The IAS Method®
Incidence of gastro-enteritis
- The purpose of the IAS® Incidence of gastro-enteritis is to perform syndrome monitoring for gastro-enteritis in France. Gastro-enteritis is a syndrome that can have several origins: bacterial (consumption of water or food contaminated by bacteria), parasitic (protozoons, amoeba) or viral (Rotavirus, Norovirus…). Gastro-enteritis is rife throughout the year, but every winter there is an epidemic of viral origin.
- The information used to build IAS® are sales data for medicines sent by member pharmacies of the Xpr-SO network to Celtipharm. Pharmacies automatically send their information every day. It is anonymous information based on sales receipts. Over 4000 pharmacies provide this information.
- Monitored medicines are a set of 256 products indicated for gastro-enteritis. They belong to 8 groups of medicines. These medicines are either dispensed with a medical prescription or are sold without prescription upon the advice of a pharmacist. The indicator is standardised by two internal references and validated by comparison with data from the Sentinel network (continuous monitoring of acute diarrhoea by a network of 339 volunteer doctors). On a national level, cross-correlation with the Sentinel network is high (0.88) which validates the relevance of IAS®.
- The first internal reference is total activity of network pharmacies (in Euro). It allows a change in the network over time to be taken into account (growth of the number of pharmacies since 2008, any delays in pharmacies submitting data) and variations in activity depending on the day of the week. The second reference is the percentage of turnover, related to monitored medicines, at the end of spring and at the start of summer (weeks 19 to 28), when gastro-enteritis is less frequent. This second reference expresses the IAS® as a base 100 index, it is calculated for each micro zone in order to take any´changes in local prescribing and medicine consumption practices into account. The index is 100 or a little below, the quieter parts of the year. Over the past five years, the values of the index found at the winter epidemic peak varied from 167 to 284.
- The large density in observation points (approximately 4600 pharmacies at the end of 2013) enables fine geographic use of IAS data®. The map is done by combining two mapping concepts: maps and geographic smoothing.
- A population map is a deformation of a standard map in proportion to the population. Therefore, if a region accounts for 20% of the French population, it will occupy 20% of the map area. This approach in « population areas » is particularly suited in epidemiology because it makes it easier to see the fraction of the population that is affected. The data used to make the map are INSEE 2010 data; it is the last year available and corresponds to surveys conducted between 2008 to 2012.
- With geographic smoothing a weighted average for values found around it is represented for a given area. Weightings are decreasing according to the distance up to which the distance called distance of interest for smoothing is cancelled. The higher the distance of interest, the more standardised are colours on the maps. The smoothing distance here is 100 km. This enables results to be anonymous and improves the legibility of the map.
- In addition, maps for weekly temporal series since January 2008 can be downloaded. These series are available at a national and departmental level. In both cases, the index for the zone is the median of values at its various points.
- Data on the OpenHealth.fr website are wholly open data and are freely reusable. They are published under ODbL licence.
- Additional information about the mapping methodology used is available at sur le site Openhealth.fr.