- Email: [email protected]
The Early Predictive Ability of Myoglobin versus Troponin in Patients Presenting to the Emergency Department With Suspected Acute Coronary Syndrome
Research Forum Abstracts Table 1. Group analysis of hemoglobin level derived by point of care HemoCue and laboratory CBC hemoglobin measurement. Group
# Patients
# Visits
HemoCue
CBC Hemoglobin
Difference
95% CI
30 77
41 612
8.559 (SD 2.023) 9.187 (SD 2.018)
8.707 (SD 1.866) 9.185 (SD 1.823)
0.148 (p¼0.730) 0.002 (p¼0.990)
1.004, 0.707 0.214, 0.217
Low-User (1 or less visit/year) High-User (>1 visit/year)
hemoglobin measurements is 0.008, and it is not significant with p¼0.9394. The 95% confidence interval of the mean difference is (0.217, 0.201). There were 30 patients in the low-user group and 77 patients in the high-user group. The HemoCue and CBC hemoglobin values are shown in the Table. There is no significant difference detected in either group between the hemoglobin level determined by HemoCue or CBC hemoglobin measurement. Conclusions: Hemoglobin measurements obtained from point-of-care HemoCue testing are highly comparable to hemoglobin values obtained from a CBC in the setting of sickle cell disease vaso-occlusive crisis.
104 105
Withdrawn Time to Treatment and Mortality in Acute Promyelocytic Leukemia
Haq SM, Willis Merriman K, Todd K/MD Anderson Cancer Center, Houston, TX
Background: Acute promyelocytic leukemia (APL), a subtype of acute myelogenous leukemia, is uniformly fatal if untreated. However, advances in treatment have markedly improved outcomes for this form of leukemia. Effective therapy is based on prompt recognition of the disease and urgent administration of all-trans retinoic acid (ATRA), a differentiating agent which leads to apoptosis of the malignant cells. National guidelines recommend immediate initiation of ATRA based treatment prior to confirmatory genetic testing. However, evidence is lacking as to whether APL treatment with ATRA is truly time sensitive. Study Objective: To evaluate the effect of time to ATRA treatment on 30-day mortality of patients with APL. Methods: All APL patients over a 5-year period, from 2007 to 2011, presenting to the emergency department (ED) at MD Anderson Cancer Center were identified using the International Classification of Diseases of Oncology histology code 98663. Patients with prior APL treatment or prior malignancies were excluded. Each patient’s chart was reviewed for administration of ATRA. Times from symptom onset to ED triage and the first dose of ATRA were recorded. Patient charts were also reviewed for age at diagnosis, sex, year of diagnosis, white blood cell count, platelet count, and coagulation times. 30 day mortality was recorded for each patient. Results: Eighty patients with a new diagnosis of APL as the first malignancy presented to the MD Anderson ED between 2007 and 2011, and all patients underwent treatment with ATRA. Four were excluded from analysis as time of symptoms onset was not available. The average patient age at diagnosis was 44 and 45% were male. The average times from onset of symptoms to ED triage and from ED triage to ATRA treatment were 17 days and 16.3 hours, respectively. Eighty-three percent of patients received ATRA treatment within 24 hours of ED presentation. Four patients died within 30 days, 3 from pulmonary hemorrhage and 1 from intracranial hemorrhage. When comparing patients who died within 30 days to the rest of the cohort, patients with early death had symptoms for four additional days. No additional differences were noted between the two groups. Conclusions: Time from symptoms onset to ED presentation was found to be associated with 30-day mortality. Strategies to detect patients at higher risk and education regarding symptoms may be of benefit, although identifying these variables would be difficult.
