This chapter was reproduced from the Emergency Medicine Clerkship Primer, 2008, Chapter 10, with the permission of the Editor, for ease of viewing on mobile devices.
Advances across all aspects of medicine have taken place in the last 25 years. These advances have resulted in improved quality of life for the US public, as well as more timely diagnosis of many medical conditions. The healthcare provided in the emergency department has benefited greatly from advances in diagnostic testing. Just 25 years ago it was difficult to obtain more than basic labs and plain radiographs for patients cared for in the emergency department. Today, because of advances in technology and the greater availability of diagnostic tests, definitive care can be initiated much earlier in the course of a patient’s presentation.
The emergency department in some ways can be viewed as a diagnostic testing center in which a physician can coordinate a patient’s care and the results of most diagnostic tests are known within a few hours. The importance of diagnostic imaging to the efficient practice of emergency medicine cannot be overstated. In most centers, advanced diagnostic imaging, ultrasounds, CT scans, and even MRI studies can be obtained quickly and have a pronounced effect on patient care by providing a prompt definitive diagnosis for conditions such as intracranial hemorrhage, pulmonary embolus, aortic aneurysm or dissection, and more.
However, with the ease in obtaining these studies comes a price. According to the American College of Radiology, diagnostic imaging is the fastest growing medical expenditure in the United States, with an annual growth rate of 9% (Bettmann and Weinreb, 2007). This is three times that of general medical expenditures. In emergency departments across the country from 1995 to 2004, the number of MRI and CT scans performed nearly quadrupled, and the number of ultrasounds more than doubled (JAMA, 2006). Remember that the ease of obtaining certain diagnostic tests, including advanced imaging studies, should not preclude or serve as a surrogate for performing a thorough H&PE.
Before ordering a diagnostic test, ask yourself a few questions: “What am I going to do with the test results?” “How is this test going to help me confirm or exclude the diagnosis?” “How will the test result affect my diagnostic strategy, management, or final disposition?” These simple yet important questions can help guide the proper use of a number of diagnostic tests ranging from relatively inexpensive blood tests to very expensive diagnostic imaging studies.
Diagnostic tests should primarily be ordered to rule in or rule out a particular condition based on the differential diagnosis generated from the patient’s H&PE. However, diagnostic tests are often ordered for a variety of reasons, including clinical suspicion of disease, dogma, perceived standard of care, a consultan’s or primary care physician’s request, a patient’s request, and risk management concerns. Remember, in the real world there are times when you may need to order a diagnostic test that is not necessarily evidence based. Sometimes, what we do is based on pattern recognition and anecdotal experience. Experience may be the only pseudo-science behind the diagnostic workup of certain patients.
As mentioned previously, a consultant physician may influence the diagnostic tests that are ordered for a particular patient. An example may be the patient presenting with abdominal pain. Your consultant may request that you order a complete blood cell count and an obstruction series. If you have a high clinical suspicion for acute appendicitis, it should not be lessened by a normal white cell blood count or a nondiagnostic obstruction series. Consultant interactions may also serve as an opportunity to educate your colleagues when differences of opinion are raised as to the utility of a particular diagnostic test.
Before ordering a diagnostic test, ask yourself a few questions:
The medical literature has reported validated approaches for diagnostic testing of certain conditions. Clinical decision rules, such as the Ottawa ankle and foot rules, the Nexus criteria the Canadian cervical spine rules, and others can be implemented to assist with the proper use of diagnos-tic tests (Steill et al., 1992, 1993, 2000; Hoffman, 2000). However, it must be remembered that the practice of medicine is an art as well as a science. With this in mind, there will be times when a particular diagnostic test is not indicated. To ensure that your patient expectations are being addressed, communicate your thought process with your patient as to why you do not feel that a particular test is indicated. As a student or physi-cian, use this opportunity to educate your patients. Remember, the value of communication with patients cannot be overemphasized.
