders, family problems or personal loss, and drug or alcohol intoxication (Ibishi et al., 2013). Finally, regards protective factors, the only sociological study for suicide by Nishori (2009) stated protective factors: dignity, Albanian male stoicism, religion, and cultural values toward violence.
To conclude, it must be acknowledged that it is quite difficult to summarize findings of these studies with different designs and various periods of times in one coherent model, which will be a descriptive model for suicide in Kosovo.

Discussion and Conclusion
The aim of the present study was to provide a comprehensive literature review on suicide in Kosovo. Research showed that the majority of studies were from the psychological/psychiatric fields and they had many methodological shortcomings, such as construction of sampling, non-validation of measures, unclear definition of categories, lack of criteria based on the accepted models or theories, and so on. Nonetheless, several aspects need to be summarized and discussed. The review of epidemiological data on suicide in Kosovo suggests that despite suggestions on increasing suicide rates, this country might be classified within the category of countries with a low suicide coefficient in Europe (e.g., Mediterranean countries such as Greece, Macedonia, Italy, and Albania with a rate less than 10/100,000). Indeed, the rate for Kosovo (3.0/100,000 to 4.0/100,000) is comparable with that for Albania (4.0/100,000 in 2003) or Greece (3.6/100,000 in 2005). One aspect that should be noted is that suicide rates are low even for Albanian minorities in the region (e.g., Albanians in Serbia), which have rate below 10 (4.9/100,000) in contrast to other parts of the country which have a rate of 20/100,000 (Penev & Stanković, 2007). These findings have been explained in terms of ethnical features and religiosity.
Regarding gender differences, data from Kosovo are consistent with those of other countries (male gender is at a higher risk). Also the most frequent suicide methods by gender (e.g., hanging, firearms) are similar to those reported for other countries such as Croatia, Slovenia, Italy, and Greece (Injac-Stevović, Jašović-Gašić, Vuković, Peković, & Terzić, 2011). However, there are differences by country in terms of age groups; for instance in Serbia, Montenegro, Slovenia, the elderly make up the highest number of suicide cases (Penev & Stanković, 2007), whereas in Kosovo, the majority of cases were 34 years old and younger.
As regards the frequency of adolescents' suicidal ideation, data suggest higher rates for Kosovo as compared with other countries (37.6% to 43%; Rutter et al., 2008). Finally, as regards the risk/protective factors, the suggested variables are not new and comparable with other countries. They included family mental health and suicide history, war trauma/PTSD, hopelessness, depression, and the expression of social factors such as unemployment, low income per capita, and family dis-cohesion (Rutter et al., 2008). Nonetheless, in the context of protective factors, traditional life (patriarchal society) and religion seem to be important. These findings might be explained in terms of family structure; for instance, extended families in Kosovo (e.g., several generations living together and having close relationships with each other) might provide the individual with emotional and financial support, which in turn might serve as protective factors for suicide. Also in cases of psychiatric disorders related to suicide (such as depression or PTSD), the supportive role of the extended family might be important. However, the influence of religious beliefs against the act of suicide might also be important (the value of life given by God is preached by Islam). Nonetheless, the above explanations are just tentative and future research in the field should examine these specific factors. To conclude, it might be said that although there are many valuable data of epidemiological nature of the suicide phenomenon in Kosovo, it is very difficult to develop a model for suicide in Kosovo. Despite the fact that war is mentioned in almost all studies, there is little evidence linking its consequences with suicide. Also none of these studies noticed the focus on specific cultural elements, although it seems they are important in the context of protective factors. Future research should consider this particular aspect, especially in developing preventive strategies for suicide in Kosovo.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research and/or authorship of this article.