Under-resuscitation of near-lethal uncontrolled hemorrhage: effects on mortality and end-organ function at 72 hours.

Laboratory studies of uncontrolled hemorrhage demonstrate that under resuscitation (UR) improves short-term survival, but at the expense of tissue perfusion. The long-term effects of UR have not been studied. The purpose of this study was to evaluate survival and the incidence of end-organ injury (EOI), 3 days following moderate and severe UR of uncontrolled hemorrhage. Thirty-four swine (14-24 kg) were assigned to 4 groups: Groups I, II, and III were hemorrhaged to a pulse pressure = 5 mmHg in the presence of a 4-mm aortic tear: Group I (control; n = 6) was not resuscitated; Group II (n = 11) was severely under resuscitated (MAP [mean arterial pressure] = 40 mmHg) for 75 min; Group III (n = 9) was moderately under resuscitated (MAP = 60 mmHg) for 75 min. After 75 min, the aortotomy was repaired, and animals were resuscitated to baseline physiologic parameters. Group IV (sham; n = 8) was instrumented, but not hemorrhaged. Seventy-two-hour mortality was 100%, 36%, 22%, and 0% for Groups I through IV (P = .001 Fisher's exact). Cardiac indices, serum bicarbonate, and systemic oxygen delivery were significantly lower in Group II as compared to Group III during the 75 min of UR (P < 0.05; repeated measures ANOVA). By 72 h, physiologic parameters in surviving animals had returned to baseline levels. Measures of kidney, liver, neurologic, and pulmonary function did not change from baseline. There was no histologic evidence of EOI. In this model, 75 min of UR did not result in EOI. There was a trend toward greater survival, and tissue perfusion was better preserved with moderate as compared to severe UR.

Comparison of the effects of bolus vs. slow infusion of 7.5% NaCl/6% dextran-70 in a model of near-lethal uncontrolled hemorrhage.

Bolus infusion of of 7.5% NaCl/6% dextran-70 (HSD-B) improves outcome from controlled hemorrhage. In contrast, HSD-B during uncontrolled hemorrhage increases bleeding and short-term mortality. The purpose of this study was to compare the effects of bolus vs. slow infusion of HSD in a near-fatal vascular injury hemorrhage model. Sixteen (15-20 kg) swine with 4-mm aortic tears were hemorrhaged to a pulse pressure of 5 mmHg. An ultrasonic flow probe was placed proximal to the aortic tear for continuous blood flow (AF) measurements. Group I (slow infusion; n = 8) was resuscitated with 8 mL/kg of HSD at 0.4 mL/kg/min. Group II (bolus infusion; n = 8) was resuscitated with 8 mL/kg of HSD at 1.33 mL/kg/min. In both groups, HSD infusion was followed by administration of 30 mL/kg of shed blood at 3 mL/kg/min. Hemorrhage volume and 90-min mortality were greater in group II (79+/-11 mL/kg; 75%) compared with group I (43+/-9 mL/kg; 12.5%) (P(Hem) < 0.001; P(Mort) = 0.04). Mean arterial pressure (MAP) and AF were greater in group II compared with group I during the first 15 min of resuscitation. In group I, MAP, AF, cardiac indices, and O2 delivery gradually returned to baseline levels and were significantly greater than group II at 30 min and throughout the remainder of the protocol. In this model of near-lethal uncontrolled hemorrhage, slow infusion of HSD restored cardiodynamics while minimizing hemorrhage volume and mortality. Resuscitation regimens that cause early increases in blood flow and pressure may result in greater hemorrhage and mortality than those regimens that yield comparable flow and pressure increases late in resuscitation.

Resident perception of academic skills training and impact on academic career choice.

