Investigation of a Cluster of Severe Respiratory Disease Referred from Uganda to Kenya,February 2017

Abstract

On February 22, 2017, Hospital X-Kampala and US CDC-Kenya reported to the Uganda Ministry of Health a respiratory illness in a 46-year-old expatriate of Company A. The patient, Mr. A, was evacuated from Uganda to Kenya and died. He had recently been exposed to dromedary camels (MERS-CoV) and wild birds with influenza A (H5N6). We investigated the cause of illness, transmission, and recommended control. We defined a suspected case of severe acute respiratory illness (SARI) as acute onset of fever (≥38°C) with sore throat or cough and at least one of the following: headache, lethargy, or difficulty in breathing. In addition, we looked at cases with onset between February 1 and March 31 in a person with a history of contact with Mr. A, his family, or other Company A employees. A confirmed case was defined as a suspected case with laboratory confirmation of the same pathogen detected in Mr. A. Influenza-like illness was defined as onset of fever (≥38°C) and cough or sore throat in a Uganda contact, and as fever (≥38°C) and cough lasting less than 10 days in a Kenya contact. We collected Mr. A's exposure and clinical history, searched for cases, and traced contacts. Specimens from the index case were tested for complete blood count, liver function tests, plasma chemistry, Influenza A(H1N1)pdm09, and MERS-CoV. Robust field epidemiology, laboratory capacity, and cross-border communication enabled investigation.

Respiratory infections can be caused by a wide range of pathogens,¹ many of which can spread easily from person to person by droplet or aerosol.² Recently, novel respiratory pathogens have emerged, including zoonotic influenza viruses such as H7N9 and H5N1,^3,4 severe acute respiratory syndrome due to coronavirus (SARS-CoV), and Middle East respiratory syndrome coronavirus (MERS-CoV).⁵ Most populations are immunologically naïve to novel pathogens because they have not previously been exposed; thus, these infections can cause severe disease with high case-fatality rates. Cases or clusters of unusually severe respiratory disease, especially with relevant exposures, should be aggressively investigated to rule out novel pathogens.

Beyond their local effects, an outbreak of one of these pathogens potentially threatens global health security.⁴ To reduce the risk from these disease threats, the global public health community has in recent years focused on improving disease prevention, detection, and response.⁶ These efforts have included improvement of in-country capacities in laboratory diagnosis and outbreak response as stipulated by the International Health Regulations, which also emphasize data sharing between countries and cooperation to control outbreaks.⁷

On February 22, 2017, the Uganda Ministry of Health was notified by Hospital X in Kampala and by the US Centers for Disease Control and Prevention (US CDC) in Kenya of a fatal case of severe acute respiratory illness (SARI) in a 46-year-old male expatriate (Mr. A) who had been medically evacuated with his family members from Kampala, Uganda, to Nairobi, Kenya. At the time of notification, Mr. A had just died in Nairobi, and family members also reported respiratory symptoms. Preliminary investigation indicated that, during a recent trip to Dubai, Mr. A had been exposed to dromedary camels, which are known sources of MERS-CoV.^8-10 He had also recently been exposed to wild aquatic birds in Lake Mburo National Park in Uganda, at a time when highly pathogenic H5N6 had been reported in Uganda.^3,11 Alarmed by the possibility that this cluster could be due to MERS-CoV or a novel influenza virus infection, the ministries of health of Uganda and Kenya, in collaboration with partners, rapidly initiated an investigation and response on February 22, when the index case died.

Methods

Case Definition and Case Identification

After an initial communication to Uganda by the US CDC in Kenya, 2 parallel investigations began, the first to investigate the illness in Mr. A and identify any exposed or ill contacts in Kenya. This was led by the Kenya Ministry of Health, assisted by the US CDC in Kenya. The second investigation was to conduct extensive contact tracing in Uganda of individuals who may have been exposed before Mr. A's medical evacuation to Kenya; this investigation was led by the Uganda Public Health Fellowship Program and assisted by the US CDC in Uganda and the Uganda Ministry of Health.

In the Kenya investigation, we defined a case of SARI as acute onset of fever (≥38°C) plus sore throat or cough, with at least 1 of the following symptoms: headache, lethargy, or difficulty in breathing. We looked at cases with onset between February 1 and March 31 in people with a history of contact with Mr. A or his ill family member, or other employees at his company (Company A) in Kampala. In Uganda, we used the case definition for influenza-like illness (ILI) for case searching and contact tracing, since the early stages of SARI may present as ILI, and we aimed to capture all possible cases. We defined ILI as onset of fever (≥38°C) and cough or sore throat in a person who had contact with Mr. A or with any of his ill contacts. In Kenya, for family members who were contacts of Mr. A, we used a modified case definition of ILI: fever ≥38°C and cough lasting fewer than 10 days.

