Derivation and validation of a machine learning record linkage algorithm between emergency medical services and the emergency department.

OBJECTIVE: Linking emergency medical services (EMS) electronic patient care reports (ePCRs) to emergency department (ED) records can provide clinicians access to vital information that can alter management. It can also create rich databases for research and quality improvement. Unfortunately, previous attempts at ePCR and ED record linkage have had limited success. In this study, we use supervised machine learning to derive and validate an automated record linkage algorithm between EMS ePCRs and ED records. MATERIALS AND METHODS: All consecutive ePCRs from a single EMS provider between June 2013 and June 2015 were included. A primary reviewer matched ePCRs to a list of ED patients to create a gold standard. Age, gender, last name, first name, social security number, and date of birth were extracted. Data were randomly split into 80% training and 20% test datasets. We derived missing indicators, identical indicators, edit distances, and percent differences. A multivariate logistic regression model was trained using 5-fold cross-validation, using label k-fold, L2 regularization, and class reweighting. RESULTS: A total of 14 032 ePCRs were included in the study. Interrater reliability between the primary and secondary reviewer had a kappa of 0.9. The algorithm had a sensitivity of 99.4%, a positive predictive value of 99.9%, and an area under the receiver-operating characteristic curve of 0.99 in both the training and test datasets. Date-of-birth match had the highest odds ratio of 16.9, followed by last name match (10.6). Social security number match had an odds ratio of 3.8. CONCLUSIONS: We were able to successfully derive and validate a record linkage algorithm from a single EMS ePCR provider to our hospital EMR.

Effect of Intravenous Versus Intraosseous Access in Prehospital Cardiac Arrest.

OBJECTIVE: The prevailing standard of care in prehospital emergency medical services (EMS) is that either intravenous (IV) or intraosseous (IO) access is an acceptable route for obtaining vascular access and delivery of resuscitation medications and volume expanders in cardiac arrest patients. The aim of this study was to evaluate the effectiveness of IV access versus IO access in terms of return of spontaneous circulation (ROSC) for patients suffering from cardiac arrest. METHODS: A retrospective chart review examining cardiac arrest data with a single advanced life support EMS agency over a 4-year period was performed. Cardiac arrest patients were identified from a quality assurance database. Exclusion criteria included trauma arrest, pediatrics, pregnancy, and obvious signs of death. RESULTS: A total of 795 patients remained after applying the exclusion criteria. A total of 183 (45.1%) out of 406 cardiac arrest patients achieved ROSC who had an IV placed. A total of 389 cardiac arrest patients had an IO placed with ROSC in 100 (25.7%). CONCLUSIONS: Higher ROSC rates were achieved with IV access versus IO access. Limitations include the small sample size, a single EMS agency, and the retrospective nature of the study. Future studies should further evaluate the effectiveness of IO versus IV access in cardiac arrest and other low perfusion states.

The Absence of Fever Is Associated With Higher Mortality and Decreased Antibiotic and IV Fluid Administration in Emergency Department Patients With Suspected Septic Shock.

OBJECTIVE: This study evaluates whether emergency department septic shock patients without a fever (reported or measured) receive less IV fluids, have decreased antibiotic administration, and suffer increased in-hospital mortality. DESIGN: This was a secondary analysis of a prospective, observational study of patients with shock. SETTING: The study was conducted in an urban, academic emergency department. PATIENTS: The original study enrolled consecutive adult (aged 18 yr or older) emergency department patients from November 11, 2012, to September 23, 2013, who met one of the following shock criteria: 1) systolic blood pressure less than 90 mm Hg after at least 1L IV fluids, 2) new vasopressor requirement, or 3) systolic blood pressure less than 90 mm Hg and IV fluids held for concern of fluid overload. The current study is limited to patients with septic shock. Patients were grouped as febrile if they had a subjective fever or a measured temperature >100.4°F documented in the emergency department; afebrile patients lacked both. MEASUREMENTS AND MAIN RESULTS: Among 378 patients with septic shock, 207 of 378 (55%; 50-60%) were febrile by history or measurement. Afebrile patients had lower rates of antibiotic administration in the emergency department (81% vs 94%; p < 0.01), lower mean volumes of IV fluids (2,607 vs 3,013 mL; p < 0.01), and higher in-hospital mortality rates (33% vs 11%; p < 0.01). After adjusting for bicarbonate less than 20 mEq/L, lactate concentration, respiratory rate greater than or equal to 24 breaths/min, emergency department antibiotics, and emergency department IV fluids volume, being afebrile remained a significant predictor of in-hospital mortality (odds ratio, 4.3; 95% CI, 2.2-8.2; area under the curve = 0.83). CONCLUSIONS: In emergency department patients with septic shock, afebrile patients received lower rates of emergency department antibiotic administration, lower mean IV fluids volume, and suffered higher in-hospital mortality.

Assessing The Predictive Value of Clinical Factors Used to Determine The Presence of Sepsis Causing Shock in the Emergency Department.

