In their newest best practice guidelines in the Annals of Internal Medicine, the American College of Physicians practically begs clinicians to stop chasing phantom pulmonary emboli. Nothing super new here, but they do explicitly call out the big offenders:
Best Practice Advice 1: Clinicians should use validated clinical prediction rules to estimate pretest probability in patients in whom acute PE is being considered.
Best Practice Advice 2: Clinicians should not obtain d-dimer measurements or imaging studies in patients with a low pretest probability of PE and who meet all Pulmonary Embolism Rule-Out Criteria.
Best Practice Advice 3: […] Clinicians should not use imaging studies as the initial test in patients who have a low or intermediate pretest probability of PE.
When I cover the ER, I am routinely impressed in the low diagnostic yield of a PE CT (for actual PE). When I review the chart in protocoling/interpreting these studies, it’s obvious that a significant portion of these patients are being imaged inappropriately, either because there is already a better diagnostic explanation from the initial history/workup, PE is clinically extremely unlikely, or because a positive d-dimer is being chased out of context. Until recently, this profligate waste was a winner to all involved parties.
- The ordering clinician could feel their anxiety and liability washed away.
- The patients could feel that they were getting a complete and thorough workup and were relieved when their tests were negative.1
- The radiologist and hospital got paid.
Nagging concerns of radiation and systemic waste aside, everybody wins. And over time, the d-dimer turned into a bludgeon against reason, and the ready availability of CT made it psychologically and medicolegally more sensible to image aggressively.
The d-dimer was never intended as a screening test for every single patient with chest pain in the emergency room. A positive dimer in an inappropriately risk-stratified patient should not mandate a follow-up CTA. This is especially the case when the test is originally ordered by a nurse as part of a standing order protocol and not by physician who is actually responsible for the patient’s ultimate care. In my brief two-month stint doing clinical medicine in the ER as an intern, I often absorbed patients from the waiting room who already had an EKG, chest radiograph, and labs including troponins and a dimer. Then we were “forced” to get a PE protocol CT to “work-up” the dimer, even in patients who had obvious other explanations for the test results (e.g. an obvious pneumonia on the radiograph). Not everyone practices this way, but it’s easier to practice thoroughly (defensively) in most of the same ways it’s easier to give antibiotics for viral illnesses.
There is one important and misleading exception to premise of the ACP report. And that’s the notion that CTs ordered in the context of “suspected” PE are exclusively obtained to evaluate for PE (i.e. PE CTAs don’t have diagnostic value outside of evaluating for PE). Some of these patients have clinical symptoms without radiographic findings, and the ordering providers are obtaining imaging to further evaluate the lung parenchyma for signs of occult infection (as well a rib fractures, anything else). CT is a troubleshooting modality in cases where the clinical picture is cloudy. So the angiographic component of the CTA may be partially a “why-not” inclusion to exclude a potentially life threatening PE in a patient that was destined for imaging anyway.
That said, I still feel like I almost diagnose more PE incidentally on abdominal imaging than I do on dedicated PE studies.