|
|
|
|
PHR
|
An ‘attack’ is defined as a violent assault upon a facility resulting in any destruction, damage or loss of the facility’s function, equipment, or medical supplies.
|
|
SAMS
|
Any intentional act that may result (directly or indirectly) in:
- Damage to the health facility or reduction in its functionality;
- Loss or damage to health equipment, assets, transportation;
- Harm to the health workers.
|
|
WHO
|
Any act of verbal or physical violence or obstruction or threat of violence that interferes with the availability, access and delivery of curative and/or preventive health services during emergencies.
|
Looking at the three definitions presented in Table 4, the reporting of AHCC incidents by these mechanisms could produce dramatically different results given the same data points. With a focus on facilities, the PHR definition emphasizes forms of violence that are most likely to be caused by heavy weapons. (Other forms of violence would be captured in their separate, personnel-focused reporting tool.) In contrast, WHO’s definition is wider and includes not only physical violence but also verbal violence, which may be difficult to determine objectively. WHO’s definition expands further by referring to preventative health services, which implies a broader definition of who might count as a health worker and what might count as an AHCC in a humanitarian emergency. The SAMS definition includes a notion of intentionality on violence that directly or indirectly affects health facilities, supplies, vehicles or personnel. Intentionality, as a concept, has a legal value in establishing culpability for a crime, suggesting this definition was formulated with a sense of justice in mind.
Despite the broad range of categories of attack type provided in the WHO mechanism, some aspects of its categorization system can obscure understanding of an incident.
A standard definition of AHCC should cover all incidents that affect healthcare by impacting the availability, functionality or accessibility of curative or preventative health services. This effect could be a result of any of the following immediate impacts of such incidents:
- Damage to health facilities, assets, supplies or vehicles;
- Deaths or injuries among medical workers;
- Blocking access to medical equipment and supplies; or
- Loss of medical training.
Taxonomical discrepancies become more obvious when it comes to the classification and categorization of violence and impact. Each mechanism uses its own categorization, which is influenced largely by the purpose of the reporting. For example, both PHR and SAMS use their own detailed categorizations of modalities of attacks, based on the used weapons or violent behaviour. The WHO-led MVH mechanism has a less detailed typology of attacks, whereas the SSA tool has a clear but broad approach towards the categorization of incident type. As indicated in Table 4 above, the SSA uses the WHO definition of AHCC, which is: ‘any act of verbal or physical violence or obstruction or threat of violence that interferes with the availability, access and delivery of curative and/or preventive health services during emergencies’. While the same definition is used across WHO offices, the way each office determines which incidents should be reported can be slightly different, taking into consideration the local context. Attacks are placed in one of 15 categories, depending either on the type of assault or the affected health resources and using simple definitions that require no military knowledge. To address discrepancies between contexts, WHO is trying to build capacity and raise awareness in relation to AHCC definition and taxonomy.
Despite the broad range of categories of attack type provided in the WHO mechanism, some aspects of its categorization system can obscure understanding of an incident. The most prominent example raised from the Syria crisis is the category of ‘violence with a heavy weapon’. That is defined as ‘violence with a weapon that requires more than one person to use such as firearms, tanks, missiles, bombs, mortars[…]’. Under this definition, it is not possible to differentiate between an airstrike, shelling from small mortars, a blast from a tank cannon, or an attack with a high-calibre machine gun mounted in the back of a truck. Information that helps the user to understand if, for example, an AHCC incident was a result of mortar fire from the front line of conflict vs a barrel bomb dropped directly on a health facility in an urban area could be vital in providing understanding related to grave breaches of IHL, as well as the identity of the perpetrators. It would also prove useful for strategizing better risk mitigation measures related to patterns of attacks. In the case of Syria, some types of weaponry were used by only one side in the conflict – e.g. anything delivered by aircraft was deployed by the Syrian government or its allies. WHO considers this level of information outside of its mandate and the capability of health staff, but the information is often either already available in the public sphere or documented by agencies with rigorous practices for doing so, such as the VDC. In the case of the Syria conflict, the use of the term ‘violence with a heavy weapon’ actually obscures the picture of what is happening on the ground, contrary to the intent of the SSA. Indeed, in the Syrian context this categorization of types of attacks works against the uptake of the SSA system by parties such as front-line healthcare workers, who demand accountability for crimes perpetrated against them and their colleagues. This example suggests that the ‘incident type’ categories could be further expanded and better contextualized to fulfil the SSA’s mission of creating a global mechanism. Alternatively, the SSA could work with agencies producing verified information with additional details.
Data collection and verification practices
The data collection and verification processes examined in this study were influenced by field location and the data sources considered by each reporting agency. Mechanisms led by agencies with operations at the field level receive, firstly, incident alerts based on primary data sources, followed by verifying alerts, either through external partners or secondary sources. In contrast, mechanisms that do not have field presence initiate incident alerts based on secondary data, such as a media report, and then verify these alerts either through assessing primary data or other means of verifying secondary data, such as satellite imagery.
To offer an example, the process of collecting and verifying data in the WHO’s MVH (and later the SSA) involves using a variety of both primary and secondary sources. It can be initiated by WHO staff, health partners, eyewitnesses, media reports or any other general sources. The data are then verified through field observations, interviews with eyewitnesses, health partners’ reports, media content (e.g. photographs, videos) or satellite images. The verification process establishes a ‘certainty level’, ascribing higher levels of certainty to direct observation than to rumours or hearsay. When an incident is identified, a report form is submitted either by a WHO field office or by a partner organization; the WHO country office then verifies each incident using the methods outlined above. There are four certainty levels for each incident, each requiring different follow-up protocols:
- Rumour: e.g. claims made on social media, without additional proof. Such incidents will not be reported externally. The report will be kept in the internal dataset only and will be flagged as rumour.