INTRODUCTION: Differentiating shock etiologies is a challenging task in the Emergency Department (ED); even the strongest clinical predictors leave some diagnostic uncertainty. This study sought to establish an evidence base for using clinical covariates in the diagnostic evaluation of septic shock. METHODS: We conducted a prospective, observational study of consecutive ED patients with shock from November 11, 2012 to September 23, 2013. We included all patients at least 18 years old with shock, defined as new vasopressor requirement, systolic blood pressure less than 90 mmHg after at least 1 L of crystalloid or 2 units packed red blood cells, or systolic blood pressure less than 90 mmHg and fluids withheld due to concern for fluid overload. Multivariate logistic regression and recursive partitioning models were constructed, predicting septic cause of shock. The logistic regression model was derived using first 500 patients, and validated with the subsequent 200 patients. RESULTS: In the derivation cohort, 55.6% (95% confidence interval: 51.2%-60.0%) were septic, and 20.8% (17.2%-24.4%) died during hospitalization. The multivariate model (derivation area under the curve = 0.88, validation area under the curve = 0.89) identified predictors of septic shock, including temperature more than 100.4°F (odds ratio 4.6, 2.3-9.2) and history of fever (odds ratio 9.2, 4.4-19.2); however, only 153 of 277 (55.3%, 49.5%-61.2%) patients with septic shock had either of these findings. In the recursive partitioning model, if all predictors were absent, the probability of sepsis causing shock was 21% (16.6%-25.6%). CONCLUSIONS: Clinical data can predict the presence of sepsis causing shock in the ED in most patients. The remaining diagnostic uncertainty provides an opportunity for adding novel diagnostic testing.

Derivation and Validation of Predictive Factors for Clinical Deterioration after Admission in Emergency Department Patients Presenting with Abnormal Vital Signs Without Shock.

INTRODUCTION: Strategies to identify high-risk emergency department (ED) patients often use markedly abnormal vital signs and serum lactate levels. Risk stratifying such patients without using the presence of shock is challenging. The objective of the study is to identify independent predictors of in-hospital adverse outcomes in ED patients with abnormal vital signs or lactate levels, but who are not in shock. METHODS: We performed a prospective observational study of patients with abnormal vital signs or lactate level defined as heart rate ≥130 beats/min, respiratory rate ≥24 breaths/min, shock index ≥1, systolic blood pressure <90 mm/Hg, or lactate ≥4 mmole/L. We excluded patients with isolated atrial tachycardia, seizure, intoxication, psychiatric agitation, or tachycardia due to pain (ie: extremity fracture). The primary outcome was deterioration, defined as development of acute renal failure (creatinine 2× baseline), non-elective intubation, vasopressor requirement, or mortality. Independent predictors of deterioration after hospitalization were determined using logistic regression. RESULTS: Of 1,152 consecutive patients identified with abnormal vital signs or lactate level, 620 were excluded, leaving 532 for analysis. Of these, 53/532 (9.9±2.5%) deteriorated after hospital admission. Independent predictors of in-hospital deterioration were: lactate >4.0 mmol/L (OR 5.1, 95% CI [2.1-12.2]), age ≥80 yrs (OR 1.9, CI [1.0-3.7]), bicarbonate <21 mEq/L (OR 2.5, CI [1.3-4.9]), and initial HR≥130 (OR 3.1, CI [1.5-6.1]). CONCLUSION: Patients exhibiting abnormal vital signs or elevated lactate levels without shock had significant rates of deterioration after hospitalization. ED clinical data predicted patients who suffered adverse outcomes with reasonable reliability.

Incidence and recognition of elevated triage blood pressure in the pediatric emergency department.

OBJECTIVES: This study aimed to determine the incidence of elevated triage blood pressure (BP) in pediatric emergency patients and to evaluate its recognition by health care practitioners. METHODS: This retrospective review randomly selected patients seen in a large academic pediatric emergency department for 13 months. Triage and subsequent BP measurements were recorded and categorized as normal or elevated (≥ 90th to < 95th, ≥ 95th-99th percentile plus 4 mm Hg, and ≥ 99th percentile plus 5 mm Hg). Physician recognition of elevated BP, training level, and specialty were collected. Demographic information and possible confounding variables (weight, pain, medications, and triage level) were also collected and analyzed. Exclusions included known hypertension or related conditions and those patients without a triage BP measurement. RESULTS: Of the 978 charts reviewed, 907 were included for study (17.5% infants, 82.5% children 1 year and older to 18 years; 50% male, 50% female; 77% African American, 16% white, 4% Hispanic, and 3% other). Fifty-five percent (n = 497) had elevated triage BP (≥ 90th percentile) with only 1% (n = 7) recognized by practitioners as having elevated triage BP. Further, 152 (20%) of the 748 children 1 year and older to 18 years had severely elevated BP with only 5 recognized. CONCLUSIONS: In this study, more than half of the patients had elevated triage BP (≥ 90th percentile), which was rarely recognized by emergency department practitioners regardless of specialty or experience. Early recognition of elevated triage BP offers opportunities for diagnosis of hypertension and related disorders but is challenging to accomplish.