106
Accuracy of Noninvasive and Invasive Point-of-Care Hemoglobin Measurement in the Emergency Department
Wager A, Thomas AW, Sewatsky TP, Boas S, O‘keefe MF, Freeman K/University of Vermont, Burlington, VT
Study Objectives: Hemoglobin measurement is one of the most common tests performed in the emergency department (ED). Rapid, point-of-care measurement of hemoglobin with a finger sensor to detect light attenuation characteristics, may provide
Volume 62, no. 4s : October 2013
benefits to ED patients if the accuracy of this noninvasive measurement is comparable to invasive testing. The objective of this study was to compare noninvasive hemoglobin measurement with the Pronto-7 (SpHb) and capillary point-of-care measurement with HemoCue to reference values from a laboratory hematology analyzer (Hb). Methods: We studied the performance characteristics of point-of-care hemoglobin measurement in ED patients requiring a complete blood count for standard care. Eligible patients were hemodynamically stable and able to provide informed consent for enrollment prior to phlebotomy. A venous sample, analyzed with a hematology analyzer (Beckman Coulter LH 50, Beckman Coulter Inc.) and two capillary samples from a single finger stick, analyzed with a point-of-care device (Hemocue 201þ, HemoCue Inc.), were obtained from each patient. Additionally, two consecutive, SpHb spot check measurements were taken with an appropriately sized (small, medium or large) reusable sensor (rainbow 4D sensor, rev F) connected to a Pronto-7 (sw version 2317, Masimo Corp.). Blood collection for venous and capillary tests was performed within 15 min of noninvasive testing. The mean of the two values obtained from the Pronto-7 and HemoCue spot check devices were compared to the value from the hematology analyzer (Hb). Bias1 standard deviation (SD) and 95% limits of agreement were calculated for the SpHb and capillary measurement. Results: Of the 71 patients enrolled, 14 were excluded from analysis due to missing values (6 capillary, 5 Hb, 1 SpHb) or protocol deviations (2). Subjects ranged from 18 to 88 yrs and 59% were female. Subjects provided 57 sets of hemoglobin values. Hb values ranged from 7.7 to 16.6 g/dL. BiasSD compared to Hb were 1.11.3 g/dL for SpHb and 1.31.0 for capillary measurement. Limits of agreement were 3.6 to 1.4 g/dL for SpHb and -3.9 to 1.2 g/dL for capillary measurement. There were no significant differences in the bias or the limits of agreement between methods. Conclusions: Noninvasive hemoglobin measurement with the Pronto-7 had similar performance characteristics as capillary HemoCue measurements in ED patients. Noninvasive measurement provides benefits of patient comfort and reduced exposure of health care providers to biohazards such as sharps injury and blood borne pathogens. Our results suggest that the Pronto-7 may provide a clinically acceptable method to diagnose anemia, which may be useful for triage, screening, or continuous hemoglobin monitoring in the ED.
107
The Early Predictive Ability of Myoglobin versus Troponin in Patients Presenting to the Emergency Department With Suspected Acute Coronary Syndrome
Nale J, Lee D, Sud P, Chen A, Ward M, Kintzer E, Boccio E, Medairos R, Bilello L, Wu M/North Shore University Hospital, Manhasset, NY
Background: Myoglobin is a component of cardiac muscle that is released upon injury and has been used as a biomarker for cardiac injury. Older literature reported that myoglobin concentrations may rise prior to other commercially available biomarkers in the setting of an acute myocardial infarction and may be more sensitive in detecting recent cardiac injury. In our institution, we routinely include myoglobin assays along with assays for troponin and creatinine phosphokinase as a panel test as suggested by McCord in 2003. However, with the increasing sensitivity of troponin testing over the past decade, many institutions no longer use myoglobin testing to detect recent cardiac injury. Study Objective: To determine the utility of serum myoglobin levels versus troponin levels in detecting acute myocardial injury in patients presenting to the emergency department (ED) with symptoms suggestive of acute coronary syndrome. Methods: We performed an IRB-approved retrospective chart review of all ED patients who had cardiac biomarker tests performed over a nine-month period in 2012 at a suburban tertiary care hospital (an annual census over 85,000 patients). Subjects were identified through laboratory query of all patients who had successive cardiac biomarkers panels which included serial myoglobin and troponin tests obtained over two consecutive days. We obtained patient demographics, hospital course metrics, and discharge. We identified subjects who had an initial elevated myoglobin result with an initial normal troponin results. We specifically reviewed all charts who had elevated troponin testing within 24-48 hours. We used descriptive statistics to analyze the data.
