How to Add Intelligence to the Healthcare System

Reading Time: 3 minutes read

Jerry Zeidenberg Publisher & Editor Canadian Healthcare Technology & Technology for Doctors Online

Jerry Zeidenberg
Publisher & Editor
Canadian Healthcare Technology &
Technology for Doctors Online

Radiologists are, without a doubt, some of the smartest people we’ll ever meet. But they’re not infallible, and IT systems are now being implemented to add further intelligence to the medical system. A case in point:

In September, Torontonians were shocked when a large community hospital announced that one of its radiologists made interpretive errors – missing stomach cancer in a patient who died eight months later – and that he might have made other mistakes.

Since the announcement, other patients have come forward saying they received clean bills of health after radiological exams at the hospital, only to discover later that they had tumours. Overall, 3,500 of the radiologist’s exams are now being re-assessed by an outside team.

Over the last couple of years, mistakes made by radiologists at other hospitals across Canada have been spotted. As a result, quality has emerged as a pressing issue, and projects are now being launched to double-check the work of diagnostic imaging departments.

In Hamilton, Ont., for example, a test system is being trialled that sends DI reports to a second radiologist for review. Neither physician knows the identity of the other, so professional embarrassment is avoided.

If a problem is noted, it is corrected before the results are sent to the referring physician. Similar tests are being run in British Columbia and Alberta.

The basic problem is that radiologists are only human and make errors. So the idea is to inject more checks and balances into the system and to catch mistakes early.

Another way of catching errors is to make greater use of computer-aided detection (CAD) systems. Some radiologists have pooh-poohed the technology as being inaccurate and only available for a few specialties, such as lung, breast and colon cancer.

However, these types of tumours are the top killers when it comes to cancer – any assistance in identifying them, as early as possible, would be a boon to patients.

What’s more, new CAD systems have steadily emerged, and there are new methods of dealing with the false-positives that tend to proliferate with CAD. Rather than writing-off CAD software holus-bolus, radiologists could be trained to better recognize the types of lesions that are identified by the automated systems. In this way, they can learn how to separate false-positives from actual tumours. A study by Nishikawa et al., reported in the March 2012 edition of the American Journal of Roentgenology, found the use of CAD in breast cancer screening increased the identification of tumours by 10 per cent. The authors concluded that additional benefits would accrue if radiologists learned how to better respond to the alerts given out by CAD systems.

On another front, medical attention can sometimes kill the patient with kindness – for example, too many X-rays, or doses that are too high, can result in radiation poisoning and illness.

That’s why both the Canadian Association of Radiologists (CAR) and the Canadian Association of Medical Radiation Technologists (CAMRT) have been advocates of the ‘image wisely, image gently’ movement.

Vendors have been responding with systems that can measure the X-ray doses received by patients from any number of machines. It’s a great way to identify patients who are getting too much ionizing radiation.

Moreover, the systems use analytics to determine when radiologists and technologists are using protocols that deliver too much radiation – managers can then provide feedback and constructive criticism to improve the way exams are delivered.

Analytics, of course, can also identify logjams and bottlenecks in patient flow – giving managers the opportunity to troubleshoot and improve workflow. The result? Better economics for DI departments and clinics. It also provides greater ‘customer satisfaction’ when the patient no longer waits as long for tests.

No one likes making mistakes, and if a computerized solution can reduce the number of medical errors that are made, reduce X-ray doses, and speed up the flow of patients, who can argue?

In the end, the public benefits, and the healthcare system achieves its goals of helping more patients and achieving higher quality ratings. It’s a win-win for patient and provider.

Jerry Zeidenberg is Publisher and Editor of Canadian Healthcare Technology, based in Toronto.


This site uses Akismet to reduce spam. Learn how your comment data is processed.