Andrea Freyer Dugas

Influenza Forecasting with Google Flu Trends

Objective We sought to develop a practical influenza forecast model, based on real-time, geographically focused, and easy to access data, to provide individual medical centers with advanced warning of the number of influenza cases, thus allowing sufficient time to implement an intervention. Secondly, we evaluated how the addition of a real-time influenza surveillance system, Google Flu Trends, would impact the forecasting capabilities of this model. Introduction Each year, influenza results in increased Emergency Department crowding which can be mitigated through early detection linked to an appropriate response. Although current surveillance systems, such as Google Flu Trends, yield near real-time influenza surveillance, few demonstrate ability to forecast impending influenza cases. Methods Forecasting models designed to predict one week in advance were developed from weekly counts of confirmed influenza cases over seven seasons (2004 – 2011) divided into training and out-of-sample verification sets. Forecasting procedures using classical Box-Jenkins, generalized linear, and autoregressive methods were employed to develop the final model and assess the relative contribution of external variables such as, Google Flu Trends, meteorological data, and temporal information. Models were developed and evaluated through statistical measures of global deviance and log-likelihood ratio tests. An additional measure of forecast confidence, defined as the percentage of forecast values, during an influenza peak, that are within 7 influenza cases of the actual data, was examined to demonstrate practical utility of the model. Results A generalized autoregressive Poisson (GARMA) forecast model integrating previous influenza cases with Google Flu Trends information provided the most accurate influenza case predictions. Google Flu Trend data was the only source of external information providing significant forecast improvements (p = 0.00002). The final model, a GARMA intercept model with the addition of Google Flu Trends, predicted weekly influenza cases during 4 out-of-sample outbreaks within 7 cases for 80% of estimates (Figure 1). Conclusions Integer-valued autoregression of influenza cases provides a strong base forecast model, which is enhanced by the addition of Google Flu Trends confirming the predictive capabilities of search query based syndromic surveillance. This accessible and flexible forecast model can be used by individual medical centers to provide advanced warning of future influenza cases.

The frequency of influenza and bacterial coinfection: a systematic review and meta‐analysis

Coinfecting bacterial pathogens are a major cause of morbidity and mortality in influenza. However, there remains a paucity of literature on the magnitude of coinfection in influenza patients.

Revitalizing a Vital Sign: Improving Detection of Tachypnea at Primary Triage

STUDY OBJECTIVE This study evaluates the accuracy of emergency department (ED) triage respiratory rate measurement using the usual care method and a new electronic respiratory rate sensor (BioHarness, Zephyr Technology Corp.), both compared to a criterion standard measurement. METHODS This is a cross-sectional study with convenience sampling conducted in an urban academic adult ED, including 3 separate respiratory rate measurements performed at ED triage: usual care measurement, electronic BioHarness measurement, and criterion standard measurement. The criterion standard measurement used was defined by the World Health Organization as manual observation or auscultation of respirations for 60 seconds. The resultant usual care and BioHarness measurements were compared with the criterion standard, evaluating accuracy (sensitivity and specificity) for detecting tachypnea, as well as potential systematic biases of usual care and BioHarness measurements using a Bland Altman analysis. RESULTS Of 191 analyzed patients, 44 presented with tachypnea (>20 breaths/min). Relative to criterion standard measurement, usual care measurement had a sensitivity of 23% (95% confidence interval [CI] 12% to 37%) and specificity of 99% (95% CI 97% to 100%) for tachypnea, whereas BioHarness had a sensitivity of 91% (95% CI 80% to 97%) and specificity of 97% (95% CI 93% to 99%) for tachypnea. Usual care measurements clustered around respiratory rates of 16 and 18 breaths/min (n=144), with poor agreement with criterion standard measurement. Conversely, BioHarness measurement closely tracked criterion standard values over the range of respiratory rates. CONCLUSION Current methods of respiratory rate measurement at ED triage are inaccurate. A new electronic respiratory rate sensor, BioHarness, has significantly greater sensitivity for detecting tachypnea versus usual care method of measurement.

Evaluation of the Xpert Flu Rapid PCR assay in High-Risk Emergency Department Patients

ABSTRACT We prospectively evaluated the performance of Cepheids GeneXpert Xpert Flu assay in a target population of 281 adults presenting to the emergency department with an acute respiratory illness who met Centers for Disease Control and Prevention (CDC) criteria for recommended antiviral treatment. Compared with the Prodesse ProFlu+ assay, Xpert Flu had an overall sensitivity of 95.3% and specificity of 99.2%.

