By Stacey Kusterbeck
In some medical malpractice claims, computer decision aid or artificial intelligence (AI) findings on radiology studies and ECGs become the central focus. Ryan Shuirman, JD, has seen many such cases recently. In a typical case, a chest pain patient visits the ED and undergoes an ECG, which shows non-specific changes and no obvious ST-elevation myocardial infarction. The computer interprets a bundle branch block. “That may not have any bearing on the acute presentation. But it could be something that requires follow-up later and clear documentation that it was recognized so subsequent treating physicians would be aware of the finding,” says Shuirman, a partner in the Raleigh, NC, office of Cranfill Sumner LLP.
If the interpreting emergency provider disagrees about bundle branch block, then this should be documented, according to Shuirman. Emergency physicians (EPs) should outline the reasons why they disagree. “For radiology, the issue could be an incidental finding by the radiologist, or something that is attributed to ‘artifact’ that then garners little attention from the ED provider,” Shuirman explains.
If something concerning appears, which may or may not be related to the patient’s acute presentation, then it is likely worth the emergency provider commenting on it. That way, everyone involved would know that it was considered and effectively ruled out as an explanation for the acute presentation.
“The radiologist should — and by extension, the ED provider should — clearly articulate why she disagrees with the AI interpretation,” Shuirman says. If for no other reason, this would shut down the plaintiff attorney’s ability to argue the computer alerted the EP defendant to the exact problem that killed the patient, but the physician ignored it. “At least clear documentation [indicating] that the finding was considered but deemed insignificant would demonstrate taking the time to thoroughly work through a differential and could demonstrate to retrospective reviewers of a case that the prospective thinking was sound,” Shuirman offers.
With radiology studies, some software programs identify abnormalities that may or may not be present. That may or may not be significant to the radiologist. “The difficulty for the ED provider is determining whether the radiologist has addressed whatever the computer-generated diagnosis has been, and whether she agrees or disagrees with what the computer has found,” Shuirman explains.
If the radiologist does not address the computer finding directly, then the ED provider is left to make assumptions about what the radiologist has found significant or insignificant. If radiologists are not routinely addressing computer findings, then resources will be spent by emergency providers attempting to sift through reports and images, trying to rule in or out what the computer has found. “Most helpful would be radiologists who acknowledge computer findings, and then comment on why or why not the computer finding is accurate and significant to the patient’s care,” Shuirman offers.
ED clinicians often interpret their own studies before radiologists can generate a report. “To the extent ED providers see computer-generated findings before the radiologist has interpreted the imaging, then the ED provider would be well-served by directly addressing whether she agrees with the computer findings — and if not, why not,” Shuirman says.
Similarly, the emergency provider generally interprets ECGs without cardiology involvement. In Shuirman’s experience, it is common for ECG computer readings to include non-specific or insignificant findings that do not correlate with the patient’s presentation.
“From a risk management standpoint, though, an argument can be made that ED providers should address each computer finding in their medical decision-making, and outline why the computer finding is inapplicable to the case,” Shuirman says.
Even if the emergency provider is ultimately wrong, at least there would be good documentation indicating the problem was considered and thoughtfully addressed.
“Rather than appearing defensive, such documentation would demonstrate a conscientiousness and thorough understanding of the patient’s issues. That might ameliorate any concerns a patient or family might have in a retrospective analysis of a given case,” Shuirman says.
If the radiologist does not address computer findings directly, the ED clinician is left to make assumptions about what the radiologist has found significant or insignificant. If radiologists are not routinely addressing computer findings, emergency providers will spend resources attempting to sift through reports and images, trying to rule in or out what the computer has found. Radiologists should acknowledge computer findings, and comment on why or why not the finding is accurate and significant to the patient’s care.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content