- Email: [email protected]
Potential Adverse Events in Children Treated in the Emergency Department
Research Forum Abstracts
191
What a Bite: Review of Snakebites in Children
Feng S, Stephan M, UT Southwestern Medical Center/Children’s Medical Center of Dallas, Dallas, TX
Study Objectives: Approximately 2000 children less than 17 years of age are victims of crotalid envenomation each year. Multiple controversies exist in the management of these bites. We present the largest pediatric review of snakebite envenomations. Methods: This is a retrospective chart review of pediatric patients presenting with venomous snakebites. Data was abstracted for; demographics, the severity of the bite, circumstances of the snakebite, lab values, method and dose of antivenom administered, VS, length of hospital stay, adverse outcomes, cost and disposition. Results: 49 patients were identified. Ages ranged 22 mos -16.8 years, 67.3% male. The envenomations consisted of: 65% copperheads, 8% cottonmouth, 27% unknown. 47% were moderate level 2 envenomations, 36.7% were mild level 1. Systemic symptoms were found in 26.8% with predominance of vomiting. Initial coagulation evaluation revealed 26.5% abnormal CBCs: 7 pts had WBCO15K, 4 had thrombocytopenia. 30.6% had abnormal PT, 10.1% abnormal PTT. 1 patient had abnormal fibrinogen and d-dimer results (cottonmouth envenomation). 31.6% had other labs performed predominantly renal group(35%). Subsequent labs were significant for persistently abnormal percentage PTT and d-dimer (16.3%, 4.1%) Antivenin was used in 34.6%: 82.3% received CroFab (CF) with 1 adverse reaction described as facial flushing. Vital signs remained stable during all infusions. 18.4% received subsequent CF dosing. No reactions were noted on subsequent dosing. 22% had progression of symptoms described as edema during hospitalization. 49% had a tox consult. Antivenin use in all snakebites with tox consult (62.5%) is statistically significant (p\0.004) using Pearson Chi-Square. 50% of copperhead bites with tox consult received no antivenin. 43% had 2 day hospital stay. 59% received antibiotics. 2 patients were readmitted for surgical debridement of hemorrhagic blisters on digits. Conclusion: Copperhead envenomations are prevalent in Northeast Texas as reflected in our data. The use of Wyeth Polyvalent Antivenin and CF appear to be safe and effective. Antivenom use still remains controversial in copperhead envenomations. We present the largest descriptive analysis of snakebite envenomations in the pediatric population.
192
Is the Combination of Ibuprofen and Acetaminophen a Better Pediatric Antipyretic Than Ibuprofen Alone?
Tenison M, Eberhardt M, Pellett N, Heller M, St Lukes Hospital, Bethlehem, PA
Study Objectives: Acetaminophen and ibuprofen are often combined for the treatment of pediatric fever despite the lack of data to support or refute this practice. The purpose of this prospective, controlled, randomized, double-blind clinical investigation is to determine whether the use of these two agents in combination is a more effective antipyretic than ibuprofen alone. Methods: Febrile (O101 C) children age 3 to 10 were enrolled during a visit to a community hospital ED or its affiliated pediatric clinic. Subjects were randomized to treatment with one of two regimens: a combination of oral ibuprofen (10mg/kg) and acetaminophen (15mg/kg) (study) or ibuprofen (10mg/kg) and identically-appearing placebo (control). Initial oral temperatures were recorded during the visit; 2, 4 and 6-hour oral temperatures were later reported via phone or mail-in form by the patient’s caregiver. Results: In this predefined interim analysis 28 febrile children were enrolled with at least 2 and 4-hour temperatures available for 20 subjects. Initial temperatures ranged from 101.2 to 104.2 (mean 102.5) and were similar between groups (p=0.42). Mean temperature change between 0 and 2 hours was also similar (-3.1 degC for the both groups; p=0.66). Mean temperature decrease between 0 and 4 hours however, was significantly greater in the study group (-4.4 deg C for the study group and -3.7 deg C for controls; p=0.05). At 6 hours this difference was even more pronounced (-4.8 deg C for the study group versus -1.3 deg C for controls; p=0.02). Conclusion: Although the combination of ibuprofen and acetaminophen was no more effective than ibuprofen alone at 2 hours, it may offer a more prolonged antipyretic effect.