OBJECTIVES: 1) To evaluate residents' perceptions of the quality of training in basic academic skills and the availability and quality of research resources during residency; 2) to evaluate the association between these attitudes and choice of an academic career; and 3) to assess residents' attitudes toward the importance of postgraduate fellowship training for success in an academic career. METHODS: A 15-item survey was administered to all U.S. emergency medicine (EM) residents in conjunction with the February 1997 American Board of Emergency Medicine (ABEM) In-service Examination. The survey assessed resident interest in a career in academic EM, and resident perception of the general quality of training in academic (research and teaching) skills. Residents were also asked to rate the quality of their training in the following specific academic skills: medical and grant writing, bedside teaching, lecturing, the use of computers, study design, statistics, and the use of audiovisual aids. Resident perceptions of the availability of the following resources were also assessed: teaching and research role models, data collection and analysis support, laboratory facilities, financial support of research, research fundamentals lectures, and computers. RESULTS: The response rate was 93%. Forty-four percent of the respondents were interested in academic EM, 36.6% were undecided, and 19.6% were not interested in an academic career. On a scale of 1 (unprepared) to 5 (well prepared), the residents rated their overall preparedness for an academic career fairly high (3.97 [0.86]). In contrast, they perceived the quality of their training in the specific academic skill areas assessed and research resource availability to be only fair. Despite resident perception of relatively inadequate training in basic academic skills, only 24% of the respondents indicated that they believed fellowship training was important for success in an academic career. Logistic regression analyses demonstrated that participation in a research project in medical school, the length of the training program (4- vs 3-year), being a first-year resident, and a better perception of one's overall academic skill preparation were factors independently associated with having a greater interest in an academic career. CONCLUSIONS: A relatively high percentage of residents initially express an interest in an academic career, but this interest wanes as residency progresses. A minority of residents believe that their training provides them with the specific skills needed to succeed in academics, or with adequate exposure to research resources or mentors. Emergency medicine may be able to increase the number of qualified academic faculty by recruiting medical students with prior research experience, and providing residents with better research training and role models.

Resuscitation of severe uncontrolled hemorrhage: 7.5% sodium chloride/6% dextran 70 vs 0.9% sodium chloride.

OBJECTIVES: Resuscitation studies of hypertonic saline using controlled and uncontrolled hemorrhage models yield conflicting results with regard to efficacy. These disparate results reflect the use of models and resuscitation regimens that are not comparable between studies. This study evaluated the effects of comparable and clinically relevant resuscitation regimens of 7.5% sodium chloride/6% dextran 70 (HSD) and 0.9% sodium chloride (NS) in a near-fatal uncontrolled hemorrhage model. METHODS: Thirty-six swine (14.2 to 21.4 kg) with 4-mm aortic tears were bled to a pulse pressure of 5 mm Hg (40-45 mL/kg). The animals were resuscitated with either NS or HSD administered in volumes that provided equivalent sodium loads at similar rates. Group II (n = 12) was resuscitated with 80 mL/kg of NS at a rate of 4 mL/kg/min. Group III (n = 12) received 9.6 mL/kg of HSD at a rate of 0.48 mL/kg/min. In both groups, crystalloid resuscitation was followed by shed blood infusion (30 mL/kg) at a rate of 2 mL/kg/min. Group I (controls; n = 12) were not resuscitated. RESULTS: One-hour mortality was significantly greater in group I (92%) as compared with group II (33%) and group III (33%) (Fisher's exact test; p = 0.004). Intraperitoneal hemorrhage was significantly greater in group II (34 +/- 20 mL/kg) and group III (31 +/- 13 mL/ kg) as compared with group I (5 +/- 2 mL/kg) (ANOVA; p < 0.05). There was no significant difference in hemodynamic parameters between groups II and III. CONCLUSION: In this model of severe uncontrolled hemorrhage, resuscitation with HSD or NS, administered in volumes that provided equivalent sodium loads at similar rates, had similar effects on mortality, hemodynamic parameters, and hemorrhage from the injury site.

Effect of initially limited resuscitation in a combined model of fluid-percussion brain injury and severe uncontrolled hemorrhagic shock.

OBJECT: Studies of isolated uncontrolled hemorrhage have indicated that initial limited resuscitation improves survival. Limited resuscitation has not been studied in combined traumatic brain injury and uncontrolled hemorrhage. In this study the authors evaluated the effects of limited resuscitation on outcome in combined fluid-percussion injury (FPI) and uncontrolled hemorrhage. METHODS: Twenty-four swine weighing 17 to 24 kg each underwent FPI (3 atm) and hemorrhage to a mean arterial pressure (MAP) of 30 mm Hg in the presence of a 4-mm aortic tear. Group I (nine animals) was initially resuscitated to a goal MAP of 60 mm Hg; Group II (nine animals) was resuscitated to a goal MAP of 80 mm Hg; and Group III (control; six animals) was not resuscitated. After 60 minutes, the aortic hemorrhage was controlled and the animals were resuscitated to baseline physiological parameters and observed for 150 minutes. Mortality rates were 11%, 50%, and 100% for Groups I, II, and III, respectively (Fisher's exact test; p = 0.002). The total hemorrhage volume was greater in Group II (69+/-32 ml/kg), as compared with Group I (41+/-18 ml/kg) and Group III (37+/-3 ml/kg) according to analysis of variance (p < 0.05). In surviving animals, cerebral perfusion pressure, cerebral blood flow (CBF), cerebral venous O2 saturation (ScvO2), and cerebral metabolic rate of O2 did not differ among groups. Although CBF was approximately 50% of baseline during the period of limited resuscitation in Group I, ScvO2 remained greater than 60%, and arteriovenous O2 differences remained within normal limits. CONCLUSIONS: In this model of FPI and uncontrolled hemorrhage, early aggressive resuscitation, which is currently recommended, resulted in increased hemorrhage and failure to optimize cerebrovascular parameters. In addition, a 60-minute period of moderate hypotension (MAP = 60 mm Hg) was well tolerated and did not compromise cerebrovascular hemodynamics, as evidenced by physiological parameters that remained within the limits of cerebral autoregulation.

