Case Definitions Used During the First 6 Months of the 10th Ebola Virus Disease Outbreak in the Democratic Republic of the Congo — Four Neighboring Countries, August 2018–February 2019

On August 1, 2018, the Democratic Republic of the Congo (DRC) declared its 10th Ebola virus disease (Ebola) outbreak in an area with a high volume of cross-border population movement to and from neighboring countries. The World Health Organization (WHO) designated Rwanda, South Sudan, and Uganda as the highest priority countries for Ebola preparedness because of the high risk for cross-border spread from DRC (1). Countries might base their disease case definitions on global standards; however, historical context and perceived risk often affect why countries modify and adapt definitions over time, moving toward or away from regional harmonization. Discordance in case definitions among countries might reduce the effectiveness of cross-border initiatives during outbreaks with high risk for regional spread. CDC worked with the ministries of health (MOHs) in DRC, Rwanda, South Sudan, and Uganda to collect MOH-approved Ebola case definitions used during the first 6 months of the outbreak to assess concordance (i.e., commonality in category case definitions) among countries. Changes in MOH-approved Ebola case definitions were analyzed, referencing the WHO standard case definition, and concordance among the four countries for Ebola case categories (i.e., community alert, suspected, probable, confirmed, and case contact) was assessed at three dates (2). The number of country-level revisions ranged from two to four, with all countries revising Ebola definitions by February 2019 after a December 2018 peak in incidence in DRC. Case definition complexity increased over time; all countries included more criteria per category than the WHO standard definition did, except for the "case contact" and "confirmed" categories. Low case definition concordance and lack of awareness of regional differences by national-level health officials could reduce effectiveness of cross-border communication and collaboration. Working toward regional harmonization or considering systematic approaches to addressing country-level differences might increase efficiency in cross-border information sharing.

On August 1, 2018, the Democratic Republic of the Congo (DRC) declared its 10th Ebola virus disease (Ebola) outbreak in an area with a high volume of cross-border population movement to and from neighboring countries. The World Health Organization (WHO) designated Rwanda, South Sudan, and Uganda as the highest priority countries for Ebola preparedness because of the high risk for cross-border spread from DRC (1). Countries might base their disease case definitions on global standards; however, historical context and perceived risk often affect why countries modify and adapt definitions over time, moving toward or away from regional harmonization. Discordance in case definitions among countries might reduce the effectiveness of cross-border initiatives during outbreaks with high risk for regional spread. CDC worked with the ministries of health (MOHs) in DRC, Rwanda, South Sudan, and Uganda to collect MOH-approved Ebola case definitions used during the first 6 months of the outbreak to assess concordance (i.e., commonality in category case definitions) among countries. Changes in MOH-approved Ebola case definitions were analyzed, referencing the WHO standard case definition, and concordance among the four countries for Ebola case categories (i.e., community alert, suspected, probable, confirmed, and case contact) was assessed at three dates (2). The number of country-level revisions ranged from two to four, with all countries revising Ebola definitions by February 2019 after a December 2018 peak in incidence in DRC. Case definition complexity increased over time; all countries included more criteria per category than the WHO standard definition did, except for the "case contact" and "confirmed" categories. Low case definition concordance and lack of awareness of regional differences by national-level health officials could reduce effectiveness of cross-border communication and collaboration. Working toward regional harmonization or considering systematic approaches to addressing country-level differences might increase efficiency in cross-border information sharing.
Ebola case definitions provided by the MOHs in DRC, Rwanda, South Sudan, and Uganda were compared during the first 6 months of the DRC outbreak. Because Rwanda, South Sudan, and Uganda had no reported cases at the time, their case definitions were for the preparedness phase of emergency response. Three dates for comparison were chosen to assess definitions: the start of the DRC outbreak (August 1, 2018), the period before the peak (November 15, 2018), and 6 months into the outbreak (February 1, 2019).
Criteria for five Ebola case definition categories (community alert, suspected, probable, confirmed, and case contact) were reviewed, accommodating minor wording differences. For example, a confirmed case category might have had three criteria: "suspected case, laboratory-confirmed by reverse transcriptionpolymerase chain reaction (RT-PCR)," "suspected case, laboratory-confirmed by IgM (immunoglobulin M) antibody presence," or "suspected case with a positive laboratory test." Alerts at points of entry (where travelers were screened for Ebola) were considered an additional category; point of entry alerts were either an independent category or described within the "community alert" category.
The number of criteria present for each country was divided by the total number of possible criteria listed by all of the four countries to calculate the percentage of criteria present per category. The category percentage concordance (overlap) across the four countries was calculated by dividing the number of criteria used by all countries by the total number of possible criteria for that category. Countries that did not use a category were excluded from that category's analysis. Rwanda, South Sudan, and Uganda, the three countries bordering DRC, reported changing their Ebola case definitions in response to the context and perceived risk for an outbreak, especially during the early months of the DRC outbreak. During the 6 months from August 1, 2018, through February 1, 2019, each country revised its case definitions two to four times. The interval between revisions varied from 1 month to 5 months. All four countries revised their definitions in January 2019 after DRC's Ebola incidence peaked in December 2018. Uganda did not include the probable category throughout the 6 months, and South Sudan removed that category by November 2018 (Table 1). Rwanda's case definition did not include a community alert category until January 2019. Only Uganda included the case contact category consistently throughout the 6 months, although other countries defined case contact at one of the three dates or might have listed criteria in other surveillance documents. Most countries listed nonbleeding symptoms of Ebola as criteria in suspected and community alert categories, except Uganda, where Ebola symptom criteria were limited to signs of unusual bleeding (Table 2). Some definitions had no thresholds for fever, and some had two thresholds concurrently in operation; where thresholds were defined, they ranged from ≥100°F (≥37.8°C) to 101.3°F (38.5°C). By the end of the 6 months, the suspected category included a higher proportion of feverdependent criteria (65%) than did the community alert category (30%). The proportion of fever-dependent criteria for the community alert category declined from 56% at the start of the outbreak to 30% 6 months later. Except for the case contact and confirmed categories, all countries included more criteria per category than did the WHO definitions. From August 2018 to February 2019, the total number of criteria in the suspected category increased from 12 to 20. Concurrently, concordance decreased from 18% to 5% (Figure). The most consistent criteria for a suspected case among all countries were fever, unexplained bleeding, sudden death, and prior contact with a person with suspected, probable, or confirmed Ebola.
However, fever was the only criterion consistently standing alone or upon which other criteria depended.
The probable category, present at least once for all countries except Uganda, always included the criterion deceased persons with a suspected case with an epidemiologic link to a case. For the DRC case definition, a probable case could be in a person alive with an epidemiologic link, and for DRC (at all three dates) and South Sudan (only at the start of the outbreak) a probable case could be based on a clinician's suspicion, even without an epidemiologic link. Percentage concordance for the probable category ranged from 33% to 100% (Figure). Laboratory confirmation by RT-PCR or detection of IgM antibody against Ebola virus in a suspected case were typically required for a confirmed case designation (Table 2).
Concordance in the criteria for case contact remained consistently low, increasing from 17% to 20% over the 6-month period (Figure). Initially there was zero concordance for community alerts; concordance increased to 25% in November 2018 but declined back to zero by February 2019. DRC defined point of entry criteria within the community alert category at the start of the outbreak. Rwanda and Uganda added point of entry categories by February 2019; Rwanda's  point of entry criteria were very similar to its criteria for community alert, and Uganda's were the same. South Sudan briefly included point of entry alert in October 2018 but removed it by February 2019.