S54 Annals of Emergency Medicine
193
Attitudes and Practices of Emergency Medicine Residents Regarding Family Presence During Pediatric Procedural and Resuscitation Scenarios
Barata IA, LaMantia J, Riccardi D, Mayerhoff RM, Litroff A, D’Abbracci P, Livote E, Ward MF, Sama A, North Shore University Hospital, Manhasset, NY
Objective: Determine attitudes and practices of emergency medicine (EM) residents regarding family presence (FP) during hypothetical pediatric procedural and resuscitation scenarios. Methods: This was a prospective, anonymous survey of residents in ACGME accredited residency programs. Surveys were mailed in September 2004 for completion by residents. Descriptive statistics were employed. Results: Responses were received from 61 programs with the following geographic distribution: 38% Northeast, 37% Midwest, 18% South and 5% West Coast. A total of 521 surveys were completed by EM residents: 193 EM-1, 170 EM2, 133 EM-3, 16 EM-4, 8 other or missing; 129 residents completed an internship year other than EM: 52 IM, 18 General Surgery, 13 Family Practice, 5 Pediatrics, 41 other; 169 (32%) residents see pediatric patients primarily at a children’s hospital. EM residents reported that FP during selected procedures would interfere with their ability to perform the interventions: 25% conscious sedation, 45% spinal tap, 48% intubation, 50% resuscitation. In addition, only a small percentage of residents would always allow FP during certain procedures: 31% conscious sedation, 19% spinal tap, 11% resuscitation, 10% intubation. EM-1 residents that have completed a prior internship year or residency were as likely as EM-1 residents without prior experience to report that FP would interfere with their performance, especially in the case scenarios of intubation (59% vs.49%; p value 0.2505) and resuscitation (51% vs. 40%; p value 0.1827). When comparing the practices of emergency medicine residents regarding FP during pediatric resuscitation at children’s hospitals ED vs. a general ED it seems that a greater number of families would be allowed to be present during resuscitation at children’s hospitals: 17% vs. 9% (p value 0.0108) (7 yr old major resuscitation), 23% vs.15% (p value 0.0375) (2 month old major resuscitation with death); however, overall it is still a small percentage. Many residents felt that FP benefits parents. Conclusions: A high percentage of EM residents report FP as an interfering factor with performance of certain procedures especially intubation and resuscitation. Residents with prior internship or residency were as likely to report FP as an interfering factor in those situations. It seems that a greater number of families would be allowed to be present during resuscitation at children’s hospital ED vs. a general ED. However, overall it is still a small percentage. Complexity of the procedures performed appears to be a significant factor in decreasing residents’ acceptance of FP. However, EM residents expressed that FP benefits parents.
194
Potential Adverse Events in Children Treated in the Emergency Department
Vega RM, Mayerhoff RM, Livote E, Argota E, Ward MF, Sama A, North Shore University Hospital, Manhasset, NY; Institute for Medical Research at North Shore-Long Island Jewish, Manhasset, NY
Objective: To determine the rate of PAEs in pediatric emergency department patients over the course of a six-week period. Methods: A retrospective review of medical records of all pediatric patients admitted through the emergency department (ED) during an approximately six-week period in June and July of 2003. Variables to be studied included documentation of patient’s weight, use of milligram per kilogram (mg/kg) dosing, type of medication ordered including intravenous fluids. Study setting was a suburban Level I Trauma Center with an annual census of approximately 60,000 total and 15,000 pediatric patients. Inclusion criteria: All patients \18 years old seen in the ED and for whom at least one weight-based medication was ordered. Exclusion criteria: Patients greater than or equal to 18 years old or patients for whom no weight-based medications were ordered. Statistical analysis: A power analysis determined that the sample size needed to show a difference of 0.1 (0.23 to 0.33) was 1973. A total of 1132 medical records were reviewed with a total of 809 medications or fluids orders prescribed. Medications were categorized as to whether they required weight based dosing or not, and classified into the following categories: antibiotics, narcotics, none-narcotic analgesics, sedatives, steroids and other. Patients ages were divided into the following groups: infants, pre-school, school age, and teenage. Results: The total number of patients for whom weight-based medications were ordered was 440. The rate of PAEs within this group were as follows: overall 36.4%; by age group: infants 12.2%, pre-school 21.2%, school age 33.8%, teenage 80%. The potential adverse event rate by drug category were as follows: antibiotics 31.3%, narcotics 38.5%, non-narcotic analgesics 32.5%, sedatives 10%, and steroids 21.3%.
