Advanced Search

  • Topic
  • Specialty
  • Content Type


Also Related

< Back To Patient Safety
0 dislikes

< Hide

Comments For

Trouble Diagnosing Embolism


< Shrink

Add Your Voice

All comments are posted anonymously. Your comment will be attributed to: "Anonymous user."

post comment


Are you sure you want to delete this comment?

Trouble Diagnosing Embolism

By Tom A. Augello, CRICO and Ron Walls, MD, Brigham and Women's Hospital

Related to: Ambulatory, Communication, Diagnosis, Emergency Medicine, Primary Care, Other Specialties, Surgery

Despite several evaluations in the ED and PCP office, the cause of pulmonary symptoms eluded providers, leading to unnecessary surgery, fear of cancer—and a mid-range settlement.


  • Ron Walls, MD; Brigham and Women’s Hospital; Harvard Medical School; Boston, MA


The following case abstract is based on closed claims in the Harvard system. Names and some details have been changed to mask identities.

A 60-year-old man, a long-term smoker, saw his primary care physician after a plane trip from the United States to Australia and New Zealand, and was diagnosed with superficial phlebitis in his right leg. A week later, he underwent surgery to repair a torn rotator cuff in his right shoulder.

Within a week he presented to the Emergency Department complaining of right flank pain. He denied chest pain or shortness of breath. His vital signs were notable for a low blood pressure of 88 over 51. During his stay in the ED he had a syncopal episode in bed and was briefly unresponsive. Arterial blood gas was abnormal with a significant AA gradient of 35, but urinalysis, and CBC were normal. His BP improved, and he was discharged home with a diagnosis of pleurisy. The patient followed instructions to follow up with his PCP.

Five days later, he returned to the ER, with complaints of left flank and back pain. He appeared moderately ill, again with low BP, and now a fever of 101.8. He denied chest pain or cough. Rales were noted in the lower left lobe; he was diagnosed with pneumonia and admitted to the care of his primary care doctor. He was noted to have hemoptysis and the PCP ordered a pulmonary consult. Chest CT showed inflammatory changes and a slight pleural effusion, but no suspicious abnormalities. The pulmonologist considered embolus, but symptoms improved on antibiotics and the providers concluded that the patient had pneumonia. He was discharged home on penicillin and doxycycline.

The patient saw his PCP a week after the second ED visit with complaints of left rib pain and rales. The physician concluded he was still suffering from pneumonia. But when the pleural effusion still had not resolved a month later, he referred the patient to the pulmonologist, who performed a thoracentesis. The fluid persisted, and two months after the second ED visit, the PCP referred the patient to a surgeon for a suspicion of malignancy. The surgeon performed a thoracoscopy, and the frozen section initially suggested a malignancy, which prompted a full wedge resection. The final pathology report, however, concluded he had an pulmonary infarction and not cancer.

The patient sued his PCP, the pulmonologist, the ED attendings, and the hospital, alleging a failure to diagnose pulmonary embolus caused by deep vein thrombosis, which resulted in unnecessary lung re-section and emotional damage from the fear of cancer. The case was settled through mediation in the mid-range.

To discuss the risk management and patient safety aspects of this case, we are joined by Dr. Ron Walls, who is Professor of Medicine at Harvard Medical School and Chair of the Department of Emergency Medicine at Brigham & Women’s Hospital in Boston.

Dr. Walls, thank you for joining us.

Nice to be here.

It is always easier to connect the dots and see what could have been better in retrospect, but as we look at the facts in this case where do we start to see the care for this patient go wrong?

In the patient’s first Emergency Department visit there are lots of reasons to think just on the face of it that this could be a pulmonary embolism, looking at a 64 year old man with a recent very long plane trip and recent surgery, but obviously the workup in the Emergency Department has to take into account a lot of different considerations and a diagnosis other than pulmonary embolism could be reasonable too. What’s concerning in this case is that it appears that the physician in that first visit did not pay attention to the patient’s blood pressure of 88/51, which is a low blood pressure for a patient to present with. While in the Emergency Department the patient also had a peculiar sort of presyncopal or syncopal episode while in bed and was briefly unresponsive.

