Diagnostic accuracy of left ventricular function for identifying sepsis among emergency department patients with nontraumatic symptomatic undifferentiated hypotension

Alan E. Jones, Patrick A. Craddock, Vivek S. Tayal, Jeffrey A. Kline

Research output: Contribution to journalArticle

81 Citations (Scopus)

Abstract

The hypothesis of this study states that in emergency department (ED) patients with nontraumatic symptomatic hypotension, the presence of hyperdynamic left ventricular function (LVF) is specific for sepsis as the etiology of shock. We performed a secondary analysis of patients with nontraumatic symptomatic hypotension enrolled in a randomized, clinical diagnostic trial. The study was done in an urban tertiary ED with a census over 100,000 visits per year. Inclusion criteria were nontrauma ED patients aged >17 years, initial vital signs consistent with shock (systolic blood pressure <100 mm Hg or shock index >1.0), and agreement of two independent observers for one sign and symptom of circulatory shock. All patients underwent focused ED echocardiography (echo) during initial resuscitation. Echos were reviewed post-hoc by a blinded physician and categorized by qualitative LVF as hyperdynamic (ejection fraction [EF] >55%), normal to moderate impairment (EF 30%-55%), and severe impairment (EF <30%). Main outcome was the criterion standard diagnosis of septic shock. Analyses include the diagnostic performance of LVF, Cohen's κ for interobserver agreement of LVF, and logistic regression for independent predictors of sepsis. There were 103 echos that were adequate for analysis. The mean age was 57 ± 16.7 years, 59% were male, and the mean initial systolic blood pressure was 83 ± 11.3 mm Hg. A final diagnosis of septic shock was made in 38% (39/103) of patients. Seventeen of 103 (17%) patients had hyperdynamic LVF with an interobserver agreement of κ = 0.8. The sensitivity and specificity of hyperdynamic LVF for predicting sepsis were 33% (95% CI 19%-50%) and 94% (85%-98%), respectively. Hyperdynamic LVF had a positive likelihood ratio of 5.3 for the diagnosis of sepsis and was a strong independent predictor of sepsis as the final diagnosis with an odds ratio of 5.5 (95% CI 1.1-45). Among ED patients with nontraumatic undifferentiated symptomatic hypotension, the presence of hyperdynamic LVF on focused echo is highly specific for sepsis as the etiology of shock.

Original languageEnglish (US)
Pages (from-to)513-517
Number of pages5
JournalShock
Volume24
Issue number6
DOIs
StatePublished - Jan 2006

Fingerprint

Left Ventricular Function
Hypotension
Hospital Emergency Service
Sepsis
Shock
Blood Pressure
Vital Signs
Censuses
Resuscitation
Signs and Symptoms
Echocardiography
Randomized Controlled Trials
Physicians

Keywords

  • Hypotension
  • Sensitivity
  • Sepsis
  • Shock
  • Specificity
  • Ultrasound

ASJC Scopus subject areas

  • Emergency Medicine
  • Critical Care and Intensive Care Medicine

Cite this

Diagnostic accuracy of left ventricular function for identifying sepsis among emergency department patients with nontraumatic symptomatic undifferentiated hypotension. / Jones, Alan E.; Craddock, Patrick A.; Tayal, Vivek S.; Kline, Jeffrey A.

In: Shock, Vol. 24, No. 6, 01.2006, p. 513-517.

Research output: Contribution to journalArticle

@article{fadb254d189741fa96a500eeab988cde,
title = "Diagnostic accuracy of left ventricular function for identifying sepsis among emergency department patients with nontraumatic symptomatic undifferentiated hypotension",
abstract = "The hypothesis of this study states that in emergency department (ED) patients with nontraumatic symptomatic hypotension, the presence of hyperdynamic left ventricular function (LVF) is specific for sepsis as the etiology of shock. We performed a secondary analysis of patients with nontraumatic symptomatic hypotension enrolled in a randomized, clinical diagnostic trial. The study was done in an urban tertiary ED with a census over 100,000 visits per year. Inclusion criteria were nontrauma ED patients aged >17 years, initial vital signs consistent with shock (systolic blood pressure <100 mm Hg or shock index >1.0), and agreement of two independent observers for one sign and symptom of circulatory shock. All patients underwent focused ED echocardiography (echo) during initial resuscitation. Echos were reviewed post-hoc by a blinded physician and categorized by qualitative LVF as hyperdynamic (ejection fraction [EF] >55{\%}), normal to moderate impairment (EF 30{\%}-55{\%}), and severe impairment (EF <30{\%}). Main outcome was the criterion standard diagnosis of septic shock. Analyses include the diagnostic performance of LVF, Cohen's κ for interobserver agreement of LVF, and logistic regression for independent predictors of sepsis. There were 103 echos that were adequate for analysis. The mean age was 57 ± 16.7 years, 59{\%} were male, and the mean initial systolic blood pressure was 83 ± 11.3 mm Hg. A final diagnosis of septic shock was made in 38{\%} (39/103) of patients. Seventeen of 103 (17{\%}) patients had hyperdynamic LVF with an interobserver agreement of κ = 0.8. The sensitivity and specificity of hyperdynamic LVF for predicting sepsis were 33{\%} (95{\%} CI 19{\%}-50{\%}) and 94{\%} (85{\%}-98{\%}), respectively. Hyperdynamic LVF had a positive likelihood ratio of 5.3 for the diagnosis of sepsis and was a strong independent predictor of sepsis as the final diagnosis with an odds ratio of 5.5 (95{\%} CI 1.1-45). Among ED patients with nontraumatic undifferentiated symptomatic hypotension, the presence of hyperdynamic LVF on focused echo is highly specific for sepsis as the etiology of shock.",
keywords = "Hypotension, Sensitivity, Sepsis, Shock, Specificity, Ultrasound",
author = "Jones, {Alan E.} and Craddock, {Patrick A.} and Tayal, {Vivek S.} and Kline, {Jeffrey A.}",
year = "2006",
month = "1",
doi = "10.1097/01.shk.0000186931.02852.5f",
language = "English (US)",
volume = "24",
pages = "513--517",
journal = "Shock",
issn = "1073-2322",
publisher = "Lippincott Williams and Wilkins",
number = "6",

