BACKGROUND: In allocating resources to improve patient care, two decisions must be addressed: 1) whether to issue guidance on the use of a health technology considering its cost-effectiveness, and 2) whether to invest in the active implementation of such guidance (e.g. via education, incentives or regulation), as its clinical use does not necessarily follow. Since these two decisions are related, resource allocation can be based on either sequential (1-by-1) or integral analysis. OBJECTIVES: To identify the preferred approach to the analysis for addressing decisions about issuing and actively implementing clinical guidance. APPLICATION: The application relates to the allocation of resources to metastatic hormone-refractory prostate cancer (mHRPC) in the UK. METHODS: An integrated Bayesian approach to decision modeling and evidence synthesis is adopted. Evidence on the costs and QALYs of all plausible treatment regimens is combined with estimates of treatment usage and population size. Implementation costs and effects are assumed to vary between treatment options. Both sequential and integral analyses of resource allocation are performed. RESULTS: (preliminary): For cost-effectiveness thresholds £25,000-£32,000 per QALY, decisions about how best to improve care for mHRPC patients differ between the alternative analytic approaches. Based on sequential analysis, Mitoxantrone+Prednisone is deemed cost-effective, without actively implementing this guidance. An integral analysis reveals that active implementation of Docetaxel+Prednisone (3 weekly) is the cost-effective option, yielding an additional 0.05 QALY per patient. By combining uncertain evidence on treatment regimens, their usage and active implementation, allowance is made for all uncertainty associated with resource allocation in mHRPC. CONCLUSIONS: For the analysis of the related decisions about 1) issuing, and 2) actively implementing clinical guidance, an integral approach is preferred over a sequential one. As the application in mHRPC demonstrates, integral analysis provides better options for improving patient management, more comprehensive insight in decision uncertainty and, consequently, an efficient allocation of resources.