Septic Shock and Adequacy of Early Empiric Antibiotics in the Emergency Department

BACKGROUND Antibiotic resistance is an increasing concern for Emergency Physicians. OBJECTIVES To examine whether empiric antibiotic therapy achieved appropriate antimicrobial coverage in emergency department (ED) septic shock patients and evaluate reasons for inadequate coverage. METHODS Retrospective review was performed of all adult septic shock patients presenting to the ED of a tertiary care center from December 2007 to September 2008. Inclusion criteria were: 1) Suspected or confirmed infection; 2) ≥ 2 Systemic Inflammatory Response Syndrome criteria; 3) Treatment with one antimicrobial agent; 4) Hypotension requiring vasopressors. Patients were dichotomized by presentation from a community or health care setting. RESULTS Eighty-five patients with septic shock were identified. The average age was 68 ± 15.8 years. Forty-seven (55.3%) patients presented from a health care setting. Pneumonia was the predominant clinically suspected infection (n = 38, 45%), followed by urinary tract (n = 16, 19%), intra-abdominal (n = 13, 15%), and other infections (n = 18, 21%). Thirty-nine patients (46%) had an organism identified by positive culture, of which initial empiric antibiotic therapy administered in the ED adequately covered the infectious organism in 35 (90%). The 4 patients who received inadequate therapy all had urinary tract infections (UTI) and were from a health care setting. CONCLUSION In this population of ED patients with septic shock, empiric antibiotic coverage was inadequate in a small group of uroseptic patients with recent health care exposure. Current guidelines for UTI treatment do not consider health care setting exposure. A larger, prospective study is needed to further define this risk category and determine optimal empiric antibiotic therapy for patients.

Cost-Utility of Rapid Polymerase Chain Reaction-Based Influenza Testing for High-Risk Emergency Department Patients

STUDY OBJECTIVE We evaluate the cost-effectiveness of polymerase chain reaction (PCR)-based rapid influenza testing and treatment for influenza in adult emergency department (ED) patients who are at high risk for or have evidence of influenza-related complications. METHODS We developed a cost-utility decision analysis model that assessed adult patients presenting to the ED with symptoms of an acute respiratory infection, who met the Centers for Disease Control and Prevention criteria for recommended antiviral treatment. Analysis was performed from the societal perspective, with incremental comparisons of 4 influenza testing and treatment strategies: treat none, treat according to provider judgment, treat according to results of a PCR-based rapid diagnostic test, and treat all. RESULTS Treating no patients with antivirals was dominated by all other strategies that increased in both cost and benefit in the following order: treat according to provider judgment, treat according to results of a PCR-based rapid diagnostic test, and treat all. As influenza prevalence increases, treating all patients eventually dominated all other options. CONCLUSION The economic benefit of incorporating use of rapid PCR-based influenza testing for ED patients at risk of developing influenza-related complications depends on influenza prevalence; treatment guided by physician diagnosis or rapid testing, and treatment of all patients is more effective and less costly than no treatment.

Diagnosis of spinal cord compression in nontrauma patients in the emergency department.

OBJECTIVES The objectives were to evaluate the presenting signs and symptoms of spinal cord and cauda equina compression (SCC) and to determine the incidence of emergency department (ED) misdiagnosis. METHODS This was a retrospective chart review at an urban, tertiary care hospital of patients discharged from an inpatient stay (April 2008 through July 2009) with an International Classification of Diseases, Ninth Revision (ICD-9) code indicating spinal disease, who had visited the ED for a related complaint within the previous 30 days, and who had a final diagnosis of new SCC. Trauma and transferred patients were excluded. The authors defined a misdiagnosis as no ED-documented diagnosis of SCC and failure to perform an appropriate diagnostic study either prior to arrival, in the ED, or immediately upon admission. RESULTS Of 1,231 charts reviewed, 63 met inclusion criteria. The most common presenting symptoms in patients with SCC were pain (44, 70%), difficulty ambulating (38, 60%), and weakness (35, 56%). On physical examination, motor deficits (45, 71%) were more common than sensory deficits (27, 43%); however, 15 (24%) patients had no motor or sensory deficit, and 13 (23%) patients only had unilateral findings. Impaired gait was present in 14 patients of only 20 tested, three of whom had no associated motor or sensory deficit. SCC was misdiagnosed in 18 (29%, 95% confidence interval [CI] = 19% to 41%) cases, which resulted in a significant delay to diagnosis (median = 54 hours, interquartile range [IQR] = 38 to 77 vs. 5.3 hours, IQR = 3.0 to 15) in these patients. CONCLUSIONS SCC can have a subtle presentation with absent or unilateral motor and sensory deficits, but gait ataxia may be an additional finding. ED misdiagnosis of SCC in nontrauma patients is common.

Google Flu Trends Spatial Variability Validated Against Emergency Department Influenza-Related Visits.