Within a day of Mr. A's death, we conducted an interview in Kenya with a proxy of Mr. A about his clinical course, potential exposures before his disease onset, and other relevant circumstances. We initiated active case-finding by searching for possible contacts of Mr. A among family members with whom he was evacuated from Kampala to Nairobi.

In Uganda, we traced contacts of both Mr. A and of any contact that became a case of SARI or ILI with no etiological confirmation. We searched for cases among the 124 workers who worked in the same office as Mr. A at Company A while he was symptomatic, by conducting group interviews with staff. In addition, we asked the human resources manager at Company A to alert workers to the possibility of circulating influenza disease and to report any worker absenteeism or illness due to suspected SARI or ILI.

The 10-year-old son of Mr. A (Child B) also reported respiratory disease symptoms; therefore, in Uganda, we initiated monitoring the health of pupils at the school that Child B attended. Teachers at the school were asked to report pupils with suspected ILI or SARI. We placed regular telephone calls to the school between February 26 and March 20 to ensure that no secondary cases had been identified. If a pupil developed symptoms consistent with ILI or SARI, we visited the contact's home to conduct a case investigation. We also established the reporting of illnesses meeting the case definitions for ILI and SARI among healthcare workers at Hospitals X and Y in Uganda. Lastly, we searched medical records in the only 3 intensive care units in Kampala (Hospitals V, W, and Y) to see if there had been any increase in SARI cases in the city, using the above case definition, over a 3-week period. Hospital X had no intensive care unit.

Specimen Collection and Testing

We collected nasopharyngeal swabs and a broncho-alveolar lavage sample from Mr. A, and a nasopharyngeal swab, broncho-tracheal aspirate, and sputum from Child B. We also collected nasopharyngeal swabs from family contacts in Nairobi who traveled with Mr. A and who reported some mild respiratory illness that did not meet the case definition for ILI or SARI. A broncho-alveolar lavage sample was collected from Mr. A to improve chances of diagnosis of MERS-CoV, if present, by real-time reverse transcription polymerase chain reaction (rRT-PCR).¹² An autopsy was performed on Mr. A on February 22, 2017, in Nairobi. We collected nasopharyngeal swabs, oropharyngeal swabs, and venous blood samples from contacts in Uganda who met the ILI or SARI case definitions.

We tested respiratory specimens from Mr. A and Child B and family members by rRT-PCR for MERS-CoV and influenza A, B, or C, and we subtyped any influenza A positives using seasonal (H1, H3) and novel influenza primers (H5, H7). We used the US CDC laboratory standard operating procedures for MERS-CoV and influenza molecular diagnosis for rRT-PCR, which are methods developed and validated by the US CDC. For MERS-CoV, we used the targets UpE, N2, and N3. For influenza rRT-PCR positive for influenza A, we subtyped with both influenza A(H1N1)pdm09 and H3, which in the index case was H1N1pdm09 positive. The US CDC Real-time RT-PCR (rRT-PCR) Protocol for Detection and Characterization of Influenza includes a panel of oligonucleotide primers and dual-labeled hydrolysis (Taqman^®) probes to be used in real-time RT-PCR assays for the in vitro qualitative detection and characterization of human influenza viruses in respiratory specimens and viral cultures.¹³ The influenza A and B primer and probe sets are designed for universal detection of type A and type B influenza viruses. Influenza A subtyping primer and probe sets are designed to specifically detect contemporary human A/H1, human A/H3, and A/H1N1pdm09 influenza viruses.

We tested the venous blood sample from Mr. A for viral hemorrhagic fever viruses by rRT-PCR. We tested the venous blood samples from Child B by rRT-PCR for viral hemorrhagic fever viruses, West Nile virus, Salmonella typhi, Rickettsia prowazekii, Leptospira interrogans, and Legionella pneumophila, since these organisms cause respiratory symptoms similar to SARI. We tested for hemorrhagic fevers because patients with hemorrhagic fevers present with high fever; our cases did not present with purpura or any other form of hemorrhage. We tested Child B's blood samples for malaria using the CareStart^TM malaria RDT test kit specific for Plasmodium falciparum histidine rich protein 2 antigen. We performed complete blood count tests using venous blood from Mr. A, Child B, and Mr. C.

