Opinions expressed in the Correspondence section are those of the authors, and not necessarily of the editors, ACER or SAEM. The editor reserves the right to edit and publish letters as space permits. Letters not meeting submission criteria will not be considered for publication. See "Instructions for Authors."
CORRESPONDENCE Bias in Searching ED Patients To the Editor: T h e a r t i c l e by G o e t z et al, "Weapons Possession by Patients in a University Emergency Department" [January 1991;20:8-10], adds additional data to the important but extremely limited study of emergency department violence. Unfortunately, their data do not convincingly support the conclusion "that psychiatric patients were no more likely to be carrying a weapon than were medical patients." While this was true among searched patients, the searches were not conducted randomly. If one were to assume, for example, that uns e a r c h e d p a t i e n t s c a r r i e d no weapons, then the absolute percentage of weapon-carrying "psychiatric" patients would be 1.92% (60 of 3,131), versus 0.07% (24 of 35,874) for "medical" patients.
R o b e r t J Rothstein, MD, FACEP - Section Editor Bethesda, Maryland
The consequential implication, that "bias" toward searching psychiatric patients may therefore be unfounded, should also be questioned. Because patients were searched on the basis of medical or security personnel "suspicion," it must be presumed that most searches resulted from an assessment of patient behavior. Until much harder data are presented, it remains logical to suspect that irrational, intoxicated, or aggressive and p r e v i o l e n t i n d i v i d u a l s w o u l d be m o s t l i k e l y to c a r r y weapons into, and (much more importantly) use them in, the ED. Unfortunately, the percentage of such patients treated by psychiatric emergency services is dangerously high. The relative percentage of "psychia t r i c " versus " m e d i c a l " p a t i e n t s searched is therefore likely an acceptable and expected reflection of these behavioral clues. Until detailed patient data are presented from those institutions elect20:9 September 1991
ing to subject all patients to metal detection or personal search, behavioral clues must remain the indications for search, and "psychiatric" patients will continue to be searched more frequently than "medical" patients. In fact, the data presented by Goetz and colleagues clearly refute unfounded bias. Because the percentage of searched patients harboring w e a p o n s was c o m p a r a b l e a m o n g "medical" and "psychiatric" groups, the searches must have been appropriately instituted based on behavioral suspicion and not on diagnosis or specialty service.
Frank W Lavoie, MD, FACEP Department of Emergency Medicine University of Louisville Louisville, Kentucky
In Reply: We appreciate the points raised by Dr Lavoie and support several of his conclusions. Data such as ours will need to be measured by the gold standard of patients screened purely on unbiased cause, such as metal detection. Barring such a device, behavioral cues t h a t m e e t all f o u r t h amendment requirements must be the basis for a searching emergency d e p a r t m e n t patients. We strongly agree that searches should not be based on diagnosis or specialty service. Because data meeting these requirements are sparse, we hoped to contribute to the discussion with our data, which were the result of a biased search premise. They only indirectly allowed identification of the bias. We appreciate Dr Lavoie's scrutiny of our conclusion regarding the likelihood of psychiatric patients in our sample to be carrying weapons. We disagree with his assumptions that unsearched patients in both groups, medical and psychiatric, carried no weapons, or with his assumption that our patients were searched according to a shared "suspicion." Had our patients during the study time been searched according to strict assessment behavior, we believe that the percentages of those searched in both groups would not Annals of Emergency Medicine
have been so wildly disparate. It seems very unlikely that our psychiatric patients were more than 25 times more likely than medical patients to display such behavior. Indeed, discussions with our public safety officers supported the understanding that psychiatric patients were searched with a lower threshold because the personnel involved in this administratively distinct psychiatry service supported searching more than did the emergency medical providers. If the bias is indeed unclear, many interpretations would seem possible. The hypothesis Dr Lavoie suggests, that unsearched patients in both groups carried no weapons, is only one alternative. An additional hypothesis would expand on the above perception of our bias and state that unsearched patients only in the psychiatric group carried no weapons, while medical patients carried weapons in at least the percentage of their searched peers. This would yield yet other percentages. We believe that our abstraction of similar percentages of weapons possession for both groups, while not proven by the data, is the most plausible i n t e r p r e t a t i o n , barring additional, more controlled data. We are in the process of developing and implementing behavioral guidelines for the search of patients so emphasized by the above ambiguities and eloquently advocated in the letter. We look forward to additional research that will help us serve our patients better and more objectively while protecting ED staff.
Rupert Goetz, MD Joseph Bloom, MD Sherry Chenell, RN John Moorhead, MD Oregon Health Sciences University Portland
UC Irvine Graded Response System To the Editor: We read with interest Dr Siegel's article, "Code 9: A Systematic Approach for Responding to Medical 1047/169