In reply to Commentary: “What Are the Unintended Consequences of Changing the Diagnostic Paradigm for Subarachnoid Hemorrhage After Brain Computed Tomography to Computed Topographic Angiography in Place of Lumbar Puncture?”
Year: 2010
Type: editorial
Abstract: We welcome the commentary by Dr. Edlow regarding the unintended consequences of changing the diagnostic paradigm for subarachnoid hemorrhage. In his second to last paragraph, Dr. Edlow recognizes that we did identify and share the concerns that he addresses in his commentary. Our purpose was to answer the question: Can computed tomograph/computed tomography angiography (CT/CTA) replace CT/lumbar puncture (LP) to diagnose aneurysmal subarachnoid hemorrhage? Our article affirmatively answers that question using a mathematical model.1 However, we close the paper with the very important next question: should it? Dr. Edlow elaborates on this question and provides accurate real-world considerations that all clinicians must weigh when choosing a diagnostic strategy for subarachnoid hemorrhage. We believe that we have made the argument that CT/CTA can be a diagnostic alternative that the physician can pursue to rule out subarachnoid hemorrhage. It is another tool in the diagnostic armamentarium. Historically, CT/LP was the only truly accepted or defensible diagnostic algorithm. Clinicians now have some ability to vary that algorithm depending on individual patient needs and concerns. Specifically, when an LP is unsuccessful or the patient absolutely refuses, CT/CTA is a reasonable alternative strategy. Dr. Edlow’s comments about “easy” technology and concern of testing “creep” are accurate and real. This is a concern in all areas of diagnostic medicine. With the use of observation units, partially because it is “easier” than the traditional admission process, we tend to observe and run diagnostic studies on much lower risk chest pain patients. Some of these patients have false-positive initial screening tests and undergo “unnecessary” invasive diagnostic testing including coronary angiograms. This is a real concern but it must be countered by the success in finding the unexpected patient who truly has significant coronary artery disease. One could argue that the CT/LP study is 100% sensitive,2 and therefore nothing is gained by an alternative diagnostic strategy. However, this is only true on the patients in whom the testing is conducted. A recent study by Vermeulen and Schull3 shows that 5% of subarachnoid hemorrhages are still initially misdiagnosed. As a result of the real-world hurdles with regard to LP, an easier strategy may have captured these patients at a critical juncture in their presentation. We agree with the concerns about the potential downstream negative effects of diagnosing an asymptomatic aneurysm. Further, any additional contrast CT studies do come with the risk of nephrotoxicity and cancer. What our paper could not address and what needs to be pursued further is a method to identify those patients truly at risk for subarachnoid hemorrhage. This disease process and the testing implications highlight the need to develop adequate clinical decision rules so that only those who require testing get it. We did not engage in a complete discussion regarding informed consent for LP. This is a complex issue and could easily warrant an entire paper. We did not mean to imply that the risks should not be discussed further. However, when obtaining informed consent for LP the question is also often asked: Is there a less invasive option? In direct response to that question, all of the implications of CT/CTA should be discussed. Finally, we agree with Dr. Edlow that a clinical trial that proves this diagnostic strategy and potential consequences should be undertaken. We stress that it would require 3000 patients because that is a real-world limitation. The practice of CT/CTA is already in use in some centers and a study of this size would take years to accomplish. It should be conducted. Unfortunately, pursuing it does not meet the current clinical needs. It should also be noted that, in all likelihood, studies will be published that are underpowered. This has been the case in the subarachnoid hemorrhage/CT literature in the past in which favorable results were limited by wide confidence intervals.4,5 An underpowered study strikes us as a riskier method on which to change clinical practice when compared with a mathematical model. We agree with the issues raised in this commentary. Which diagnostic algorithm to pursue should remain a choice based on an individual patient’s presentation. Clinicians must have an awareness that all testing algorithms have significant shortcomings or potential consequences. Again, our goal was to show that CT/CTA has diagnostic equivalence similar to CT/LP. Whether or not it should replace the criterion standard remains to be seen. If an easier testing strategy captures those 5% of patients with “missed” subarachnoid hemorrhage, it may offset the unintended consequences. This will only be determined by further study. Supervising Editor: David C. Cone, MD.
Source: Academic Emergency Medicine
Authors: Robert F. McCormack, Alan Hutson
Language: English
Subfield: Neurology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W1561609801