Our investigation identified a single contact who met the ILI case definition (Mr. C), with illness onset on February 24, 2017. In addition to testing Mr. C for influenza A(H1N1)pdm09, influenza A(H5), influenza A(H7), and influenza B, we tested his samples by rRT-PCR for MERS-CoV, viral hemorrhagic fevers, malaria, West Nile virus, salmonellosis, rickettsiosis, leptospirosis, and Legionnaire's disease, using US CDC protocols at the reference laboratory in Uganda.¹⁴

We tested the nasopharyngeal swab from Mr. C for 32 etiologies of respiratory illnesses by US CDC laboratory standard operating procedures for rRT-PCR, comprising influenza A/B, influenza A(H1N1)pdm09, influenza A(H3), influenza A(H5), enterovirus, human bocavirus, Mycoplasma pneumoniae, human respiratory syncytial virus, human metapneumovirus, human adenovirus, rhinovirus, human parainfluenza viruses 2/3/4, human parechovirus, coronavirus 43/229, human coronavirus 1, parainfluenza virus 1, Pneumocystis carinii, Legionella pneumophila, Klebsiella pneumoniae, Salmonella typhimurium, Moraxella catarrhalis, Bordatella pertussis, Streptococcus pneumoniae, Staphylococcus aureus, Chlamydia pneumoniae, and Haemophilus influenza B. These organisms can cause symptoms similar to SARI or ILI.

Mr. A and his son, Child B, received special medical care and had a battery of laboratory tests from the hospitals in Uganda and Kenya because they presented with a grave illness of no obvious etiology and possibly of epidemic potential; they were foreign, with a recent history of travel to Dubai (where they had been exposed to MERS); they had comprehensive health insurance coverage. Thus, all possible efforts were made to save their lives through healthcare services available at the time. A wide array of laboratory tests were done on specimens from Mr. A, Child B, and Mr. C to identify the etiological agent of their illnesses and to assist with epidemiologic investigations in order to institute evidence-based control measures. Mr. C, who was native to Uganda and a workplace close contact of Mr. A, also was tested comprehensively beyond the routine level of care for patients on suspicion that he had contracted the same illness as Mr. A and Child B.

Hospitals X and Y in Uganda received laboratory data from the testing lab in Uganda and subsequently relayed the information to the investigating team at the Ministry of Health and CDC Uganda in Kampala. Hospital Z in Nairobi also received laboratory information from their testing laboratory while treating and caring for the suspected cases.

CDC Kenya, with relevant health workers in Kenya, provided additional demographic, clinical, and laboratory information regarding the suspected and confirmed SARI cases, to CDC Uganda in order to aid investigations in Uganda.

Results

Clinical Course and Exposures

On February 6, Mr. A, a 46-year-old man, and his family, including his 10-year-old son (Child B), visited Lake Mburo National Park, Uganda, when avian influenza H5N8 had been recently reported in migrating wild bird populations along the shores of Lake Victoria. Mr. A and family (including his wife, a son, and 3 daughters) then traveled from Kampala to Dubai during February 11 to 18 for a vacation, during which Mr. A reportedly had contact with a dromedary camel, although we could not establish the details of this interaction. He returned to Kampala on February 19 and developed an influenza-like illness on February 20, with cough and fever. The next day, he rapidly developed severe symptoms (fever, cough, difficulty breathing). Mr. A worked in his office on February 20 and 21; during these 2 days, he shared rooms in meetings and shook hands with several co-workers. He also shared meals and accommodation and had other contact with his family members at home.

On February 21, Mr. A was hospitalized briefly at Hospital X in Kampala, where he was diagnosed with pneumonia (in addition to fever ≥38°C, cough, sore throat, difficulty in breathing, headache), before being airlifted to a Hospital Z in Nairobi, Kenya, the same day. He died of multiple organ failure on February 22. Mr. A had no known underlying illnesses or medical conditions.

On the evening of February 22, Mr. A's son, Child B, also developed respiratory symptoms (fever ≥38°C, cough, sore throat, difficulty in breathing, headache) and was hospitalized in Hospital Y in Kampala. Within 4 hours, he was flown to Hospital Z in Nairobi, where he was treated for cough in intensive care as a precautionary measure and discharged on February 24.

Contact Tracing

Our active case finding identified 1 secondary case, Mr. C, among Mr. A's contacts. Mr. C was a co-worker of Mr. A who reported that he had shaken hands with Mr. A while Mr. A was symptomatic. On February 24, Mr. C developed ILI (fever ≥38°C, cough, headache) and was admitted to Hospital Y in Kampala. He was treated with antibiotics and given other supportive treatments and was discharged on February 27 (Table 1).

Table 1.