Annals of Emergency Medicine S41
Research Forum Abstracts Results: During the nine-month period, 240 subjects had two sets of cardiac biomarker measured. In 7 of the 240 cases (2.9%), troponin levels were elevated in subsequent testing and myoglobin levels remained abnormal. None of the 7 had electrocardiograms that triggered immediate thrombolytics or immediate angiography. Only 1 of these cases underwent coronary angiography and was subsequently found to have a significant occlusion (100% occlusion of an obtuse marginal artery). The remaining 6 cases either underwent coronary angiography and no new lesion was found or they were not candidates for catheterization. Conclusions: A previous study by McCord in 2003 found a rate of 16.6% (14 of 84 cases) for patients presenting to the ED that had elevated myoglobin levels without elevated troponin levels and had a final diagnosis of myocardial infarction. Our study demonstrates a rate of 2.9% for patients with a positive myoglobin in the setting of an initially negative troponin. These results suggest improved sensitivity of serum troponin tests for detecting myocardial injury and raise the question of whether serum myoglobin levels should be routinely measured in cases of suspected acute coronary syndrome.
108
Subarachnoid Hemorrhage in the Emergency Department: A Systematic Review and Meta-Analysis of Diagnostic Accuracy in History, Physical Exam Findings, and Testing
Hussain A, Carpenter CR, Sivilotti M, Pines JM/George Washington University, Washington, DC; Washington University at St. Louis, St. Louis, MO; Queens University, Kingston, ON, Canada
Study Objectives: Acute subarachnoid hemorrhage (SAH) is a rare, serious etiology of atraumatic headache (HA) in emergency department (ED) patients. We conducted a meta-analysis of history, physical exam, imaging studies, and lumbar puncture (LP) results for adult ED patients presenting with HA suspicious for SAH. Methods: We conducted a systematic review/meta-analysis of studies reporting data on ED HA patients with suspected SAH where 2x2 tables could be constructed. We searched 1966-2012 in PUBMED, EMBASE, SCOPUS and additional studies from research abstracts in 5 EM and neuro journals. QUADAS-2 was used to assess study quality/bias. When 2 similar studies were identified, meta-analysis was conducted using Meta-DiSc. Outcomes were summary sensitivity, specificity, and positive and negative likelihood ratios (LRþ and LR). Results: In 3,274 citations, 126 underwent full-text review; 16 were included. Across studies, SAH definitions were highly variable, specifically how positive LP was defined. Clinical follow-up varied considerably. In QUADAS-2, study quality was variable; however, most had a low-risk of biases. Conclusions: Non-contrast head CT is highly sensitive for SAH and has perfect specificity. Certain history and physical exam findings are more suggestive including unconsciousness, AMS, focal neuro deficits, nuchal rigidity, blurred vision, and exertion at onset; however none are sufficiently specific to rule out SAH.
109 EMF
Point-of-Care Testing in Setting of Nitromethane and Methanol Co-Ingestion Will Not Mask True Creatinine, Anion Gap, or Osmolar Gap
Cao D, Maynard SM, Mitchell-Smith AM, Kerns WP II, Beuhler MC/Carolinas Medical Center, Charlotte, NC; Indiana University School of Medicine, Indianapolis, IN; Carolinas Poison Center, Charlotte, NC
Study Objectives: Co-ingestion of racing fuel containing nitromethane and methanol can produce significant toxicity. Nitromethane interferes with plasma creatinine measured via the Jaffé reaction by mimicking the spectrographic properties of creatinine picric acid complex. The interference presents diagnostic and management challenges with nitromethane co-ingestion. Point-of-care (POC) testing with the i-Stat Chem8+ measures creatinine by an enzymatic methodology, which we hypothesize will provide an accurate measure of the true creatinine when compared with the Jaffé measurement. We further hypothesize that the anion and osmolar gaps will remain unchanged by either method with nitromethane co-ingestion. Methods: Study enrolled healthy volunteers with predicted normal baseline renal function. The primary investigators were blinded to patient information and sex. Nitromethane was added to whole blood to achieve 5 concentrations (0, 0.25, 0.5, 1, and 2 mmol/L). Samples were prepared for basic metabolic panel (BMP) with Jaffé reaction, POC Chem8+, plasma osmolarity, and gas-chromatography (GC) at each concentration. Remaining blood samples were refrigerated at 2-8 C and analyzed by GC after 7 days for nitromethane. Anion and osmolar gaps were calculated for each patient sample at each concentration. Data collected for Jaffé creatinine, POC creatinine, osmolar gap, anion gap, and nitromethane recovery were analyzed using single factor ANOVA (p < 0.05). Linear regression was used to determine the magnitude of nitromethane interference on apparent creatinine. GC data was used to verify the proportional recovery of nitromethane in each patient sample and to estimate degradation of nitromethane under standard storage conditions. Results: Two male and two female samples were used for analysis. All participants had creatinine within the normal range for analysis method with mean of 0.9 mg/dL by POC analysis and 0.76 mg/dL by Jaffé reaction. Jaffé creatinine showed significant difference across nitromethane concentrations (p ¼ <0.0001). The mean apparent creatinine was a linear function of the nitromethane concentration (R2 ¼ 1): Apparent creatinine (mg/dL) ¼ 7.1 * nitromethane (mmol/L) þ 0.79. No significant difference was detected for POC creatinine across nitromethane concentrations (p ¼ 0.99). Similarly, no significant difference was detected for anion gap using POC (p ¼ 0.3), anion gap using BMP (p ¼ 0.31), osmolar gap using POC (p ¼ 0.52), and osmolar gap using BMP (p ¼ 0.36). Nitromethane peaks were detected at all concentrations using GC on day 0. Detection of nitromethane at the 0.25 mmol/L concentration was inconsistent on day 7 and was not recovered in all patient samples. Linear recovery of nitromethane was detected on day 0 (R2 ¼ 0.9996) and day 7 (R2 ¼ 0.9979). Nitromethane degradation over 7 days was most pronounced at the 2 mmol/L concentrations with a mean 81% recovery.
Table. Meta-Analysis of Diagnostic Accuracy. Test/Finding
# Studies
Non-contrast CT LP Xanthochromia Nausea Vomiting Photophobia Stiff Neck Blurred Vision Exertion at Onset Exploding Onset Instantaneous Onset 2-60 seconds Onset 1-5 minutes Female Nuchal Rigidity Focal Neuro Deficit Altered Mental Status (AMS) Unconscious
S42 Annals of Emergency Medicine
4 4 5 2 4 4 2 2 2 2 2 2 3 2 3 4 2
Sensitivity (95% CI) 0.94 0.74 0.60 0.61 0.16 0.66 0.11 0.30 0.29 0.63 0.23 0.07 0.57 0.30 0.37 0.25 0.17
(0.91-0.96) (0.57-0.88) (0.52-0.68) (0.53-0.68) (0.10-0.24) (0.58-0.73) (0.02- 0.28) (0.23-0.37) (0.19-0.42) (0.53-0.72) (0.16-0.32) (0.03-0.13) (0.50-0.63) (0.22-0.38) (0.25-0.50) (0.16-0.35) (0.11-0.24)
Specificity (95% CI) 1.00 0.81 0.55 0.73 0.79 0.70 0.96 0.89 0.66 0.44 0.73 0.86 0.41 0.94 0.91 0.91 0.96
(1.00-1.00) (0.79-0.84) (0.51-0.59) (0.71-0.75) (0.75 to 0.82) (0.68-0.72) (0.93 to 0.98) (0.87-0.90) (0.58-0.73) (0.40-0.49) (0.68-0.77) (0.82-0.89) (0.39-0.43) (0.93 0.95) (0.86-0.94) (0.87-0.94) (0.95-0.96)
LR+ (95% CI) 6.64 1.22 1.64 1.08 2.55 3.65 2.03 1.28 1.12 1.08 0.26 1.03 2.53 3.13 2.22 3.78
N/A (1.63-27.06) (0.91-1.64) (0.89-3.04) (0.52-2.24) (1.54-4.22) (0.24-54.75) (1.53-2.68) (0.89-1.83) (0.81-1.54) (0.47-2.47) (0.01-6.26) (0.83-1.26) (0.24-26.68) (1.63-6.02) (1.35-3.65) (2.51-5.71)
LR- (95% CI) 0.08 0.36 0.85 0.58 1.03 0.54 0.84 0.83 0.92 0.86 1.00 1.13 1.00 0.88 0.75 0.87 0.87
(0.03-0.23) (0.16-0.83) (0.48-1.51) (0.48-0.71) (0.91-1.16) (0.36-0.80) (0.42-1.66) (0.70-0.98) (0.67-1.25) (0.52-1.42) (0.78-1.24) (1.08-1.18) (0.82-1.22) (0.58-1.32) (0.61-0.94) (0.78-0.98) (0.81-0.93)
Volume 62, no. 4s : October 2013
# Patients
# Visits
HemoCue
CBC Hemoglobin
Difference
95% CI
30 77
41 612
8.559 (SD 2.023) 9.187 (SD 2.018)
8.707 (SD 1.866) 9.185 (SD 1.823)
0.148 (p¼0.730) 0.002 (p¼0.990)
1.004, 0.707 0.214, 0.217
Low-User (1 or less visit/year) High-User (>1 visit/year)
hemoglobin measurements is 0.008, and it is not significant with p¼0.9394. The 95% confidence interval of the mean difference is (0.217, 0.201). There were 30 patients in the low-user group and 77 patients in the high-user group. The HemoCue and CBC hemoglobin values are shown in the Table. There is no significant difference detected in either group between the hemoglobin level determined by HemoCue or CBC hemoglobin measurement. Conclusions: Hemoglobin measurements obtained from point-of-care HemoCue testing are highly comparable to hemoglobin values obtained from a CBC in the setting of sickle cell disease vaso-occlusive crisis.
