Surveillance of food poisoning and salmonella infections.

SIR,-Dr David A Rands (27 September, p 854) reported a death due to anaphylaxis following desensitisation for allergic rhinitis. In the past, hyposensitisation was widely used in the treatment of allergic disorders; however, impressions of benefit were often based on anecdote and where controlled trials have been performed results are variable. For example, in studies of desensitisation to the house dust mite, Dermatophagoides pteronyssinus, in patients with asthma or allergic rhinitis no benefit was found in one study'; the British Thoracic Association study2 showed marginal benefit (not statistically significant) in patients not on steroids and no benefit in patients on steroids; and two other studies showed improvement.3 Today newer drugs (local steroids and sodium cromoglycate) are available which control allergic symptoms in the majority of patients. These drugs, together with measures to reduce exposure to the allergens, are effective and safe and constitute the treatment of choice. Hyposensitisation, which is potentially dangerous and often ineffective, should rarely be required. Although anaphylaxis is a rare complication of hyposensitisation, Dr Rands's article provides a tragic reminder of the consequences of a potentially dangerous form of therapy in a mild disease.