Demographics, Timelines, and Clinical Details of Mr. A, Child B, and Mr. C

Features	Mr. A, Index Case	Child B, Son of Index Case	Mr. C, Co-worker of Index Case
Age in years	46	10	24
Sex	male	male	male
Race	Caucasian	Caucasian	Black African
History of travel outside Uganda 3 days before onset of symptoms	Yes, spent time in Dubai, exposed to dromedary camels Feb. 11-18, 2017	Yes, spent time in Dubai Feb. 11-18, 2017	No; co-worker of Mr. A; shook hands while Mr. A was symptomatic
Date of onset	Feb. 20, 2017	Feb. 22, 2017	Feb. 24, 2017
Date of hospitalization	Feb. 20, 2017	Feb. 22, 2017	Feb. 24, 2017
Date of discharge	Fatal case (died Feb. 22, 2017)	Feb. 24, 2017	Feb. 27, 2017
Symptoms at admission	Sore throat, cough, fever >38°C, difficulty in breathing (pneumonia), headache	Sore throat, cough, fever >38°C, difficulty in breathing, headache	Cough, fever >38°C, headache
Site of hospitalization	Hospital X in Kampala (Uganda), referred to Hospital Z in Nairobi (Kenya)	Hospital Y in Kampala (Uganda), referred to Hospital Z in Nairobi (Kenya)	Hospital Y in Kampala, Uganda
Initiation of clinical and laboratory investigations	Feb. 20, 2017	Feb. 22, 2017	Feb. 24, 2017
Lab confirmation of influenza A(H1N1)pdm09	Feb. 22, 2017	Feb. 23, 2017	Not applicable
Initiation of epidemiologic investigations	Feb. 26, 2017	Feb. 26, 2017	Feb. 26, 2017

The index case (Mr. A) and the only fatal case was a 46-year-old Caucasian male with a history of recent travel to Dubai, where he was exposed to dromedary camels. Epidemiologic investigations commenced after laboratory confirmation of influenza A(H1N1)pdm09 in the index case and, in Child B, influenza A(H1N1).

Follow-up in Uganda included intensive care staff of Hospitals V and W, 465 hospital staff (72 in Hospital X and 393 in Hospital Y, where cases were treated), 124 employees of the company where Mr. A and Mr. C worked, the 40 classmates of Child B at the school the child attended, and 4 additional family members of Mr. A. This follow-up did not identify additional cases of ILI or SARI. We monitored staff in Hospital Z in Nairobi who were involved with the treatment of Mr. A and his son and found no additional cases. Family members of the index case were primary contacts, while hospital contacts and school contacts were mostly secondary contacts. The attack rate was 2/1,000.

The reproduction or transmissibility (R0) rate for any infectious agent refers to the average number of secondary cases of disease generated by a typical primary case in a susceptible population; an R0 rate of 1.0 would indicate no transmission, where the disease declines naturally. R0 values more than 1 means the disease can cause an epidemic. We confirmed only 1 secondary case (Child B) who lived in the same household as the index case (Mr. A) despite investigating several co-workers and hospital workers in various hospitals in Uganda and Kenya over a 3-week period. Influenza transmissibility rate ranges from 1 to 3 in a space of 2 to 3 days,¹⁵ and in this case it was 1 (confirmed secondary case)

Laboratory Test Results

Prior to Mr. A's being evacuated to Kenya, results of lab tests done on him in Uganda (Hospital X) revealed nothing beyond lymphocytopenia, elevated liver enzymes, and elevated D-dimer, all of which pointed to a viral respiratory infection given his signs and symptoms at the time (Table 2). From the same hospital in Uganda, Mr. A's urinalysis, red blood cell indices, and blood chemistry were normal, while a malaria rapid diagnostic test was negative for Plasmodium falciparum.

Table 2.

Laboratory and Pathology Investigation Results of 3 Cases: Mr. A, Child B, and Mr. C, February 2017, Kampala, Uganda, and Nairobi, Kenya

Case	Specimen Collected	Laboratory Test Done	Result of Test
Mr. A, 46-year-old index case	Broncho-alveolar lavage fluid, Naso/oropharyngeal (NP/OP) swab	rRT-PCR for Influenza A(H1, H3, H5, H7), subtyped	Positive: influenza A(H1N1)pdm09
	Broncho-alveolar lavage fluid, naso/oropharyngeal swab	rRT-PCR for MERS-CoV	Negative
	Serum	rRT-PCR for viral hemorrhagic fevers	Negative
	Serum	RT-PCR chikungunya	Negative