104 105
Withdrawn Time to Treatment and Mortality in Acute Promyelocytic Leukemia
Haq SM, Willis Merriman K, Todd K/MD Anderson Cancer Center, Houston, TX
Background: Acute promyelocytic leukemia (APL), a subtype of acute myelogenous leukemia, is uniformly fatal if untreated. However, advances in treatment have markedly improved outcomes for this form of leukemia. Effective therapy is based on prompt recognition of the disease and urgent administration of all-trans retinoic acid (ATRA), a differentiating agent which leads to apoptosis of the malignant cells. National guidelines recommend immediate initiation of ATRA based treatment prior to confirmatory genetic testing. However, evidence is lacking as to whether APL treatment with ATRA is truly time sensitive. Study Objective: To evaluate the effect of time to ATRA treatment on 30-day mortality of patients with APL. Methods: All APL patients over a 5-year period, from 2007 to 2011, presenting to the emergency department (ED) at MD Anderson Cancer Center were identified using the International Classification of Diseases of Oncology histology code 98663. Patients with prior APL treatment or prior malignancies were excluded. Each patient’s chart was reviewed for administration of ATRA. Times from symptom onset to ED triage and the first dose of ATRA were recorded. Patient charts were also reviewed for age at diagnosis, sex, year of diagnosis, white blood cell count, platelet count, and coagulation times. 30 day mortality was recorded for each patient. Results: Eighty patients with a new diagnosis of APL as the first malignancy presented to the MD Anderson ED between 2007 and 2011, and all patients underwent treatment with ATRA. Four were excluded from analysis as time of symptoms onset was not available. The average patient age at diagnosis was 44 and 45% were male. The average times from onset of symptoms to ED triage and from ED triage to ATRA treatment were 17 days and 16.3 hours, respectively. Eighty-three percent of patients received ATRA treatment within 24 hours of ED presentation. Four patients died within 30 days, 3 from pulmonary hemorrhage and 1 from intracranial hemorrhage. When comparing patients who died within 30 days to the rest of the cohort, patients with early death had symptoms for four additional days. No additional differences were noted between the two groups. Conclusions: Time from symptoms onset to ED presentation was found to be associated with 30-day mortality. Strategies to detect patients at higher risk and education regarding symptoms may be of benefit, although identifying these variables would be difficult.
