UK flusurvey participants have been reporting how many people they meet over the course of a day. We ask about encounters that involve conversations, and those that include some physical contact, and we also ask about the age of the people met.

Below we show some of the results that have emerged from the contact questionnaire.

Contact patterns and holiday periods 

The pattern of school holidays drove the swine-flu epidemic. When schools closedover the summer, the epidemic declined, when they re-opened in September the second wave started. Even half-term caused a wobble inthe incidence in late October. What actually happens when schoolsclose? Presumably children make fewer contacts, but how many fewer? Do they compensate by making more contact with other age groups (e.g.their grandparents)? The data that Flusurvey users have provided allowed us to answer these questions. As the figure below shows, when schools close the children make (on average) about 15 fewer contacts per day.Other age groups do not appear to change the number of contacts that they make.

Lookingat the data for children more closely (below), it is apparent that thischange of contacts results from a decline in "conversational" contactswith other children. There appears to be very little evidence of anycompensation - i.e. increase in contacts with other age groups.

National characters

When looking at the results of the contact questionnaire broken down by country, some clear differences emerge. Participants from Northern Ireland reported, on average, many more contacts (both conversational and physical) than participants from England, Scotland or Wales.

This doesn't necessarily mean that the Irish are more sociable than anyone else, merely that those who take part in the flusurvey seem to be; however, maybe it does pick up real differences in national characters. Less controversially, the large number of contacts with children reported by Irish participants suggests that the results we see might merely be a consequnce of recruiting more teachers from Northern Ireland. 

Place of Contact

The contact questionnaire showed us that on average people have 22 conversational contacts and 6 physical contacts each day. However, that "average" conceals a lot of detail. Most people have far fewer contacts than these numbers suggest. This happens because a few people with many more contacts than everyone else drag the average upwards. For example, although the average number of conversational contacts is 22, half of the people surveyed report 11 or fewer daily contacts

Of course, even for one person, the number of contacts changes from day to day. Because flusurvey participants have filled in the contact questionnaire several times in the past few months, we can start to explore this in more detail. 4 examples are shown here: in each case the daily numbers of reported contacts are shown as crosses,and the average as a coloured square. Although each of these people has about the same average number of daily contacts, we can see a range of behaviours. The individual on the left, in red, has a fairly consistent routine; the person in yellow sometimes meets many people but sometimes hardly anyone at all; while the personin blue usually has low numbers of contacts with occasional busy days.

The majority of conversational encounters happen at work or school, with only about an eighth occuring at home. Physical encounters are split evenly between home and work/school. This might go some way to explaining why infection spreads most easily at home - close encounters are likely to be more risky.

Risk of infection

Contact with children seems to be a major risk factor for having reported flu-like-symptoms over the course of this summer (see below). Children themselves are more susceptible to swine flu and so it is not surprising that contact with them puts others at greater risk.

Perhaps more surprisingly, use of public transport does not appear to be a risk factor. The figure on the right shows the fraction of people who reported flu-like-symptoms broken down by use of public transport. There appears to be no difference at all. So we can't see any evidence that sitting on the bus makes anyone more likely to get swine flu.

Who gets ill?

One purpose of the UK flusurvey is to workout who is at risk of getting infected. Here we take a brief look at the reported behaviour of people who have ever reported flu-like symptoms and compare them to people who have never reported flu-like symptoms.

Number of contacts We might expect that people who have lots of social encounters are at greater risk of infection. This seems to be true, but the effect isn't huge. The results may be complicated by the fact that once infected people probably stay at home and have relatively few contacts.
	Household size We also see that people who get ill live in larger households. In particular, they tend to live with approximately 50% more children under the age of 5 than people who have not reported flu-like symptoms.

Meeting risk groups We asked whether people spend time with a few potential risk groups -children, the elderly, patients - or with large numbers of people. We can see very clearly that contacts with patients or the elderly are not associated with greater risk of infection, but meeting large groups ofpeople or children are. This reinforces the message that infection spreads through social contacts and that, in particular, children are at high risk of infection.

Similarly, those who report having flu tend to live in larger households, in particular sharing a house with more children.

So far in the swine flu epidemic, there have been many more cases in children than in any other age group and our results showed that women appear to be slightly more at risk of getting infected than men; part of the reason for this might be that women have more contacts with children than men do.

Encountering Infection

We asked participants to tell us whether they met anyone with flu-like symptoms recently. No study had asked this before, so we weren't sure that it would give us useful information. The data collected so far seem logical: people who report flu-like symptoms are much more likely to record that they have met someone with flu-like symptoms. This suggests that people may be able to identify who they caught their infection from, which would be really useful information for understanding better how swine flu spreads.