Methods

Outbreak setting
The outbreak investigation was conducted from June 25 to 30, 2018 in Nkawkaw, capital of the Kwahu West Municipality in the Eastern Region of Ghana Figure 1. This was the first reported AI outbreak in the municipality. The municipality is one of the twenty-six (26) districts in the Eastern Region of Ghana [12]. The projected population for Kwahu West Municipality is about 115000 in 2018. About half of the population live in the rural areas [13]. Nearly 44% of households in the municipality are engaged in agricultural activities including crop farming, livestock rearing and fish farming.
Livestock rearing is the second most important agricultural activity accounting for about 25% of all agricultural activities. About 61% of livestock kept are birds with 95% being chicken [12]. The birds are mainly raised on a small-scale backyard basis with few commercial farms. Farm capacities range between 100 and 5,000 birds. About 47,000 commercial layers and 70,000 domestic or rural birds were raised in the district in 2018. There were three live bird markets located within the Nkawkaw Central Market near the roadside, where trading of birds and eggs is mainly done [14].
Only one veterinary clinic provides veterinary services to livestock farmers in the municipality. Fifty health facilities provide human healthcare services in the municipality. There is no routine surveillance for Influenza-like illnesses in the Kwahu West Municipality.
 
Study design 
We conducted a survey involving interviews with poultry farmers, human contacts of affected farms and veterinary personnel. The interviews with the farmers were conducted using a structured questionnaire detailing farm demographic characteristics, flock history, farm biosecurity, farm management practices, contacts information and other risk factors to HPAI (S2 File). We actively searched for case farms within a 6 kilometer radius of the index-case farm and identified six farms for assessment. We defined a suspected HPAI case-farm as a poultry farm (domestic or commercial) in Nkawkaw with sudden death of birds, with or without clinical signs of HPAI on from June 1 to July 10, 2018. Six (6) farm owners/managers were interviewed. This includes the managers of the two affected and four other unaffected farms. All twenty-three contacts of the case farms were followed up daily for 14 days. The contacts were interviewed with a structured questionnaire to assess for any clinical manifestations of influenza. We defined a primary human contact as any person who has had physical contact with the case farms, and/or the birds or eggs of case farms from June 15 to 26 2018. A secondary human contact was defined as any person exposed by physical contact with a primary contact. We also conducted environmental assessment of the immediate vicinity of both farms to identify risk factors to the spread of the AI infection including the presence of wild birds and water bodies close to pens as well as the biosecurity measures practiced on the farms.
 
Laboratory investigations 
Six whole bird carcasses were collected from the two affected farms. The fresh carcasses were transported in appropriate double-layered plastic bags inside a leak-proof plastic container to the Accra Veterinary laboratory within four hours of collection. In this Biosafety Level 3 laboratory, viral ribonucleic acid was extracted from cloacal and oropharyngeal swabs taken from the carcasses and diagnosed using Reverse Transcription Polymerase Chain Reaction (RT-PCR). The samples were tested for influenza viruses adhering to all standard protocols [15]. Appropriate personal protective equipment were donned in handling the specimen.
 
Data analysis 
We performed descriptive analysis of the outbreak data by person, place and time. We calculated overall mortality rate and farm specific mortality rates. Continuous variables including age of human contacts were expressed with appropriate measures of central tendency and dispersion. The results were presented as frequencies and relative frequencies in tables and graphs. We drew a column graph to describe the magnitude and the course of the death of birds. We used quantum geographic information system tools [16] to map the poultry farms in the outbreak setting. Data was analyzed using Stata version 15.0. The interviews with the affected farm owners were transcribed and coded deductively, noting the dates and key events that occurred during the outbreak. The results were presented as a narrative supported by a flow graph of dates of key events during the outbreak.
 
Coordination 
The response to the outbreak was organized through multi-sectoral collaboration in a One Health approach. Field epidemiologists from the Ghana Field Epidemiology and Laboratory Training Programme supported the Municipal Health Management Team to respond to the outbreak. The team constituted by the Kwahu West Municipal Health Management Team included staff of the Ghana Health Service, Veterinary Services Department, and Municipal Assembly. The Municipal Fire Service Department, National Disaster Management Organization, and Ministry of Food and Agriculture supported the investigation team in the response to the outbreak.
 
Ethical considerations
The investigation was considered a response to a public health emergency by the Ghana Health Service and Veterinary Services Department and therefore did not receive formal review by Ethical Review Committees. We sought and obtained permission from the Eastern Regional Health Directorate, Eastern Regional Veterinary Services Directorate, Kwahu West Municipal Assembly, Kwahu West Municipal Veterinary Department and Kwahu West Municipal Health Management Team before commencement of the investigation. All respondents provided consent and were assured of confidentiality. We anonymized all information collected on the affected farms and human contacts and stored them securely as confidential records. The use of a water-based foam in the depopulation exercise allowed for a humane and most effective way of euthanizing the birds on the affected farms.