106
Accuracy of Noninvasive and Invasive Point-of-Care Hemoglobin Measurement in the Emergency Department
Wager A, Thomas AW, Sewatsky TP, Boas S, O‘keefe MF, Freeman K/University of Vermont, Burlington, VT
Study Objectives: Hemoglobin measurement is one of the most common tests performed in the emergency department (ED). Rapid, point-of-care measurement of hemoglobin with a finger sensor to detect light attenuation characteristics, may provide
Volume 62, no. 4s : October 2013
benefits to ED patients if the accuracy of this noninvasive measurement is comparable to invasive testing. The objective of this study was to compare noninvasive hemoglobin measurement with the Pronto-7 (SpHb) and capillary point-of-care measurement with HemoCue to reference values from a laboratory hematology analyzer (Hb). Methods: We studied the performance characteristics of point-of-care hemoglobin measurement in ED patients requiring a complete blood count for standard care. Eligible patients were hemodynamically stable and able to provide informed consent for enrollment prior to phlebotomy. A venous sample, analyzed with a hematology analyzer (Beckman Coulter LH 50, Beckman Coulter Inc.) and two capillary samples from a single finger stick, analyzed with a point-of-care device (Hemocue 201þ, HemoCue Inc.), were obtained from each patient. Additionally, two consecutive, SpHb spot check measurements were taken with an appropriately sized (small, medium or large) reusable sensor (rainbow 4D sensor, rev F) connected to a Pronto-7 (sw version 2317, Masimo Corp.). Blood collection for venous and capillary tests was performed within 15 min of noninvasive testing. The mean of the two values obtained from the Pronto-7 and HemoCue spot check devices were compared to the value from the hematology analyzer (Hb). Bias1 standard deviation (SD) and 95% limits of agreement were calculated for the SpHb and capillary measurement. Results: Of the 71 patients enrolled, 14 were excluded from analysis due to missing values (6 capillary, 5 Hb, 1 SpHb) or protocol deviations (2). Subjects ranged from 18 to 88 yrs and 59% were female. Subjects provided 57 sets of hemoglobin values. Hb values ranged from 7.7 to 16.6 g/dL. BiasSD compared to Hb were 1.11.3 g/dL for SpHb and 1.31.0 for capillary measurement. Limits of agreement were 3.6 to 1.4 g/dL for SpHb and -3.9 to 1.2 g/dL for capillary measurement. There were no significant differences in the bias or the limits of agreement between methods. Conclusions: Noninvasive hemoglobin measurement with the Pronto-7 had similar performance characteristics as capillary HemoCue measurements in ED patients. Noninvasive measurement provides benefits of patient comfort and reduced exposure of health care providers to biohazards such as sharps injury and blood borne pathogens. Our results suggest that the Pronto-7 may provide a clinically acceptable method to diagnose anemia, which may be useful for triage, screening, or continuous hemoglobin monitoring in the ED.
107
The Early Predictive Ability of Myoglobin versus Troponin in Patients Presenting to the Emergency Department With Suspected Acute Coronary Syndrome
Nale J, Lee D, Sud P, Chen A, Ward M, Kintzer E, Boccio E, Medairos R, Bilello L, Wu M/North Shore University Hospital, Manhasset, NY
Background: Myoglobin is a component of cardiac muscle that is released upon injury and has been used as a biomarker for cardiac injury. Older literature reported that myoglobin concentrations may rise prior to other commercially available biomarkers in the setting of an acute myocardial infarction and may be more sensitive in detecting recent cardiac injury. In our institution, we routinely include myoglobin assays along with assays for troponin and creatinine phosphokinase as a panel test as suggested by McCord in 2003. However, with the increasing sensitivity of troponin testing over the past decade, many institutions no longer use myoglobin testing to detect recent cardiac injury. Study Objective: To determine the utility of serum myoglobin levels versus troponin levels in detecting acute myocardial injury in patients presenting to the emergency department (ED) with symptoms suggestive of acute coronary syndrome. Methods: We performed an IRB-approved retrospective chart review of all ED patients who had cardiac biomarker tests performed over a nine-month period in 2012 at a suburban tertiary care hospital (an annual census over 85,000 patients). Subjects were identified through laboratory query of all patients who had successive cardiac biomarkers panels which included serial myoglobin and troponin tests obtained over two consecutive days. We obtained patient demographics, hospital course metrics, and discharge. We identified subjects who had an initial elevated myoglobin result with an initial normal troponin results. We specifically reviewed all charts who had elevated troponin testing within 24-48 hours. We used descriptive statistics to analyze the data.
