North Carolina Disease Event Tracking and Epidemiologic Collection Tool (NC DETECT) and the National Hospital Ambulatory Medical Care Survey (NHAMCS): comparison of emergency department data

EMS injury cause codes more accurate than emergency department visit ICD-10-CM codes for firearm injury intent in North Carolina

BACKGROUND

The timeliness, accuracy, and completeness of data for firearm injury surveillance is crucial for public health surveillance efforts and informing injury prevention measures. While emergency department (ED) visit data can provide near real-time information on firearms injuries, there are concerns surrounding the accuracy of intent coding in these data. We examined whether emergency medical service (EMS) data provide more accurate firearm injury intent coding in comparison to ED data.

METHODS

We applied a firearm injury definition to EMS encounter data in NC's statewide syndromic surveillance system (NC DETECT), from January 1, 2021, through December 31, 2022. We manually reviewed each record to determine intent, and the corresponding manual classifications were compared to the injury cause codes entered in the EMS data and to ED visit records where EMS-ED record linkage was possible. We then calculated the sensitivity, specificity, positive and negative predictive values for each intent classification in SAS 9.4 using the manually reviewed intent classifications as the gold standard.

RESULTS

We identified 9557 EMS encounters from January 1, 2021, through December 31, 2022 meeting our firearm injury definition. After removing false positives and duplicates, 8584 records were available for manual injury classification. Overall, our analysis demonstrated that manual and EMS injury cause code classifications were comparable. However, for the 3401 EMS encounters that could be linked to an ED visit record, sensitivity of the ED ICD-10-CM codes was low for assault and intentional self-harm encounters at 18.2% (CI 16.5-19.9%) and 22.2% (CI 16-28.5%), respectively. This demonstrates a marked difference in the reliability of the intent coding in the two data sources.

CONCLUSIONS

This study illustrates both the value of examining EMS encounters for firearm injury intent, and the challenges of accurate intent coding in the ED setting. EMS coding has the potential for more accurate intent coding than ED coding within the context of existing hospital-based coding guidance. This may have implications for future firearm injury research, especially for nonfatal firearm injuries.

Linking Emergency Medical Services and Emergency Department Data to Improve Overdose Surveillance in North Carolina

Introduction

Linking emergency medical services (EMS) data to emergency department (ED) data enables assessing the continuum of care and evaluating patient outcomes. We developed novel methods to enhance linkage performance and analysis of EMS and ED data for opioid overdose surveillance in North Carolina.

Methods

We identified data on all EMS encounters in North Carolina during January 1–November 30, 2017, with documented naloxone administration and transportation to the ED. We linked these data with ED visit data in the North Carolina Disease Event Tracking and Epidemiologic Collection Tool. We manually reviewed a subset of data from 12 counties to create a gold standard that informed developing iterative linkage methods using demographic, time, and destination variables. We calculated the proportion of suspected opioid overdose EMS cases that received International Classification of Diseases, Tenth Revision, Clinical Modification diagnosis codes for opioid overdose in the ED.

Results

We identified 12 088 EMS encounters of patients treated with naloxone and transported to the ED. The 12-county subset included 1781 linkage-eligible EMS encounters, with historical linkage of 65.4% (1165 of 1781) and 1.6% false linkages. Through iterative linkage methods, performance improved to 91.0% (1620 of 1781) with 0.1% false linkages. Among statewide EMS encounters with naloxone administration, the linkage improved from 47.1% to 91.1%. We found diagnosis codes for opioid overdose in the ED among 27.2% of statewide linked records.

Practice Implications

Through an iterative linkage approach, EMS–ED data linkage performance improved greatly while reducing the number of false linkages. Improved EMS–ED data linkage quality can enhance surveillance activities, inform emergency response practices, and improve quality of care through evaluating initial patient presentations, field interventions, and ultimate diagnoses.

Emergency Department Visits by Older Adults with Mental Illness in North Carolina

Introduction

We analyzed emergency department (ED) visits by patients with mental health disorders (MHDs) in North Carolina from 2008–2010 to determine frequencies and characteristics of ED visits by older adults with MHDs.

