J M Portnoy1. 1. Section of Allergy, Asthma & Immunology, Children's Mercy Hospitals and Clinics, Kansas City, MO 64108, USA. jportnoy@cmh.edu
Abstract
INTRODUCTION: Many studies have demonstrated efficacy of immunotherapy (IT) for asthma. Two meta-analyses reached the same conclusion. Because some studies failed to demonstrate efficacy, it is instructive to review reasons why that may have been the case. METHODS: Two meta-analyses were used to identify studies that did and did not demonstrate efficacy. These were reviewed to identify the sources of variability. RESULTS: Most studies are able to demonstrate a treatment effect of IT although they have wide confidence intervals. The studies were heterogeneous with respect to selection of subjects, underlying populations, the main treatment protocol, concomitant treatments, the duration of treatment and followup, and the effect of the protocol on concurrent treatment. There were also problems with the blinding. In most cases, it was difficult to determine how the doses of extract given compared between studies. CONCLUSIONS: Studies that fail to demonstrate a treatment effect of IT may do so for many reasons. Given the heterogeneity of IT clinical trials as a group, it is not surprising that variable results occur. The use of multiple studies in a meta-analysis with its underlying assumption of homogeneity is problematic.
INTRODUCTION: Many studies have demonstrated efficacy of immunotherapy (IT) for asthma. Two meta-analyses reached the same conclusion. Because some studies failed to demonstrate efficacy, it is instructive to review reasons why that may have been the case. METHODS: Two meta-analyses were used to identify studies that did and did not demonstrate efficacy. These were reviewed to identify the sources of variability. RESULTS: Most studies are able to demonstrate a treatment effect of IT although they have wide confidence intervals. The studies were heterogeneous with respect to selection of subjects, underlying populations, the main treatment protocol, concomitant treatments, the duration of treatment and followup, and the effect of the protocol on concurrent treatment. There were also problems with the blinding. In most cases, it was difficult to determine how the doses of extract given compared between studies. CONCLUSIONS: Studies that fail to demonstrate a treatment effect of IT may do so for many reasons. Given the heterogeneity of IT clinical trials as a group, it is not surprising that variable results occur. The use of multiple studies in a meta-analysis with its underlying assumption of homogeneity is problematic.