- Email: [email protected]
236 Withdrawn
Research Forum Abstracts (OR¼3.81), hematology-oncology disorders (OR¼3.79), digestive system, (OR¼1.46), and other disorders (OR¼1.27), had a higher odds of being transferred as compared to trauma/injury and poisoning; while patients with disorders related to genitourinary (OR¼0.93), respiratory (OR¼0.78), musculoskeletal (OR¼0.59), skin (OR¼0.47), infectious and parasitic diseases (OR¼0.23), and eyes/ears/nose/ throat (OR¼0.09), had a lower odds of being transferred as compared to trauma/ injury and poisoning. Conclusion: Children younger than 1 year had relatively higher transfer rates. Transfers also differed by payer; patients covered by Medicaid and self-pay had the lowest likelihood of transfer. Transfer rates varied significantly by condition and several of the high-transfer diagnostic categories were related to circulatory, endocrine, nervous, hematology-oncology, and mental disorders as well as congenital anomalies which may be related to a lack of ED or inpatient resources to care for children with problems that require more complex care.
235
Predictive Variables for Abnormal Comprehensive Metabolic Panel Testing and Potential Cost Savings in Children Receiving Pediatric Emergency Department Care
Huckaby MD, Freeman S, Thurmond C, Cooper M, Losek J/Louisiana State University Health Science Center Shreveport, Shreveport, LA; Medical University of South Carolina, Charleston, SC
Study Objectives: To determine clinical variables predictive of abnormal comprehensive metabolic panel (CMP) test results in pediatric emergency department (PED) patients and then use these predictive variables to determine the potential cost savings of ordering a basic metabolic panel (BMP) versus a CMP. Methods: Retrospective cross-sectional descriptive study of children (< 18 years) at an urban academic PED (annual census of 22,000) who had CMP testing. Demographic and clinical data included 12 clinical variables: present illness (right upper quadrant pain, overdose and emesis), past medical history (liver disorder, malignancy, heart disease and bleeding disorder), and physical examination findings (jaundice, right upper quadrant tenderness, hepatomegaly, ascites/peripheral edema and shock) and the 6 CMP (6-CMP) test results not included in a BMP (alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, total bilirubin, total protein, and albumin). Patients with normal vs abnormal 6-CMP tests were compared to determine the predictive value of 12 clinical variables. Results: There were 207 children in the study population. The mean age (months) was 95.4 and range 2 to 206 months. There were 106 (51%) males. Clinical variables significantly associated with an abnormal 6-CMP result were history of liver disease (LR 3.57 and P ¼ .007), history of heart disease (LR 2.92 and P ¼ .040), jaundice (LR >10 and P ¼ .045), and hepatomegaly (LR 5.67 and P ¼ .048). The presence of at least one variable had a LR 1.55, P < .001, and sensitivity of 84.1%. The false negative rate (failure of having at least one of the 12 clinical variables identify a patient with an abnormal 6-CMP) was nearly 16%. However, of the10 patients for whom this false negative rate remained true, the 6-CMP values were marginally abnormal and performance of further investigation of these abnormal results was minimal to none. There were 66 patients with no clinical variables and normal CMP results. With a cost difference of $21 between BMP and CMP, this gives a potential savings of 66 x $21 ¼ $1386 over a 71-day period and $7,125 if extrapolated over 1 year in a relatively low volume PED. Conclusion: Limiting testing to a BMP for patients with none of the 12 clinical variables has the potential annual cost savings of $7,125.
236 237
Withdrawn Retrospective Review of Symptoms and Signs of Intussusception Present on Initial Evaluation
Wolford RW, Browning B/OSF Saint Francis Medical Center, Peoria, IL
Study Objectives: Intussusception is the second leading cause of acute pediatric abdominal emergencies, after appendicitis. The classic triad of pain, vomiting, and bloody stools occurs variably (7% to 60%). We diagnosed a 7-month old male with
S86 Annals of Emergency Medicine
intussusception, whose only symptom/sign was emesis. A literature search was unable to determine the frequency of isolated vomiting as the presentation of intussusception. Our objective is to determine the frequency and patterns of symptoms/signs of pediatric intussusception. Methods: A retrospective chart review of patients (<18 years of age) with the discharge diagnosis of intussusception, from May 1, 2010 to November 30, 2014, was conducted. A single investigator reviewed and abstracted the data (month of presentation, age, sex, time since onset, 9 specific symptoms/signs). The study was approved by the institutional review board. Results: A total of 110 patients were identified (56% male, median age 26 months, range 1 - 191 months). The 3 most common symptoms/signs were: pain (82%), emesis (72%), and irritable/fussy (33%). Only 8 (7%) had the “classic” triad. Eleven (10%) patients had only 1 symptom/sign: 9 (8%) pain and 2 (2%) emesis. Twenty-seven (25%) presented with 2 symptoms/signs: 10 (9%) pain and emesis, 12 (11%) pain and other (not already listed, 3 (3%) emesis and lethargy, and 2 (2%) emesis and bloody stool. Seventy-two (65%) patients presented with 3 symptoms/signs. All patients had either pain or vomiting. Conclusion: Emesis as the only symptom or sign of intussusception is rare and the majority of patients have 2 symptoms/signs. The absence of pain and vomiting makes the diagnosis of intussusception unlikely and other etiologies of the symptoms and signs should be considered. This may help guide the efficient use of laboratory testing and imaging. A prospective multicenter study would be of value.