Volume 46, no. 3 : September 2005
Research Forum Abstracts Conclusions: Compared to a national rate of 1.81 - 2.96 per 100 discharges, there were no documented adverse events in the patients studied. Documentation and medication prescribing practices need to be improved to minimize the risks to pediatric patients. A future direction would be to compare the number of documentation errors in conventional medical records to that in electronic medical records.
195
Cost Effective Quality Care for Transient Ischemic Attack
Stead LG, Gilmore RM, Boie ET, Geerdes P, Decker WW, Mayo Clinic, Rochester, MN
Study Objectives: Backround the need to promptly investigate a transient ischemic attack (TIA) is based on the risk of ischemic infarct being highest soon after a TIA, at approximately 5% within the first 2 days. The management of TIA is complicated by the difficulty in making the diagnosis at times, which argues for a protocol to address this high takes diagnosis. Furthermore, cost containment has become a reality for most institutions, as historically, the medicare reimbursement for TIA has been poor. Together, these factors led us to pilot an expedited Emergency Department Observation Unit protocol for TIA (EDOU-TIA). Objective: To study the feasibility, and cost effectiveness of an EDOU protocol for patients presenting with TIA. Methods: The study design was an observational one and consisted of patients who presented to the Emergency Department (ED) with TIA during the one year period between April 1, 2004 and March 31, 2005. Inclusion criteria were: 1) Patients 18 years and older with symptoms suggestive of TIA; 2)Head CT negative for mass, bleed, shift or fracture; systolic blood pressure within 110-220 mmHg; 4) asymptomatic. Exclusion criteria were: 1) patients with acute stroke; 2) patients who would be admitted to hospital regardless of EDOU workup; 3) patients who are unstable in any way (hypotense, ongoing chest pain, ongoing cerebrovascular ischemia symptoms). The EDOU workup consisted of carotid ultrasonography, brain ischemia laboratory panel, electrocardiogram, and evaluation by the neurology consultation service. All patients were given a prescription for an antithrombotic agent upfront. Additionally, the EDOU group was given educational material on stroke risk factors and the option to watch a video on stroke prevention. For each of the two groups, (EDOU and Inpatient), total hospital direct costs were calculated, including individual costs for ED, inpatient medical service, pharmacy, radiology, laboratory medicine, and patient care services. Since it was a pilot, some patients were admitted to the hospital for workup while others were put through the EDOU TIA protocol. This allowed for a direct cost comparison of those who were admitted solely for workup (hospital length of stay [HLOS] of one day) versus those who got their workup in the ED. The 3rd group, which consisted of patients who were admitted with HLOS R 2 days, were excluded from the analysis, as they had other co-morbidities that required hospitalization for purposes other than workup of TIA. Results: A total of 144 patients were seen with an ultimate discharge diagnosis of TIA. Of these, 68 were admitted with HLOS R 2 days and were excluded from further analysis, as they had other co-morbidities that required hospitalization for purposes other than workup of TIA. Of the remaining 76 patients, 47 were admitted to the hospital for workup and 29 went through the EDOU TIA protocol. The protocol functioned smoothly, and served to decompress the main area of the ED. Baseline characteristics between the two groups were similar. The total direct costs were significantly lower for the ED group when compared to the inpatient group ($1322 vs. $3251). Conclusion: Implementation of an EDOU TIA protocol is feasible, and appears to be more cost effective than performing the workup in an inpatient setting.