	Whole blood Whole bloodWhole bloodUrineClinical examination	WBC totalRBC indicesBlood chemistryLiver function tests:Alkaline phosphatase 173 IU/l (ref 51-117); S-ALT 36 IU/L (ref: 7-35), S-AST 42 IU/L (ref: 13-35)Malaria RDTD-dimerUrinalysisChest x-ray	0.93x10⁹/L (normal 4-12x10⁹/L), lymphopenia with relative neutrophiliaNormalNormalElevated Negative for P. falciarum3,568.8ng/mL (elevated)NormalPulmonary abnormalities
	Autopsy	Autopsy	Fulminant pneumonia, multi-organ failure
Child B, 10-year-old son of index case	Naso/oropharyngeal swab, sputum, broncho-tracheal aspirate	rRT-PCR for influenza A & B, subtyped	Positive: influenza A(H1N1)pdm09^*
	Naso/oropharyngeal swab	rRT-PCR for MERS-CoV	Not done
	Naso/oropharyngeal swab	rRT-PCR for bacterial and viral respiratory etiologies	Negative
	Serum	rRT-PCR for viral hemorrhagic fevers: Ebola, Marburg, Lassa, dengue, yellow fever virus	Negative
	Serum	rRT-PCR chikungunya	Negative
	Whole bloodUrine	WBC totalUrinalysis	4.37x10⁹/L, (normal 4-12x10⁹/L)Normal
Mr. C, 24-year-old colleague of Mr. A	Naso/oropharyngeal swab	rRT-PCR for influenza A & B, subtyped	Negative
	Naso/oropharyngeal swab	rRT-PCR for MERS-CoV	Negative
	Naso/oropharyngeal swab	rRT-PCR for bacterial and viral respiratory etiologies^*	Negative
	Naso/oropharyngeal swab	rRT-PCR for viral hemorrhagic fevers: Ebola, Marburg, Lassa, dengue, yellow fever virus	Negative
	Serum	rRT-PCR for Ebola virus	Negative
	Serum	rRT-PCR chikungunya	Negative
	Whole blood	WBC total	13.25x10⁹/L^* (normal 4-12x10⁹/L)
	Urine	Differential white cell countOther hematology testsLiver and renal function testsHemoparasites microscopyUrinalysis	Neutrophilia 79%^*NormalNormalNo hemoparasites seenNormal

Significant test results. rRT-PCR is real time reverse transcription polymerase chain reaction. Child B was tested for the following bacterial etiologies of respiratory tract infection: Chikungunya virus, West Nile virus, Salmonella typhi, Rickettsia prowazekii, Leptospira interrogans, and Legionella pneumophila. Mr. C was tested for the following bacterial and viral etiologies of respiratory tract infection: influenza A(H1N1)pdm09, influenza B, H5 and H7 influenza viruses, MERS-CoV infection, viral hemorrhagic fevers, dengue fever, chikungunya fever, malaria, West Nile virus disease, salmonellosis, rickettsiosis, leptospirosis, and legionnaires' disease. Candidate bacterial etiologies of respiratory tract infections were tested for since they can cause infections that present as SARI.

In Kenya (Hospital Z), rRT-PCR tests on Mr. A's respiratory specimens and those of his son, Child B, tested positive for influenza A (H1N1)pdm09 virus infection. Tests were negative for other etiologies, including MERS-CoV (Table 2). An autopsy done on Mr. A revealed fulminant pneumonia and multi-organ failure. Family members of Mr. A tested negative for influenza A/B by RT-PCR. The nasopharyngeal and oropharyngeal swabs from Mr. C were negative for all respiratory pathogens tested, including influenza A (H1N1)pdm09 and MERS-CoV. However, his white blood cell differential count demonstrated neutrophilia, suggesting a bacterial infection (Table 2). Mr. C should have been tested more thoroughly for bacterial chest infection by sputum culture, but this was not done before he was treated empirically with antibiotics, given his white blood count profile.

Discussion

Here we report an investigation of a cluster of respiratory disease with a fatality, involving both Uganda and Kenya health authorities, most likely caused by influenza A(H1N1)pdm09.¹⁶ To our knowledge, this is the first systematically reported epidemiologic investigation of a cluster outbreak of influenza A(H1N1)pdm09 in Uganda involving cross-border communication and data sharing with a neighboring country, with unique challenges and lessons learned. Importantly, the initial exposure history of this case pointed to several other epidemic pathogens, such as novel influenza A or MERS-CoV, and this investigation exemplifies the importance of information sharing between 2 neighboring countries in East Africa to rapidly investigate a disease of epidemic potential. Timely communication between the countries, with significant inputs by the US CDC, allowed for a full investigation in both countries. Moreover, this investigation highlights how investments in laboratory and epidemiologic capacity can result in effective international investigation with extensive contact tracing. This allows for rapid identification of the etiology and any secondary cases, and for the ruling out of novel pandemic threats, which is key to effective prevention. Although this cluster of cases did not spread, this type of collaboration is critical to prevent spread in a similar future situation, with influenza or other pathogens.

We value the role of contact tracing as a major contribution to ensuring that no primary or secondary contacts succumbed to SARI. Both Uganda and Kenya have surveillance systems for influenza and other disease, with the US CDC and the World Health Organization (WHO) playing major roles in strengthening systems, but this incident reveals opportunities for improvement in event-based surveillance across the affected national frontiers, including diseases spread from person to person.¹⁷ In this case, potentially fatal cases posed threats of large outbreaks of influenza A (H1N1)pdm09 or of other novel influenza strains affecting humans and domestic and wild animals across the geographical boundaries of Kenya and Uganda.