Methods

We extracted ED visit data from the North Carolina Disease Event Tracking and Epidemiologic Collection Tool (NC DETECT). We defined mental health visits as visits with a mental health ICD-9-CM diagnostic code, and organized MHDs into clinically similar groups for analysis.

Results

Those ≥65 with MHDs accounted for 27.3% of all MHD ED visits, and 51.2% were admitted. The most common MHD diagnoses for this age group were psychosis, and stress/anxiety/depression.

Conclusion

Older adults with MHDs account for over one-quarter of ED patients with MHDs, and their numbers will continue to increase as the “boomer” population ages. We must anticipate and prepare for the MHD-related needs of the elderly.

Utility of a near real-time emergency department syndromic surveillance system to track injuries in New York City

Background

The New York City emergency department (ED) syndromic surveillance (SS) system provides near real-time data on the majority of ED visits. The utility of ED SS for injury surveillance has not been thoroughly evaluated. We created injury syndromes based on ED chief complaint information and evaluated their utility compared to administrative billing data.

Methods

Six injury syndromes were developed: traffic-related injuries to pedal cyclists, pedestrians, and motor vehicle occupants; fall-related injuries; firearm-related injuries; and assault-related stabbings. Daily injury counts were compared for ED SS and the administrative billing data for years 2008–2010. We examined characteristics of injury trends and patterns between the two systems, calculating descriptive statistics for temporal patterns and Pearson correlation coefficients (r) for temporal trends. We also calculated proportions of demographic and geospatial patterns for both systems.

Results

Although daily volume of the injuries varied between the two systems, the temporal patterns were similar (all r values for daily volume exceeded 0.65). Comparisons of injuries by time of day, day of week, and quarter of year demonstrated high agreement between the two systems—the majority had an absolute percentage point difference of 2.0 or less. Distributions of injury by sex and age group also aligned well. Distribution of injury by neighborhood of residence showed mixed results—some neighborhood comparisons showed a high level of agreement between systems, while others were less successful.

Conclusions

As evidenced by the strong positive correlation coefficients and the small absolute percentage point differences in our comparisons, we conclude that ED SS captures temporal trends and patterns of injury-related ED visits effectively. The system could be used to identify changes in injury patterns, allowing for situational awareness during emergencies, timely response, and public messaging.

Electronic supplementary material

The online version of this article (doi:10.1186/s40621-015-0044-5) contains supplementary material, which is available to authorized users.

Use of ED and hospital services for patients with acute leukemia after induction therapy: one year follow-up

Previous studies have documented use of health care services by oncology patients in the Emergency Department (ED), but little is known about the utilization of health services of patients with acute leukemia after induction therapy. The aim of this study was to examine chief reasons for ED and hospital use by patients newly diagnosed with acute leukemia patients after induction therapy up to one year after discharge. A retrospective, longitudinal study of all visits to the ED or unplanned hospital admissions at a single institution for patients with acute leukemia was conducted. Inclusion criteria were patients ≥18 years of age at time of diagnosis, a confirmed diagnosis of AML or ALL, and received and discharged from induction treatment between 2007 and 2010. Donabedian's structure-process-outcome framework guided this study examining health services utilization and assessing patient outcomes. 80 patients met the inclusion criteria; 52 had AML and 28 had ALL; median age was 48 (range: 18-76) and 29% (n=23) were non-Caucasian. 70% (n=56) were discharged from induction in remission. 81% (n=65) had at least 1 ED or hospitalization event, and 44% (n=35) had 2 or more events. Of 137 events in 65 patients, the most common reason was neutropenic fever/infection (55%), bleeding (12%), and GI problems (11%). Mean number of events for ALL was 2.43 compared to 1.33 for AML patients (p=0.02), and 2.23 for <50 years of age compared to 1.20 for those older (p=0.002). 20 patients died within one year of diagnosis. Findings from this study can help inform health services delivery and utilization among patients with acute leukemia after induction therapy. Oncology providers can anticipate discharge needs and enhance follow-up care for those at higher risk for problems needing hospitalization.

