The need for an enhanced focus on evidence-based decision making in public health settings is critical for several reasons. First, we need scientific information on the programs that are most likely to be effective in promoting health.1,2 An array of evidence-based interventions (EBIs) is now available from numerous sources including the Guide to Community Preventive Services.3 Second, to translate science to practice, we need to marry information on EBIs from the peer-reviewed literature with the realities of a specific real-world environment.4,5 To do so, we need to better define processes that lead to evidence-based decision making. This process of evidence-based decision making can include the so-called "administrative evidence-based practices," which are agency (health department)- and work unit-level structures and activities that are positively associated with performance measures (eg, achieving core public health functions, carrying out EBIs).6 Finally, we need a stronger body of literature on effective approaches for scaling up interventions of proven effectiveness more consistently at the state and local levels. This process often involves extending the reach of an intervention by replicating it in other localities, cities, or states (the so-called "horizontal scale-up").7

The article by DeGroff and colleagues,8 published recently in the Journal of Public Health Management & Practice, illustrates several of the key issues and gaps related to evidence-based decision making in cancer control. In particular, the authors examined the use of EBIs related to breast and cervical cancer early detection. By surveying grantees in the National Breast and Cervical Cancer Early Detection Program, the authors were able to document the use of 5 EBIs. They found that EBIs are being used widely with clients and providers, yet gaps remain related to evaluation and training.

The DeGroff and colleagues study and related literature9-12 illustrate several important implications for practitioners:

* While the literature is increasingly showing which EBIs are being implemented, there is continued need for attention to how EBIs are being implemented and the factors needed for scaling up EBIs, particularly for lower resource settings.

* Data on implementation alone are not sufficient. DeGroff and colleagues found that less than half of respondents conducted evaluation of EBIs, suggesting that greater attention on real-world evaluation is needed.13

* As in the DeGroff and colleagues study, nearly all data on the use of EBIs are self-reported and subject to recall bias. The medical record audit is a "gold standard" for validation in health services research.14,15 We need to identify some parallel system for validating data on EBIs in public health settings.

* Early data on the extent of mis-implementation (ie, ending effective programs and policies or continuing ineffective ones)9 show that while cancer control programs tend to fare better than most other programmatic areas, there is room for improvement across all public health topics.

REFERENCES