SIIM: CDS transforms radiology from chump to champion
thumbs up - 247.78 Kb

ORLANDO, Fla.—Brigham & Women’s Hospital in Boston, home to the Center for Evidence-Based Imaging, is a clinical decision support (CDS) pioneer and may be the original CDS success story. For more than a decade, the hospital has been building and implementing CDS systems, and its architects have learned more than few lessons along the way.

Imaging is plagued by an appropriateness challenge, said Ramin Khorasani, MD, MPH, director of the Center for Evidence-Based Imaging. “When we started this program in 1998, multidisciplinary expert panels told us five to 15 percent of studies were inappropriate.” Recent data are less encouraging and suggest between 10 and 30 percent of all imaging studies ordered are likely inappropriate.

Khorsani spoke about the CDS program during the annual meeting of the Society for Imaging Informatics in Medicine (SIIM).

Another facet of the problem is the robust growth rate in advanced diagnostic imaging, which has outpaced growth in other services among Medicare beneficiaries. At the same time, the political dialogue is focused on cost reduction. The upshot? Imaging is a target, said Khorasani. Policymakers are weighing expanded payor-based pre-authorization and radiology benefits management as cost-containment mechanisms.

BWH, however, scored a win when it convinced its payors to green-light advanced imaging exams without pre-authorization. The lever BWH used to sway payors was evidence-based CDS.

Khorasani shared a few tricks of the trade. “Change must address people, process and technology.” Khorasani emphasized the practical nature of decision support, and explained that rules should be enabled for physicians who would most benefit from them. Khorasani, for example, does not tangle with neurosurgeons over head CT, preferring to focus on providing decision support where there is clear evidence and collaborative decision-making among stakeholders.

Specifically, evidence needs to be unambiguous and actionable. If national or local best practices have not clearly identified a decision support rule for a specific application, it is not applied in the BWH program. Khorasani, in fact, turns off more rules than he applies in the decision support program, keeping it focused, specific and meaningful.

In addition, the BWH team is quick to respond to changing evidence. If new data overturn prior practice, BWH can go live within four to eight weeks. “Otherwise, we will lose the support and trust of physicians,” said Khorasani. In contrast, in conventional practice it takes five to 14 years to implement evidence into practice, he noted.

Effective decision support extends beyond the clinical, according to Khorasani. Consider the case of the primary care provider who wants to order a chest CT to screen for lung-cancer in a low-risk patient. The first pop-up focuses on the financial i.e. the payor will not cover the study, which may be a key patient priority. If the provider attempts to proceed, he or she is then directed to evidence from the National Lung Cancer Screening Trial.

The BWH decision support tool provides physicians with a pager number to consult with a radiologist if he or she chooses to ignore evidence, an effort that supports communication and collaboration. It also incorporates a duplicate study mechanism, alerting physicians to a prior study for the same indication acquired in the previous 90 days.

Effective decision support combines carrots and sticks. The largest driver of decision support at BWH was online scheduling, said Khorasani. That is, radiology enticed referring physicians to comply with decision support by providing the incentive of online scheduling, which improves efficiency of front desk staff.

Sometimes, however, a stick such as a policy change requiring compliance is necessary.  A final key to decision support success is measurement. “Measure everything,” said Khorasani, who hinted at promising findings of in-progress studies at BWH examining the impact of policies and other levers.

Decision support is a win-win, said Khorasani. “It enables radiology to be more relevant, improve quality and reduce waste to create value.”