Write a Report...

Reports

Decipher My Data (Flu): Final report on research findings

As part of our carrying out of the Decipher My Data (Flu) project at our school, we analysed the relationship between the illness absence data from our school (sourced from SIMS), and the nationwide influenza data provided by the Royal College of General Practitioners as a group of seventeen year 12 students. The nationwide influenza data was defined as the number of individuals across England and Wales going to see their General Practitioner with Influenza Like Illness (per 100,000 of the population), and is regarded as one of the most reliable indicators for the prevalence of influenza in a population. Overall, from the graphs shown on the Decipher My Data website, it appeared that our school’s rate of illness absences were slightly lower than the England average for illness absences in schools from September to February, except for some periods in November.

Our school is located in Kingston upon Thames in outer London, and has a relatively affluent pupil intake, owing to relatively affluent areas being located in South West London, such as Richmond upon Thames and Kingston upon Thames, which are some of the least deprived areas in London and England. Our school also has a relatively high educational attainment rate, with 99-100% of students achieving 5-A*-Cs at GCSE, which can be attributed to the entrance exams that prospective pupils must take in order to enter the school. As there is no catchment area utilised in the admissions process for our school, this means that pupils can come from non-local areas via public transport. This could provide an extra opportunity for the transmission of influenza due to the high numbers of people in close proximity on public transport, which would aid the spread of disease via airborne droplets or the many surfaces that people would come into contact with.

In terms of the national trend of illness absence rates, the rate increases steadily from the start of term in September to the end of September to 15 days missed per 100 students due to illness, before continuing to fluctuate around 10-15 days missed per 100 students until the beginning of December, before increasing in the two weeks before Christmas to almost 30 days missed per 100 students due to illness (almost twice the previous number of days missed). There was then a two week gap due to holidays, and the number of days missed had returned to around 13 days per 100 students, before increasing to around 25 days per 100 students.

Our school’s data broadly followed the national trend, but generally had lower absences due to illness than the national average, which may be because of the importance of high grades and maintaining a high standard of academic work that is very much promoted at our school, which may mean that students which have influenza may still be attending school because they feel under pressure to attend classes and keep up with their studies, which contributes to what could be described as a culture of ‘presenteeism’ at the school. This may mean that relying on illness absence data may not be so reliable as a method of identifying the number of students with influenza. School absence data can also be affected by a number of other factors, such as exams. Students may either embrace the culture of ‘presenteeism’ and attend school even when they are ill, or they may take days off and feign illness in order to revise for upcoming January exams, which is in effect a form of study leave. However, the trend for Year 11 students at our school, which had formal GCSE January exams, showed that the number of days missed per 100 students was not significantly higher than the school aggregate value, which suggests that most of the year chose to come in for lessons instead. However, the added pressure of exams, which could cause illness due to additional stress in the run up to exams, also needs to be taken into account. Of course, data from SIMS is wholly reliant on the honesty of pupils and parents, and obviously the reliability of the data may be negatively affected if there are a large enough minority of students ‘playing truant’.

In terms of influenza illness data, this remains fairly stable at around 5 individuals per 100,000 going to see their GP about influenza from 08/10/12 to 05/11/12, before increasing from 12/11/12 onwards to a peak of just over 30 per 100,000 people on week commencing 24/12/12. However, we were unable to directly compare this peak with school absence data, as this occurred in the middle of the Christmas holidays, so there would have been no absence data available for analysis. Although we can see an increase in illness absences from 03/12/12 to 17/12/12 which corresponds with a rise in the number of individuals going to see their GP about influenza, our school’s illness absence data continued to fluctuate through October and November even when rates of national influenza like illness remained fairly steady, which could suggest that there are other factors that are affecting school illness absence data, such as the possibility that the students are ill with a different infectious disease. Also, there are a number of drawbacks with using GP data, as this data relies on individuals attending the GP surgery themselves and organising an appointment, and as not all individuals that have influenza-like symptoms will visit their GP, the rate provided by GPs may not be entirely representative of the total number of individuals that may have influenza.

