Strengths and limitations
This review was conducted by a task force of diverse clinicians, allied health professionals, public representatives, teachers and researchers. This was a strength because it meant that interventions and outcomes were considered on clinical and methodological grounds, with robust checks by multiple experts.
The review provides an up-to-date summary of research, with two thirds of the included studies being published in the past decade. However, it has several limitations. The available evidence is heterogeneous and mostly at moderate or high risk of bias. Meta-analysis was not appropriate because the interventions and outcomes varied greatly and there were too few studies with similar outcomes. A number of studies examined outcomes that may not be the most helpful when seeking to assess effectiveness, such as whether people carry autoinjectors or short-term changes in quality of life. Very few studies reported in detail on mortality, admissions, preferences or resource use. There was also a lack of evidence about emergency management outside hospital.
Not all available interventions are included in the review because data from registry studies, cohort studies and similar were not included. These designs have often been used to explore educational interventions or to track the value of preventive approaches.