Annals of Emergency Medicine
Volume 39, Issue 2 , Pages 192-194, February 2002

Feedback: Computed tomography and lumbar puncture for the diagnosis of subarachnoid hemorrhage: Evidence, action, and error☆

Division of Emergency Medicine, Harvard School of Medicine, Department of Emergency Medicine, Beth Israel Deaconess Medical Center, Boston, MA Department of Medicine, Columbia University, College of Physicians and Surgeons, New York Presbyterian Hospital, New York, NY

Address for correspondence: Jonathan A. Edlow, MD, Department of Emergency Medicine, Beth Israel Deaconess Medical Center, CC-2-204, One Deaconess Place, Boston, MA 02215; E-mail jonathan_edlow@hms.harvard.edu.

Article Outline

Abstract 

[Edlow JA, Wyer PC. Feedback: computed tomography and lumbar puncture for the diagnosis of subarachnoid hemorrhage: evidence, action, and error [response]. Ann Emerg Med. February 2002;39:192-194.]

 

Back to Article Outline

In response: 

The commentary by Dr. Schwartz,1 in addition to calling attention to the importance of precise criteria for interpretation of diagnostic tests, raises 2 relevant issues pertaining to the adequacy of cranial computed tomography (CT) in ruling out subarachnoid hemorrhage (SAH) that were not addressed in our original review.2 Both issues pertain to the process through which a clinical estimate of disease likelihood is modified by a diagnostic test result in a way that influences the actions we recommend on behalf of our patients. Both issues also highlight the ways that evidence-based medicine, as a dimension of clinical thinking, differs from purely academic or research-oriented approach to clinical practice.

Dr. Schwartz points out that the CT scan result that was treated as “false-negative” by Morgenstern et al3 and in our review2 was not, in fact, normal. As illustrated by Dr. Schwartz, that scan, done within hours after the onset of headache, revealed a communicating hydrocephalus, a finding known to be consistent with a diagnosis of SAH. It did not show blood. The scan result is therefore ambiguous to the extent that a “positive” scan for SAH had been defined as direct visualization of blood in the subarachnoid space. Morganstern et al do not tell us how this scan was interpreted at the time it was performed. Their protocol called for performance of an LP if the scan did not show intracranial hemorrhage.3 Finding hydrocephalus only a few hours after onset of headache caused by SAH is distinctly unusual. Certainly, its absence should not be relied on to exclude SAH in patients whose scans do not show blood.

In accepting Morganstern et al's3 treatment of this scan result as false-negative, we were guided by principles of clinical reasoning. In particular, we followed a principle well known to emergency clinicians caring for potentially critically ill patients: when ambiguity arises, err on the side of patient safety. Exercise of this principle in routine clinical decisionmaking is also known as clinical judgment.

Clinical judgment emphatically pertains to the process of applying information from health care research to patient care. Stroke trials have illustrated graphically how difficult it can be to achieve uniformity in the interpretation of cranial CT scans, even under carefully prepared research conditions.4 In North America, scans are usually read by general radiologists. We believed that a subtle finding such as hydrocephalus on a cranial CT scan is particularly likely to be missed or misinterpreted in a nonresearch setting when first encounter caretakers are looking for direct evidence of bleeding in a neurologically intact patient with headache. We therefore decided to adhere to the assessment of the scan in question as a “false-negative” study, thereby lowering the estimated performance of CT in identifying patients with SAH.

Similar clinical reasoning led us to reject the apparently superior performance of cranial CT reported by Van der Wee et al.5 As Dr. Schwartz points out, the patients in that study were stated to be neurologically intact. However, Van der Wee et al admitted all of their study patients to hospital and found 68% of them to be positive for SAH. Because this practice and disease prevalence vary so markedly from what we ourselves are accustomed to, we decided that a much more detailed description of the patient population studied by Van der Wee et al than that provided by the authors would be required for us to be sure that a significant difference in spectrum of disease was not at play.6 Once again, we exercised the equivalent of clinical judgment in deciding to prefer the results observed by Morganstern et al3 to those reported by Van der Wee et al. Once again, we elected to err on the side of patient safety in an ambiguous situation.

Back to Article Outline

Making action thresholds explicit 

A second point raised by Dr. Schwartz's commentary has to do with the effect of a diagnostic test result on the decision to activate or not to activate specific diagnostic or management strategies. When specific posttest disease probabilities can be identified, above or below which such strategies are mandated, these are often termed “action thresholds.” This concept was introduced in a previous “Skills” article in the Annals of Emergency Medicine Evidence-Based Emergency Medicine series.7 The hypothetical patient around whom our review was centered identified an action threshold of less than 1% probability of SAH.2 This occurred in a context in which the hypothetical patient had a stated aversion to undergoing a lumbar puncture (LP). Certainly, not all patients manifest such an aversion, nor will all patients be able to respond to such an issue in a numerically precise way. We believe that this aspect of our scenario reflected a reality commonly faced by emergency practitioners. Ten patients, comprising nearly 10% of the study population in the study by Morganstern et al,3 refused LP outright. Our patient was willing to entertain it, but only if a negative CT findings left the likelihood of SAH above a level that she psychologically perceived to be low.