Emergency Medical Text Classifier: New system improves processing and classification of triage notes

Objective

Automated syndrome classification aims to aid near real-time syndromic surveillance to serve as an early warning system for disease outbreaks, using Emergency Department (ED) data. We present a system that improves the automatic classification of an ED record with triage note into one or more syndrome categories using the vector space model coupled with a ‘learning’ module that employs a pseudo-relevance feedback mechanism. Materials and Methods: Terms from standard syndrome definitions are used to construct an initial reference dictionary for generating the syndrome and triage note vectors. Based on cosine similarity between the vectors, each record is classified into a syndrome category. We then take terms from the top-ranked records that belong to the syndrome of interest as feedback. These terms are added to the reference dictionary and the process is repeated to determine the final classification. The system was tested on two different datasets for each of three syndromes: Gastro-Intestinal (GI), Respiratory (Resp) and Fever-Rash (FR). Performance was measured in terms of sensitivity (Se) and specificity (Sp). Results: The use of relevance feedback produced high values of sensitivity and specificity for all three syndromes in both test sets: GI: 90% and 71%, Resp: 97% and 73%, FR: 100% and 87%, respectively, in test set 1, and GI: 88% and 69%, Resp: 87% and 61%, FR: 97% and 71%, respectively, in test set 2. Conclusions: The new system for pre-processing and syndromic classification of ED records with triage notes achieved improvements in Se and Sp. Our results also demonstrate that the system can be tuned to achieve different levels of performance based on user requirements.

Primary care availability and emergency department use by older adults: a population-based analysis

OBJECTIVES

To assess the relationship between the number of primary care providers (PCPs) in an area and emergency department (ED) visits by older adults.

DESIGN

Population-based cross-sectional observational study.

SETTING

Nonfederal EDs in North Carolina in 2010.

PARTICIPANTS

All older adults (n = 640,086) presenting to a nonfederal ED in North Carolina in 2010.

MEASUREMENTS

The primary outcome was the number of ED visits by older adults in each ZIP code per 100 adults aged 65 and older living in that ZIP code. A secondary outcome was the number of ED visits not resulting in hospital admission per 100 older adults. The primary predictor variable was the number of PCPs per 100 older residents for each ZIP code. Covariates included those representing healthcare need (Medicare hospitalizations, nursing home beds), predisposing factors for healthcare use (race, education, population density of older adults), and enabling factors (distance to the nearest ED).

RESULTS

In a multivariable regression model corrected for spatial clustering, ZIP code characteristics associated with ED visits included more hospitalizations by Medicare beneficiaries, more nursing home beds, and closer proximity to an ED. Number of PCPs per 100 older adult residents in each ZIP code was not associated with ED use, and the 95% confidence limit indicates at most a small effect of PCP availability on ED use.

CONCLUSION

These findings suggest that primary care availability has at most a limited effect on ED use by older adults in North Carolina.

Advancing a framework to enable characterization and evaluation of data streams useful for biosurveillance

In recent years, biosurveillance has become the buzzword under which a diverse set of ideas and activities regarding detecting and mitigating biological threats are incorporated depending on context and perspective. Increasingly, biosurveillance practice has become global and interdisciplinary, requiring information and resources across public health, One Health, and biothreat domains. Even within the scope of infectious disease surveillance, multiple systems, data sources, and tools are used with varying and often unknown effectiveness. Evaluating the impact and utility of state-of-the-art biosurveillance is, in part, confounded by the complexity of the systems and the information derived from them. We present a novel approach conceptualizing biosurveillance from the perspective of the fundamental data streams that have been or could be used for biosurveillance and to systematically structure a framework that can be universally applicable for use in evaluating and understanding a wide range of biosurveillance activities. Moreover, the Biosurveillance Data Stream Framework and associated definitions are proposed as a starting point to facilitate the development of a standardized lexicon for biosurveillance and characterization of currently used and newly emerging data streams. Criteria for building the data stream framework were developed from an examination of the literature, analysis of information on operational infectious disease biosurveillance systems, and consultation with experts in the area of biosurveillance. To demonstrate utility, the framework and definitions were used as the basis for a schema of a relational database for biosurveillance resources and in the development and use of a decision support tool for data stream evaluation.