Our article also demonstrates the importance of strong laboratory and clinical systems for the investigation and clinical management of SARI for routine patient care, and the need to capture the information in influenza surveillance systems. The laboratory testing and capacity was sufficient and rapid enough in Uganda and Kenya, given that we used influenza reference laboratories in the 2 countries with state-of-the-art facilities, equipment, quality assurance, and personnel to ensure standard services.

Considering the onset dates for the 2 cases, the known incubation period of influenza A, and the fact that Child B was a close household contact of Mr. A during his illness, Child B was likely infected by Mr. A. Although influenza A virus is highly contagious, especially within enclosed environments such as households,^18,19 it seems that the index case infected only his son, whose onset date was 2 days after that of his father, and no further spread occurred to family, professional, or school contacts.

The death of Mr. A was unusual; most fatalities associated with seasonal influenza occur in very young children, the elderly, or in people with underlying health conditions. However, this death due to influenza A (H1N1)pdm09 virus infection of a middle-aged man with no known underlying medical conditions serves as a reminder that adults in good general health also can be vulnerable to severe influenza disease. Indeed, people of all ages without underlying medical conditions can die from severe influenza A(H1N1)pdm09,¹⁸ and the speed of the disease progression in Mr. A to death in 2 days likely indicates a fulminant primary viral pneumonia, possibly in combination with some other undiagnosed etiologies. Influenza A(H1N1)pdm09 virus is more often associated with viral pneumonia than other influenza viruses.²⁰

A limitation of our investigation is that we did not ascertain the vaccination status of Mr. A or his family members, which might have further explained the spread of illness among contacts and the severity of illness in the index case. Moreover, we did not collect blood and sputum cultures, and so some potential pathogens and the diseases they cause (eg, plague, melioidosis) were not investigated. Burkholderia pseudomallei, the etiological agent of melioidosis that presents as severe pneumonia, was not tested for in the index case.²¹ However, Burkholderia pseudomallei and bacteria in general would have been an unlikely etiology. since the index case had severe leucopenia attributed to viral pneumonia.

The investigation highlights the role of private hospitals in disease management, surveillance, and reporting, as was evident in the communication of clinical, laboratory, and epidemiological information that was initiated between 2 hospitals, one in Uganda (Hospital X) and the other in Kenya (Hospital Z), while arranging for the referral of the index case. A similar pattern was repeated while handling the second case, Child B. It appears that this was an opportunity for the ministries of health in Uganda and Kenya to initiate an international response, but this did not occur immediately, mainly because of the independent mode of operation of most private hospitals in Africa.

International travelers in Uganda, as well as individuals covered by health insurance, such as Mr. A and his family, often prefer to use the services of private hospitals that subsequently share information with relevant government authorities and partners. Because of better health insurance coverage that enabled cross-border referral from Uganda to Kenya, the index case and his son received clinical care that went beyond the routine level of care for patients; this makes the investigation largely unreproducible for patients without health insurance.

Logistically, the investigation was facilitated through the cooperation of private hospitals that shared information between themselves and with their respective ministries of health in Uganda and Kenya, with the help of the US CDC based in both countries. It therefore became unnecessary for the epidemiology investigators to travel across borders to obtain information, given the ample data shared. We observed that communication between private hospitals and government in this instance can critically affect the speed of the response. Infectious disease surveillance and communication of the information in Uganda and Kenya are guided by WHO-AFRO Integrated Disease Surveillance and Response (IDSR) strategy, developed in 1998 as a comprehensive strategy to improve disease surveillance and response in WHO member states in Africa and adopted in Uganda in 2000.^22-25 The current IDSR guidelines for Uganda provide for recommendations for surveillance and responses to alert thresholds for influenza-like illnesses.²⁶

We propose that the governments of Uganda and Kenya should incorporate IDSR training into preservice curricula for health workers and strengthen regular support supervision in public and private hospitals. This is already a proven strategy known to maintain and strengthen communication between private hospitals and public health authorities regarding the surveillance of diseases with epidemic potential.^23,25,27 It is legally allowable for health ministries to share information regarding diseases of epidemic importance within the East African Community, an economic organization to which Uganda and Kenya belong. Proper channels were followed throughout the investigation.

The global nature of public health threats is now well recognized, and efforts to bolster global health security and stop diseases at their source are an increasing priority, particularly since the recent West African Ebola outbreak.²⁸ For respiratory pathogens, efforts such as the US CDC's DaRRE (Detection and Response to Respiratory Events),²⁹ which focuses on expanding and improving surveillance for respiratory diseases to reduce the risk of outbreaks, are critical to achieving improved global health security.