The impact of environmental factors on emergency medicine resident career choice.

OBJECTIVE: To evaluate the impact of environmental factors on emergency medicine (EM) resident career choice. METHODS: Program directors of all U.S. EM residencies were surveyed in November 1997. A 22-item questionnaire assessed resources allocated to research, fellowship availability, academic productivity of faculty and residents, and career choices of residency graduates. RESULTS: The response rate was 83%. The program director (mean+/-SD) estimates of resident career choice were as follows: 27.8+/-19.1% pursued academic positions with emphasis on teaching, 5.4+/-9.8% pursued academic positions with emphasis on research, and 66.8+/-23.1%, pursued private practice positions. In addition, 5.70+/-6.13% of the residency graduates were estimated to seek fellowship training. Univariate analyses demonstrated that increasing departmental funding for research, having substantial resource availability (defined as having at least two of the following: dedicated laboratory space; support for a laboratory research technician/assistant, a clinical research nurse or study coordinator, a statistician, or an assistant with a PhD degree), a greater number of peer-reviewed publications by residents (r = 0.22; p = 0.08), and a greater number of peer-reviewed publications by faculty (r = 0.26; p = 0.04) positively correlated with the percentage of graduates who pursue academic research careers. Using multiple regression, however, increasing intramural funding and the presence of substantial resource availability were the only variables predictive of resident pursuit of an academic research career. CONCLUSION: Modification of the EM training environment may influence the career choices of graduates. Specifically, greater commitment of departmental funds and support of resources for research may enhance the likelihood of a trainee's choosing an academic research career.

Funding strategies for emergency medicine research.

The importance of adequate funding for sustaining research efforts cannot be overemphasized. This article addresses funding strategies for emergency physicians, including the necessity of establishing a research track record, developing a well-written grant proposal, and anticipating the grant review process. Funding sources are reviewed with an emphasis on federal institute support and private foundations (including the Emergency Medicine Foundation) in the United States. Sources of current grant support information available from the Internet are provided. Recommendations for enhancing research funding in emergency medicine (EM) are made, including enhancement of formal research training, promotion of EM research and investigators, federal study section membership, and collaboration with established investigators.

Funding strategies for emergency medicine research.

The importance of adequate funding for sustaining research efforts cannot be overemphasized. This article addresses funding strategies for emergency physicians including the necessity of establishing a research track record, developing a well-written grant proposal, and anticipating the grant review process. Funding sources are reviewed with an emphasis on federal institute support and private foundations (including the Emergency Medicine Foundation) in the United States. Sources of current grant support information available from the Internet are provided. Recommendations for enhancing research funding in emergency medicine are made, including enhancement of formal research training, promotion of emergency medicine research and investigators, federal study section membership, and collaboration with established investigators.

Effect of supplemental perfluorocarbon administration on hypotensive resuscitation of severe uncontrolled hemorrhage.