Relationship of opioid prescription sales and overdoses, North Carolina

BACKGROUND

In the United States, fatal drug overdoses have tripled since 1991. This escalation in deaths is believed to be driven primarily by prescription opioid medications. This investigation compared trends and patterns in sales of opioids, opioid drug overdoses treated in emergency departments (EDs), and unintentional overdose deaths in North Carolina (NC).

METHODS

Our ecological study compared rates of opioid sales, opioid related ED overdoses, and unintentional drug overdose deaths in NC. Annual sales data, provided by the Drug Enforcement Administration, for select opioids were converted into morphine equivalents and aggregated by zip code. These opioid drug sales rates were trended from 1997 to 2010. In addition, opioid sales were correlated and compared to opioid related ED visits, which came from a Centers for Disease Control and Prevention syndromic surveillance system, and unintentional overdose deaths, which came from NC Vital Statistics, from 2008 to 2010. Finally, spatial cluster analysis was performed and rates were mapped by zip code in 2010.

RESULTS

Opioid sales increased substantially from 1997 to 2010. From 2008 to 2010, the quarterly rates of opioid drug overdoses treated in EDs and opioid sales correlated (r=0.68, p=0.02). Specific regions of the state, particularly in the southern and western corners, had both high rates of prescription opioid sales and overdoses.

CONCLUSIONS

Temporal trends in sales of prescription opioids correlate with trends in opioid related ED visits. The spatial correlation of opioid sales with ED visit rates shows that opioid sales data may be a timely way to identify high-risk communities in the absence of timely ED data.

Development and validation of a prioritization rule for obtaining an immediate 12-lead electrocardiogram in the emergency department to identify ST-elevation myocardial infarction

BACKGROUND

Current guidelines recommend an immediate (eg, <10 minutes) 12-lead electrocardiogram (ECG) to identify ST-elevation myocardial infarction (STEMI) among patients presenting to the emergency department (ED) with chest pain. Yet, one third of all patients with myocardial infarction do not have chest pain. Our objective was to develop a practical approach to identify patients, especially those without chest pain, who require an immediate ECG in the ED to identify STEMI.

METHODS

An ECG prioritization rule was derived and validated using classification and regression tree analysis among >3 million ED visits to 107 EDs from 2007 to 2008.

RESULTS

The final study population included 3,575,178 ED patient visits; of these, 6,464 (0.18%) were diagnosed with STEMI. Overall, 1,413 (21.9%) of patients with STEMI did not present to the ED with chest pain. Major predictors of those requiring an immediate ECG in the ED included age ≥30 years with chest pain; age ≥50 years with shortness of breath, altered mental status, upper extremity pain, syncope, or generalized weakness; and those with age ≥80 years with abdominal pain or nausea/vomiting. When the ECG prioritization rule was applied to a validation sample, it had a sensitivity of 91.9% (95% CI 90.9%-92.8%) for STEMI and a negative predictive value 99.98% (95% CI 99.98%-99.98%).

CONCLUSION

A simple ECG prioritization rule based on age and presenting symptoms in the ED can identify patients during triage who are at high risk for STEMI and therefore should receive an immediate 12-lead ECG, often before they are seen by a physician.

Why do patients with cancer visit emergency departments? Results of a 2008 population study in North Carolina

PURPOSE

Emergency departments (EDs) in the United States are used by patients with cancer for disease or treatment-related problems and unrelated issues. The North Carolina Disease Event Tracking and Epidemiologic Collection Tool (NC DETECT) collects information about ED visits through a statewide database.

PATIENTS AND METHODS

After approval by the institutional review board, 2008 NC DETECT ED visit data were acquired and cancer-related visits were identified. Descriptive statistics and logistic regressions were performed. Of 4,190,911 ED visits in 2008, there were 37,760 ED visits by 27,644 patients with cancer.