Improved surveillance as well as strong international collaboration are vital to reducing the risk from novel as well as known respiratory pathogens. In recent years, sub-Saharan African countries have built considerable capacities for detecting, responding to, and controlling outbreaks at their sources.³⁰ The most important measure to prevent influenza spread is vaccination, and despite WHO recommendations,^31,32 very few African countries have adopted influenza vaccination policies; the coverage is very low in Uganda according to a 2014 report.³³ Future cases of SARI could be better managed through rapid laboratory testing methods and thorough screening of individuals with recent travel to influenza-endemic countries. To improve communication and information sharing between Uganda and Kenya regarding disease outbreaks and epidemics, the coordinating offices in the respective ministries of health should establish open communication channels that bypass slow or bureaucratic processes.

The newly established Africa Centres for Disease Control and Prevention (Africa-CDC) and improved regional cooperation through development (such as the East African Community) and strengthening of national public health institutes and emergency operations centers will be key to improving disease surveillance, responding to future events, and mitigating risks of full blown epidemics that threaten global health security.

Footnotes

Acknowledgments

We thank the administrators at the hospitals where case-patients were admitted and the medical teams who made it possible to investigate this cluster of cases and provided support. We appreciate the African Field Epidemiology Network (AFENET) for funding a 1-week technical manuscript writing workshop in September 2017, in which draft 0 of this manuscript was developed. We also appreciate the clinicians and others in the various hospitals where cases were treated for the valuable information they provided to us during the investigation. The Ministry of Health of Uganda gave the directive and approval to investigate this outbreak, and so the need for ethical approval was waived by the institutional review board of Makerere University School of Public Health. In agreement with the International Guidelines for Ethical Review of Epidemiological Studies by the Council for International Organizations of Medical Sciences (1991), the Office of the Associate Director for Science, US CDC/Uganda, determined that this activity was not human subjects research and its primary intent was public health practice or a disease control activity (specifically, epidemic or endemic disease control activity). Verbal informed consent was obtained from the participants before the start of each interview. The purpose and nature of the investigation were explained to all participants. Participants were also informed that their involvement was entirely voluntary and their refusal to answer any or all of the questions would not result in any negative consequences. Participants identified as patients were referred for treatment in Uganda or Kenya. To protect participants' confidentiality, personal information was de-identified during data analysis, and the interview forms were locked up.

This project was supported by the President's Emergency Plan for AIDS Relief (PEPFAR) through the US Centers for Disease Control and Prevention Cooperative Agreement number GH001353-01, through Makerere University School of Public Health to the Uganda Public Health Fellowship Program, Uganda Ministry of Health. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the US Centers for Disease Control and Prevention, the Agency for Toxic Substances and Disease Registry, the US Department of Health and Human Services, Makerere University School of Public Health, or the Uganda Ministry of Health.

References

Hunt

, ed. Microbiology and Immunology On-line. University of South Carolina; 2000. https://www.microbiologybook.org/. Accessed February 19, 2020.

Pourbohloul

, Meyers

, Skowronski

, Krajden

, Patrick

, Brunham

. Modeling control strategies of respiratory pathogens. Emerg Infect Dis, 2005; 11(8):1249-1256.

, Cao

. Pandemic and avian influenza A viruses in humans: epidemiology, virology, clinical characteristics, and treatment strategy. Clin Chest Med, 2017; 38(1):59-70.

Widdowson

, Bresee

, Jernigan

. The global threat of animal influenza viruses of zoonotic concern: then and now. J Infect Dis, 2017; 216(Suppl 4):S493-S498.

Wang

, Sun

, Zhu

, Zhao

. Current understanding of Middle East respiratory syndrome coronavirus infection in human and animal models. J Thorac Dis, 2018; 10(Suppl 19):S2260-S2271.

Fidler

, Gostin

. The new International Health Regulations: an historic development for international law and public health. J Law Med Ethics, 2006; 34(1):85-94.

World Health Organization. International Health Regulations (2005). 3d ed. Geneva: WHO; 2016. https://www.who.int/ihr/publications/9789241580496/en/. Accessed February 19, 2020.

Koçer

, Krauss

, Zanin

, et al. Possible basis for the emergence of H1N1 viruses with pandemic potential from avian hosts. Emerg Microbes Infect, 2015; 4(7):e40.

Mackay

, Arden

. MERS coronavirus: diagnostics, epidemiology and transmission. Virol J, 2015; 12(1):222.

10.

Mohran

, Farag

, Reusken

, et al. The sample of choice for detecting Middle East respiratory syndrome coronavirus in asymptomatic dromedary camels using real-time reverse-transcription polymerase chain reaction. Rev Sci Tech, 2016; 35(3):905-911.

11.