Recent animal studies of acute hemorrhage in the presence of a vascular injury have demonstrated improved survival and decreased hemorrhage volume with hypotensive resuscitation, but this has occurred at the expense of tissue perfusion. It was hypothesized that addition of an oxygen-carrying perfusate would improve tissue oxygen delivery during hypotensive resuscitation. Hypotensive resuscitation of severe uncontrolled hemorrhage was compared with and without supplementation with Oxygent HT, an emulsion of perflubron (perfluorooctylbromide; PFOB; Alliance Pharmaceutical Corporation, San Diego, CA), an oxygen-carrying perfusate. Fifteen swine (15 to 22 kg) with 4-mm aortic tears were bled to a pulse pressure of 5 mm Hg and then resuscitated (estimated blood loss, 40 to 50 mL/kg). All animals were resuscitated with normal saline (6 mL/kg/min) infused as needed to maintain a mean arterial pressure of 40 mm Hg. One group (PFC) of animals also received an infusion of 6 mL/kg perfluorooctylbromide emulsion. Another group served as controls and received an equal volume of placebo (normal saline). Animals were observed for 120 minutes or until death. Data were compared using repeated measures analysis of variance (ANOVA) the Student's t test, and Fisher's exact. A P value < .05 was considered significant. Two-hour mortality rates were 12.5% and 43% for PFC-treated animals and controls, respectively (P > .05; 95% confidence interval [95% CI] for this difference in mortality is -13% to 74%). Oxygen content and delivery were significantly greater in the treatment group. In conclusion, administration of an oxygen-carrying perfusate significantly improves oxygen delivery in hypotensive crystalloid resuscitation of severe uncontrolled hemorrhage.

Multiple resuscitation regimens in a near-fatal porcine aortic injury hemorrhage model.

OBJECTIVE: To compare early and delayed blood administrations in animals subjected to near-fatal hemorrhage in the presence of a vascular injury and resuscitated to different mean arterial pressures (MAPs). METHODS: Fifty-four immature swine with 4-mm infrarenal aortic tears were bled to a pulse pressure of 5 torr and then resuscitated (estimated blood loss 40 to 45 mL/kg). Groups I, II, and III were resuscitated with shed blood at a rate of 2 mL/kg/min, followed by normal saline at a rate of 6 mL/kg/min. Groups IV, V, and VI received the same fluids in reverse order. The fluids were infused intermittently to maintain MAPs of 40, 60, and 80 torr. The animals were observed for 60 minutes or until death. RESULTS: The animals resuscitated to a MAP of 80 torr experienced significantly higher intraperitoneal hemorrhage volumes and mortality than did the animals intentionally maintained hypotensive, regardless of whether blood or normal saline was administered first. There was no significant difference in mortality or hemorrhage volumes between any of the groups intentionally maintained hypotensive. The animals maintained at a MAP of 60 torr were significantly less acidotic than were the animals resuscitated with the same fluid regimen but to a MAP of 40 torr. Early blood administration also minimized the acidosis associated with hypotensive resuscitation. CONCLUSION: In this model of near-fatal hemorrhage with a vascular injury, maintenance of the hypotensive state produced comparable improvements in one-hour survival and reductions in hemorrhage volume regardless of whether blood or saline was administered first. Although hypotensive resuscitation resulted in improved outcome, it was associated with significant acidosis. This effect was minimized with moderate rather than severe underresuscitation and early blood administration.

A comparison of the response of near-fatal acute hemorrhage models with and without a vascular injury to rapid volume expansion.

Recent studies in which animals were bled from a vascular injury rather than an intravascular catheter demonstrate increased blood loss and mortality with rapid volume expansion. The purpose of this study was to better define the importance of incorporating a vascular injury in animal models of acute hemorrhage. We directly compared the response to resuscitation from hemorrhage of comparable severity in animals with and without a vascular injury. Thirty-four immature swine (14.6 to 23.2 kg) were instrumented and subjected to severe blood loss (40 to 46 mL/kg). Groups I and II were hemorrhaged from a femoral artery catheter only. Groups III and IV were initially bled in the same manner; however, when the mean arterial pressure (MAP) decreased to 30 mm Hg, a 4-mm tear was created in the infrarenal aorta, allowing free intraperitoneal hemorrhage. In all groups, the catheter hemorrhage was discontinued once the pulse pressure reached 5 mm Hg. Groups II and IV were resuscitated with normal saline (NS) infused at a rate of 6 mL/kg/min followed by shed blood at a rate of 2 mL/kg/min. The resuscitation fluids were infused as needed to maintain a MAP of 80 mm Hg. Groups I and III served as controls and were not resuscitated. All animals were observed for 60 minutes or until death. The data were compared using repeated measures analysis of variance with a post hoc Tukey Kramer and the Fisher's exact test. Mortality was 100%, 0%, 88%, and 78% for groups I, II, III, and IV, respectively (P < .05 for group II vs groups I, III, and IV).(ABSTRACT TRUNCATED AT 250 WORDS)

Effect of blood pressure on hemorrhage volume and survival in a near-fatal hemorrhage model incorporating a vascular injury.

