The Importance of Clinical Correlation and Impact of Testing Choices on Clinical Care and Outcome
Ms. J was a very healthy, 63-year-old woman in the prime of her life until 2 days prior to admission, when she developed a sharp pleuritic chest pain on the left side. She denied any shortness of breath or upper respiratory symptoms, and had no pain or swelling in her lower extremities. She had traveled to Las Vegas, a 3-hour plane trip, 3 weeks earlier and had what appeared to be a viral illness at that time that improved within a few days. She was not taking any medications, and reported no other changes in her daily routine or abilities. Thinking that there must be some acute problem, she went to the emergency department. The physical examination demonstrated no leg swelling or tenderness; she did complain of some minor discomfort on percussion of the left lower lung field, the same side as her pleuritic chest pain. She had normal vital signs other than a respiratory rate of 20 breaths per minute, and a normal pulse oximetry reading. All of her laboratory values were within normal range, other than for a microcytic anemia thought secondary to her having a history of “heavy menses” for many years. She was still premenopausal.
The emergency room physician was concerned that she had a pulmonary embolus (PE), and ordered a spiral computed tomography (CT) scan of her lungs as the first diagnostic test. This was read by the on-call radiologist as being positive for a small PE on the right side in a small distal vessel. The left lung had no PE noted, but a small pleural effusion was described at the left lung base. The patient was started on anticoagulation therapy for a presumed PE. The next morning upon further review and consideration, the diagnosis of a PE was questioned because her initial symptoms were, in fact, on the side opposite the reported radiological finding, and a post-viral pleurisy was being suggested as the most likely cause of her chest pain. This was further supported when the spiral CT scan was re-read by another radiologist as well as a consulting pulmonologist, both of whom thought the CT was less than definitive for a PE. Upon hearing this, the patient’s attending physician chose to get a V/Q scan to help assess this situation further. The scan showed a small ventilation-perfusion mismatch at the right lung base, at the same location where a defect was noted on the previously performed spiral CT scan; the diagnosis of PE was now thought to be confirmed. There were no other mismatches identified.
Ms. J remained on anticoagulation therapy. Her left-sided pleuritic chest pain continued to improve, and her pleural effusion was felt worthy of being followed over time to make sure it resolved. If not, or if it was to grow in size, further testing would be performed. The physicians now believed that Ms. J had two problems—a post-viral pleurisy on the left side and a PE on the right side, with the latter found as a by-product of the work-up chosen and likely unrelated to her presenting chest pain.
This case illustrates several issues. First, patients may have more than one problem (eg, post-viral pleurisy and PE). What is more illustrative, however, is the fact that the symptoms that were responsible for this patient seeking medical assistance in the first place were on the side opposite the supposed PE. If this patient had gotten a simple chest radiograph upon presentation, and the pleural effusion was noted in the exact place where the chest pain was being reported, the physician likely would not have pursued the diagnosis of PE any further at that time. Was the finding on the initial spiral CT scan a “false positive” or merely a coincidental finding, given the fact that pulmonary emboli are likely to occur more commonly than we might like to think? Regardless, Ms. J is now committed to months of anticoagulation therapy, not without some risk in itself, given her history of anemia and heavy menses. Needless to say, she is currently undergoing an extensive work-up to find the source of the presumed PE.
I found this case of interest because it illustrates the importance of clinical correlation and the impact that our choice of testing has on clinical care and outcome. Medicine is not a black-and-white science, and at times, even what may appear to be an inappropriate test may return with “positive” findings that will alter treatment. We can only do our best to choose the “right” initial test and to follow our clinical judgment. As we develop new technology, however, and a greater number of diagnostic tests become more readily available, I can only think we will have similar situations to what I have described above. Knowing the sensitivity and specificity of each testing modality will help us somewhat, but medical-legal concerns also play a role in our decision-making process. Few physicians would have denied anticoagulation therapy to Ms. J given the findings described, even in the face of her initial clinical presentation. The question remains: What percentage of the general population would have a similar diagnostic picture on any given day in the absence of related clinical findings? For now, we can only do our best and approach each problem based on accurate data, sound knowledge, proper judgment, and proven evidence-based medical practice. The additional “cost” to society and to the patient can be great if we choose wrongly.
Send comments to Dr. Gambert at firstname.lastname@example.org