Annals of Emergency Medicine S41
Research Forum Abstracts Results: During the nine-month period, 240 subjects had two sets of cardiac biomarker measured. In 7 of the 240 cases (2.9%), troponin levels were elevated in subsequent testing and myoglobin levels remained abnormal. None of the 7 had electrocardiograms that triggered immediate thrombolytics or immediate angiography. Only 1 of these cases underwent coronary angiography and was subsequently found to have a significant occlusion (100% occlusion of an obtuse marginal artery). The remaining 6 cases either underwent coronary angiography and no new lesion was found or they were not candidates for catheterization. Conclusions: A previous study by McCord in 2003 found a rate of 16.6% (14 of 84 cases) for patients presenting to the ED that had elevated myoglobin levels without elevated troponin levels and had a final diagnosis of myocardial infarction. Our study demonstrates a rate of 2.9% for patients with a positive myoglobin in the setting of an initially negative troponin. These results suggest improved sensitivity of serum troponin tests for detecting myocardial injury and raise the question of whether serum myoglobin levels should be routinely measured in cases of suspected acute coronary syndrome.
108
Subarachnoid Hemorrhage in the Emergency Department: A Systematic Review and Meta-Analysis of Diagnostic Accuracy in History, Physical Exam Findings, and Testing
Hussain A, Carpenter CR, Sivilotti M, Pines JM/George Washington University, Washington, DC; Washington University at St. Louis, St. Louis, MO; Queens University, Kingston, ON, Canada
Study Objectives: Acute subarachnoid hemorrhage (SAH) is a rare, serious etiology of atraumatic headache (HA) in emergency department (ED) patients. We conducted a meta-analysis of history, physical exam, imaging studies, and lumbar puncture (LP) results for adult ED patients presenting with HA suspicious for SAH. Methods: We conducted a systematic review/meta-analysis of studies reporting data on ED HA patients with suspected SAH where 2x2 tables could be constructed. We searched 1966-2012 in PUBMED, EMBASE, SCOPUS and additional studies from research abstracts in 5 EM and neuro journals. QUADAS-2 was used to assess study quality/bias. When 2 similar studies were identified, meta-analysis was conducted using Meta-DiSc. Outcomes were summary sensitivity, specificity, and positive and negative likelihood ratios (LRþ and LR). Results: In 3,274 citations, 126 underwent full-text review; 16 were included. Across studies, SAH definitions were highly variable, specifically how positive LP was defined. Clinical follow-up varied considerably. In QUADAS-2, study quality was variable; however, most had a low-risk of biases. Conclusions: Non-contrast head CT is highly sensitive for SAH and has perfect specificity. Certain history and physical exam findings are more suggestive including unconsciousness, AMS, focal neuro deficits, nuchal rigidity, blurred vision, and exertion at onset; however none are sufficiently specific to rule out SAH.
109 EMF
Point-of-Care Testing in Setting of Nitromethane and Methanol Co-Ingestion Will Not Mask True Creatinine, Anion Gap, or Osmolar Gap
Cao D, Maynard SM, Mitchell-Smith AM, Kerns WP II, Beuhler MC/Carolinas Medical Center, Charlotte, NC; Indiana University School of Medicine, Indianapolis, IN; Carolinas Poison Center, Charlotte, NC