STUDY HYPOTHESIS: In a model of near-fatal hemorrhage that incorporates a vascular injury, stepwise increases in blood pressure associated with aggressive crystalloid resuscitation will result in increased hemorrhage volume and mortality. DESIGN: This study used a swine model of potentially lethal hemorrhage in the presence of a vascular lesion to compare the effects of resuscitation with mean arterial pressures of 40, 60, and 80 mm Hg. Twenty-seven fully instrumented immature swine (14.8 to 20 kg), each with a surgical-steel aortotomy wire in place, were bled continuously from a femoral artery catheter to a mean arterial pressure of 30 mm Hg. At that point the aortotomy wire was pulled, producing a 4-mm aortic tear and uncontrolled intraperitoneal hemorrhage. When the animal's pulse pressure reached 5 mm Hg, the femoral artery hemorrhage was discontinued and resuscitation was begun. INTERVENTIONS: Saline infusion was begun at 6 mL/kg/min and continued as needed to maintain the following desired endpoints: group 1 (nine) to a mean arterial pressure of 40 mm Hg, group 2 (nine) to a mean arterial pressure of 60 mm Hg, and group 3 (nine) to a mean arterial pressure of 80 mm Hg. After 30 minutes or a total saline infusion of 90 mL/kg, the resuscitation fluid was changed to shed blood infused at 2 mL/kg/min as needed to maintain the desired mean arterial pressure or to a maximum volume of 24 mL/kg. Animals were observed for 60 minutes or until death. MEASUREMENTS AND MAIN RESULTS: Data were compared using repeated-measures analysis of variance with a post hoc Tukey-Kramer, Fisher's exact test, and Kruskal-Wallis. Mortality was significantly greater in group 3 (78%) compared with either group 1 (11%; P = .008) or group 2 (22%; P = .028). Mean survival times were significantly shorter in group 3 (44 +/- 12 minutes) compared with either group 1 (58 +/- 6 minutes; P = .007) or group 2 (59 +/- 3 minutes; P = .006). The average intraperitoneal hemorrhage volumes were 13 +/- 14 mL/kg, 20 +/- 25 mL/kg, and 46 +/- 11 mL/kg for groups 1, 2, and 3, respectively (group 1 versus 2, P = .425; group 1 versus 3, P < .001; group 2 versus 3, P = .014). Group 2 animals demonstrated significantly greater oxygen deliveries compared with groups 1 and 3. CONCLUSION: In a model of near-fatal hemorrhage with a vascular injury, attempts to restore blood pressure with crystalloid result in increased hemorrhage volume and markedly higher mortality.

Comparison of midazolam and diazepam for conscious sedation in the emergency department.

STUDY OBJECTIVE: To compare the efficacy of diazepam and midazolam when used for conscious sedation in emergency department patients. DESIGN: Prospective, randomized, double-blind, multicenter trial. SETTING: Three university EDs. TYPE OF PARTICIPANTS: Patients requiring one of the following procedures: abscess drainage, joint reduction, extensive suturing, chest tube insertion, or lumbar puncture. INTERVENTIONS: Diazepam (2.5 mg/mL) or midazolam (1 mg/mL) was administered until the desired level of sedation was achieved to a maximum of 5 mL. Fentanyl citrate was administered if needed for pain. MEASUREMENTS AND MAIN RESULTS: Thirty-three patients received diazepam and 36 received midazolam. Patients receiving midazolam had a greater degree of early sedation (P < .05), a higher 90-minute alertness scale score (P < .05), more patients ready for discharge at 90 minutes (P = .05), significantly less recall for the procedure (P < .02), and less pain on injection (P < .01) than patients who were given diazepam. CONCLUSIONS: Diazepam and midazolam are both effective for conscious sedation in ED patients. Midazolam causes less pain on injection, a significantly greater degree of early sedation, and a more rapid return to baseline function.

Improved outcome with early blood administration in a near-fatal model of porcine hemorrhagic shock.

