patients with a suspected cerebrovascular accident (CVA). The author concludes that routine blood glucose measure- ment is not always indicated in these patients; we disagree with this conclusion and believe that there are several issues that warrant further discussion.

First, a number of flaws in the study may have led to a deceptively low number of recorded hypoglycemic epi- sodes. The broad exclusion criteria leave out patients whose potential diagnosis includes hypoglycemia and a CVA (eg, unconscious/unresponsive persons, generalized weakness/ dizziness, seizure). More importantly, the author defines hypoglycemia as a blood glucose concentration of 60 mg/dL or less. It is known that patients with poorly controlled diabetes may manifest symptoms of hypoglycemia at blood glucose concentrations of 78 F 5 mg/dL [2]. Finally, because the development of focal neurologic deficits in the setting of hypoglycemia is rare, the study population is inappropriately small.

Based on this flawed data set, the author recommends that blood glucose testing be reserved for patients with a history suggestive of pathology involving glucose homeo- stasis. This criterion excludes patients who might have ingested someone else’s diabetic medication or are hypo- glycemic for other reasons (eg, ethanol intoxication, hypothermia). In addition, it is unlikely that Prehospital personnel are able to obtain such a Complete medical history in the short time that they are with patients in the field. In fact, the study makes use of paramedic-run reports in fulfilling its inclusion criteria. However, these reports are often completed in the hospital after a definitive diagnosis has been made by emergency physicians.

The author’s rationale for embarking on this study is unclear. He seems to suggest that eliminating blood glucose measurements can help minimize time in the field. In our clinical experience, the time that it would take to obtain a complete medical history far exceeds that of per- forming a finger stick for blood glucose measurement using a glucometer.

We believe that causes of hypoglycemia are varied and not always easily determined in the field. Although hypo- glycemia can have devastating results, it is easily treated with intravenous dextrose. Whether or not a blood glucose concentration is obtained in the field, this diagnosis should be considered in all patients with neurologic findings and treated promptly. We do not believe that obtaining a rapid blood glucose concentration will delay care, increase morbidity, or substantially alter the cost of care of patients with a suspected CVA.

Beth Y. Ginsburg MD Robert S. Hoffman MD Department of Emergency Medicine

New York City Poison Control Center New York University School of Medicine

New York, NY 10016, USA

Laura Spano MD Department of Emergency Medicine The Brooklyn Hospital Center Brooklyn, NY 11201, USA

DOI of original article: 10.1016/j.ajem.2006.01.004 doi:10.1016/j.ajem.2006.02.002

References

1. Abarbanell NR. Is prehospital blood glucose measurement necessary in suspected cerebrovascular accident patients? Am J Emerg Med 2005; 23:823 – 7.
2. Boyle PJ, Schwartz NS, Shah SD, et al. Plasma glucose concentra- tions at the outset of hypoglycemic symptoms in patients with poorly controlled diabetes and in nondiabetics. N Engl J Med 1988; 318:1487 – 92.

Reply

To the Editor,

Although Dr Ginsberg ably presents her personal beliefs with regard to medical practice, she appears to have missed the basic premise of this research.

This study was completed to establish an epidemiologic database defining prehospital management of suspected Cerebrovascular accidents (CVAs) with attention to blood glucose measurement in hope of developing recommenda- tions for further Treatment protocols. This study consisted of 9495 paramedic run reports, which yielded 185 persons presenting with signs/symptoms suggestive of CVAs. Defin- itive parameters were given for patient inclusion [1-4] in this study, based on the medical literature, and it was carefully stated throughout the text that any recommendations made were relevant only given similar patient populations and emergency medical services system characteristics. Study limitations were recognized and discussed. Paramedic docu- mentation of the word CVA/stroke on ambulance run reports, or patient complaint/examination suggestive of CVA, that is, hemiparesis, hemiplegia, select cases of speech/visual dis- turbances, confusion and/or incoordination, met inclusion [1-4] criteria for this study, therefore decreasing the potential for selection bias. To minimize discrepancies in [1,2] differential diagnosis, thus reducing the chance of false positives, unconscious/unresponsive persons, as well as those presenting with exclusive complaints of syncope, seizures, generalized weakness/dizziness, obvious intoxication, known hypoglycemia/diabetic emergencies, or trauma-associated neurologic deficits were excluded from the study population. Paramedic charting was thorough and complete. Para- medics documented medical histories in 100% of cases. Hypoglycemia was defined as blood glucose measurement of

60 mg/dL or less, given paramedic standard operating procedures for the Saint Francis Emergency Medical Services System, Saint Francis Hospital, Evanston, Ill, and existing

adult advanced cardiac life support, pediatric advanced life support, and advanced trauma life support [5-9] guidelines. All paramedic standard operating procedures are subject to review by the State of Illinois, Department of Public Health. The readers questioning of these issues is unfounded. One hundred eighty-five persons (1.95%) presented with signs and/or symptoms suggestive of CVAs. All patients were hemodynamically stable. One hundred fifty-five persons (83.78%) had no risk factors for hypoglycemia. Thirty persons (16.22%) presented with concomitant pathology suggestive of abnormalities in glucose homeostasis. Five persons (2.70%), all of whom had a concurrent history of medication-controlled diabetes, were found to be hypogly- cemic. No cases noted involved hypoglycemia in the absence of a suggestive setting. Therefore, the likelihood that a given patient with apparent CVA lacking known abnormalities in glucose homeostasis will in fact be hypoglycemic is 0.00% (95% confidence interval, 0.00%-2.35%). Should risk factors for hypoglycemia be present, the likelihood of having a low blood glucose level is 16.67% (95% confidence interval, 5.65%-34.72%).

