Isolated Case of Marburg Virus Disease, Kampala, Uganda, 2014

In September 2014, a single fatal case of Marburg virus was identified in a healthcare worker in Kampala, Uganda. The source of infection was not identified, and no secondary cases were identified. We describe the rapid identification, laboratory diagnosis, and case investigation of the third Marburg virus outbreak in Uganda.

(812601) was generated from the clinical specimen after a single passage in cell culture (Vero E6).
After laboratory confirmation, a multidisciplinary team from UVRI and CDC-Uganda performed the initial outbreak investigations at Mengo Hospital and Mpigi Health Center IV. The investigation team provided the outbreak case definition, and details of case contacts were obtained from both health facilities. An investigation was also performed in Kasese District, where the health worker was taken for burial. In addition, an ecologic investigation was conducted to identify roosting sites nearby for the potential presence of R. aegyptiacus bats.
We created an outbreak database using the Epi Info Viral Hemorrhagic Fever outbreak management application (10). All case and contact data were managed by the Uganda Ministry of Health Public Health Emergency Operations Center. We identified 197 close contacts, who were followed for 21 days. During the course of followup, 33 (16.2%) of the 197 contacts converted to suspected case-patients by exhibiting clinically compatible signs or symptoms matching the outbreak case definition. Blood samples from suspected case-patients were tested at the UVRI/CDC laboratory in Entebbe; all were negative for MARV by RT-PCR and serologic analysis.

Conclusions
We describe the second single-case filovirus outbreak detected in Uganda; a case of infection with Sudan ebolavirus was reported in Luwero District in 2011 (11). The investigation was unable to identify a conclusive source of infection, including evidence of the natural reservoir host near where the infected patient was working or residing or potential cases in persons who visited health centers before this case was confirmed. No secondary cases arose from contact with the initial case-patient.
The patient tested positive for malaria but later tested positive for MARV by RT-PCR and serologic analysis. These findings indicate that co-infection of viral hemorrhagic fever (VHF) and other tropical infectious diseases can confound diagnosis and delay early detection, potentially resulting in large outbreaks that are much harder to control, as was seen during the 2014-2015 Ebola virus outbreak in West Africa (12). This finding emphasizes the need for continued surveillance and awareness even when other, more common pathogens are initially suspected.
The finding of no secondary cases in this investigation can be attributed, in part, to use of infection control practices and personal protective equipment when first encountering a suspected VHF case. Because Uganda has experienced 10 VHF outbreaks since 2011, increased awareness and use of personal protective equipment and infection control practices have greatly limited secondary transmission, especially in the healthcare setting. Routine use of gloves, protective gowns, and chlorine is now more common in lower-level healthcare facilities in Uganda; these protective products were used at both Mengo Hospital and Mpigi Health Center IV. The full genomic sequence of this MARV isolate (812601; Genbank accession no. KP985768) falls into a cluster that consists mostly of MARV sequences isolated from bats. The closest related sequence was obtained from MARV isolated from a juvenile male R. aegyptiacus bat (Q843) captured in August 2009 in Python Cave, Queen Elizabeth National Park (Figure 2). Other viral sequences from bat specimens in this clade were from bats collected from either Python Cave or the Kitaka mine during 2007-2009. The most closely related human sequence (01Uga07) was from a miner who worked in the Kitaka mine in July 2007 (3,7). Because of the wide genetic diversity of MARVs, there is no definitive way to identify where this patient may have become infected.
We do not know why the patient did not transmit MARV to any of his close contacts. We can assume that none of the contacts had substantial exposure to the patient while he was infectious. The relatively small size of this and previous filovirus outbreaks in Uganda can be attributed to enhanced VHF surveillance, rapid case identification, laboratory testing, and investments from the global health security agenda in rapid sample transportation to the national VHF reference laboratory for diagnostic testing (13). VHF surveillance continues to be a top priority for Uganda, and the VHF surveillance program continues to play a crucial role in detecting, responding to, and helping to control these outbreaks.