The end of a war does not end the tension and division between ethnic groups, nor does it eliminate its psychological and physical effects. Unresolved issues of ethnic conflict and identity in the past are reflected in every clash in the present. Ethnic-based aggression and defensive hostility continue to exist for decades within the social fabric of societies coming out of a conflict, and many individuals who have suffered from violence continue to suffer both physically and mentally [1].

The Kosovo war ended in June 1999 and during the last decade Kosovo was administered by the United Nations Mission in Kosovo (UNMIK). Security is provided by the NATO-led Kosovo Force (KFOR). A majority of the Serb population fled during the war to the north of Kosovo or to Serbia. All ethnic groups have continued to be exposed to ethnically-generated violence in the north of Kosovo and Serbian enclaves in the eastern part of Kosovo since 1999. In February 2008, violence escalated in Mitrovicë district following the declaration of independence by the Albanian majority in Kosovo.

The study in Kosovo forms the second part of a multi-country epidemiological study on massive exposure to violence and its health impact among the affected population. The first study was implemented in Meherpur district of Bangladesh (score 4 on Political Terror Scale [14]) in February-March 2008 [15, 16]. A further study will be implemented at a third site in 2010. The key components of this methodology are: 1) collection of statistical data and mapping information; 2) a fact-finding mission and key informant interviews; 3) a population-based study consisting of two components: a household survey followed by detailed screening of selected victims of OPV and human rights violations at mobile clinics. Statistical data were collected from the Ministry of Health of Kosovo and the Organization for Security and Co-operation in Europe (OSCE) mission in Kosovo. Mapping was excluded from the Kosovo study since we were unable to obtain the vector layer.

While it was shown earlier that the population in Mitrovicë district had more experience with traumatic events than other districts in Kosovo [13], our study demonstrated that Mitrovicë district was severely affected by ethnical conflict. Almost 90% of households experienced at least two categories of violence and human rights violations. Forced evacuation and displacement were frequently mentioned, and from the key informant interviews we learned that most people became traumatised after their return to Kosovo because they found their houses and property completely or partially destroyed. More than half (55%) of the houses were burned. They often found that close relatives or friends had been killed or were missing.

However, we did not intend for our results to represent the overall situation in Kosovo. The results of a district-level study, carried out 10 years after the war, cannot be directly compared with those based on national surveys [3, 4, 23–25], carried out during the war or some years ago. For instance, the prevalence of violence and human rights violations among Albanian refugees and civilians reported from other national-wide surveys was high. Cardozo et al estimated the prevalence of abuse and torture experiences at 48.9% in 1999, although it dropped to 11.7% in another survey in 2000 [3, 4]. The authors of above studies assumed that the discrepancy could be due to war survivors' failure to recall painful past events. However, one should not exclude the possibility that the torture experience reported by refugees or civilians at war time could be exaggerated owing to a wish to attract international aid or to gain asylum status. In 2005, KRCT found in a national-wide study in Kosovo that only 10.7% of the population reported being tortured [13], although "torture" was not clearly defined. In our study we found an even lower rate of 7% in Mitrovicë district. This could be explained because we used a strict definition of a torture case based on the UN convention. It could also be due to survivors experiencing avoidance or having memory block 10 years after the event.

The effects of violence are not confined to individuals. Kosovo is a collectivistic society; therefore it was appropriate for us to use the household as a unit to study the impact of collective exposure to violence and human rights violations. For example, if we considered experience of torture based on the UN definition, the figures for torture in Mitrovicë district were low. But if a wide variety of abusive treatments was considered, we found that over 30% of households were affected. Family members who are not directly attacked can become "secondary victims" because they witnessed what was happening to their loved ones. Emotional disturbances were reported among the war survivors and their children after the war [4, 26]. Emotional disturbance can be contagious among family members. Injury and disablement do not disappear with the end of a war, and taking care of an injured or a disabled person at home affects the whole family. When someone lives with persistent pain, everybody in the household shares the burden of the resulting anxiety, stress, sadness and depression. Family members often feel helpless and hopeless about providing care. Healthy family members are overloaded by assuming the duties of the person in pain, and they may leave their jobs to care for the victim, or quit school to meet the financial needs of the family as expenses increase and income declines (Table 6). Eventually more and more family members experience physical or mental pain and the entire family is haunted by the heavy emotional and financial burden. Non-specific pain could be the result of somatic and psychological expressions of emotional distress. Pain complaints increase with age and reach a plateau between the ages of about 45-75 years [27] or mostly of 55-85 years in a general population [28]. We also noted that pain complains increase with age from 35 years in a conflict-affected or repressed population [15]. Longitudinal evaluation of pain prevalence and the economic impact due to financial burden or job loss in post-war settings is needed. Assessment of quality-of-life of individual victim may provide similar information, but the prevalence of non-specific pain associated with the loss of property or financial burden on a family has never been assessed in a war-affected population, uprooted from the homeland or re-settled in the home country.

