Improving the Health-care Transaction

If there is one key thing that patients want from health care, this is it: that they come away from the encounter in better shape. Payors—including the government, insurers, employers, and, increasingly, patients—are now demanding that these encounters be more affordable, and they are less tolerant of mistakes. It wasn’t planned this way, but by happy accident, this issue simultaneously hits on three requisite ingredients for the safe delivery of reasonably priced radiology: quality, efficiency, and appropriateness. Our cover story explores the use of lean-manufacturing techniques in radiology. The specialty, one radiologist says, is perfectly suited to standardization, a key instrument in the lean-services toolbox. Another article looks at the considerable investment that one radiology practice has made in defining and measuring quality. A third article examines the need to manage the appropriate utilization of imaging and the results of Massachusetts General Hospital’s deployment of decision-support software to ensure that every imaging study ordered was clinically indicated. It has been 11 years since the publication of the seminal report¹ from the US Institute of Medicine (IOM) calling for a heightened commitment to safety from the medical community. The report estimated that medical error killed 44,000 to 98,000 people annually, and, naturally, the initial response of the medical community was somewhat defensive: The numbers are inflated; those nurses are killing a lot of people; if only those surgeons would learn to tell their left from their right. It took a few years before most physicians thought about how their decisions could contribute to a safer health-care environment. In an article in Health Affairs, Wachter² partially attributes to the IOM report the ensuing lack of attention to diagnostic (as opposed to medication) error, largely because medication error responds better to the systems solutions that are the focus of the authors. The other reason for this inattention to diagnostic error is that the solutions available are not as airtight as systems solutions, which are based largely on measurement and standardization. The solutions that do exist for preventing diagnostic error are identified as appreciating the risks of cognitive shortcuts (heuristics) and using IT, both to aid decision making and to filter and organize clinical information better. Clearly, diagnostic error has not been ignored because it is a small problem. The muted response is more a factor of the difficulty of the solutions. Wachter points out that the Harvard Medical Practice Survey (on which the IOM’s startling mortality estimates were based) identified diagnostic error as responsible for 17% of these events—far more than the percentage of deaths caused by medication error. Wachter offers five suggestions: First, encourage research. Second, get regulators and accrediting bodies to promote those activities proven to reduce errors. Third, make those proven solutions that involve health IT a part of the definition of meaningful use. Fourth, improve the teaching of diagnostic reasoning in medical school. Fifth, make this skill a focus of board certification. An excellent new report³ from Pennsylvania Patient Safety Authority, produced by ECRI Institute (Plymouth Meeting, Pennsylvania) and the Institute for Safe Medication Practices (Horsham, Pennsylvania), cites a 2003 autopsy review that found diagnostic-error rates of 4% to 50%, with a median rate of 24%. The report also cites a 2008 meta-analysis of diagnostic-error studies; it shows that diagnostic error is encountered, across specialties, at an average rate of 10% to 15%—but at an average rate of less than 5% in specialties that rely heavily on visual interpretation (radiology, pathology, and dermatology). The report looks at commonly misdiagnosed conditions; common causes of diagnostic error, including cognitive-processing errors, communication issues, and other system-related issues; and strategies for decreasing diagnostic errors. It includes several tables and a self-assessment tool. Diagnostic error is not the sole province of radiology, but because imaging is a primary tool in diagnostic decisions, it is central to radiology and is likely to be the subject of growing interest in years to come. The work of such researchers as Elizabeth Krupinski, PhD, the current board chair of the Society for Imaging Informatics in Medicine, is a great foundation. Radiology’s next frontier well may be the discovery of new solutions to diagnostic error.