OBJECTIVES: In health technology assessment, decisions about reimbursement of new health technologies are largely based on effectiveness estimates. These estimates are sometimes also used to predict the health-related quality of life outcomes, such as EQ-5D, as part of economic evaluation. However, sometimes these effectiveness estimates are not readily available. When many alternative instruments measuring these outcomes are being used (and are not all reported) or an extended follow-up time of clinical trials is needed to evaluate long-term endpoints (and drug development is at an early stage), data on relevant outcomes may be limited. The aim of this study was to develop methodology that would allow synthesis of all available evidence to assess interventions early and reduce uncertainty around relevant outcomes. METHODS: Bayesian multivariate meta-analysis methods have been developed for synthesis of diverse sources of evidence: multiple outcomes (including surrogate, potentially short-term endpoints) and other external evidence. These methods were applied to an example in rheumatoid arthritis where outcomes such as the Health Assessment Questionnaire (HAQ), the Disease Activity Score (DAS-28) and the American College of Rheumatology (ACR20) are synthesized. External information about correlations between the outcomes was included in the form of informative prior distributions. Estimates of HAQ were then mapped onto EQ-5D. Also in an alternative approach, the multivariate framework was applied to model jointly the utility estimates and the clinical effectiveness outcomes. RESULTS: The use of multivariate meta-analysis led to reduced uncertainty around the effectiveness and utility estimates. Combining the HAQ with DAS-28 gave a 19% reduction in the uncertainty around the estimate of HAQ and also 16% around the estimate of EQ-5D. CONCLUSIONS: By allowing all relevant data to be incorporated in economic evaluations of new health technologies, this multivariate approach to meta-analysis can lead to reduced uncertainty and hence more efficient decision-making in health care.