Current recommendations for the preoperative management of hemorrhagic shock include the initial infusion of 2 L of isotonic crystalloid regardless of the severity of hemorrhage. While this approach may be adequate for patients who experience only mild to moderate hemorrhagic insults, it has never been tested in a clinically relevant model of severe life-threatening hemorrhage. The authors used a porcine model of rapidly fatal hemorrhage with a reproducible and relevant physiologic end-point, the absence of vital signs, to test the hypothesis that even brief delays in blood replacement may result in higher mortality rates and worsen hemodynamic and metabolic responses to hemorrhage. Twenty-four immature swine (11-17 kg) were bled continuously at a decelerating rate until the following criteria were met: (1) respiratory arrest, (2) a pulse pressure of 0 and, (3) a slowing of cardiac electrical activity of 15% or more. Resuscitation was begun 1 minute later. The animals were randomly assigned to one of three resuscitation regimens. Group A (n = 8) received shed blood at a rate of 3 mL/kg/min for 10 minutes followed by normal saline (NS) at a rate of 3 mL/kg/min for 10 minutes. Group B (n = 8) received NS at a rate of 3 mL/kg/min for 10 minutes followed by shed blood at a rate of 3 mL/kg/min for 10 minutes. Group C, controls, (n = 8) received NS at a rate of 3 mL/kg/min for 20 minutes. Animals were observed for 30 minutes after resuscitation or until death. Mortality was 25%, 37.5%, and 100% for groups A, B, and C, respectively (P < .05 for group C versus group A or B).(ABSTRACT TRUNCATED AT 250 WORDS)

Prospective evaluation of the scapular manipulation technique in reducing anterior shoulder dislocations.

STUDY OBJECTIVE: To evaluate the speed, efficacy, and safety of the scapular manipulation technique in reducing acute anterior shoulder dislocations. DESIGN: Prospective study. SETTING: Urban emergency department with an annual census of 65,000 patients. PARTICIPANTS: Forty-eight adult patients with acute anterior shoulder dislocation. INTERVENTIONS: Patients had an initial neurovascular and radiographic evaluation performed. They were sedated with IV fentanyl and midazolam. The shoulder was reduced using the scapular manipulation technique. The patient was re-evaluated for any evidence of complication. The total dose of analgesic required and time to reduction were recorded. RESULTS: The scapular manipulation technique was successful in 46 of 48 (96%) cases. The average time to reduction was 6.05 minutes, and no complications were detected. Average doses of 1.83 mg midazolam and 204 micrograms fentanyl were required for reduction. CONCLUSION: The scapular manipulation technique is a very fast, effective, safe method of reducing anterior shoulder dislocations in the ED. [Kothari RU, Dronen SC: Prospective evaluation of the scapular manipulation technique in reducing anterior shoulder dislocations.

Acute myocardial infarction presenting as flank pain and tenderness: report of a case.

Studies have demonstrated that 4% to 10% of patients with chest pain and acute myocardial infarction (AMI) are discharged from the emergency department. The patient with an atypical presentation of AMI is difficult to diagnose and has been demonstrated to have an associated increased risk of morbidity and mortality. A case is reported of a patient with AMI presenting to the emergency department with flank pain and tenderness, mimicking acute renal colic.

The scapular manipulation technique for the reduction of acute anterior shoulder dislocations.

Anterior shoulder dislocations are a common occurrence in busy emergency departments. Numerous techniques for treating this problem have been reported. The majority of these techniques use traction and leverage of the humerus, often requiring considerable force and causing significant patient discomfort. We report a simple, relatively painless, and atraumatic method of shoulder reduction involving manipulation of the scapula as well as the humerus. This technique has been used in our emergency department with considerable success and no complications.

Midazolam use in the emergency department.

Midazolam is the first water-soluble benzodiazepine. As with other benzodiazepines it has amnestic, sedative, hypnotic, anxiolytic, and anticonvulsant properties. Midazolam is about two to four times more potent than diazepam. Midazolam has been extensively used for a variety of outpatient procedures, but there has been no documentation of its safety in emergency department patients. The authors retrospectively reviewed all patients receiving midazolam during a 2-year period at the University of Cincinnati Center for Emergency Care. The study population consisted of 389 patients (men 56%; women 44%) with an average age of 33.3 years. Midazolam was used intravenously for sedation before a wide variety of painful procedures and for agitation control. The average dose was 3.86 mg, with a range of 0.5 mg to 20.0 mg. The majority of patients (79.2%) received narcotics or sedative/hypnotic agents in addition to midazolam. There was an overall complication rate of 1.0%. Two patients (0.5%) developed clinically significant respiratory depression after midazolam use. Both patients had also received fentanyl citrate and the respiratory depression was reversed with naloxone. Two patients (0.5%) receiving several other drugs developed short periods of hypotension. There were no apparent long term sequelae. The authors conclude that midazolam can be safely used in the emergency department setting. Careful dosing and titration to the desired clinical effects is mandatory. Patients should be closely monitored to maximize safety.