Selected Papers from the 1999 Society for Academic Emergency Medicine Meetings


Clinical Practice

Abdominal Pain in the HIV-Positive Patient

This was a retrospective chart review of consecutive patients presenting with a chief complaint of abdominal pain to a high-volume ED with a large population of HIV-positive patients. Of 108 patients comprising the study population, 72% had acute (lasting fewer than 7 days) abdominal pain. The leading ED diagnosis was abdominal pain of unknown etiology (19%), followed by the usual assortment of abdominal pathologies (e.g, gastroenteritis, peptic ulcer disease, appendicitis, etc.). Interestingly, AIDS-associated opportunistic infections (OIs) were diagnosed in 6% of patients, with only 3% later being diagnosed with AIDS OIs after admission or in follow-up. Thirty-five percent of HIV patients with abdominal pain were admitted, as opposed to 18% of the general ED population (P < 0.001).

Comment by Richard Harrigan, MD, FAAEM, FACEP

This study seems to indicate what we should already know, but must remember—patients with HIV can develop the same abdominal problems that HIV-negative patients do, and our work-ups should be done with this in mind. Of course, OIs are important additions to the differential diagnosis, and should be considered in light of the patients’ CD4 count if it is known (in this study, the mean CD4 count was 263 ± 224 mm3). As with the general population,1 the most common diagnosis at discharge was nonspecific abdominal pain, a diagnosis that I suspect was underutilized in this study, as some patients left with diagnoses such as "peptic ulcer disease," which is at best a presumptive diagnosis in the ED. (Source: Yoshida DK, Caruso M. Abdominal pain in the HIV positive patient [abstract]. Acad Emerg Med 1999;6:470.)


1. Lukens TW, et al. The natural history and clinical findings in undifferentiated abdominal pain. Ann Emerg Med 1993;22:690-696.

The most common discharge diagnosis in HIV-infected patients with abdominal pain in Yoshida and colleagues’ study was:

a. gastroenteritis.

b. lymphoma.

c. cytomegalovirus.

d. abdominal pain of unknown etiology.

Estimating the Probability of PE

The purpose of this prospective, observational study was to examine the concordance and accuracy of the pretest estimate of having a pulmonary embolus (PE). Emergency physicians (EP) made a choice for estimated pretest probability of low, mid, or high for patients receiving a diagnostic study for PE. When possible, a second EP completed a similar form that was blinded to the first EP’s assessment. A PE was considered present with a positive angiogram, CT, or MRA; high probability V/Q without contradictory evidence; or an intermediate probability V/Q with evidence for deep vein thrombosis.

In 142 cases, a primary EP recorded a pretest estimate and PE was diagnosed in 29 (20%). A second estimate was available in 34 cases, and the agreement was only fair (k = 0.42). The positive predictive value for PE of the first EP’s assessment was 0.27 for high, 0.22 for mid, and 0.09 for low pretest estimates.

Comment by Stephanie Abbuhl, MD, FACEP

Ever since the landmark PIOPED study was published in 1990, most physicians use a combination of pretest probability assessment and the results of VQ scan to determine a "post-test" probability of the likelihood of PE.1 PIOPED showed us that in a given patient with a low probability VQ scan, the risk of PE could vary from 4% to 40% depending on the pretest probability assessment. This preliminary study attempts to expose an underappreciated problem in the diagnosis of PE—how does one determine the pretest probability that is so critical to interpretation of a VQ scan? Does low clinical suspicion translate into no risk factors? Are all risk factors created equally? Despite volumes of literature on PE, very little has focused on this aspect of most testing algorithms. This pilot study reminds us that the determination of pretest probability is not, at this point, a structured, validated process and more research needs to be done to define explicit criteria to determine pretest probability. (Source: Jackson RE, et al. Emergency physician (EP) assessment of the pretest probability of pulmonary embolism (PE) [abstract]. Acad Emerg Med 1999;6:437.)


1. PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism. JAMA 1990;263:2753-2759.