Statistical Considerations Related to Diagnostic Testing
Sensitivity and Specificity
As much as we would like otherwise, many diagnostic tests are not abso-lute, and interpretation of the results may not be as simple as yes or no or black or white. A number of the tests have a range of probability for the correct answer. Many tests have the possibility of giving us a true-positive (TP), false-positive (FP), true-negative (TN), or false-negative (FN) result. Therefore, it is important to understand the statistical concepts of sensitivity and specificity. Sensitivity refers to the likelihood of a test being positive or abnormal in the presence of disease.
Mathematically, sensitivity is expressed as follows:
sensitivity = TP/(TP + FN)
The higher the sensitivity of a test is, the lower the number of false neg-atives will be. A test with a high sensitivity also means that a negative result has a high probability of truly being negative because of the low number of false negatives. Highly sensitive tests are able to reasonably rule out disease. A test that has poor sensitivity has a high likelihood of false-negative results.
Specificity refers to the likelihood of the test being negative or normal in the absence of disease.
It is represented mathematically as follows:
specificity = TN/(TN + FP)
A test that has high specificity means that it has a low rate of reporting false positives. A test that has poor specificity has a high likelihood of false-positive results.
The sensitivity and specificity of a test is not affected by the prevalence (the proportion of diseased patients in the population). However, the predictive value of a test is affected by the prevalence of disease in the population. Positive predictive value (PPV) refers to the likelihood of the patient truly having the disease when the test is positive or abnormal.
PPV is represented mathematically as follows:
PPV = TP/(TP + FP)
Negative predictive value (NPV) refers to the likelihood that the patient does not have the disease when the test is negative or normal.
NPV is represented as follows:
TN/(TN + FN)
Predictive values are highly affected by prevalence of disease in a given population. The lower the prevalence of a disease is, the higher the risk of a false-positive result will be and the lower the positive predictive value will be for a particular test.
Another important point to consider is the diagnostic testing theory of pretest and posttest probability. Probability relates to the concern that you have regarding a particular patient’s having an illness or condition and how that concern may or may not be affected by the diagnostic test results. An example of using pretest–posttest probability in the evalua-tion of chest pain is given in the box on this page. When assessing prob-ability, emergency physicians take into account trends and risk factors. A series of articles published in the Journal of the American Medical Association titled “The Rational Clinical Exam” reviews the current literature regarding a number of clinical conditions and provides evidence-based recommendations as to the usefulness of certain diagnostic tests in ruling in or ruling out a particular condition (JAMA, 2008).
Case Study—Examining Probability
For example, your patient is a 24-year-old, previously healthy, athletic male who presents with sharp fleeting chest pain lasting 5 minutes without shortness of breath while jogging earlier today. In the emergency department, he is without complaints. His ECG shows sinus rhythm with T-wave inversions V1–V3. Another patient is a 63-year-old man. He smokes cigarettes and has a history of poorly controlled hypertension. Earlier today, he experienced 20 minutes of left-sided chest pain with exertion, radiating to his left shoulder, associated with shortness of breath and diaphoresis. His ECG shows normal sinus rhythm and is unchanged from an earlier ECG. The first patient has a relatively low pretest probability for cardiac chest pain. His ECG is not normal, but in a healthy young man, the T-wave inversions are nonspecific at best. These findings have been commonly noted to be a normal variant in young healthy adults. The posttest probability (after the ECG was performed) that the 24-year-old patient has cardiac chest pain is still very low and is essentially unchanged by his abnormal ECG. However, the case of the 63-year-old patient with chest pain is very different. The pretest probability that he is having cardiac chest pain is relatively high. He has multiple risk factors for coronary artery disease, and his history is suggestive of cardiac chest pain. Although his ECG is normal, his posttest probability of having cardiac chest pain is no less concerning.