196
A Primary Care Physician Seeing Patients in a Fast Track Area: Effect on Effectiveness, Efficiency, and Perceived Care Quality
Jimenez S, Miro O, De la Red G, Bragulat E, Coll-Vinent B, Sanchez M, Hospital Clinic, Barcelona, Spain
Study Objectives: In Spain, tertiary care hospitals usually have emergency departments (ED) divided into independent medicine, surgery, and orthopedics units. To make lighter external pressure and overcrowding, some units contain fast track areas (FTA). The aim was to know the effect that the presence of one primary care physician (PCP) had on FTA effectiveness, efficiency and perceived care quality. Methods: Prospective interventional study carried out in an ED medicine unit of a tertiary care hospital. Our FTA is opened at 8 am and ideally closed at 12 am, and
Volume 46, no. 3 : September 2005
staffed by 2 residents (1 PGY-1 and 1 PGY-2). Intervention: 8 PGY-2 hours (from 4 pm to 12 am) were substituted by 8 PCP hours. The study period was August, 2002 (PCP presence), and the control period, October, 2002. From each period, 10 days and 100 patients were randomly selected. From each day, FTA census (DC), percentage of revisits and patients leaving without being seen, elapsed time to actual FTA closing (TC), percentage of patients moved to the observation area, and percentage of admissions were recorded. From each patient, demographic characteristics, waiting time to be seen (WT), number of tests performed, elapsed time to first treatment (TT), and length of stay (LOS) were collected along with the number of patients finally discharged without specialist consultation and those with no test ordered. Perceived care quality was assessed by a telephone survey. Three effectiveness indexes were calculated: DC/WT (E1), DC/TC (E2), and perceived care quality/perceived WT (E3). Finally, costs (C) from both periods were calculated, and cost-effectiveness analysis performed. Results: Periods showed no differences regarding DC and patient characteristics. In the study period (PCP presence), all time variables significantly improved: 20% reduction in WT, 25% in TT, 36% in LOS, and 17,8% in TC. A decrease in the number of tests ordered (41% less), in the percentage of patients moved to the observation area (78% less), and in the revisit rate (75% less) was also significantly noted. Finally, E1 improved in 77% and E2 in 51%. Cost-effectiveness analysis clearly supported the study period, showing a significant decrease in C/E1 (55% less), in C/E2 (33% less), and in C/E3 (6% less). From telephone survey, no differences between periods were detected but a perceived WT within the study period less than that within the control period. Conclusion: PCP leads to an improvement in FTA effectiveness, with a modestly positive effect on perceived quality. Besides, this presence is also efficient. Therefore, this intervention could be taken into account by administrators to better manage ED resources.
197
Resident Sex and the Pelvic Exam: The Effect of Physician Sex on Selection and Management of Patients Who May Require a Pelvic Exam
Mohart AE, Banet GA, Katz ED, Washington University in St Louis, St. Louis, MO
Study Objectives: 1. To determine if resident sex predicts the likelihood of selecting a patient whose demographics and chief complaint infer a high likelihood of the need for a pelvic exam. 2. To determine if resident sex affects the frequency with which these patients receive a pelvic exam. Methods: A continuous, retrospective chart review was performed on all emergency department charts at a high volume, tertiary care, university hospital from 11/1/2004 through 04/06/2005. The inclusion criteria were that the ED patient be female, 12-55 years old, have a chief complaint suggestive of the need for pelvic exam (abdominal pain, pelvic pain, urinary or vaginal complaint, pregnancy complication, or sexual assault), and be primarily cared for by an emergency medicine resident. Charts were excluded if the patient declined or refused a pelvic exam. A sole physician abstractor recorded the patient’s age, race, and the occurrence of pelvic examination. The section of the ED the patient was in was also recorded to allow for the impact of different patient populations in different patient care areas. The primary resident physician was recorded by sex and post graduate year of training. Chi-square test and Student’s t test were used for statistical analysis. Results: Of 32,202 charts reviewed, 889 (2.8%) met inclusion criteria. During the period under review, 61% of shifts were covered by male residents while 39% were covered by female residents. Male residents signed up for 580 (65.2%, 95% CI 62.1-68.3)) while females signed up for 309 (34.8%, 95% CI 31.7-38.0)) which was not a statistically significant difference from the ratio of male to female residents. 