The findings from the study and how they were used in the case against Ongwen illustrate four key things. First, the extent of serious violations of IHL experienced during the brief massacres in the three IDP camps in 2004 are far greater than what the average household in war affected Lango and Aholi sub-region experienced across the entire 20 + years of war. Second, the possible impacts of IHL violations are compounding and can last for at least a decade after the event, in a multi-generational manner. Third, despite the association of IHL violations and physical disability and psychosocial wellbeing, the affected population has less access to the health services they need, impeding long-term recovery. And fourth, getting data on the association of serious violations of IHL with key outcomes is not just possible but absolutely necessary to represent the experiences of the victims and determine appropriate reparations and national health policy.

Across the 20 + years of war, a representative survey of Acholi and Lango sub-region (using the same questions as our survey), found that on average, individuals experienced 0.34 serious violations of IHL, with those who experienced a serious violation averaging 2.5 violations [20]. In contrast, as a result of the LRA attacks against the three IDP camps in 2004, victims experienced an average of 6.9 serious violations of IHL. To put it more starkly, the VP experienced, on average, 20 times more serious violations of IHL compared to the average war affected person in Acholi and Lango, with children experiencing 14 times more violations than individuals in the sub-region representative survey. On average, each of the VP households experienced 22 serious violations, compared to the average household in Acholi and Lango where members experienced 2.3 violations per household over the 20 years of war [20].

Few studies look at the role and availability of health care for affected populations. Research among IDPs in Gulu and Amuru districts in northern Uganda shows that being ill without medical care had the strongest association with post-traumatic stress disorder and depression, with over half of all respondents having reported these symptoms; less than one-third being able to get the appropriate care [28]. Another study in Gulu district identified a lack of services available to victims of gender-based violence (GBV), specifically limited qualified staff and medical supplies to detect and manage survivors and services offered without ensuring confidential treatment and counseling [29]. A more recent review (2021) found that almost two-thirds of medical care professionals dealing with GBV in Uganda needed additional training and nearly half were uncertain or disagreed that there were clear protocols for care of survivors [30]. A study in three districts in northern Uganda shows how the conflict itself can directly affect the health services required by the affected population with direct attacks on health facilities, looting of medical supplies, and abduction of health providers [31]. While health care access was poor before the war, the limited evidence to date points to an even further weakened health care system that is unable to meet the physical and mental health care needs of the war-affected population. In northern Uganda, Betancourt et al. (2009) and Porter (2016) investigated how local government health providers dealt with mental health problems. They have provided important insights into how culture mediates what constitutes ill-health, its sources and manifestations, and solutions people seek to restore their health [32, 33].

The reports on serious violations of IHL shows that individuals and their households affected by a massacre are significantly different from the general war-affected population, in both their experience of IHL violations [20] and the possible impact of those violations. The VP are significantly worse off in terms of their psychosocial and physical well-being; these are directly associated with lower wealth and worse food insecurity, further exacerbated by less access to and lower rates of utilization of health services. The complexity of physical and mental health needs of the war-affected population is rarely addressed and redressed in the northern Uganda context [34]. The data also show that the association of IHL violations with worse outcomes is not only for the individual, but also for those in the households where individuals live, making recovery even more difficult. When it comes to the VP households, the association of IHL violations with worse outcomes is apparent in lower rates of school attendance by children who were not even alive at the time of the attacks. Not only does the continued association with worse outcomes point to possible multi-generational impact, but also exacerbates a possible ‘poverty trap’ (the poor cannot escape their poverty and with lack of resources only get poorer) for affected households created in part by their lower educational attainment [35]. The experiences of serious violations of IHL are compounded, with worse outcomes associated with more IHL violations.

The VP report longer travel time to health facilities, are less likely to have access (measured as a combination of travel, cost, and availability of services) to the services they need for routine or serious health problems and are more likely to report that the health services and medications that they need were not available. The difference in ‘access’ to, availability of, and travel time to appropriate health services between the GP and VP is not likely related to the use of different health centers, given the geographical proximity, but rather an indication of the significantly greater needs of the VP population. The IHL violations experienced by the VP populations is associated with far more complex psychosocial and therapeutic medical needs than services available at already poor and understaffed health centers. An in-depth analysis of the war-wounded within the same population in Uganda shows that there is a lack of the necessary treatments required for their ailments at the health centers, leading affected individuals to become disillusioned and discouraged from seeking the help they need [36]. Previous research in Uganda shows that when health services are targeted, they are much more likely to benefit combatants compared to their non-combatant peers as a result of programming by nongovernmental agencies focus on youth combatants [18].

The study also indicates that data on the association of IHL violations with key outcomes can be collected and serve as a new type of evidence to present before the ICC and to use to improve government policy. Data on the effects of war and serious violation of IHL are critical to give voice to the affected population in international cases and to affect national health policy. Witness or expert testimonies in court cases are traditionally limited to one aspect of the serious violations of IHL that take place. This study provides holistic evidence about the immediate and long-term association between the conflict and worse outcomes. The mix of quantitative data and testimonials allows us to demonstrate the effect of the human rights abuses on the lives of the victims, offer them a means to represent themselves, and offer the court tangible data for determining reparations. Based on the findings, the research team was able to give specific recommendations on how to target services and provide support to the most affected, strengthen psychosocial support and disability support, provide specialized therapeutic health services, provide education support, and the need for physical and monetary compensation for destroyed assets and livelihoods.

The findings have direct implications to international and government actors working in previously conflict affected contexts, not only in Uganda, but across the multitude of protracted humanitarian contexts driven in part by conflict. Once the initial conflict subsides, there is a tendency by development partners to treat populations as `post-conflict’, as though internal differentiation, including varying experiences of serious violations of IHL and related experience of disability and trauma, is unimportant. The findings from this study support the argument that post conflict national health actors and development partners cannot safely assume that everyone is recovering or recovering equally. Instead, these data expose disadvantages previously unappreciated, including persistent health inequality and multi-generational trauma to individuals, households, and communities.

Our study confirms that in post-conflict settings, war continues to be associated with long lasting and profound negative outcomes that need to be directly addressed in rebuilding war-damaged healthcare systems and treating the war-wounded. More attention should be given to the psychosocial and physical health needs of civilians suffering from IHL violations and disability, and the association with their disadvantage in receiving therapeutic treatment over time. Knowledge of the prevalence and negative relationship of war crimes with civilians’ mental and physical health, disability and access to health should be used to help develop more responsive post conflict health and psychosocial policies and services.

Because data collection occurred 13 years after the massacres the sample is only representative of the individuals who survived and did not migrate from the area. Thus, the sample prevalence of serious violations of IHL is likely biased, either downward because individuals who were murdered or disappeared are not included in the sample, or upwards because those who experienced fewer violations might have been able to migrate since the massacre. Because men were overwhelmingly more likely to be affected by direct combat, taken as child soldiers, or disappeared [37], our sample, purposely stratified by sex, is again not representative of the real sex-based experience of IHL violations. Another limitation is that the data is self-reported. Although the findings point to a multi-generational association between IHL violations and worse outcomes, having nutrition and health data collected from the children under five (using standard anthropometrical methods) would have been critical to prove this hypothesis. The analysis also operates under the assumption that prior to the attacks the respondents in the VA survey were more likely to resemble the general population and hence attributes differences in 2018 to the 2004 attack. Without data prior to the attack, we cannot confirm that this is the case; however, given that the GP sample is representative of the area from which the VA population was sampled and that the attacks were not targeted based on household characteristics, we feel confident that we are comparing two populations that more or less resembled each other prior to the massacre.