238 239
Moved to 16
The Accuracy of the Yale Observation Scale Score and Unstructured Clinician Suspicion to Identify Febrile Infants Aged £60 Days With Serious Bacterial Infections
Nigrovic LE, Mahajan PV, Tzimenatos L, Alpern ER, Rogers AJ, Simmons T, Casper C, Ramilo O, Kuppermann N/Boston Children’s Hospital, Boston, MA; Children’s Hospital of Michigan, Detroit, MI; University of California, Davis, Sacramento, CA; Lurie Children’s Hospital of Chicago, Chicago, IL; University of Michigan, Ann Arbor, MI; Pediatric Emergency Care Applied Research Network (PECARN) Data Coordinating Center, Salt Lake, UT; Nationwide Children’s Hospital, Columbus, OH
Study Objectives: Both the Yale Observation Scale (YOS) score and unstructured clinician suspicion have been used to predict the risk of serious bacterial infections (SBIs) in young febrile infants. However, the number of infants 60 days studied has been limited. Our objective was to assess the ability of the YOS score and unstructured clinician suspicion to identify febrile infants 60 days with SBIs. Methods: We performed a planned secondary analysis of a prospective cohort study of non-critical febrile infants 60 days presenting to one of 26 emergency departments (EDs) in PECARN who had blood cultures obtained. We defined a SBI as urinary tract infection (UTI), bacteremia, or bacterial meningitis and only included infants whose SBI status was known. ED faculty and fellow clinicians applied the YOS score and also estimated the risk of SBI ( 50%) using unstructured clinician suspicion. We used previously described cut-points for the YOS (“perfect” score of 6 and “normal” score of 10). We then compared the performance of the dichotomized YOS scores to unstructured clinician suspicion ( 1% and > 5%) for the prediction of SBIs. Results: We enrolled 4778 patients, of whom 4592 (96%) had known SBI status. Of these 4592 infants, 445 (9.7%) had SBIs of whom 348 (78% of SBIs) had UTIs. 1467 (32%) were 28 days of age. The median YOS score was no different in infants with SBIs than those without (6 vs 6; P ¼ .55). Performance of the YOS score and clinician suspicion for SBI is given in the table below. Of the 24 infants with bacterial meningitis, the YOS scores ranged from 6 to 26, with 2 (38%) having a scores of 6 and 14 (58%) having a scores 10. Clinician suspicion varied from 50% with 2 (8%) having risk assessments of < 1% and 10 (42%) having risk assessments of 5%. Conclusion: Neither the YOS score nor unstructured clinician suspicion could reliably exclude SBIs in febrile infants 60 days, including meningitis, in febrile infants 60 days of age. An accurate clinical prediction rule to identify infants at low and high risk for SBIs is needed.
Volume 66, no. 4s : October 2015
235
Predictive Variables for Abnormal Comprehensive Metabolic Panel Testing and Potential Cost Savings in Children Receiving Pediatric Emergency Department Care
Huckaby MD, Freeman S, Thurmond C, Cooper M, Losek J/Louisiana State University Health Science Center Shreveport, Shreveport, LA; Medical University of South Carolina, Charleston, SC
Study Objectives: To determine clinical variables predictive of abnormal comprehensive metabolic panel (CMP) test results in pediatric emergency department (PED) patients and then use these predictive variables to determine the potential cost savings of ordering a basic metabolic panel (BMP) versus a CMP. Methods: Retrospective cross-sectional descriptive study of children (< 18 years) at an urban academic PED (annual census of 22,000) who had CMP testing. Demographic and clinical data included 12 clinical variables: present illness (right upper quadrant pain, overdose and emesis), past medical history (liver disorder, malignancy, heart disease and bleeding disorder), and physical examination findings (jaundice, right upper quadrant tenderness, hepatomegaly, ascites/peripheral edema and shock) and the 6 CMP (6-CMP) test results not included in a BMP (alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, total bilirubin, total protein, and albumin). Patients with normal vs abnormal 6-CMP tests were compared to determine the predictive value of 12 clinical variables. Results: There were 207 children in the study population. The mean age (months) was 95.4 and range 2 to 206 months. There were 106 (51%) males. Clinical variables significantly associated with an abnormal 6-CMP result were history of liver disease (LR 3.57 and P ¼ .007), history of heart disease (LR 2.92 and P ¼ .040), jaundice (LR >10 and P ¼ .045), and hepatomegaly (LR 5.67 and P ¼ .048). The presence of at least one variable had a LR 1.55, P < .001, and sensitivity of 84.1%. The false negative rate (failure of having at least one of the 12 clinical variables identify a patient with an abnormal 6-CMP) was nearly 16%. However, of the10 patients for whom this false negative rate remained true, the 6-CMP values were marginally abnormal and performance of further investigation of these abnormal results was minimal to none. There were 66 patients with no clinical variables and normal CMP results. With a cost difference of $21 between BMP and CMP, this gives a potential savings of 66 x $21 ¼ $1386 over a 71-day period and $7,125 if extrapolated over 1 year in a relatively low volume PED. Conclusion: Limiting testing to a BMP for patients with none of the 12 clinical variables has the potential annual cost savings of $7,125.