498/889 patients (56.0%, 95% CI 52.7-59.3) had a pelvic exam documented. Of the patients seen by male residents, 291/580 (50.2%, 95% CI 46.1-54.2) had a pelvic exam performed, while female residents did pelvic exams on 207/309 patients (67.0%, 95% CI 61.6-72.0%), a highly significant difference (p\.001, chi-square 23.1) Variation was found in patient population between the two groups. Male residents cared for a higher percentage of AfricanAmerican patients compared to female residents (73% vs. 62%, p=.01). There was a trend (p=.051) towards males selecting younger patients with female physicians’ patients’ mean age 32.2 years, (95% CI = 30.9-33.4) and male physicians’ patients’ mean age 30.6 years (95% CI=29.7-31.5). Race was not related to whether a pelvic exam was done (54.2% for African-Americans versus 56.8% for Caucasians, p=.486). Younger age (\28 years) was highly associated with greater likelihood of having a pelvic exam (71% versus 39%, p\.001). Multi-variate analysis will be presented at conference. Conclusion: Sex was not found to predict resident selection of patients with a potential need for a pelvic exam. However, female residents were far more likely to
Annals of Emergency Medicine S55
191
What a Bite: Review of Snakebites in Children
Feng S, Stephan M, UT Southwestern Medical Center/Children’s Medical Center of Dallas, Dallas, TX
Study Objectives: Approximately 2000 children less than 17 years of age are victims of crotalid envenomation each year. Multiple controversies exist in the management of these bites. We present the largest pediatric review of snakebite envenomations. Methods: This is a retrospective chart review of pediatric patients presenting with venomous snakebites. Data was abstracted for; demographics, the severity of the bite, circumstances of the snakebite, lab values, method and dose of antivenom administered, VS, length of hospital stay, adverse outcomes, cost and disposition. Results: 49 patients were identified. Ages ranged 22 mos -16.8 years, 67.3% male. The envenomations consisted of: 65% copperheads, 8% cottonmouth, 27% unknown. 47% were moderate level 2 envenomations, 36.7% were mild level 1. Systemic symptoms were found in 26.8% with predominance of vomiting. Initial coagulation evaluation revealed 26.5% abnormal CBCs: 7 pts had WBCO15K, 4 had thrombocytopenia. 30.6% had abnormal PT, 10.1% abnormal PTT. 1 patient had abnormal fibrinogen and d-dimer results (cottonmouth envenomation). 31.6% had other labs performed predominantly renal group(35%). Subsequent labs were significant for persistently abnormal percentage PTT and d-dimer (16.3%, 4.1%) Antivenin was used in 34.6%: 82.3% received CroFab (CF) with 1 adverse reaction described as facial flushing. Vital signs remained stable during all infusions. 18.4% received subsequent CF dosing. No reactions were noted on subsequent dosing. 22% had progression of symptoms described as edema during hospitalization. 49% had a tox consult. Antivenin use in all snakebites with tox consult (62.5%) is statistically significant (p\0.004) using Pearson Chi-Square. 50% of copperhead bites with tox consult received no antivenin. 43% had 2 day hospital stay. 59% received antibiotics. 2 patients were readmitted for surgical debridement of hemorrhagic blisters on digits. Conclusion: Copperhead envenomations are prevalent in Northeast Texas as reflected in our data. The use of Wyeth Polyvalent Antivenin and CF appear to be safe and effective. Antivenom use still remains controversial in copperhead envenomations. We present the largest descriptive analysis of snakebite envenomations in the pediatric population.
192
Is the Combination of Ibuprofen and Acetaminophen a Better Pediatric Antipyretic Than Ibuprofen Alone?
Tenison M, Eberhardt M, Pellett N, Heller M, St Lukes Hospital, Bethlehem, PA
Study Objectives: Acetaminophen and ibuprofen are often combined for the treatment of pediatric fever despite the lack of data to support or refute this practice. The purpose of this prospective, controlled, randomized, double-blind clinical investigation is to determine whether the use of these two agents in combination is a more effective antipyretic than ibuprofen alone. Methods: Febrile (O101 C) children age 3 to 10 were enrolled during a visit to a community hospital ED or its affiliated pediatric clinic. Subjects were randomized to treatment with one of two regimens: a combination of oral ibuprofen (10mg/kg) and acetaminophen (15mg/kg) (study) or ibuprofen (10mg/kg) and identically-appearing placebo (control). Initial oral temperatures were recorded during the visit; 2, 4 and 6-hour oral temperatures were later reported via phone or mail-in form by the patient’s caregiver. Results: In this predefined interim analysis 28 febrile children were enrolled with at least 2 and 4-hour temperatures available for 20 subjects. Initial temperatures ranged from 101.2 to 104.2 (mean 102.5) and were similar between groups (p=0.42). Mean temperature change between 0 and 2 hours was also similar (-3.1 degC for the both groups; p=0.66). Mean temperature decrease between 0 and 4 hours however, was significantly greater in the study group (-4.4 deg C for the study group and -3.7 deg C for controls; p=0.05). At 6 hours this difference was even more pronounced (-4.8 deg C for the study group versus -1.3 deg C for controls; p=0.02). Conclusion: Although the combination of ibuprofen and acetaminophen was no more effective than ibuprofen alone at 2 hours, it may offer a more prolonged antipyretic effect.
