That’s a lot of data and obviously the next logical step is to think about whether we can answer the project’s research question, which was: can school absence data detect flu peaks early?

The first thing we needed to do was establish the baseline level of school illness absence so that we would know when something abnormal was happening such as a flu peak. Our excellent set of data shows us that for the participating schools it is between 10 to 14 days per 100 pupils a week.

Knowing that set us up nicely to be able to detect a peak if and when one arose. I say if, because as you’ll have seen this year has seen an abnormally low number of cases of flu. This makes detecting a peak much trickier obviously, because like in the real world, looking for something small like your keys is obviously much harder than looking for the car to put them in.

The national influenza surveillance data peaked in week 7, which was actually a school holiday so we weren’t collecting data for that week and is one of the major limitations of this project. However, all is not lost and an early look at the data is promising as school absences peaked in the previous week 6 which is very encouraging because even though we don’t have data for the highest week, the timings are definitely right.

We’ll be doing lots more analyses over the coming weeks and I’ll be writing more blogs about this. We’ll also be deciding what to do next. One thing is certain, which is that scientists like to be able to repeat results as many times as possible to check they get the same results each time.  So if we think schools are still interested in taking part in the project, and if we can get the money to run it again, then we’ll be back again next year before you know it!

So what does this mean for you and your students? Firstly the lack of a strong peak does make it more difficult for you to analyse your data and report it back to me. Because of that we’re not expecting anything from you, but if you do have something you want me to look at then please email it through.

Secondly it means that we are unlikely to attempt to get a paper published. We’ve got lots of data from our schools including many interesting findings worth publishing but we’ll need to get data from a more typical year before we’d be ready to write it up for an academic paper. The current plan is to try to repeat the work next year and if, as expected, there is a stronger peak then we can better answer to the original research question.

Scientific experiments don’t always turn out quite as planned and we certainly didn’t think we’d see such low levels of flu this year. We hope this might be a useful learning point for the students  - we don’t always get the data, but this doesn’t mean we’ve failed, it’s just the nature of science. We now have to pick ourselves up, dust ourselves down and get ready for next season.

I hope that you and students will take part again. We’re going to make some changes to simplify the project and make it easier to do. You should hear about that later in the Summer term.

Once again thank you for every piece of data you uploaded. Things may not have gone exactly as expected, but it has been a very useful science project for us and I hope for you.

Credit for the discovery of the flu virus is shared. Some experiments were done in France around 1918 by two men called, Nicolle and Le Bailly, who took lung secretions (basically snot!) from people with flu symptoms. First they put the snot into the nose and eyes of a monkey, which became sick, and then they injected it under the skin of a man, who also got flu symptoms (you can read an extract from the 1918 paper here: http://jama.ama-assn.org/content/71/26/2154.extract). This study was used to suggest that flu is caused by a virus that can be transmitted, but the study was limited (this means there are problems with the methods) and experiments like this wouldn’t be allowed today!

The main people involved were a vet in America called, Koen, who found flu-like illness in pigs in 1918, and a scientist called Richard Shope who proved that flu could be transmitted between pigs in the 1920′s. Curiously, UK scientists working with ferrets in 1933 found their animals became unwell after one of the researchers got flu. Eighty years later, ferrets are still used for experiments (instead of humans) to understand flu virus and vaccines.

You can read more about the discovery of flu here: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1145139/

For more information see the Health Protection Agency ‘Influenza factsheet for schools’

http://www.hpa.org.uk/Topics/InfectiousDiseases/InfectionsAZ/SeasonalInfluenza/InformationForThePublic/influinfluenzafactsheet/

It’s because we can’t forecast the start of the season that we carry out projects like Decipher my data! Flu!

Does the timing of the school term affect the timing of flu pandemics? Image by Nikolay Olkhovoy

We can at least detect flu and so far this year the Health Protection Agency has only found low levels of flu circulating in the community. Our school data is consistent with this finding and it’s been hovering at around 10-16 days missed due to illness per 100 pupils per week. Knowing this baseline figure allows us to establish the levels of school illness when there isn’t much flu around.

Some of things we think trigger the start of a flu season include temperature and humidity. Others include how people mix with each other, and during the swine flu season in the USA, the timing of the school term was found to be associated with the start of the pandemic.  Our immune systems might also play a role, as our ability to fight viruses probably fluctuates throughout the year.

The most recent HPA surveillance report tells us that although levels of flu are still low, some of the measures have shown a slight increase. Not many schools have uploaded data for week 4 yet, and although at the time of writing this it’s our highest level so far, at 16 it’s still within the baseline (i.e. normal) levels. Given the recent cold weather and the fact that there are small amounts of flu circulating in the community, I’ll be watching the data closely over the next few weeks. I hope you’ll be joining me and don’t forget to send us your observations via the LabLogs.

One way to think about diseases is to decide whether they are caused by infections or not. Some of the really nasty diseases are caused by infections, so I’m going to focus on infections in humans to answer your question about the most dangerous disease out there.

One way to measure the danger caused by diseases is to measure case fatality, or what proportion of people die if they get the infection. Here are some of the more impressive examples I could think of:

Ebola virus occurs in Africa and is transmitted in body fluids. Initially it causes a flu-like illness with fever, muscle ache, stomach cramps, headache, confusion and sometimes coma. In some cases it causes bleeding from the body orifices. Case fatality is around 70%. However, there has never been a sustained outbreak in the world (it probably kills people too quickly!) and it only occurs in Africa, so it is unlikely to be dangerous to people in the UK.