}

TY - JOUR

T1 - Diagnostic accuracy of left ventricular function for identifying sepsis among emergency department patients with nontraumatic symptomatic undifferentiated hypotension

AU - Jones, Alan E.

AU - Craddock, Patrick A.

AU - Tayal, Vivek S.

AU - Kline, Jeffrey A.

PY - 2006/1

Y1 - 2006/1

N2 - The hypothesis of this study states that in emergency department (ED) patients with nontraumatic symptomatic hypotension, the presence of hyperdynamic left ventricular function (LVF) is specific for sepsis as the etiology of shock. We performed a secondary analysis of patients with nontraumatic symptomatic hypotension enrolled in a randomized, clinical diagnostic trial. The study was done in an urban tertiary ED with a census over 100,000 visits per year. Inclusion criteria were nontrauma ED patients aged >17 years, initial vital signs consistent with shock (systolic blood pressure <100 mm Hg or shock index >1.0), and agreement of two independent observers for one sign and symptom of circulatory shock. All patients underwent focused ED echocardiography (echo) during initial resuscitation. Echos were reviewed post-hoc by a blinded physician and categorized by qualitative LVF as hyperdynamic (ejection fraction [EF] >55%), normal to moderate impairment (EF 30%-55%), and severe impairment (EF <30%). Main outcome was the criterion standard diagnosis of septic shock. Analyses include the diagnostic performance of LVF, Cohen's κ for interobserver agreement of LVF, and logistic regression for independent predictors of sepsis. There were 103 echos that were adequate for analysis. The mean age was 57 ± 16.7 years, 59% were male, and the mean initial systolic blood pressure was 83 ± 11.3 mm Hg. A final diagnosis of septic shock was made in 38% (39/103) of patients. Seventeen of 103 (17%) patients had hyperdynamic LVF with an interobserver agreement of κ = 0.8. The sensitivity and specificity of hyperdynamic LVF for predicting sepsis were 33% (95% CI 19%-50%) and 94% (85%-98%), respectively. Hyperdynamic LVF had a positive likelihood ratio of 5.3 for the diagnosis of sepsis and was a strong independent predictor of sepsis as the final diagnosis with an odds ratio of 5.5 (95% CI 1.1-45). Among ED patients with nontraumatic undifferentiated symptomatic hypotension, the presence of hyperdynamic LVF on focused echo is highly specific for sepsis as the etiology of shock.

AB - The hypothesis of this study states that in emergency department (ED) patients with nontraumatic symptomatic hypotension, the presence of hyperdynamic left ventricular function (LVF) is specific for sepsis as the etiology of shock. We performed a secondary analysis of patients with nontraumatic symptomatic hypotension enrolled in a randomized, clinical diagnostic trial. The study was done in an urban tertiary ED with a census over 100,000 visits per year. Inclusion criteria were nontrauma ED patients aged >17 years, initial vital signs consistent with shock (systolic blood pressure <100 mm Hg or shock index >1.0), and agreement of two independent observers for one sign and symptom of circulatory shock. All patients underwent focused ED echocardiography (echo) during initial resuscitation. Echos were reviewed post-hoc by a blinded physician and categorized by qualitative LVF as hyperdynamic (ejection fraction [EF] >55%), normal to moderate impairment (EF 30%-55%), and severe impairment (EF <30%). Main outcome was the criterion standard diagnosis of septic shock. Analyses include the diagnostic performance of LVF, Cohen's κ for interobserver agreement of LVF, and logistic regression for independent predictors of sepsis. There were 103 echos that were adequate for analysis. The mean age was 57 ± 16.7 years, 59% were male, and the mean initial systolic blood pressure was 83 ± 11.3 mm Hg. A final diagnosis of septic shock was made in 38% (39/103) of patients. Seventeen of 103 (17%) patients had hyperdynamic LVF with an interobserver agreement of κ = 0.8. The sensitivity and specificity of hyperdynamic LVF for predicting sepsis were 33% (95% CI 19%-50%) and 94% (85%-98%), respectively. Hyperdynamic LVF had a positive likelihood ratio of 5.3 for the diagnosis of sepsis and was a strong independent predictor of sepsis as the final diagnosis with an odds ratio of 5.5 (95% CI 1.1-45). Among ED patients with nontraumatic undifferentiated symptomatic hypotension, the presence of hyperdynamic LVF on focused echo is highly specific for sepsis as the etiology of shock.

KW - Hypotension

KW - Sensitivity

KW - Sepsis

KW - Shock

KW - Specificity

KW - Ultrasound

UR - http://www.scopus.com/inward/record.url?scp=33646773996&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33646773996&partnerID=8YFLogxK

U2 - 10.1097/01.shk.0000186931.02852.5f

DO - 10.1097/01.shk.0000186931.02852.5f

M3 - Article

C2 - 16317380

AN - SCOPUS:33646773996

VL - 24

SP - 513

EP - 517

JO - Shock

JF - Shock

SN - 1073-2322

IS - 6

ER -