S54 Annals of Emergency Medicine
193
Attitudes and Practices of Emergency Medicine Residents Regarding Family Presence During Pediatric Procedural and Resuscitation Scenarios
Barata IA, LaMantia J, Riccardi D, Mayerhoff RM, Litroff A, D’Abbracci P, Livote E, Ward MF, Sama A, North Shore University Hospital, Manhasset, NY
Objective: Determine attitudes and practices of emergency medicine (EM) residents regarding family presence (FP) during hypothetical pediatric procedural and resuscitation scenarios. Methods: This was a prospective, anonymous survey of residents in ACGME accredited residency programs. Surveys were mailed in September 2004 for completion by residents. Descriptive statistics were employed. Results: Responses were received from 61 programs with the following geographic distribution: 38% Northeast, 37% Midwest, 18% South and 5% West Coast. A total of 521 surveys were completed by EM residents: 193 EM-1, 170 EM2, 133 EM-3, 16 EM-4, 8 other or missing; 129 residents completed an internship year other than EM: 52 IM, 18 General Surgery, 13 Family Practice, 5 Pediatrics, 41 other; 169 (32%) residents see pediatric patients primarily at a children’s hospital. EM residents reported that FP during selected procedures would interfere with their ability to perform the interventions: 25% conscious sedation, 45% spinal tap, 48% intubation, 50% resuscitation. In addition, only a small percentage of residents would always allow FP during certain procedures: 31% conscious sedation, 19% spinal tap, 11% resuscitation, 10% intubation. EM-1 residents that have completed a prior internship year or residency were as likely as EM-1 residents without prior experience to report that FP would interfere with their performance, especially in the case scenarios of intubation (59% vs.49%; p value 0.2505) and resuscitation (51% vs. 40%; p value 0.1827). When comparing the practices of emergency medicine residents regarding FP during pediatric resuscitation at children’s hospitals ED vs. a general ED it seems that a greater number of families would be allowed to be present during resuscitation at children’s hospitals: 17% vs. 9% (p value 0.0108) (7 yr old major resuscitation), 23% vs.15% (p value 0.0375) (2 month old major resuscitation with death); however, overall it is still a small percentage. Many residents felt that FP benefits parents. Conclusions: A high percentage of EM residents report FP as an interfering factor with performance of certain procedures especially intubation and resuscitation. Residents with prior internship or residency were as likely to report FP as an interfering factor in those situations. It seems that a greater number of families would be allowed to be present during resuscitation at children’s hospital ED vs. a general ED. However, overall it is still a small percentage. Complexity of the procedures performed appears to be a significant factor in decreasing residents’ acceptance of FP. However, EM residents expressed that FP benefits parents.
194
Potential Adverse Events in Children Treated in the Emergency Department
Vega RM, Mayerhoff RM, Livote E, Argota E, Ward MF, Sama A, North Shore University Hospital, Manhasset, NY; Institute for Medical Research at North Shore-Long Island Jewish, Manhasset, NY
Objective: To determine the rate of PAEs in pediatric emergency department patients over the course of a six-week period. Methods: A retrospective review of medical records of all pediatric patients admitted through the emergency department (ED) during an approximately six-week period in June and July of 2003. Variables to be studied included documentation of patient’s weight, use of milligram per kilogram (mg/kg) dosing, type of medication ordered including intravenous fluids. Study setting was a suburban Level I Trauma Center with an annual census of approximately 60,000 total and 15,000 pediatric patients. Inclusion criteria: All patients \18 years old seen in the ED and for whom at least one weight-based medication was ordered. Exclusion criteria: Patients greater than or equal to 18 years old or patients for whom no weight-based medications were ordered. Statistical analysis: A power analysis determined that the sample size needed to show a difference of 0.1 (0.23 to 0.33) was 1973. A total of 1132 medical records were reviewed with a total of 809 medications or fluids orders prescribed. Medications were categorized as to whether they required weight based dosing or not, and classified into the following categories: antibiotics, narcotics, none-narcotic analgesics, sedatives, steroids and other. Patients ages were divided into the following groups: infants, pre-school, school age, and teenage. Results: The total number of patients for whom weight-based medications were ordered was 440. The rate of PAEs within this group were as follows: overall 36.4%; by age group: infants 12.2%, pre-school 21.2%, school age 33.8%, teenage 80%. The potential adverse event rate by drug category were as follows: antibiotics 31.3%, narcotics 38.5%, non-narcotic analgesics 32.5%, sedatives 10%, and steroids 21.3%.