Based on this data, and the pathophysiology of hypogly- cemia, routine blood glucose measurement of patients with suspected CVA must be subject to question, as problems involving glucose homeostasis presenting with focal neuro- logic deficits occur only rarely and then can be suspected based on patient medical history. Provided a similar patient population and prehospital environment, a selective approach to blood glucose measurement given the presence of risk factors for hypoglycemia or rescuer inability to obtain adequate patient historical information must be considered.

Routine prehospital blood glucose measurement in persons with suspected CVA may be an unnecessary practice, lacking clinical justification. Only by engaging in clinical research can physicians improve the scientific basis of our specialty. We all need to challenge our personal experience when it conflicts with data and scientific observation.

Neal Robert Abarbanell MD

Division of Emergency Medicine, Department of Medicine

University of Miami School of Medicine

Jackson Memorial Hospital Miami, FL 33101, USA

DOI of original article:10.1016/j.ajem.2006.02.002 doi:10.1016/j.ajem.2006.02.003

References

1. Scott PA, Barsan WG. Stroke, transient ischemic attack, and other central focal conditions. In: Tintinalli JE, Kelen GD, Stapczynski JS

practice of emergency medicine. Philadelphia (Pa)7 Lippincott; 1991.

p. 1084 – 9.

1. Katzman IL, Furlan AJ, Lloyd LE. Use of tissue-type plasminogen activator for acute ischemic stroke: the Cleveland area experience. JAMA 2000;283:1151 – 8.
2. Adams HP, Brott TG, Furlan AJ, et al. Guidelines for thrombolytic therapy for acute stroke: a supplement to the guidelines for the management of patients with acute ischemic stroke. Circulation 1996; 94:1167 – 74.
3. Advanced cardiac life support provider manual. Dallas (Tex)7 American

Heart Association; 2001. p. 1 – 225.

1. Emergency Cardiac Care Committees and Subcommittees, American Heart Association. Guidelines for cardiopulmonary resuscitation and emergency cardiac care. Parts I-IX. JAMA 1992;268:2171 – 298.
2. Advanced cardiac life support. Dallas (Tex)7 American Heart Associ- ation; 1997. p. 1 – 377.
3. American Heart Association, American Academy of Pediatrics. Pediatric advanced life support. Dallas (Tex)7 American Heart Association; 1997. p. 1 – 148.
4. American College of Surgeons Committee on Trauma. Advanced trauma life support for doctors. Chicago (Ill)7 American College of Surgeons; 1997. p. 1 – 444.

Trends in the use of ipecac in the ED setting: data from the 1992-2002 NHAMCS-ED (letter)^B

To the Editor,

As recently as the late 1980s, syrup of ipecac (ipecac) was frequently administered to both adults and children to induce gastric decontamination after ingestion of real or potential toxins within the setting of the ED [1,2]. Beginning in the early 1990s, concern was raised over the safety and effectiveness for use of ipecac in the ED, although its use in the home per recommendation of a physician or poison control center was still valued [3]. Further studies and practice guidelines now suggest the frequency of adminis- tration of ipecec within the ED to be bextremely limitedQ [4] or brareQ [5].

To assess the impact of recent studies and practice guidelines for use of ipecac on clinical practice, we examined the National Hospital Ambulatory Medical Care Survey , specifically that section of NHAMCS which includes services provided in hospital EDs (NHAMCS-ED), for trends in the administration of ipecac in the ED. As background, the NHAMCS-ED begins within selected hospitals where medical record abstractors com- plete patient record instruments for a systematic random sample of patient visits. Information collected includes up to 3 reasons for which the patient presented for care in the ED and up to 5 medications injected, administered, supplied, or prescribed. For the purposes of this study, any visit containing a code for ipecac was considered as a visit for which ipecac was administered to the patient. Of interest, NHAMCS-ED does not capture any use of ipecac (or other

editors. Emergency medicine: a comprehensive study guide. 5th ed.

New York (NY)7 McGraw-Hill; 2000. p. 1430 – 40.

1. Ordog GJ, Wasserberger J, Zehr TW. Cerebrovascular disease. In: Harwood-Nuss A, Linden CH, Luten RC, et al, editors. The clinical

^B Disclaimer: The views expressed are those of the authors and do not necessarily represent those of the Food and Drug Administration or imply its endorsement.