236 237
Withdrawn Retrospective Review of Symptoms and Signs of Intussusception Present on Initial Evaluation
Wolford RW, Browning B/OSF Saint Francis Medical Center, Peoria, IL
Study Objectives: Intussusception is the second leading cause of acute pediatric abdominal emergencies, after appendicitis. The classic triad of pain, vomiting, and bloody stools occurs variably (7% to 60%). We diagnosed a 7-month old male with
S86 Annals of Emergency Medicine
intussusception, whose only symptom/sign was emesis. A literature search was unable to determine the frequency of isolated vomiting as the presentation of intussusception. Our objective is to determine the frequency and patterns of symptoms/signs of pediatric intussusception. Methods: A retrospective chart review of patients (<18 years of age) with the discharge diagnosis of intussusception, from May 1, 2010 to November 30, 2014, was conducted. A single investigator reviewed and abstracted the data (month of presentation, age, sex, time since onset, 9 specific symptoms/signs). The study was approved by the institutional review board. Results: A total of 110 patients were identified (56% male, median age 26 months, range 1 - 191 months). The 3 most common symptoms/signs were: pain (82%), emesis (72%), and irritable/fussy (33%). Only 8 (7%) had the “classic” triad. Eleven (10%) patients had only 1 symptom/sign: 9 (8%) pain and 2 (2%) emesis. Twenty-seven (25%) presented with 2 symptoms/signs: 10 (9%) pain and emesis, 12 (11%) pain and other (not already listed, 3 (3%) emesis and lethargy, and 2 (2%) emesis and bloody stool. Seventy-two (65%) patients presented with 3 symptoms/signs. All patients had either pain or vomiting. Conclusion: Emesis as the only symptom or sign of intussusception is rare and the majority of patients have 2 symptoms/signs. The absence of pain and vomiting makes the diagnosis of intussusception unlikely and other etiologies of the symptoms and signs should be considered. This may help guide the efficient use of laboratory testing and imaging. A prospective multicenter study would be of value.
238 239
Moved to 16
The Accuracy of the Yale Observation Scale Score and Unstructured Clinician Suspicion to Identify Febrile Infants Aged £60 Days With Serious Bacterial Infections
Nigrovic LE, Mahajan PV, Tzimenatos L, Alpern ER, Rogers AJ, Simmons T, Casper C, Ramilo O, Kuppermann N/Boston Children’s Hospital, Boston, MA; Children’s Hospital of Michigan, Detroit, MI; University of California, Davis, Sacramento, CA; Lurie Children’s Hospital of Chicago, Chicago, IL; University of Michigan, Ann Arbor, MI; Pediatric Emergency Care Applied Research Network (PECARN) Data Coordinating Center, Salt Lake, UT; Nationwide Children’s Hospital, Columbus, OH
Study Objectives: Both the Yale Observation Scale (YOS) score and unstructured clinician suspicion have been used to predict the risk of serious bacterial infections (SBIs) in young febrile infants. However, the number of infants 60 days studied has been limited. Our objective was to assess the ability of the YOS score and unstructured clinician suspicion to identify febrile infants 60 days with SBIs. Methods: We performed a planned secondary analysis of a prospective cohort study of non-critical febrile infants 60 days presenting to one of 26 emergency departments (EDs) in PECARN who had blood cultures obtained. We defined a SBI as urinary tract infection (UTI), bacteremia, or bacterial meningitis and only included infants whose SBI status was known. ED faculty and fellow clinicians applied the YOS score and also estimated the risk of SBI ( 50%) using unstructured clinician suspicion. We used previously described cut-points for the YOS (“perfect” score of 6 and “normal” score of 10). We then compared the performance of the dichotomized YOS scores to unstructured clinician suspicion ( 1% and > 5%) for the prediction of SBIs. Results: We enrolled 4778 patients, of whom 4592 (96%) had known SBI status. Of these 4592 infants, 445 (9.7%) had SBIs of whom 348 (78% of SBIs) had UTIs. 1467 (32%) were 28 days of age. The median YOS score was no different in infants with SBIs than those without (6 vs 6; P ¼ .55). Performance of the YOS score and clinician suspicion for SBI is given in the table below. Of the 24 infants with bacterial meningitis, the YOS scores ranged from 6 to 26, with 2 (38%) having a scores of 6 and 14 (58%) having a scores 10. Clinician suspicion varied from 50% with 2 (8%) having risk assessments of < 1% and 10 (42%) having risk assessments of 5%. Conclusion: Neither the YOS score nor unstructured clinician suspicion could reliably exclude SBIs in febrile infants 60 days, including meningitis, in febrile infants 60 days of age. An accurate clinical prediction rule to identify infants at low and high risk for SBIs is needed.
Volume 66, no. 4s : October 2015