Ultrasound Testing by Emergency Physicians
In the past decade, the use of bedside ultrasound by emergency medicine residents and faculty in academic medical centers has become common-place. The application of this technology is also becoming more available to community emergency physicians as well. The use of ultrasound as a diagnostic modality has been incorporated into the 2007 Model of the Clinical Practice of emergency medicine, a comprehensive document representing the essential information and skills necessary for the clinical practice of emergency medicine by board certified emergency physicians. The incorporation of bedside ultrasound by trained emergency physicians has been shown to improve patient outcomes and quality and efficiency of patient care provided in the emergency department.
The need to improve patient care has been the single most important driving factor responsible for the migration of this diagnostic imaging modality out of the walls of the department of radiology. Clinical decisions regarding specialty consultation, operative management, and disposition can now at times be made in minutes as opposed to hours. The use of ultrasound by emergency physicians does not in any way supplant the need for adequate follow up diagnostic imaging or confirmatory studies in selected patients. The role or focus of emergency medicine bedside ultrasound (EMBU) is set forth to answer a few simple straightforward questions. EMBU is often used in selected clinical presentations to confirm or exclude conditions such as hemoperitoneum, pericardial effusion, abdominal aortic aneurysm, cholelithiasis, and intrauterine gestation.
Diagnostics, including point of care testing in the emergency department, continues to evolve. As our technology continues to advance, we will undoubtedly have greater access to the results of a multitude of diagnostic studies in a timely fashion. We must still continue to strive to practice medicine in a cost-effective manner that benefits our patients and does not overburden them and the health care system with unnecessary, and at times overused, testing.
- Bettmann, MA, Weinreb, JC. New Initiatives for ACR Appropriateness Criteria. Reston, Va: American College of Radiology; 2005. Available at: http://www.acr.org/SecondaryMainMenuCategories/NewsPublications/FeaturedCategories/ACRBulletin/Archives/2005.aspx. Accessed March 25, 2008.
- This statement discusses the appropriate use of radiographic imaging.
- Hoffman JR, Mower W, Wolfson AB, et al. Validity of a set of clinical criteria to rule out injury to the cervical spine in patients with blunt trauma. National Emergency X-Radiography Utilization Study Group. N Engl J Med. 2000;343:94–99.
- This article reports the results of the National X-Radiography Utilization Group.
- JAMA. Quickstats: number of emergency department (ed) visits with diagnostic imaging ordered or performed—United States, 1995 and 2004. JAMA. 2006;296:762. Available at: http://jama.ama-assn.org/cgi/ content/full/296/7/762. Accessed February 7, 2008.
- This article reports the number of emergency department visits with diagnostic imaging ordered or performed.
- JAMA. The rational clinical exam. Available at: http://jama.ama-assn.org/ cgi/collection/rational_clinical_exam. Accessed February 7, 2008.
- This is an ongoing review of the current literature on a number of clinical conditions that provides evidence-based recommendations of the usefulness of various diagnostic tests in ruling in or ruling out a particular condition.
- Stiell IG, Greenberg GH, McKnight RD, et al. A study to develop clinical decision rules for the use of radiography in acute ankle injuries. Ann Emerg Med. 1992;21:384–390.
- This study reports the development of a clinical decision rule for the use of radiography in acute ankle and foot injuries.
- Stiell IG, Greenberg GH, McKnight RD, et al. Decision rules for the use of radiography in acute ankle injuries. Refinement and prospective validation. JAMA. 1993;269:1127–1132.
- This study further refines and validates a clinical decision rule for the use of radiography in acute ankle and foot injuries.
- Stiell IG, Wells GA, Vandemheen KL, et al. The Canadian C-spine rule for radiography in alert and stable trauma patients. JAMA. 2001;286:1841– 1848.
- This article describes the Canadian C-spine rule and its application in alert stable trauma patients.
- 2007 Model of the Clinical Practice of Emergency Medicine. Available at: https://www.abem.org/public/_Rainbow/Documents/2007%20EM%20Model.pdf. Accessed March 6, 2008.
- This document was generated by a taskforce charged with revising the core content of emergency medicine and includes an extensive list of conditions seen and treated in the emergency department, subdivided into one of three acuity levels.