The crude mortality rate and under-five mortality rate are both indicators of the general health of a community. In Kosovo, both rates were slightly higher than the average for Serbia in 2007, but were compatible with a reasonable standard of community health. The Political Terror Scale showed that the level of terror declined from the highest level: 5 in 1999-2000 to the lower level: 2 [14] in 2007-2008. Restoring the security and the rule of law in the northern Kosovo can be expected to lower the mortality rate as well. Under-five mortality rate was lower than the crude mortality rate. It could be credited to the country's advanced material and child health care and high coverage of vaccination [22, 29]. This pattern is also observed in Croatia and Serbia (http://unstats.un.org/unsd/default.htm).

Kosovo has a long history of ethnic conflict; it could be difficult to isolate the impact of war in 1999 on a population in post-war setting although the experience of violence and human rights violations is endogenous, embedded within a complex web of personal, socio-economic and political factors before the war and ten years after the war. The results of the household survey showed that the social or political participation of an Albanian family could mean that family members are more likely to have been targets of the Serbian paramilitary or law enforcement agencies. Political involvement as a risk factor for OPV and human rights violations has also been reported in other locations [30–33]. On another hand, it is notable that in the politically involved families, the proportion of family members with pain complaints within two weeks was lower. It has been shown in other settings that patterns of social or political participation of a victim of politically-motivated violence can affect the rehabilitation process [16, 34]. Dimensions of empowerment such as self-determination, social inclusion and participation, as well as ability to influence or control one's life and environment have implications for the general health and well-being of a population affected by ethnic or political conflicts. Ethnic, political, and spiritual perspectives may affect symptoms, coping patterns and healing process as well [10]. The Kosovo's unilateral declaration of independence in February 2008 was regarded as a final victory over Serbia. Anyone who made sacrifices for their political claims has been rewarded by this act to some extent and their family members could be emotionally inspired as well. There is also the possibility that families involved with the ruling political party may have better access to financial resources or humanitarian aid.

The Serb population was probably under-sampled in our survey. It is hard to estimate this effect precisely, since there is no reliable figure for the size of the Serb population in Kosovo. We did find that many Serbs who were registered in the census had apparently left "mixed villages" in Albanian-dominated areas. Non-Serb interviewers were chased out from an isolated Serbian village, Bivolak, surrounded by the Albanian villages in Vushtrri municipality. Although Serb victims were invited, no Serbs attended the subsequent mobile clinics, perhaps because they were only deployed in Albanian-dominated areas. Ethnic segregation in Kosovo post-war healthcare system has been observed [35]. While some of the politically motivated Kosovo Albanians seem to be inspired by the change of the political landscape, it is unclear what level of difficulty and anxiety the Serbs may have in adjusting from being a majority to becoming a minority, with the resulting change in identity, during the Kosovo nation-building process. However, policy makers need to be aware of what factors might pose a challenge to reconciliation and reintegration and what can be a trigger for future violent clashes. Further study focusing on mental health of Serb population living in the post-war Kosovo is needed.

The study examined the long-lasting effects of ethnical conflict on health and financial burden from the family and collective aspects rather than from the individual aspect. Crude mortality rate, under-five mortality rate and annual injury rate in Mitrovicë district were low. Nevertheless, one fifth of the population had pain complaints reflecting the long-term impact. We discussed the risk factors for victimisation during an ethnic conflict, and we identified a variety of factors related to the proportion of family members with pain complaints. Higher level of lifetime exposure to violence, higher proportion of family members reporting lifetime violence-related injury and pain complaints and family's financial burden were found to be inter-correlated.

It is a big challenge to conduct research on conflict-related subjects in a society that is still polarised. In Mitrovicë district, both sides were still full of anger and often made accusations against each other, even 10 years after the war. It is essential to conduct population-based studies which can reflect and observe the changes in general health indicators, public attitude, physical, psychological or emotional needs of all the groups involved while implementing rehabilitation and re-integration programs in a post-war setting.