Each year the World Health Organisation gathers data on influenza strains and recommends which strains countries should include in formulating vaccines for the next year.
Many countries around the world administer these vaccines because the influenza virus - known as the flu virus - is a significant contributor to respiratory illness.
In Africa only a few countries issue the vaccine annually: Ivory Coast, Egypt, Libya, Mauritius, Tunisia and South Africa.
Kenya is one of the countries on the continent that doesn't have a flu vaccination programme. The vaccine is available in the private sector but isn't affordable for most Kenyans. Because there is no national programme, children also don't receive it as part of their regular immunisation schedule in public hospitals. This is despite the fact that a quarter of Kenyan patients who seek medical care for respiratory symptoms have the flu virus in their systems.
The reason there isn't a programme is because, until recently, there were no updated figures on the national burden of flu across age groups in Kenya. In addition, for many years data was only collected at a regional level. And when national data was available, it was collected when there was a flu pandemic in the country in 2009. The data from this period was not helpful for policymakers to establish what the national burden of seasonal flu would be in a period when there wasn't a pandemic.
We set out to get the first breakdown of Kenya's flu burden on a national scale after the 2009 flu pandemic period, to inform the decision to approve and roll out an effective seasonal flu vaccine programme.
Based on our findings we conclude that children under the age of two should be prioritised for vaccination.
Understanding the burden
We obtained data from health facilities across the country that are part of the national influenza surveillance programme. We collected samples from about 10 000 patients with a cough and fever who were admitted to either Kenyatta National Hospital or five county referral hospitals between January 2012 and December 2014. The samples were tested to confirm the presence of the virus.
The virus was found in 9% of the patients who were tested. We found flu was associated with 50 000 cases where people were severely ill, and about 10 000 admissions each year across the country.
We were able to draw several conclusions from our study. Firstly, the risk of children being admitted to hospital varied with age. Children under the age of five were 17 times more likely to be admitted with flu than older children and adults. Among children, those under the age of two were most likely to get sick from the flu.
The rates of severe flu were also high in elderly people and primary school children between the ages of five and 14. But this was not to the same degree as babies.
Secondly, there were differences in the rates of disease across regions in Kenya. The Rift Valley region, for example, recorded the highest rates of people who were admitted to hospital for flu related symptoms. Nairobi recorded the lowest.
These variations indicate that there are differences in risk factors for severe respiratory illness between regions in Kenya. These include the prevalence of malnutrition, overcrowding within homes, non-exclusive breastfeeding in children, household pollution and HIV.
The North Eastern region of the country had the highest rates of severely sick people who had not been hospitalised for their illness. What this reveals is that people in this region were unable to access health facilities as easily as their counterparts in the rest of the country.
Kenya's warm climate means that it's not naturally considered to have a flu problem. Flu has usually been shown to be of concern in more temperate climates where there are clear summer and winter seasons.
But our statistics reveal that Kenya has a higher rate of disease than many other countries which have a temperate climate where we would expect the burden of disease to be higher.
If a flu vaccine is to be introduced in Kenya it would have to be administered annually. The vaccine's effectiveness each year would depend on how well the vaccine strain matches the circulating flu strains.
To address these fluctuations, the vaccine development world is looking into the viability of a universal flu vaccine. This would be effective against all circulating flu strains. The universal vaccine would require vaccination either once in a lifetime or once every few years depending on how long the protection lasts.
In the Kenyan setting there are several considerations policymakers need to think about when deciding whether or not to introduce a vaccine programme. This includes evaluating whether it would be cost-effective in the country, what the staffing and cold chain requirements would be and how best it would be incorporated in the current immunisation schedule.
In the meantime, our study provides the Kenyan government with the impetus it needs to seriously start considering this annual vaccination programme, particularly for young children.