Background Influenza is a deadly and costly public health problem. Variations in its seasonal patterns cause dangerous surges in emergency department (ED) patient volume. Google Flu Trends (GFT) can provide faster influenza surveillance information than traditional CDC methods, potentially leading to improved public health preparedness. GFT has been found to correlate well with reported influenza and to improve influenza prediction models. However, previous validation studies have focused on isolated clinical locations. Objective The purpose of the study was to measure GFT surveillance effectiveness by correlating GFT with influenza-related ED visits in 19 US cities across seven influenza seasons, and to explore which city characteristics lead to better or worse GFT effectiveness. Methods Using Healthcare Cost and Utilization Project data, we collected weekly counts of ED visits for all patients with diagnosis (International Statistical Classification of Diseases 9) codes for influenza-related visits from 2005-2011 in 19 different US cities. We measured the correlation between weekly volume of GFT searches and influenza-related ED visits (ie, GFT ED surveillance effectiveness) per city. We evaluated the relationship between 15 publically available city indicators (11 sociodemographic, two health care utilization, and two climate) and GFT surveillance effectiveness using univariate linear regression. Results Correlation between city-level GFT and influenza-related ED visits had a median of .84, ranging from .67 to .93 across 19 cities. Temporal variability was observed, with median correlation ranging from .78 in 2009 to .94 in 2005. City indicators significantly associated (P<.10) with improved GFT surveillance include higher proportion of female population, higher proportion with Medicare coverage, higher ED visits per capita, and lower socioeconomic status. Conclusions GFT is strongly correlated with ED influenza-related visits at the city level, but unexplained variation over geographic location and time limits its utility as standalone surveillance. GFT is likely most useful as an early signal used in conjunction with other more comprehensive surveillance techniques. City indicators associated with improved GFT surveillance provide some insight into the variability of GFT effectiveness. For example, populations with lower socioeconomic status may have a greater tendency to initially turn to the Internet for health questions, thus leading to increased GFT effectiveness. GFT has the potential to provide valuable information to ED providers for patient care and to administrators for ED surge preparedness.

Machine-Learning-Based Electronic Triage More Accurately Differentiates Patients With Respect to Clinical Outcomes Compared With the Emergency Severity Index

Study objective Standards for emergency department (ED) triage in the United States rely heavily on subjective assessment and are limited in their ability to risk‐stratify patients. This study seeks to evaluate an electronic triage system (e‐triage) based on machine learning that predicts likelihood of acute outcomes enabling improved patient differentiation. Methods A multisite, retrospective, cross‐sectional study of 172,726 ED visits from urban and community EDs was conducted. E‐triage is composed of a random forest model applied to triage data (vital signs, chief complaint, and active medical history) that predicts the need for critical care, an emergency procedure, and inpatient hospitalization in parallel and translates risk to triage level designations. Predicted outcomes and secondary outcomes of elevated troponin and lactate levels were evaluated and compared with the Emergency Severity Index (ESI). Results E‐triage predictions had an area under the curve ranging from 0.73 to 0.92 and demonstrated equivalent or improved identification of clinical patient outcomes compared with ESI at both EDs. E‐triage provided rationale for risk‐based differentiation of the more than 65% of ED visits triaged to ESI level 3. Matching the ESI patient distribution for comparisons, e‐triage identified more than 10% (14,326 patients) of ESI level 3 patients requiring up triage who had substantially increased risk of critical care or emergency procedure (1.7% ESI level 3 versus 6.2% up triaged) and hospitalization (18.9% versus 45.4%) across EDs. Conclusion E‐triage more accurately classifies ESI level 3 patients and highlights opportunities to use predictive analytics to support triage decisionmaking. Further prospective validation is needed.

Emergency Physicians' Adherence to Center for Disease Control and Prevention Guidance During the 2009 Influenza A H1N1 Pandemic.

Introduction: Little is known regarding compliance with management guidelines for epidemic influenza in adult emergency department (ED) settings during the 2009 novel influenza A (H1N1) epidemic, especially in relation to the Centers for Disease Control and Prevention (CDC) guidance. Methods: We investigated all patients with a clinical diagnosis of influenza at an inner-city tertiary academic adult ED with an annual census of approximately 60,000 visits from May 2008 to December 2009. We aimed to determine patterns of presentation and management for adult patients with an ED diagnosis of influenza during the H1N1 pandemic, using seasonal influenza (pre-H1N1) as reference and to determine the ED provider’s adherence to American College of Emergency Physicians and CDC guidance during the 2009 H1N1 influenza pandemic. Adherence to key elements of CDC 2009 H1N1 guidance was defined as (1) the proportion of admitted patients who were recommended to receive testing or treatment who actually received testing for influenza or treatment with antivirals; and (2) the proportion of high-risk patients who were supposed to be treated who actually were treated with antivirals. Results: Among 339 patients with clinically diagnosed influenza, 88% occurred during the H1N1 pandemic. Patients were similarly managed during both phases. Median length of visit (pre-H1N1: 385 min, H1N1: 355 min, P > 0.05) and admission rates (pre-H1N1: 8%, H1N1: 11%, P > 0.05) were similar between the 2 groups. 28% of patients in the pre-H1N1 group and 16% of patients in the H1N1 group were prescribed antibiotics during their ED visits (P > 0.05). There were 34 admitted patients during the pandemic;, 30 (88%) of them received influenza testing in the ED, and 22 (65%) were prescribed antivirals in the ED. Noticeably, 19 (56%) of the 34 admitted patients, including 6 with a positive influenza test, received antibiotic treatment during their ED stay. Conclusion: During the recent H1N1 pandemic, most admitted patients received ED diagnostic testing corresponding to the current recommended guidance. Antibiotic treatment for ED patients admitted with suspected influenza is not uncommon. However, less than 70% of admitted patients and less than 50% of high-risk patients were treated with antivirals during their ED visit, indicating a specific call for closer adherence to guidelines in future influenza pandemics.