Skip to main content

Differential diagnosis not reported


A 68-year-old man came to his family physician with complaints of a “several month” history of worsening memory, confusion, difficulty sleeping, and intermittent prob­lems with his left hand and arm becoming weak and numb.

Physician action

The family physician suspected TIAs, but wanted to rule out brain cancer. He or­dered a CT scan of the head and arranged for a carotid ultrasound. The family physi­cian completed the order form for the CT, requesting the CT to rule out brain cancer, but noted possible TIAs. He also included the patient’s symptoms on the form, and asked that the patient’s medical records be forwarded to the testing facility.

The family physician’s nurse called the hospital to set up the CT scan. She later testi­fied that she read the request from the family physician as “R/O brain cancer.” The billing clerk at the hospital changed that to read “R/O METS.” This information was then sent to the hospital’s radiology technician who changed it from “R/O METS” to “METS” because “R/O METS” did not fit the Medicare codes.

When the radiologist received the request, the clinical diagnosis was “METS.” None of the family physician’s suspicions or medical records noting “TIAs, organic brain syndrome, or mental status changes,” were forwarded to the radiologist. The CT scan was performed with and without enhancement. In the initial portion of the radiolo­gist’s report, he noted that what he saw was “consistent with metastatic disease.” Later in his report, he made reference to “this metastasis” rather than “this possible metastasis.”

The day after the CT scan, the patient reported to the emergency department (ED) at another hospital. His symptoms included dizziness, weakness, memory loss, and slurred speech. The ED physician suspected a TIA and administered warfarin. The patient was then admitted to the hospital, under the care of an internal medicine physician. This physician continued the warfarin, ordered a carotid ultrasound, and contacted the radiologist regarding the previous CT scan.

The radiologist read the report to the internal medicine physician. At that time, the internal medicine physi­cian decided to discontinue the patient’s anticoagulation treatment because it was contraindicated for patients with cancer. The carotid ultrasound was also cancelled.

The internal medicine physician ordered tests to look for the tumor, but they failed to find any evidence of cancer. After two days in the hospital, the patient was discharged with a diagnosis of “metastatic brain disease, primary tumor site undetermined,” and was referred to an oncologist.

Two weeks after he left the hospital, the patient suffered a major CVA. A CT scan and MRI of the head identified multiple areas of infarction with no evidence of metastatic tumor. A carotid flow study revealed total occlusion of the left internal carotid artery. The CVA caused severe paralysis to the left side of the body. The patient currently uses a wheelchair and is unable to speak.  


A lawsuit was filed against the radiologist, alleging the following:

  • failure to report an appropriate, accurate differential diagnosis;
  • failure to suggest additional, follow-up radiological studies;
  • issuing a misleading, inaccurate CT report of metastasis; and
  • failure to clinically correlate the information in the CT report which led to a failure to diagnose the patient’s condition.

The family physician, the hospital where the CT scan occurred, and the internal medicine physician were also named in the lawsuit.

Legal implications

The defendant radiologist was adamant that his interpretation of the CT scan was correct and was consistent with the history provided on the radiology request. The statement “METS” led the radiologist to believe that a diagnosis of cerebral metastases had been established, and that he was to report whether brain metastases were present on the CT scan.

Two board-certified radiologists reviewed this case and both felt the CT scan was far more suggestive of stroke than brain metastasis. Both radiologists said they would have listed possible ischemia on the differential.

The plaintiff’s expert, also a board-certified radiologist, felt the defendant’s read of the CT scan was accurate, but the defendant’s final impressions were incorrect because he did not list ischemic disease as a possible differential diagnosis.

The case against the radiologist was further weakened by the testimony from the code­fendant physicians and their experts. They all testified that it was within the standard of care to rely on the radiologist’s review of the CT in deciding to discontinue the patient’s anticoagulation treatment.

This case was further complicated by two factors. There was a dispute between the family physician’s nurse and the hospital billing clerk over what information was relayed over the telephone about the request for the CT. Regardless of this dispute, the radiology technician changed the diagnosis from “R/O METS” to “METS” and this affected the defendant’s review of the CT. Further, when the family physician received the CT report from the radiologist, the admitting diagnosis at the top of the report said “METS.” Had this been noted, it may have alerted the family physician to the error.


The case against the radiologist was settled during trial. The hospital and family physi­cian also settled. The case against the internal medicine physician was closed without indemnity payment.

Risk management considerations

The communications in this case broke down at many levels. The physician might have improved the communication from his office by having a procedure requiring that the CT order form be faxed to the testing facility, along with the pertinent records. That way, the testing facility staff would have received his full message. Proper staff training can also help keep problems like this from occurring or recurring.

It is common for ordering physicians to read only the diagnosis section of lab and radi­ology reports, missing important information elsewhere in the document. In this case, the physician missed the fact that the admitting diagnosis was erroneously recorded as “METS.” While that may seem understandable, an attorney will argue that each person in the chain had an opportunity to correct this problem, and none did.

The carotid ultrasound was never performed. Systems that record, in a very visible place, which tests have been ordered and performed, might assist physicians in reviewing their charts for this information. Physicians who fail to review the charts to confirm that ordered tests have been completed not only put their patients at risk, but subject them­selves to potential litigation as well. If patients choose not to proceed with the physician’s recommendations, that should be accurately documented.

Changing a diagnosis because it does not meet a coding requirement is fraught with risk. If that is required, there must be some place on the same document to record the full and unaltered information. If more information is needed, or something just does not seem to be complete or “ring true,” then follow up with the ordering physician is advised.

The radiologist’s interpretation in this case was likely swayed by the implication of a pre-existing diagnosis of malignancy, without the clinical history of possible TIAs, cognitive changes, and intermittent numbness of the left arm. Some of the consultants did not feel that the films were consistent with metastasis. It is important for physicians to realize that their unbiased opinions are needed because they may be the ones to correct an inaccurate prior diagnosis. Including differential diagnoses and recommending further studies to rule out or confirm each one will help guide the ordering physician in the efforts to arrive at the correct diagnosis.