OBJECTIVES: Prior research has highlighted the importance of psychosocial factors in 'difficult' asthma. This study aimed to review the content, effectiveness and cost-effectiveness of psycho-educational interventions designed to address these factors in patients with severe and difficult asthma. DATA SOURCES: Thirty-two electronic databases and other sources were searched for studies of educational, self-management, psychosocial and multifaceted interventions. REVIEW METHODS: Abstracts were screened in duplicate, against prior definitions, to identify eligible interventions targeted to patients with forms of or risk factors for difficult asthma. Studies were classified by patient group (child, adult) and graded along two dimensions related to study design and relevance in terms of the degree to which they were judged to have targeted difficult asthma. Detailed data were extracted from studies meeting a minimum design and relevance threshold. Characteristics of studies were tabulated and results qualitatively synthesised. Where sufficiently similar studies reported adequate data about comparable outcomes, quantitative syntheses of results were undertaken using a random effects approach to calculate pooled relative risks (RR) or standardised mean differences (SMD), with 95% confidence intervals (CI). RESULTS: Searches identified over 23,000 citations. After initial screening and removal of duplicates, 4240 possibly relevant abstracts were assessed. Papers associated with 188 studies were initially obtained and classified. Fifty-seven studies including control groups and those that were judged to have at least 'possible' targeting of difficult asthma (35 in children, 21 in adults, 1 in both) were selected for in-depth review. The delivery, setting, timing and content of interventions varied considerably even within broad types. Reporting of interventions and methodological quality was often poor, but studies demonstrated some success in targeting and following up at-risk patients. Studies reporting data suitable for calculation of summary statistics were of higher quality than those that did not. There was evidence from these that, compared to usual or non-psycho-educational care, psycho-educational interventions reduced admissions when data from the latest follow-ups reported were pooled across nine studies in children (RR = 0.64, CI = 0.46-0.89) and six studies with possible targeting of difficult asthma in adults (RR = 0.57, CI = 0.34-0.93). In children, the greatest and only significant effects were confined to individual studies with limited targeting of difficult asthma and no long-term follow-up. Limited data in adults also suggested effects may not extend to those most at risk. There was no evidence of pooled effects of psycho-educational interventions on emergency attendances from eight studies in children (RR = 0.97, CI = 0.78-1.21) and four in adults (RR = 1.03, CI = 0.82-1.29). There were overall significant reductions in symptoms, similar in different sub-groups of difficult asthma, across four paediatric studies that could be combined (SMD = -0.45, CI = -0.68 to -0.22), but mixed results across individual adult studies. A few individual studies in children showed mainly positive effects on measures of self-care behaviour, but with respect to all other outcomes in adults and children, studies showed mixed results or suggested limited effectiveness of psycho-educational interventions. No studies of psychosocial interventions were included in any quantitative syntheses and it was not possible to draw clear conclusions regarding the relative effectiveness of educational, self-management and multifaceted programmes. Data on costs were very limited. Of the two well-designed economic evaluations identified, both of multifaceted interventions, one in children suggested an additional cost of achieving health gain in terms of symptom-free days. Provisional data from the other study suggested that in adults the significantly increased costs of providing an intervention were not offset by any short-term savings in use of healthcare resources or associated with improvements in health outcomes. CONCLUSIONS: There was some evidence of overall positive effects of psycho-educational interventions on hospital admissions in adults and children, and on symptoms in children, but limited evidence of effects on other outcomes. The majority of research and greatest effects, especially in adults, were confined to patients with severe disease but who lacked other characteristics indicative of difficult asthma or likely to put them at risk. A lack of good-quality research limited conclusions about cost-effectiveness. Although psycho-educational interventions may be of some benefit to patients with severe disease, there is currently a lack of evidence to warrant significant changes in clinical practice with regard to the care of patients with more difficult asthma. Further research is needed to: (1) standardise reporting of complex interventions; (2) extend and update this review; (3) improve identification of patients at risk from their asthma; (4) develop and test appropriate outcome measures for this group; and (5) design and evaluate, via the conduct of high-quality pragmatic RCTs, more powerful psycho-educational interventions that are conceptualised in terms of the ways in which psychosocial factors and asthma interact.