Challenging Nature of Evidence-Based Decision Making

Challenging Nature of Evidence-Based Decision Making What is the drug of choice for condition x? is among the most commonly asked questions in primary care. Reflecting the complexity of prescribing decisions, answering this question requires a difficult trade-off between the benefits and harms of multiple drugs for a given condition. The principles of evidence-based medicine suggest that prescribing decisions should be guided by an objective benchmark, namely scientific evidence. Such evidence is particularly important when choosing a first-line treatment among multiple alternatives. Unfortunately, existing clinical evidence on benefits and harms is rarely adequate to inform prescribing decisions. A randomized controlled trial comparing all relevant drugs would provide such information. However, clinical trials are often designed for regulatory purposes and, therefore, include selective patient populations and do not include all available comparator drugs. To obtain insight into the comparative benefits and harms of multiple drugs, prescribers turn to summaries of evidence to discern the most promising drugs from their less effective comparators. Recent methods used to synthesize existing evidence provide much-needed information on the comparative benefits and harms of multiple drugs. Network meta-analysis is one such method that allows for the combination of direct and indirect evidences from randomized trials, facilitating the comparison of all relevant drugs even when they are not directly compared with each other in clinical trials. The recent surge in the number of network meta-analyses in the general medical literature is a testament to the increasing need for comparative evidence in prescribing decisions. Even when comparative evidence from network meta-analyses exists, however, making sense of it remains a challenge. In particular, prescribers and patients often struggle to weigh the relative benefits and harms of multiple alternatives. In this proof-of-concept study, we discuss the important yet challenging role of comparative clinical evidence in guiding prescribing decisions in clinical practice. Using a recent systematic review and network meta-analysis of statins as an example, we highlight the need to adopt a more formal framework to help prescribers and patients in identifying a first-line drug among multiple alternatives. We call for combining network meta-analysis methods with decision analytic approaches, such as multicriteria decision analysis, to encourage and to facilitate shared decision making between prescribers and patients.