Volume 46, no. 3 : September 2005
Research Forum Abstracts Conclusions: Compared to a national rate of 1.81 - 2.96 per 100 discharges, there were no documented adverse events in the patients studied. Documentation and medication prescribing practices need to be improved to minimize the risks to pediatric patients. A future direction would be to compare the number of documentation errors in conventional medical records to that in electronic medical records.
195
Cost Effective Quality Care for Transient Ischemic Attack
Stead LG, Gilmore RM, Boie ET, Geerdes P, Decker WW, Mayo Clinic, Rochester, MN
Study Objectives: Backround the need to promptly investigate a transient ischemic attack (TIA) is based on the risk of ischemic infarct being highest soon after a TIA, at approximately 5% within the first 2 days. The management of TIA is complicated by the difficulty in making the diagnosis at times, which argues for a protocol to address this high takes diagnosis. Furthermore, cost containment has become a reality for most institutions, as historically, the medicare reimbursement for TIA has been poor. Together, these factors led us to pilot an expedited Emergency Department Observation Unit protocol for TIA (EDOU-TIA). Objective: To study the feasibility, and cost effectiveness of an EDOU protocol for patients presenting with TIA. Methods: The study design was an observational one and consisted of patients who presented to the Emergency Department (ED) with TIA during the one year period between April 1, 2004 and March 31, 2005. Inclusion criteria were: 1) Patients 18 years and older with symptoms suggestive of TIA; 2)Head CT negative for mass, bleed, shift or fracture; systolic blood pressure within 110-220 mmHg; 4) asymptomatic. Exclusion criteria were: 1) patients with acute stroke; 2) patients who would be admitted to hospital regardless of EDOU workup; 3) patients who are unstable in any way (hypotense, ongoing chest pain, ongoing cerebrovascular ischemia symptoms). The EDOU workup consisted of carotid ultrasonography, brain ischemia laboratory panel, electrocardiogram, and evaluation by the neurology consultation service. All patients were given a prescription for an antithrombotic agent upfront. Additionally, the EDOU group was given educational material on stroke risk factors and the option to watch a video on stroke prevention. For each of the two groups, (EDOU and Inpatient), total hospital direct costs were calculated, including individual costs for ED, inpatient medical service, pharmacy, radiology, laboratory medicine, and patient care services. Since it was a pilot, some patients were admitted to the hospital for workup while others were put through the EDOU TIA protocol. This allowed for a direct cost comparison of those who were admitted solely for workup (hospital length of stay [HLOS] of one day) versus those who got their workup in the ED. The 3rd group, which consisted of patients who were admitted with HLOS R 2 days, were excluded from the analysis, as they had other co-morbidities that required hospitalization for purposes other than workup of TIA. Results: A total of 144 patients were seen with an ultimate discharge diagnosis of TIA. Of these, 68 were admitted with HLOS R 2 days and were excluded from further analysis, as they had other co-morbidities that required hospitalization for purposes other than workup of TIA. Of the remaining 76 patients, 47 were admitted to the hospital for workup and 29 went through the EDOU TIA protocol. The protocol functioned smoothly, and served to decompress the main area of the ED. Baseline characteristics between the two groups were similar. The total direct costs were significantly lower for the ED group when compared to the inpatient group ($1322 vs. $3251). Conclusion: Implementation of an EDOU TIA protocol is feasible, and appears to be more cost effective than performing the workup in an inpatient setting.