Omrani

, Al-Tawfiq

, Memish

. Middle East respiratory syndrome coronavirus (MERS-CoV): animal to human interaction. Path Glob Health, 2015; 109(8):354-362.

12.

Arabi

, Balkhy

, Hayden

, et al. Middle East respiratory syndrome. N Engl J Med, 2017; 376(6):584-594.

13.

Centers for Disease Control and Prevention. US CDC Real-time RT-PCR (rRT-PCR) Protocol for Detection and Characterization of Influenza.

14.

Katz

, Muthoka

, Emukule

, et al. Results from the first six years of national sentinel surveillance for influenza in Kenya, July 2007-June 2013. PLoS One, 2014; 9(6):e98615.

15.

Biggerstaff

, Cauchemez

, Reed

, Gambhir

, Finelli

. Estimates of the reproduction number for seasonal, pandemic, and zoonotic influenza: a systematic review of the literature. BMC Infect Dis, 2014; 14:480.

16.

Dawood

, Iuliano

, Reed

, et al. Estimated global mortality associated with the first 12 months of 2009 pandemic influenza A H1N1 virus circulation: a modelling study. Lancet Infect Dis, 2012; 12(9):687-695.

17.

Sanicas

, Forleo

, Pozzi

, Diop

. A review of the surveillance systems of influenza in selected countries in the tropical region. Pan Afr Med J, 2014; 19:121.

18.

Heymann

, ED. Control of Communicable Diseases Manual. 20th ed. Washington, DC: American Public Health Association; 2015.

19.

Casado

, Martinez-Baz

, Burqui

, et al. Household transmission of influenza A(H1N1)pdm09 in the pandemic and post-pandemic seasons. PLoS One, 2014; 9(9):e108485.

20.

Ishiguro

, Takayanagi

, Kanauchi

, et al. Clinical and radiographic comparison of influenza virus-associated pneumonia among three viral subtypes. Intern Med, 2016; 55(7):731-737.

21.

Perumal Samy

, Stiles

, Sethi

, Lim

LHK

. Melioidosis: clinical impact and public health threat in the tropics. PLoS Negl Trop Dis, 2017; 11(5):e0004738.

22.

World Health Organization. Integrated disease surveillance in African: a regional strategy 1999-2003. Geneva: WHO; 1998. https://apps.who.int/iris/handle/10665/1749. Accessed February 19, 2020.

23.

Masiira

, Nakiire

, Kihembo

, et al. Evaluation of integrated disease surveillance and response (IDSR) core and support functions after the revitalisation of IDSR in Uganda from 2012 to 2016. BMC Public Health, 2019; 19(1):46.

24.

Phalkey

, Yamamoto

, Awate

, Marx

. Challenges with the implementation of an Integrated Disease Surveillance and Response (IDSR) system: systematic review of the lessons learned. Health Policy Plan, 2015; 30(1):131-143.

25.

Nsubuga

, Nwanyanwu

, Nkengasong

, Mukanga

, Trostle

. Strengthening public health surveillance and response using the health systems strengthening agenda in developing countries. BMC Public Health, 2010; 10(Suppl 1):S5.

26.

Republic of Uganda Ministry of Health. National Technical Guidelines for Integrated Disease Surveillance and Response. Kampala, Uganda; 2012. http://health.go.ug/sites/default/files/IDSR%20NTG%20for%20Integrated%20Disease%20Suveillance.pdf. Accessed February 19, 2020.

27.

Lukwago

, Nanyunja

, Ndayimirije

, et al., The implementation of Integrated Disease Surveillance and Response in Uganda: a review of progress and challenges between 2001 and 2007. Health Policy Plan, 2012; 28(1):30-40.

28.

Abramowitz

, Hipgrave

, Witchard

, Heymann

. Lessons from the West Africa Ebola epidemic: a systematic review of epidemiological and social and behavioral science research priorities. J Infect Dis, 2018; 218(11):1730-1738.

29.

Kenya

, Annual Report, 2014. 2017.

30.

Fitzmaurice

, Mahar

, Moriarty

, et al. Contributions of the US Centers for Disease Control and Prevention in implementing the Global Health Security Agenda in 17 partner countries. Emerg Infect Dis, 2017; 23(Suppl 1):S15.

31.

World Health Organization. Vaccines against influenza. WHO position paper—November 2012. Wkly Epidemiol Rec, 2012; 87(47):461-476.

32.

World Health Organization. Recommended composition of influenza virus vaccines for use in the 2020 southern hemisphere influenza season. Wkly Epidemiol Rec, 2019; 94(42):473-484.

33.

Duque

, McMorrow

, Cohen

. Influenza vaccines and influenza antiviral drugs in Africa: are they available and do guidelines for their use exist?. BMC Public Health, 2014; 14(1):41.