Overall, this scheme has been useful for a number of things, such as identifying absence trends within year groups at our school, and being able to compare our school’s absence data due to illness with the nationwide numbers of illness absences. However, there are limitations with the method of comparing school absence data, as crucial data that could help complete a trend is missing, such as the October half term dates and Christmas holidays. As a suggestion for the future, it may be an interesting idea to use this project in a workplace environment, as this could mean more reliable results, as employees are more likely to have a more regular working pattern, and are more likely to have fewer holiday breaks than pupils at schools. Data collection was also an issue originally, as we had planned to go round form groups and get data from form tutors and collect symptoms, but this plan had to be shelved because of lack of cooperation from some form tutors, and inaccuracies in data collection by some members of the group. Because of this, we simply extracted the data off SIMS ourselves, with the assistance of our Head of Science. Another useful addition to this project for next year would be to include the possible symptoms for illnesses as an input box when uploading data, so that it would be possible to differentiate further between influenza and other illnesses that are causing absences.

0

SAMPLE REPORT High Ability – 2011-2012

We performed a time-trend analysis to compare absence-due-to-illness levels in our school against influenza-like illness data reported by the Royal College of General Practitioners. Overall our school appeared to have lower-than-average absence-due-to-illness levels however there was a surge in absence levels alongside a surge in the national data shortly before the peak in influenza-like illness.

Background: Our school is of an average size for a secondary school in the UK and is based in the south of the country on the edge of a large town. The intake is reasonably affluent. People are prepared to travel some distance to come here, up to 45 miles, and many come on buses.

Observations: The national data seemed to settle into a background level of between 30 and 55 illness sessions per week. The times that broke from this pattern happened just after long (greater than one week) breaks and just before the February half term. It’s likely that in the time immediately after holidays, people are better rested and less prone to getting ill. We searched for information linking fatigue to the body’s immune response but were unable to find any evidence we could access to confirm if there was a link. Psycholological factors will also play a part here. Many parents and students will be keen to make a good start to a new term and they may be more likely to come into school even if they are ill. Absence levels seemed to peak just before the holidays in all cases. The opposite factor may be coming into play with parents more likely to let their children stay at home thinking that the time they are missing will be less important. There was a particularly big jump just before Christmas.

Our school data seemed on average, lower than the national data.

There appeared to be a peak in illness in our class in the middle of the second half of the Autumn term and this also appeared in the year-wide data but this did not affect the whole-school data. Looking at the year group data at this point, there was a lot more variance than there was in the first term. So although it seemed like more people were off, they were just the people that we were more aware of. There was much less variance seen in the school-wide data and the national data was even smoother as were were effectively averaging results over a much larger sample.

The most interesting thing to notice was the peak in absence levels shortly before half term in February. Many people were off and were reporting flu-like symptoms before they left and after they came back. Lots of staff seemed to be off at the time as well but we have no evidence for this apart from remembering being in a lot of lessons with cover teachers! The national flu peak in the GP data came just after this. Unfortunately we had no data for the week before the national peak and the national peak itself wasn’t very big compared to the background levels so it’s unclear how significant it is that the school and national school data all peaked just before the national GP data.

Our school data seemed on average, lower than the national data. It’s possible that the children at our affluent school, are better nourished or keep stricter sleep routines than children in other schools which could make our absence levels lower if the people at our school are better-prepared to fight an illness.

It was interesting to note that the Y13 data broke away from the rest of the school data in the early weeks of January around the time they were sitting exams. This could either be because they wanted extra study leave or it could be that they were more prone to getting ill in a period of stress. We should ensure that this anomalous pattern does not influence the school data by removing it from our averages. Year 9 seemed to be very badly affected by the illness in the run up to the February half term – we’re considering doing a survey of the year group to try to find out why they were so badly hit.

0

SAMPLE REPORT Low Ability – 2011-2012

peak in the flu data from the national school data

This is our report for decipher my data for the 2011-2012 flu season.

The graph we’ve included shows our school’s data, the data from schools across the country, and the national influenza like illness data from GPs. There wasn’t much flu around this year, which makes it pretty difficult to spot an outbreak of influenza early.

There was a bit of peak in the flu data from the national school data, and this looked like it happened just before the peak in the GP data. However, it’s a bit difficult to tell if that’s really true or not because it was such a small peak.

Our school’s illness data varied randomly a lot throughout the year and we’re not 100% sure why, but we think it might have been because of random variation. During this year we’ve learnt a lot about flu, how you catch it, how to stop it spreading, how to collect data to monitor its spread, and how to analyse the data from the project.

In one way it was good that there was not much flu around this year as it meant not many people got sick, but it did mean that we didn’t get to see whether decipher my data flu worked or not.

0