A syncopal episode while lying supine is very unusual and in the absence of a dysrhythmia, which probably would have been observed during the patient’s care. It is most likely some significant interruption in cardiac output and that plus the hypotension plus the plane trip alone should have brought people around to thinking more about pulmonary embolism and less about more benign causes of presentation. So, at the end of the visit when the physician is summating his or her knowledge what they have is a patient that came in with this right sided flank pain, hypotension albeit that resolved, a recurrent episode of a syncope or near syncope while lying in bed, and an arterial blood gas with a clearly abnormal AA gradient.  To make that into a diagnosis of pleurisy, which is a completely benign viral condition that can’t cause any of those things, represents faulty reasoning or an inability to include essential data elements in the differential diagnosis.

Okay, and then there is a discharge and there are multiple visits with a PCP back to the Emergency Department; so obviously there are some factors that we see very frequently in these failure or delayed diagnosis cases, which are these hand offs. There are multiple providers, and communication that is required in sort of seeing the big picture.

Well, there are and the important thing is that the patient is seen later by physicians who don’t have that same depth of knowledge of that first presentation that the first physician had. When the physician sees the patient the next morning in the primary care office he or she is very likely saying, “So, what did they tell you that you had?”  The patient would say “They told me I had pleurisy” and he might be on Ibuprofen or something and so the physician understands that their purpose in follow up is to check the pleurisy and see if it is doing okay, not to reopen the diagnostic question. 

The patient then returns to the Emergency Department five days after the first visit, and now he has pain on the opposite side.  This time he has a moderately ill appearance.  He is hypotensive again and he has a fever so he is diagnosed with pneumonia.  That’s not an inappropriate diagnosis for that presentation.  The patient didn’t have any cough though …and again if you put that together with his previous presentation you would have to ask yourself wait a minute why was the pain on one side five days ago and on the other side now, and am I really getting into a diagnosis of pneumonia here or am I missing something?  It is likely that the elements of history and data from the first visit like the hypotension, the near syncopal episode, and the AA gradient were either not conveyed to the second provider or could have been found by that second provider who didn’t look for them or look at them.

When we step back and look at the full picture of the care that the patient received—and the patient from his perspective was doing everything he was asked; he was showing up for the follow up visits with the PCP going into the Emergency Department when he felt that he needed to—what can we suggest as key changes in the way the care is organized or the actions or thought processes of individuals that might prevent this from happening to another patient?

One of the challenges we face particularly in emergency medicine, but throughout medicine, is how do we convey a full set of information and our thought process to the next physician who is going to care for the patient? Our system inherently is built on hand offs. Emergency physicians might see a patient for a short period of time during the course of an acute illness, but their primary care physician and other specialists may see them later. Each of these people needs to have a full set of information from the person before them. In fact, from all of the people before them. In the Emergency Department, the notes are often incomplete and may not be dictated or thoroughly written at the time the patient sees his primary care physician in follow up. This is something we’re working on with electronic medical records now so that in the very near future when a patient is seen in the Emergency Department and evaluated the follow up physician will have access not only to the laboratory results, which are now available electronically, but to all of the information from the visit including all of the data given by the patient to the providers. The same goes for consultation. There is a tendency with consultation to give the consultant really minimal information, such as fever, hemoptysis, questionable pneumonia. In reality what the consultant needs to know is everything that has happened to that patient up until then and is expected to find that out from the patient and the medical records. 

If some of that information is uniquely the domain of a physician who has seen the patient such as in this case the original emergency physician who noted the hypotension and the near syncopal episode. It is very difficult for that consultant to get the whole picture. So, they get a partial picture and then they are anchored by the things that people were thinking before they came along. 

So, what can we do to try to fix that? 

I think one of the things that may help would be more of a verbal discussion between a requesting consulting team and a providing consulting team. In other words, if I am internal medicine and I want a pulmonary consult it would be much better if I had a detailed conversation with a pulmonary consultant about the patient’s history and what my specific concerns were rather than conveying that in writing in a short sentence or two. I think that would help a lot. Secondly, if there are records that lie outside the system there has to be an easier way to get them. For example, in a hospital system usually the records that are available are just the records from hospital visits or hospital stays. In the new electronic medical record age it is possible to link primary physician records and outpatient visits into the medical records that are part of the database that the physician can access in hospital. The ability to access all of these records gives the consulting physician a much better look at the whole picture. 

May 1, 2007
0 dislikes

< Back To Patient Safety