- Possible: e.g. reports in the media. Further information is collected and corroborated for verification.
- Probable: e.g. one eyewitness or two secondary sources. Further information is collected and corroborated for verification.
- Confirmed: This category comprises incidents reported by direct observation by SSA partners, and details can be published immediately.
Once incidents are verified at WHO country office level, the information is published on the SSA webpage. The SSA team at WHO headquarters then checks the information through triangulation and cross-checking. Details of the incidents are logged at a later stage in a secure central database.
The type of data considered by each mechanism is influenced by that mechanism’s purpose. All mechanisms collect data on variables such as location, date, deaths and injuries that establish some basic facts, but other questions remain elusive and require more specialization. While mechanisms that have a legal focus tend to collect more data related to intent and responsibility for the attack, those focused on health outcomes collect more data related to the incident’s impact on healthcare. While data points related to AHCC incidents for justice and legal matters require rigorous proof, the range of data needed to measure the impact and public health implications of AHCC events is extremely heterogeneous, and it may not be possible accurately to determine the impact without additional research. Reporting systems concerned with health impact should also establish methodologies for impact measurement, to render impact data more tangible.
The use of technology is one of the main features of reporting AHCC in Syria. An excellent example of this is PHR’s use of satellite imagery to confirm an incident’s location during the verification process. At PHR’s request, the American Association for the Advancement of Science conducted an independent analysis of 15 high-resolution satellite images and was able to confirm two of four specific incidents that had been called into question. To give a further example, a collaboration took place between research teams at several academic institutions in the US and the SAMS in 2016, to develop and use a mobile application for data collection related to AHCC in Syria.
An overview of existing data collection and verification processes is provided in Table 5.
|
|
|
|
MVH
|
- Health cluster members send alerts of incidents via WhatsApp and an anonymized online data entry tool.
- Field staff seeks further information through interviews with eyewitnesses.
- Data are triangulated and flash updates are sent to partners.
- Data are sought from external partners.
- Monthly, or more frequent, verification takes place (for an incident to be verified, it needs to be reported by at least one health cluster member and one external partner).
|
|
PHR
|
- Secondary data to identify potential incidents.
- Targeted search, a systematic two-tier analysis of the credibility of both source and data content, data triangulation, comparison of multiple sources.
- Aerial and satellite imagery analysis of the location.
- Contacting medical organizations or personnel working in Syria. At least three independent sources or two credible sources, with a reviewing panel for each incident.
|
|
SAMS
|
- Health facilities and field monitors send alerts of incidents via WhatsApp and/or mobile app, email or any other available means of communication.
- Field monitors seek further information through observations and interviews.
- The incident is verified with partners working in the same area and/or staff of health facilities in the vicinity of the facility attacked.
- Verification is carried out on an incident-by-incident basis.
|
Substantial limitations continue to arise from these various processes. While there has been robust reporting of AHCC incidents in the Syrian conflict since 2016, there remains the possibility of unreported incidents. Factors that might limit AHCC reporting are related to the design of reporting mechanisms, the political sensitivity of the conflict and the lack of recognized health actors on the ground.
The political sensitivity and military complexity of the conflict presents additional complications. Reporting agencies in Syria might lack access to some areas, such as those controlled or besieged by the Syrian government or by NSAGs. It is noteworthy that there was very little reporting of AHCC incidents in the Syrian government-held areas, where the authorities control and restrict journalism and reporting, and have poor relations both with the international human rights community and with states opposed to the Assad regime. AHCC might also be under-reported in areas with a limited number of field health actors and activists. For example, very few incidents were reported in eastern governorates – such as Raqqa and Deir ez-Zor – that were controlled by ISIS between 2014 and 2017.
The purpose of reporting mechanisms
The purpose of any reporting mechanism reflects the politics of the organization or group which created it. There are two main purposes for reporting AHCC: an undisputable health-focused purpose, with the aim of improving both the humanitarian health response and the allocation of health resources; and a more politically divisive legal focus, with the aim of stopping the attacks and bringing perpetrators to justice. A third category of purpose, captured in the spirit of campaigns led by MSF, the ICRC and Safeguarding Health in Conflict cited in section 3, consists in the mobilization of political will to protect healthcare.
As stated above, the design of reporting mechanisms differs with respect to their purpose. Purpose affects everything, from the data points to be considered for inclusion and the data verification process to be followed, to communication plans and other uses of the data. Most healthcare-implementing agencies involved in reporting AHCC in Syria have aimed to serve multiple purposes, collecting information related to affected health resources as well as to intentionality and perpetrators. However, health actors are not well positioned to collect and communicate such sensitive data and, in some cases, must avoid doing so to avoid being targeted. This suggests that more collaboration is needed between legal and health actors to standardize this process.
At the opposite end of the spectrum, WHO’s SSA system is designed only to serve the health purpose of reporting. There remain some technical problems that obscure facts, and an absence of information that captures a spirit of justice. While WHO tries to preserve its neutrality by not becoming involved in sensitive data that might be politicized, the WHO-led MVH reporting did collect such data on a large scale. This was largely due to health cluster members, pushed by local health workers, who continued to collect and report this information with a conviction that it was the only hope for stopping such attacks. The preference among those reporting for justice mechanisms, together with the tensions caused by the neutering of the advocacy value of AHCC data, runs the risk of disengaging local actors. Syrian health workers, as reported during the interviews and the workshop, were disappointed with AHCC reporting because of its inability to catalyse action to prevent attacks. Despite this, many interviewees remained hopeful that the AHCC data that was gathered would, in future, provide evidence that would secure justice for the many victims of the attacks.
