Study details
The second study of diarrhoea and vomiting (D&V) in the community comprises seven separate but related studies. The structure of the research project is illustrated in figure 1. The prospective studies mirror the structure that was used in the original Infectious Intestinal Diseases (IID) study, funded by the Medical Research Council (MRC) and the Department of Health (DH), in the mid-1990s. The telephone survey is an addition.
Study 1 is a telephone survey of self-reported symptoms of D&V.
This is a retrospective study i.e. we will ask people to remember what has happened in the recent past. We will ask people to remember if they have had diarrhoea or vomiting during either the last week or the last month. This is to find out how quickly people’s memories fade (examining the effect of recall bias). The results from this study will be compared with those from the prospective studies. We will be able to estimate the community incidence (new cases per 100,000 population) of D&V over a period of one week and one month from this study. We will also be able to compare the incidence rates between the four UK nations.
Study 2 is a prospective cohort study that involves following people over time.
We will recruit a group of 8,400 people across the U.K. We will recruit them at random from 84 General Practices. We will then follow up the people at weekly intervals for a period of one calendar year to find out how many of them develop new symptoms of D&V. We will ask those people who develop D&V to complete a symptom questionnaire telling us about their illness and their contact with health services e.g. NHS Direct/NHS24, and provide a faecal sample. We will estimate the community incidence of D&V over one year and we will compare this figure with those from the telephone survey. We will also be able to compare the incidence of D&V in England in 2007/8 with the incidence in the mid 1990s. This will tell us if D&V has got better, stayed the same or got worse.
Study 3 is a prospective study of people presenting to their General Practitioner with symptoms of D&V.
This is called the GP Presentation study. The 84 practices mentioned in study 2 will be assigned randomly to take part in study 3 or study 4 (see below). Forty-two practices will take part in study 3. We will ask all people who go to see their GP with a new episode of diarrhoea (who can be different people from those taking part study 2) to complete a symptom questionnaire and provide a faecal specimen. We will use this information to estimate the incidence of D&V, and the germs that cause D&V, in people that present to general practice.
Study 4 is an audit of people presenting to their General Practitioner with symptoms of D&V.
This is called the GP Enumeration study. It will take place in the remaining 42 practices. We will ask the Research Nurse to find out from the practice records about all people who present with a new episode of D&V. S/he will record the patient’s age, sex, postcode and information about the place of consultation, admission to hospital and whether or not a faecal sample was requested. If a sample was requested s/he will record the result. From this we will be able to compare the incidence of D&V presenting to GP’s with the incidence of laboratory-confirmed infection.
Study 5 is an audit of studies 3 and 4. It is called the validation study.
Using practice records the research nurse will search for all new cases of D&V that present to the GP surgery during the 12 month period of the study. S/he will generate a list of all the cases that should, therefore, have been included in studies 3 and 4. This will tell us how many people were missed. In the original IID study, funded by the MRC, researchers found that although 1,514 people met the case definition for being included in studies 3 and 4, only 974 (64%) were actually included. If we understand what happens at Practice level to influence whether or not people are included in studies 3 and 4, we can take account of this in our calculations.
Study 6 is a laboratory-based study. This is known as the Microbiology study.
This will test all of the faeces samples from studies 2 and 3 for a wide variety of germs. All first line testing will be carried out at the Health Protection Agency Manchester Laboratory. All faecal samples and any germs found will then be sent to the Health Protection Agency Centre for Infections at Colindale for further detailed tests.
Study 7 is the national reporting study. This is an audit of official statistics.
We will use the results from studies 1 to 6 to work out by how much the official statistics differ from the community burden of D&V as measured in our studies. In the IID study funded by the MRC and DH in the 1990s the researchers found that for every person with D&V recorded in official statistics there were another 136 people who were also ill. We want to know whether this has changed or not. If we know how accurate official statistics are, we can calculate by how much to multiply them to give a better picture of D&V in the community.
The relationship between disease burden in the community (which is often large) and that which gets into national statistics (usually small) is often displayed as a pyramid (Figure 2). Each of the studies described above provides a slice through that pyramid.
