• Ray B Jones, Paul Hewson, and Edward R Kaminski.
• Faculty of Health, University of Plymouth, 3 Portland Villas, Plymouth PL4 8AA, UK. Ray.Jones@plymouth.ac.uk
• Bmc Public Health. 2010 Dec 29; 10: 790.

BackgroundAllergy is a serious and apparently increasing public health problem yet relatively little is known about the types of allergy seen in routine tertiary practice, including their spatial distribution, co-occurrence or referral patterns. This study reviewed referrals over an eleven year period to a regional allergy clinic that had a well defined geographical boundary. For those patients confirmed as having an allergy we explored: (i) differences over time and by demographics, (ii) types of allergy, (iii) co-occurrence, and (iv) spatial distributions.MethodsData were extracted from consultant letters to GPs, from September 1998 to September 2009, for patients confirmed as having an allergy. Other data included referral statistics and population data by postcode. Simple descriptive analysis was used to describe types of allergy. We calculated 11 year standardised morbidity ratios for postcode districts and checked for spatial clustering. We present maps showing 11 year rates by postcode, and 'difference' maps which try to separate referral effect from possible environmental effect.ResultsOf 5778 referrals, 961 patients were diagnosed with an allergy. These were referred by a total of 672 different GPs. There were marked differences in referral patterns between GP practices and also individual GPs. The mean age of patients was 35 and there were considerably more females (65%) than males. Airborne allergies were the most frequent (623), and there were very high rates of co-occurrence of pollen, house dust mite, and animal hair allergies. Less than half (410) patients had a food allergy, with nuts, fruit, and seafood being the most common allergens. Fifteen percent (142) had both a food and a non-food allergy. Certain food allergies were more likely to co-occur, for example, patients allergic to dairy products were more likely to be allergic to egg.There were age differences by types of allergy; people referred with food allergies were on average 5 years younger than those with other allergies, and those allergic to nuts were much younger (26 Vs 38) than those with other food allergies.There was clear evidence for spatial clustering with marked clustering around the referral hospital. However, the geographical distribution varied between allergies; airborne (particularly pollen allergies) clustered in North Dartmoor and Exmoor, food allergies (particularly nut allergies) in the South Hams, and on small numbers, some indication of seafood allergy in the far south west of Cornwall and in the Padstow area.ConclusionsThis study shows marked geographical differences in allergy referrals which are likely to reflect a combination of environmental factors and GP referral patterns. The data suggest that GPs may benefit from education and ongoing decision support and be supported by public education on the nature of allergy. It suggests further research into what happens to patients with allergy where there has been low use of tertiary services and further research into cross-reactivity and co-occurrence, and spatial distribution of allergy.

12 keywords...

hide…