BACKGROUND: Insulin-dependent diabetes mellitus, also known as type 1 diabetes, is a life-threatening condition and is the third most common chronic illness among young people. As a result of minimal or non-existent insulin production, people with diabetes must take over the normally automatic task of regulation of blood glucose levels. This is achieved by a complex regimen involving multiple, daily administrations of insulin coordinated with dietary intake and energy expenditure and monitored by blood glucose testing. OBJECTIVES: To examine the effectiveness of educational and psychosocial interventions for adolescents with type 1 diabetes designed to improve their diabetes management. Specifically, it addressed the following research questions: (1) Do educational and psychosocial interventions for adolescents with type 1 diabetes have beneficial effects on biological and psychosocial outcomes? (2) Are there types or features of interventions that have been shown to be more effective than others? (3) What evidence is there of the cost-effectiveness of interventions? METHODS: A search strategy was formulated, piloted and refined. Three journals were handsearched, 11 electronic databases were searched and personal contacts, flyers, conferences and websites were used to notify the research community of the review to access further literature. This process generated 10,535 abstracts, which, after screening, resulted in 367 articles identified for retrieval. This number was augmented by hand-searching, personal contact and exploding references, and a final total of 457 articles were scrutinised. Of these, 64 reports describing 62 studies were identified as empirical papers evaluating educational or psychosocial interventions. The relevant data were extracted from the papers and summary tables for each study were prepared. Where possible, effect sizes were computed for outcomes from studies that included a randomised control group (CG) and other relevant information. RESULTS: A descriptive analysis of the 62 studies was undertaken. Most studies (67.7%) were conducted in the USA and 41% were randomised controlled trials (RCTs), none of which were UK-based. Only 48% of the reports provided an explicit theoretical rationale for the intervention. The mean number of participants was 53.8. The studies took place in various settings, evaluated a variety of interventions, involved various interventionists, addressed various components and assessed the effects by a range of outcomes, including measures of metabolic control and psychological and behavioural outcomes. Follow-up assessments were relatively rare. RESULTS - THE EFFECTIVENESS OF INTERVENTIONS: The 25 RCTs were examined in more detail and three of the most effective were described in depth. Effect sizes could be calculated for 14 studies. The mean (pooled) effect size for psychosocial outcomes was 0.37 and 0.33 for glycated haemoglobin with outliers (0.08 without outliers), indicating that these interventions have small to medium beneficial effects on diabetes management outcomes. A narrative review of the 21 pre-post studies with no CG was performed, including evaluations of interventions conducted at summer camps, interventions for poorly controlled patients and educational interventions. All studies reported beneficial effects. RESULTS - COST-EFFECTIVENESS: Few studies addressed economic considerations associated with interventions, and the lack of information on costs and the diversity of outcomes included by investigators impeded cost- effectiveness comparisons. Shorter hospitalisation at diagnosis is at least as effective in achieving control and avoiding complications in adolescence as longer stays. Home care may result in improved outcomes but may not be cheaper than hospital care at diagnosis. Targeting poorly controlled subjects may reduce adverse events and hospitalisations and may be more cost-effective than generic interventions. There is a need for rigorous cost-effectiveness studies of educational and psychosocial interventions for adolescents with type 1 diabetes that include longer-term considerations. CONCLUSIONS: The following conclusions were drawn from this review: (1) Educational and psychosocial interventions have small to medium beneficial effects on various diabetes management outcomes. (2) Well-designed trials of such interventions are needed in the UK (no completed RCTs of educational or psychosocial interventions for adolescents with type 1 diabetes conducted in the UK were found). (3) The evidence, arising primarily from studies in the USA, provides a starting point for the design of interventions in the UK. (4) Quantitative and narrative analysis of the evidence suggested that interventions are more likely to be effective if they demonstrate the inter-relatedness of the various aspects of diabetes management. (ABSTRACT TRUNCATED)