Discussion
Because of the high volume of cross-border population movement between DRC and neighboring countries, strengthening binational and multinational public health communication and coordination is a growing priority. The four contiguous countries are currently reviewing their case definitions and developing procedures to engage in cross-border and regional collaborations that respond to and accommodate differences in case definitions to prevent cross-border transmission of Ebola. Case definitions might not move toward concordance among countries responding to an outbreak and countries in different stages of preparation for possible outbreak spread

Summary
What is already known about this topic?
Whereas countries might initially base case definitions on global standards, historical context and perceived risk often affect why countries modify and adapt disease case definitions over time, moving either toward or away from regional harmonization.
What is added by this report?
Even with a regional risk for Ebola virus disease (Ebola) (i.e., importation into three countries bordering the Democratic Republic of the Congo), Ebola case definitions became increasingly complex and less concordant during a 6-month period.
What are the implications for public health practice?
The low level of concordance in case definitions among countries, when case definitions are critical to many outbreak response and preparedness activities, indicates the need for routine evaluation of regional differences in case definitions and implementation of systematic approaches to advance harmonization.  (3,4). This analysis found a sustained low level of concordance in Ebola case definitions among DRC and three neighboring countries throughout revisions made over the first 6 months of the outbreak. As the number of criteria increased, case definitions became more complex, and concordance among countries decreased. DRC is operating in a response phase, and case definitions in the preparedness-phase countries might need to vary in sensitivity thresholds to identify cases based on available resources, perceived level of risk, and competing priorities (5). Complexity of and discordance in case definitions affect information sharing about alerts and cases across national borders. The potential risk associated with this discordance to cross-border communication and collaboration during an outbreak with a threat of cross-border spread might warrant a move toward regional harmonization or tailored binational and multinational communication strategies.

FIGURE. Percentage of concordance of
The findings in this report are subject to at least two limitations. First, although there was variation among countries in case definition sensitivity, this analysis did not evaluate the effect of discordance on surveillance; cases with in-country transmission have been limited to DRC. Second, MOHapproved case definitions are at the national level and might not represent those used by stakeholders at all levels, where local or cross-border informal information sharing might occur.
Awareness of differences in case definitions across the region provides critical fact-based support to national governments, regional or multinational bodies, and other public health stakeholders as they engage in or shift to preparedness and response initiatives with enhanced cross-border collaboration. Countries should consider routine evaluation of case definitions and implement systematic approaches to harmonization, when possible, and accommodate country-level differences when necessary. Revisiting these strategies throughout the continuum of preparedness and response might reduce the likelihood of cross-border transmission.