The US system of health care has been subject to numerous criticisms, citing costly and inefficient care delivery, fragmented and heterogeneous resource utilization, and lower life expectancy per health care dollar spent compared to other countries. The application of evidence-based medicine, using the best available scientific evidence in clinical decision making, seeks to address these criticisms. However, generating the appropriate evidence to inform clinical decision making, policy, and health care coverage remains a critical issue, especially as options for the prevention, diagnosis, and treatment of disease exponentially grow.
The Agency for Healthcare Research and Quality recommends comparative effectiveness research (CER) to generate evidence on the effectiveness, benefits, and risks of diagnostic and treatment options in a format that is immediately usable by clinicians, patients, policy makers, and health plans. Rawson frames CER for each of these stakeholders and highlights the unique challenges of CER application to diagnostic imaging, especially considering that the locus of control for CER implementation resides outside the radiology department. To conduct CER effectively requires information technology systems that facilitate data gathering, storage, analysis, and sharing. Safdar et al review technology resources that allow standard lexicography to code and report data, permit image annotation for structured indexing and retrieval, and automate information extraction in an effort to ease the constraints on the conduct of CER using diverse types of data.
Randomized controlled trials have been cited as the “gold standard” for generating evidence in CER. However, evaluation of the clinical effectiveness of diagnostic imaging differs from therapeutics in important ways. The clinical effectiveness of imaging is often measured by its effect on clinical decision making rather than on long-term outcomes. Furthermore, relevant comparisons between imaging technologies may require assessment not just of individual technologies but of the myriad ways these technologies can be combined. Therefore, randomized controlled trials may be of limited use in evidence generation for diagnostic imaging. Instead, Rutter et al summarize decision analysis and modeling, using computer simulation models to evaluate short-term outcomes and predict long-term outcomes and costs, as an alternative research methodology. One of the most important short-term outcomes in diagnostic imaging is test performance, typically measured in test sensitivity and specificity. Using the Prospective Investigation of Pulmonary Embolism Diagnosis data, Cronin and Kelly remind us of the consequences of neglecting false-positive findings when implementing a diagnostic test in populations with different levels of disease prevalence. Killeen et al and Elias et al provide examples of CER in diagnostic imaging illustrating the potential range of research studies that CER encompasses.
CER results should inform not just clinical decision making but also evidence-based coverage decisions. Fendrick et al note that value-based insurance design, whereby health plans reduce cost sharing for (and thereby economic barriers to) health care services with strong evidence of clinical benefit, can improve health care quality and efficiency. In contrast, low-value services, with little marginal benefit or even harm, should be defined and eventually not covered. Garrison et al provide an overview of the application of value-based insurance design to diagnostic imaging, while Cronin and Kelly review barriers to the implementation of value-based insurance design.
Radiology has many challenges ahead as we seek to justify our clinical value and our contribution to patient care quality. CER represents another tool to quantitatively assess this value and present the information to the diverse group of stakeholders, including patients, referring clinicians, payers, and policy makers.
Dr Carlos is a member of a physician advisory board with Philips Medical Systems (Andover, MA). Philips Medical Systems provided no support for this editorial, financial or otherwise.
References
1. Agency for Healthcare Research and Quality. What is comparative effectiveness research. Available at: http://www.effectivehealthcare.ahrq.gov/index.cfm/what-is-comparative-effectiveness-research1/ . Accessed June 2, 2011.
2. Rawson J.V.: Comparative effectiveness research in radiology: patients, physicians and policy makers. Acad Radiol 2011; 18: pp. 1067-1071.
3. Safdar N.M., Siegel M., Erickson B.J., et. al.: Enabling comparative effectiveness research with informatics: show me the data!. Acad Radiol 2011; 18: pp. 1072-1076.
4. Rutter C.M., Knudsen A.B., Pandharipande P.V.: Computer disease simulation models: integrating evidence for health policy. Acad Radiol 2011; 18: pp. 1077-1086.
5. Cronin P., Kelly A.M.: Influence of population prevalences on numbers of false positives: an overlooked entity. Acad Radiol 2011; 18: pp. 1087-1093.
6. Killeen R.P., Mushlin A.I., Johnson C.E., et. al.: Comparison of CT perfusion and digital subtraction angiography in the evaluation of delayed cerebral ischemia. Acad Radiol 2011; 18: pp. 1094-1100.
7. Elias A., Carlos R.C., Maly Sundgren P.: MR spectroscopy using normalized and non-normalized metabolite ratios for differentiating recurrent brain tumor from radiation injury. Acad Radiol 2011; 18: pp. 1101-1108.
8. Fendrick A.M., Smith D.G., Chernew M.E.: Applying value-based insurance design to low-value health services. Health Aff (Millwood) 2010; 29: pp. 2017-2021.
9. Garrison L.P., Bresnahan B.W., Higashi M.K., et. al.: Innovation in diagnostic imaging services: assessing the potential for value-based reimbursement. Acad Radiol 2011; 18: pp. 1109-1114.
10. Cronin P., Kelly A.M.: Value based insurance design: barriers to implementation in radiology. Acad Radiol 2011; 18: pp. 1115-1122.