Reimbursement authorities often require pharmaceutical companies to provide them with more than just placebo-controlled data from RCTs. Instead, they typically seek data from a wider "real-world" setting, where the focus is on generating evidence of comparative effectiveness. The natural temptation for many pharmaceutical companies is to provide this evidence from separate, post-market approval studies. However, this approach can be expensive and undoubtedly leads to delays in reimbursement. We propose that both the additional costs of evidence gathering and the delays between regulatory and reimbursement approvals could be reduced by combining the main design elements of randomised, pragmatic, and prospective observational studies into a single, integrated Phase 3/4 study. This single study approach would typically begin with a standard RCT phase where, for example, an initial cohort of patients would be randomised to receive either the investigational therapy or placebo. Either in parallel with or following this phase, a second patient cohort would be randomised under pragmatic clinical trial conditions with the aim of comparing the investigational therapy with placebo and a limited number of active comparator treatments. Lastly, a third (observational) cohort would be enrolled and allocated to a wider range of therapies, as per clinical practice. Data from the RCT cohort would be used to obtain limited regulatory approval. Following this, data from the pragmatic cohort, once available, would then be formally combined using standard statistical techniques with data from the RCT cohort in order to obtain a wider regulatory approval and possibly some form of conditional reimbursement. The pragmatic and observational cohorts would then provide the comparative effectiveness data to allow for reimbursement across different patient groups. We outline the strengths and weaknesses of this approach, and discuss its operational considerations.