196
A Primary Care Physician Seeing Patients in a Fast Track Area: Effect on Effectiveness, Efficiency, and Perceived Care Quality
Jimenez S, Miro O, De la Red G, Bragulat E, Coll-Vinent B, Sanchez M, Hospital Clinic, Barcelona, Spain
Study Objectives: In Spain, tertiary care hospitals usually have emergency departments (ED) divided into independent medicine, surgery, and orthopedics units. To make lighter external pressure and overcrowding, some units contain fast track areas (FTA). The aim was to know the effect that the presence of one primary care physician (PCP) had on FTA effectiveness, efficiency and perceived care quality. Methods: Prospective interventional study carried out in an ED medicine unit of a tertiary care hospital. Our FTA is opened at 8 am and ideally closed at 12 am, and
Volume 46, no. 3 : September 2005
staffed by 2 residents (1 PGY-1 and 1 PGY-2). Intervention: 8 PGY-2 hours (from 4 pm to 12 am) were substituted by 8 PCP hours. The study period was August, 2002 (PCP presence), and the control period, October, 2002. From each period, 10 days and 100 patients were randomly selected. From each day, FTA census (DC), percentage of revisits and patients leaving without being seen, elapsed time to actual FTA closing (TC), percentage of patients moved to the observation area, and percentage of admissions were recorded. From each patient, demographic characteristics, waiting time to be seen (WT), number of tests performed, elapsed time to first treatment (TT), and length of stay (LOS) were collected along with the number of patients finally discharged without specialist consultation and those with no test ordered. Perceived care quality was assessed by a telephone survey. Three effectiveness indexes were calculated: DC/WT (E1), DC/TC (E2), and perceived care quality/perceived WT (E3). Finally, costs (C) from both periods were calculated, and cost-effectiveness analysis performed. Results: Periods showed no differences regarding DC and patient characteristics. In the study period (PCP presence), all time variables significantly improved: 20% reduction in WT, 25% in TT, 36% in LOS, and 17,8% in TC. A decrease in the number of tests ordered (41% less), in the percentage of patients moved to the observation area (78% less), and in the revisit rate (75% less) was also significantly noted. Finally, E1 improved in 77% and E2 in 51%. Cost-effectiveness analysis clearly supported the study period, showing a significant decrease in C/E1 (55% less), in C/E2 (33% less), and in C/E3 (6% less). From telephone survey, no differences between periods were detected but a perceived WT within the study period less than that within the control period. Conclusion: PCP leads to an improvement in FTA effectiveness, with a modestly positive effect on perceived quality. Besides, this presence is also efficient. Therefore, this intervention could be taken into account by administrators to better manage ED resources.
197
Resident Sex and the Pelvic Exam: The Effect of Physician Sex on Selection and Management of Patients Who May Require a Pelvic Exam
Mohart AE, Banet GA, Katz ED, Washington University in St Louis, St. Louis, MO
Study Objectives: 1. To determine if resident sex predicts the likelihood of selecting a patient whose demographics and chief complaint infer a high likelihood of the need for a pelvic exam. 2. To determine if resident sex affects the frequency with which these patients receive a pelvic exam. Methods: A continuous, retrospective chart review was performed on all emergency department charts at a high volume, tertiary care, university hospital from 11/1/2004 through 04/06/2005. The inclusion criteria were that the ED patient be female, 12-55 years old, have a chief complaint suggestive of the need for pelvic exam (abdominal pain, pelvic pain, urinary or vaginal complaint, pregnancy complication, or sexual assault), and be primarily cared for by an emergency medicine resident. Charts were excluded if the patient declined or refused a pelvic exam. A sole physician abstractor recorded the patient’s age, race, and the occurrence of pelvic examination. The section of the ED the patient was in was also recorded to allow for the impact of different patient populations in different patient care areas. The primary resident physician was recorded by sex and post graduate year of training. Chi-square test and Student’s t test were used for statistical analysis. Results: Of 32,202 charts reviewed, 889 (2.8%) met inclusion criteria. During the period under review, 61% of shifts were covered by male residents while 39% were covered by female residents. Male residents signed up for 580 (65.2%, 95% CI 62.1-68.3)) while females signed up for 309 (34.8%, 95% CI 31.7-38.0)) which was not a statistically significant difference from the ratio of male to female residents. 498/889 patients (56.0%, 95% CI 52.7-59.3) had a pelvic exam documented. Of the patients seen by male residents, 291/580 (50.2%, 95% CI 46.1-54.2) had a pelvic exam performed, while female residents did pelvic exams on 207/309 patients (67.0%, 95% CI 61.6-72.0%), a highly significant difference (p\.001, chi-square 23.1) Variation was found in patient population between the two groups. Male residents cared for a higher percentage of AfricanAmerican patients compared to female residents (73% vs. 62%, p=.01). There was a trend (p=.051) towards males selecting younger patients with female physicians’ patients’ mean age 32.2 years, (95% CI = 30.9-33.4) and male physicians’ patients’ mean age 30.6 years (95% CI=29.7-31.5). Race was not related to whether a pelvic exam was done (54.2% for African-Americans versus 56.8% for Caucasians, p=.486). Younger age (\28 years) was highly associated with greater likelihood of having a pelvic exam (71% versus 39%, p\.001). Multi-variate analysis will be presented at conference. Conclusion: Sex was not found to predict resident selection of patients with a potential need for a pelvic exam. However, female residents were far more likely to
Annals of Emergency Medicine S55