Computer-aided detection in mammography: Perplexing medial/legal issues for new technology


View content online at: http://www.appliedradiology.com/Issues/2001/08/Articles/Computer-aided-detection-in-mammography--Perplexing-medial/legal-issues-for-new-technology.aspx

Abstract:  As technology advances, the legal landscape must address the new medical and legal developments it creates.
Loading...

Dr. Berlin is Chairman of the Department of Radiology at Rush North Shore Medical Center in Skokie, IL and a Professor of Radiology at Rush Medical College in Chicago, IL. He has authored articles and lectured on numerous medical/legal issues.

A recent report from the Physician Insurers Association of America, a nonprofit organization consisting of major malpractice insurance carriers, indicated that radiologists are the specialists most frequently sued in malpractice lawsuits involving breast cancer. 1 Further, the allegation that an error in the diagnosis of breast cancer has occurred is now the most prevalent condition precipitating medical malpractice lawsuits against all physicians. 2

While some radiologists point to a drop in the number of malpractice cases alleging a missed mammographic diagnosis of breast cancer as a positive trend, an increase in the average indemnity paid in breast cancer lawsuits has become cause for concern. 3 Accordingly, radiologists and allied professionals around the country are taking steps to improve the accuracy of mammographic interpretation in the belief that it will assist them in recognizing breast cancer at its earliest stages and help reduce mortality from the disease.

Many published articles have suggested that the accuracy rate of mammography screening can be improved by incorporating a variety of practices, including: continuing medical education, double interpretation by a second radiologist, comparing current mammograms with previous ones, establishing 2000 as the minimum number of mammograms interpreted annually by an individual radiologist, and the use of computer-aided detection (CAD). 4-7

Most radiologists agree that it is too early to know definitively if any of these suggestions will improve accuracy of interpretation, but the initial findings appear to be encouraging. This article will provide a brief overview of one of these practices, CAD, as well as some of the medical/legal questions that may develop in the wake of this promising new technology.

Computer-aided detection

CAD is a sophisticated screening tool that identifies suspicious microcalcifications and masses on mammograms. The technology is designed to assist radiologists by providing a computerized second review that draws their attention to subtle changes in tissue that may indicate the presence of cancer. CAD works by digitizing an actual mammogram, then running proprietary algorithms over the digitized image. Once complete, highlighted areas of concern are provided to the radiologist for further review.

CAD is not intended to replace a radiologist's interpretation, but was designed as a computerized second opinion to be consulted only after the radiologist completes his or her initial review of the mammograms. The radiologist still makes the final diagnosis after analyzing all available data and differentiating between benign and malignant lesions.

Considering the desirability of detecting missed cancers and the complexity of double reading, many view CAD as a viable method to increase the sensitivity of mammographic screening. In fact, according to a study presented at the 2000 Radiological Society of North America meeting, CAD can find up to 20% more cancers than screening mammography alone. 8

Medical/legal questions

While CAD has shown promise in its ability to increase the detection rate and decrease error rate, there are a number of medical malpractice issues associated with the technology that will undoubtedly be the topic of many discussions as CAD systems become more commonplace.

One of the legal uncertainties surrounding the technology involves the ramifications of a radiologist dismissing information provided by a CAD system. For example, what happens if the technology marks a density that the radiologist concludes is a false-positive finding, but months later his or her patient develops a cancer in that highlighted area? Does the fact that the radiologist held a different "opinion" from that of the CAD system put the radiologist in legal jeopardy?

This leads us to another issue, perhaps the most controversial question regarding CAD: What should radiologists do with the results provided by the CAD system indicating potential areas of concern? If they keep the results on file and are later faced with a malpractice suit, they will be asked to turn over all records pertaining to the case in question. The CAD results could then be used against them if the cancer was indeed marked. On the other hand, if a radiologist knowingly discards the results, his or her motives could potentially be questioned in a court of law.

At the moment, there is no case law or federal or state regulation regarding whether CAD results should be saved or discarded. However, laws may eventually offer more direction on this point.

As the use of CAD becomes more widely accepted, another question beginning to surface is whether or not radiologists will eventually be violating the standard of care if they don't offer and use the technology. The law defines malpractice as "a breach of the standard of care that causes injury to a patient." 3

So how is this standard defined? It is defined by what reasonable and appropriate physicians in the same specialty are doing. By extension, the standard is established from a variety of sources, including scientific literature, seminars, testimony of other physicians, legal decision, and by standards published by the American College of Radiology. At the moment, it is clear that using CAD is not the standard of care, but that could easily change in time.

In malpractice litigation, the law takes the position that the jury of laypeople is not familiar with the medical standard of care. As a result, both sides in the case produce expert witnesses who attempt to explain the standard of care and whether or not the physician on trial breached it. Unfortunately for some radiologists, past juries have demonstrated that they expect physicians to be near perfect in their diagnoses, which has made cases involving missed breast cancers difficult to defend.

Summary

Every technological breakthrough requires a period of adjustment and adaptation for it to obtain acceptance. It would appear that while initial studies regarding CAD have been favorable, the technology must continue to demonstrate its value over time.

As with many innovations, CAD has both opened the door to an exciting new frontier and has created a flood of new issues and unanswered questions. One thing is certain, however, the technology's ultimate influence on malpractice litigation has yet to be seen. AR