Study Objectives: Co-ingestion of racing fuel containing nitromethane and methanol can produce significant toxicity. Nitromethane interferes with plasma creatinine measured via the Jaffé reaction by mimicking the spectrographic properties of creatinine picric acid complex. The interference presents diagnostic and management challenges with nitromethane co-ingestion. Point-of-care (POC) testing with the i-Stat Chem8+ measures creatinine by an enzymatic methodology, which we hypothesize will provide an accurate measure of the true creatinine when compared with the Jaffé measurement. We further hypothesize that the anion and osmolar gaps will remain unchanged by either method with nitromethane co-ingestion. Methods: Study enrolled healthy volunteers with predicted normal baseline renal function. The primary investigators were blinded to patient information and sex. Nitromethane was added to whole blood to achieve 5 concentrations (0, 0.25, 0.5, 1, and 2 mmol/L). Samples were prepared for basic metabolic panel (BMP) with Jaffé reaction, POC Chem8+, plasma osmolarity, and gas-chromatography (GC) at each concentration. Remaining blood samples were refrigerated at 2-8 C and analyzed by GC after 7 days for nitromethane. Anion and osmolar gaps were calculated for each patient sample at each concentration. Data collected for Jaffé creatinine, POC creatinine, osmolar gap, anion gap, and nitromethane recovery were analyzed using single factor ANOVA (p < 0.05). Linear regression was used to determine the magnitude of nitromethane interference on apparent creatinine. GC data was used to verify the proportional recovery of nitromethane in each patient sample and to estimate degradation of nitromethane under standard storage conditions. Results: Two male and two female samples were used for analysis. All participants had creatinine within the normal range for analysis method with mean of 0.9 mg/dL by POC analysis and 0.76 mg/dL by Jaffé reaction. Jaffé creatinine showed significant difference across nitromethane concentrations (p ¼ <0.0001). The mean apparent creatinine was a linear function of the nitromethane concentration (R2 ¼ 1): Apparent creatinine (mg/dL) ¼ 7.1 * nitromethane (mmol/L) þ 0.79. No significant difference was detected for POC creatinine across nitromethane concentrations (p ¼ 0.99). Similarly, no significant difference was detected for anion gap using POC (p ¼ 0.3), anion gap using BMP (p ¼ 0.31), osmolar gap using POC (p ¼ 0.52), and osmolar gap using BMP (p ¼ 0.36). Nitromethane peaks were detected at all concentrations using GC on day 0. Detection of nitromethane at the 0.25 mmol/L concentration was inconsistent on day 7 and was not recovered in all patient samples. Linear recovery of nitromethane was detected on day 0 (R2 ¼ 0.9996) and day 7 (R2 ¼ 0.9979). Nitromethane degradation over 7 days was most pronounced at the 2 mmol/L concentrations with a mean 81% recovery.
Table. Meta-Analysis of Diagnostic Accuracy. Test/Finding
# Studies
Non-contrast CT LP Xanthochromia Nausea Vomiting Photophobia Stiff Neck Blurred Vision Exertion at Onset Exploding Onset Instantaneous Onset 2-60 seconds Onset 1-5 minutes Female Nuchal Rigidity Focal Neuro Deficit Altered Mental Status (AMS) Unconscious
S42 Annals of Emergency Medicine
4 4 5 2 4 4 2 2 2 2 2 2 3 2 3 4 2
Sensitivity (95% CI) 0.94 0.74 0.60 0.61 0.16 0.66 0.11 0.30 0.29 0.63 0.23 0.07 0.57 0.30 0.37 0.25 0.17
(0.91-0.96) (0.57-0.88) (0.52-0.68) (0.53-0.68) (0.10-0.24) (0.58-0.73) (0.02- 0.28) (0.23-0.37) (0.19-0.42) (0.53-0.72) (0.16-0.32) (0.03-0.13) (0.50-0.63) (0.22-0.38) (0.25-0.50) (0.16-0.35) (0.11-0.24)
Specificity (95% CI) 1.00 0.81 0.55 0.73 0.79 0.70 0.96 0.89 0.66 0.44 0.73 0.86 0.41 0.94 0.91 0.91 0.96
(1.00-1.00) (0.79-0.84) (0.51-0.59) (0.71-0.75) (0.75 to 0.82) (0.68-0.72) (0.93 to 0.98) (0.87-0.90) (0.58-0.73) (0.40-0.49) (0.68-0.77) (0.82-0.89) (0.39-0.43) (0.93 0.95) (0.86-0.94) (0.87-0.94) (0.95-0.96)
LR+ (95% CI) 6.64 1.22 1.64 1.08 2.55 3.65 2.03 1.28 1.12 1.08 0.26 1.03 2.53 3.13 2.22 3.78
N/A (1.63-27.06) (0.91-1.64) (0.89-3.04) (0.52-2.24) (1.54-4.22) (0.24-54.75) (1.53-2.68) (0.89-1.83) (0.81-1.54) (0.47-2.47) (0.01-6.26) (0.83-1.26) (0.24-26.68) (1.63-6.02) (1.35-3.65) (2.51-5.71)
LR- (95% CI) 0.08 0.36 0.85 0.58 1.03 0.54 0.84 0.83 0.92 0.86 1.00 1.13 1.00 0.88 0.75 0.87 0.87
(0.03-0.23) (0.16-0.83) (0.48-1.51) (0.48-0.71) (0.91-1.16) (0.36-0.80) (0.42-1.66) (0.70-0.98) (0.67-1.25) (0.52-1.42) (0.78-1.24) (1.08-1.18) (0.82-1.22) (0.58-1.32) (0.61-0.94) (0.78-0.98) (0.81-0.93)
Volume 62, no. 4s : October 2013