Commentary

Asthma education is one of the actions recommended as part of national guidelines for asthma management.¹ But systematic reviews of the evidence supporting this recommendation have been lacking until now. The reviews by Powell and Gibson, and Wolf et al presented in this issue systematically summarised the literature to determine whether asthma education reduces clinically important morbidity.

Both reviews used extensive literature searches and duplicate and independent data abstraction, and assessed the methodological quality of the individual studies. Important limitations of the studies included in the review by Powell and Gibson reduce the strength of the inferences. Substantial statistical heterogeneity was detected in some of the pooled results, but was reduced or eliminated with the use of standardised mean differences (SMDs). Some of the SMDs were calculated using imputed standard deviations, but the assumed transciption error was not corroborated by the authors of the original trials in question. A random effects model provides a more conservative estimate of the effect size when there is substantial heterogeneity, and may be more appropriate. None of the studies measured all of the outcomes, and a limited number of studies measured certain outcomes such as quality of life. Importantly, a single study in this review showed that less intense education increased unscheduled doctors’ visits.²

The review on educational interventions for asthma in children by Wolf et al was more rigorous. They used correspondence with the authors to verify data, subgroup analyses, and comparison of random and fixed effects models. Important limitations of the review included lack of quality of life as an outcome measure, statistical heterogeneity, and a limited number of studies for some important outcomes. The use of SMDs may introduce small sample bias, since the SMD is overestimated with very small sample sizes (usually <10).³ The benefits of self management programmes were manifest at 7–12 months following education initiation, after initial physiological improvements. Behaviour modification is delayed, and patients must first learn methods to control their asthma and prevent exacerbations.

Despite these caveats, self management programmes seem to be useful for managing asthma optimally. According to Powell and Gibson, just providing one—and intensively—seems to be the key, rather than the type of programme. Self management may provide patients with a sense of control over their asthma, reducing dependence on the healthcare professional and system. In particular, the use of a peak flow meter reduces school absenteeism, nocturnal symptoms, healthcare use, and admissions to hospital compared with symptom based strategies in children but not adults. Perhaps children are especially poor at estimating their bronchoconstriction until they learn to associate what they measure with their symptoms. The efficacy of self management programmes on quality of life remains uncertain.