Rabies virus is mainly spread by bites in saliva from infected animals. Rabies causes inflammation of the brain and sometimes makes people scared of water (hydrophobia), violent and manic. Case fatality for rabies is nearly 100%. Luckily for us, rabies is not very good at infecting humans and there is a vaccine to protect people who get bitten (and it is only found rarely in a few species of bats in the UK).

Bubonic plague is caused by a bacteria, called Yersinia pestis, that is transmitted by fleas which are carried around by rats. Bubonic plague used to cause some really nasty symptoms, including swelling of the lymph nodes and gangrene of the fingers, toes and other extremities – the skin turns black and people die within about four days. That is apparently why it was called, ‘Black death’!! Case fatality without treatment is about 60% and many millions of people died with this infection in the middle ages. However, the bacteria is now easily treated with antibiotics and improvements in hygiene mean that it is unlikely to cause disease in the UK.

HIV is spread by body fluid contact and is one of the newest diseases known to man. There is a great internet radio programme called Radiolab all about Patient Zero for HIV (www.radiolab.org/2011/nov/14/) that you might like. HIV is dangerous because it infects and kills immune cells, called CD4 positive T lymphocytes. Over about ten years, the immune system of someone with HIV gradually becomes more diseased until it doesn’t work anymore and can’t protect them from simple infections. When this happens they may die of the disease called AIDS (acquired immunodeficiency syndrome). Case fatality used to be very high in the 1980′s and early 1990′s, but now we have drugs that control HIV and can prevent the development of AIDS.

Another way to measure the danger caused by an infectious disease is to find out how many people it can spread to. Scientists measure transmission potential (how likely an infection is to spread) using a number called Ro (pronounced R-naught). Ro is a measure of the number of new cases to expect from each person who is infected. Below I’ve put down some examples of infections with their estimated Ro numbers:

When influenza virus infections were measured in New Zealand in 2009, Ro = 1.25

• Measles is one of the most easily transmitted infections in the world, you just have to be in the same room as someone to be infected, and Ro = 12 – 18
• Small pox has been eradicated by a vaccine programme, but it used to cause lots of deaths in the world, and Ro = 5 – 7
• Polio is also controlled by vaccines, but it causes paralysis and death when it infects people. Fortunately the last case in the UK was in 1982, and Ro = 5 – 7

Measles is probably the winner here with a Ro of 18. If I had measles, this means I might pass it on to 18 other people before my body’s immune system was able to overcome the measles virus. Of course, what really happens depends on lots of other things, including the people I meet. If I stayed at home and didn’t see anyone while I was unwell then I couldn’t pass the infection on to anyone else (I might be pretty bored though), and if everyone I met had been vaccinated they would all be protected from infection. However, even though measles is very infectious, most people who get measles get a rash and a fever and then get better.

I’m not sure which is the most dangerous disease of all, I’ll let you make your own mind up, but it probably depends on who you are, where you are and who you meet on the way!

What’s even stranger about the start of swine flu is that although all the news coverage was about Mexico, it was actually scientists in America who actually detected the new ‘swine flu’ strain in two American children. Quite often it takes a while before we get a final picture of how a new strain started and although we’re unlikely to ever find out who was the first person to be infected with this new strain, scientists think that the virus first entered humans around september 2008 – www.nature.com/news/2009/090505/full/459014a/box/1.html

We haven’t seen a peak in national flu, yet. This means school absence can’t be attributed to flu and so I’ll be waiting for a flu peak before I get too carried away analysing the XY scatter graphs.

The science team and I at UCL are still keeping a close eye on the flu! data. However any comments and observations you have will bring fresh insights, and you may well spot something we haven’t. If you’re a consented teacher or student you can now record your observations in a LabLogs. I’ll read and reply to your LabLogs, giving us a chance to discuss and analyse the data together.

You might want to tell me about some data that’s looking dodgy. For instance, if a school has entered data for the coming week, which hasn’t yet finished. Anything that looks a bit strange, write it in a LabLog and we’ll look into it.

Alternatively, you might want to add in some local knowledge to help explain your school’s absence data. Whether you notice that absence increased during a cold spell or that teachers also got sick as student absence increased, this information could provide some exciting insights.

LabLogs are also used in the time trend analysis section of Lesson Two, where students enter their own observations and potential explanations. To ensure school anonymity is protected, LabLogs are strictly moderated and only visible to a school’s teacher and students.

The purpose of the XY scatter graphs is to investigate whether school absence due to flu is linked to the four variables. So the first thing to keep in mind, is that we’ve not seen a peak in national flu, yet. This means school absence can’t be attributed to flu and so I’ll be waiting for a flu outbreak before doing this activity with my students.

When you open it, you can see how the different school-based variables correlate (or not!) with the national school absence data.

The variables are taken from my school profile and appear on the X axis of the scatter graph. Some variables have been filled in by lots of schools, whereas other variables need to be filled by more schools to make them worthwhile. Having more information for some variables compared to others isn’t all bad, as it presents an opportunity for my students to talk about data quality.

Scanning through the weeks, patterns appear and then collapse into chaos, which I believe is caused by variations in different schools’ absence data. It’ll be really cool to see whether any patterns hold up if flu peaks.

My students will have the How good is your correlation resource, when I cover Lesson Two which will show how strong any correlations in the data are. After the lesson, I’ll be asking my students to record their observations, explanations and comments in a LabLog for homework. These are kind of like a Facebook “wall” where students can post their observations.

One of my students asked about whether there is any connection between our susceptibility to flu and temperature – I passed this on to Dr Rob through the “Ask a scientists” function. It was great to see a really informed answer and I loved the fact that my year 8s could keep Dr Rob on his toes, as he had to do a literature search before he could answer!