A 1% action threshold for deferring invasive workup for SAH was suggested by another author8 cited by Dr. Schwartz. Here again, the threshold was based purely on psychology, differing from our scenario only in that clinician psychology rather than patient psychology was at play.

Can either a patient's or a clinician's intuition be taken as a legitimate basis for defining important clinical action thresholds? Neither our review nor the clinical scenario we used as a vehicle were intended to propose an action threshold of 1% for general application in connection with the clinical question at issue. Such parameters, to be applicable to broad classes of patients, are best based on established approaches to modeling clinical decisionmaking.9, 10, 11 Such approaches use objective structured methods for combining patient values or incremental costs of selected strategies with best available evidence on the likely harms and benefits associated with the adoption of those strategies. They allow specific thresholds to be identified in an objective fashion. In this context, the effect of uncertainties and ambiguities in the performance of diagnostic tests, of pretest probabilities, and of the consequences of specific courses of action can also be explored and the effect of those uncertainties on the choice of strategies consistent with objectively stated values determined. Short of such analyses in a particular area, practitioners must continue to exercise clinical judgement in arriving at informed choices for their individual patients.

In closing, we are struck that Dr. Schwartz comes to the same conclusion as did others who have responded to our review,12, 13 and indeed, the same conclusion that we ourselves arrived at: In patients with possible SAH and a negative scan, an LP is warranted. That our review inspired this much discussion and debate in a context in which all involved parties are in agreement with the conclusion encourages us in the conviction that emergency physicians actively seek better conceptual tools and skills in applying scientific knowledge to their daily practice.

Back to Article Outline

References 

  1. Schwartz DT. Feedback: computed tomography and lumbar puncture for the diagnosis of subarachnoid hemorrhage: the importance of accurate interpretation. Ann Emerg Med. 2002;39:190–192
  2. Edlow JA, Wyer PC. How good is a negative cranial computed tomographic scan result in excluding subarachnoid hemorrhage?. Ann Emerg Med. 2000;36:507–516
  3. Morgenstern LB, Luna-Gonzales H, Huber JC, et al.  Worst headache and subarachnoid hemorrhage: prospective, modern computed tomography and spinal fluid analysis. Ann Emerg Med. 1998;32:297–304
  4. Hacke W, Kaste M, Fieschi C, et al.  Intravenous thrombolysis with recombinant tissue plasminogen activator for acute hemispheric stroke. The European Cooperative Acute Stroke Study (ECASS). JAMA. 1995;274:1017–1025
  5. Van der Wee N, Rinkel GJE, Hasan D, et al.  Detection of subarachnoid haemorrhage on early CT: is lumbar puncture still needed after a negative scan?. J Neurol Neurosurg Psychiatry. 1995;58:357–359
  6. Edlow JA, Wyer PC. Feedback: computed tomography for subarachnoid hemorrhag—don't throw the baby out with the bath water. [response] Ann Emerg Med. 2001;37:680–685
  7. Hayden SR, Brown MD. Likelihood ratio: a powerful tool for incorporating the results of a diagnostic test into clinical decisionmaking. Ann Emerg Med. 1999;33:575–580
  8. Singal BM. A tap in time?. Acad Emerg Med. 1996;3:823
  9. Petitti DB. Meta-analysis, Decision Analysis and Cost-effectiveness Analysis: Methods for Quantitative Synthesis in Medicine.. 2nd ed. New York, NY: Oxford University Press; 2000;
  10. Richardson WS, Detsky AS. Users' guides to the medical literature. VII. How to use a clinical decision analysis. A. Are the results of the study valid? Evidence-Based Medicine Working Group. JAMA. 1995;273:1292–1295
  11. Richardson WS, Detsky AS. Users' guides to the medical literature. VII. How to use a clinical decision analysis. B. What are the results and will they help me in caring for my patients? Evidence-Based Medicine Working Group. JAMA. 1995;273:1610–1613
  12. Prosser RL. Which review should we believe regarding the diagnostic power of computed tomography for ruling out subarachnoid hemorrhage?. Ann Emerg Med. 2001;37:679–680
  13. Hoffman JR. Computed tomography for subarachnoid hemorrhage: what should we make of the “evidence?”. Ann Emerg Med. 2001;37:345–349

☆ Reprints not available from the authors.

PII: S0196-0644(02)97797-8

doi:10.1067/mem.2002.121642

Annals of Emergency Medicine
Volume 39, Issue 2 , Pages 192-194, February 2002

Article Tools