Our study, which was limited by small numbers, was not able to demonstrate that immunological or clinical criteria can adequately replace virological criteria for the determination of treatment failure.

The aim of this study was to review chronic idiopathic urticaria patients with positive skin prick testing to mites and the clinical relevance of this positivity. Case records of patients with chronic idiopathic urticaria who underwent skin prick testing during the years 2000 to 2007 were reviewed. The studied allergens included house dust mite allergens; Dermatophagoides pteronyssinus and farninae. A total of 172 patients were enrolled in the study. The prevalence of positive skin prick testing to mites among patients with chronic idiopathic urticaria was high (34.9%), but had little clinical relevance (3.3%) to their urticarial symptoms. Patients with mite-sensitization were more commonly male and more often had a personal or family history of atopy compared with those without mite-sensitization.

From the letter entitled Causes for low positive predictive values of CD4 counts for antiretroviral treatment failure by Eisenhut, 1 we agree with the author that based on the finding from our study of a very low positive predictive value of the CD4 count for antiretroviral treatment failure, 2 there may be other causes for low CD4 counts in Thai HIV patients on antiretroviral treatment. Although it has been demonstrated that malnutrition can affect the levels of CD4 cell counts in the African population, 3 a study from Bangkok failed to show the impact of micronutrient supplementation on CD4 cell count or plasma viral load. 4 Concerning zidovudine-induced lymphopenia, data from our cohort showed that three of 35 patients (8.6%) who had immunologic failure were taking zidovudine compared to 14 of 292 patients (4.8%) who had a good immunologic response. However, this is not statistically significant ( p = 0.342).No patient in our study was taking tenofovir and didanosine, which can cause the imbalance of adenosine metabolites. While the causes of low CD4 counts without virological failure may require further study, we still suggest that this measure should not be used to determine antiretroviral treatment failure due to an unacceptably low sensitivity. In Thailand, it is now the standard recommendation to monitor plasma HIV RNA every 6 months in HIV-infected patients on highly active antiretroviral therapy, in order to detect antiretroviral treatment failure early so that there can be a prompt switch to a suppressive antiretroviral regimen.Conflict of interest: No conflict of interest to declare.