RESULTS

Among patients, 77.2% had only one ED visit in 2008, the mean age was 64 years, and there were slightly more men than women. Among visits, the payor was Medicare for 52.4% and Medicaid for 12.1%. More than half the visits by patients with cancer occurred on weekends or evenings, and 44.9% occurred during normal hours. The top three chief complaints were related to pain, respiratory distress, and GI issues. Lung, breast, prostate, and colorectal cancers were identified in 26.9%, 6.3%, 6%, and 7.7% of visits, respectively, with diagnosis. A total of 63.2% of visits resulted in hospital admittance. When controlling for sex, age, time of day, day of week, insurance, and diagnosis position, patients with lung cancer were more likely to be admitted than patients with other types of cancer.

CONCLUSION

To the best of our knowledge, this is the first study to provide a population-based snapshot of ED visits by patients with cancer in North Carolina. Efforts that target clinical problems and specific populations may improve delivery of quality cancer care and avoid ED visits.

Patient throughput times and inflow patterns in Swedish emergency departments. A basis for ANSWER, A National SWedish Emergency Registry

OBJECTIVE

Quality improvement initiatives in emergency medicine (EM) often suffer from a lack of benchmarking data on the quality of care. The objectives of this study were twofold: 1. To assess the feasibility of collecting benchmarking data from different Swedish emergency departments (EDs) and 2. To evaluate patient throughput times and inflow patterns.

METHOD

We compared patient inflow patterns, total lengths of patient stay (LOS) and times to first physician at six Swedish university hospital EDs in 2009. Study data were retrieved from the hospitals' computerized information systems during single on-site visits to each participating hospital.

RESULTS

All EDs provided throughput times and patient presentation data without significant problems. In all EDs, Monday was the busiest day and the fewest patients presented on Saturday. All EDs had a large increase in patient inflow before noon with a slow decline over the rest of the 24 h, and this peak and decline was especially pronounced in elderly patients. The average LOS was 4 h of which 2 h was spent waiting for the first physician. These throughput times showed a considerable diurnal variation in all EDs, with the longest times occurring 6-7 am and in the late afternoon.

CONCLUSION

These results demonstrate the feasibility of collecting benchmarking data on quality of care targets within Swedish EM, and form the basis for ANSWER, A National SWedish Emergency Registry.

Automated detection of influenza-like illness using clinical surveillance markers at a Department of Veterans Affairs Medical Center

Background:

Using demographic and clinical measures from emergency department evaluations, we developed an automated surveillance system for influenza-like illness (ILI).

Methods:

We selected a random sample of patients who were seen at the Durham, NC Veterans Affairs Medical Center between May 2002 and October 2009 with fever or a respiratory ICD-9 diagnosis code and divided this into subsets for system development and validation. Comprehensive chart reviews identified patients who met a standard case definition for ILI. Logistic regression models predicting ILI were fit in the development sample. We applied the parameter estimates from these models to the validation sample and evaluated their utility using receiver-operator characteristic analysis.

Results:

The models discriminated ILI very well in the validation sample; the C-statistics were >0.89.

Conclusions:

Risk estimates based on statistical models can be incorporated into electronic medical records systems to assist clinicians and could be used in real-time surveillance for disease outbreaks.

Data based integration of critical illness and injury patient care from EMS to emergency department to intensive care unit

PURPOSE OF REVIEW

Describe the challenges and opportunities for an integrated emergency care data system for the delivery and care of critical illness and injury.

RECENT FINDINGS

Standardized data comparable across geographies and settings of care has been a critical challenge for emergency care data systems. Emergency medical services (EMS), emergency department (ED), ICU and hospital care are integrated units of service in critical illness and injury care. The applicability of available evidence and outcome measures to these units of service needs to be determined. A recently developed fully integrated, emergency care data system for quality improvement of EMS service delivery and patient care has been linked to ED, ICU and in-hospital data systems for myocardial infarction, trauma and stroke. The data system also provides a platform for linking EMS with emergency physicians, other healthcare providers, and public health agencies responsible for planning, disease surveillance, and disaster preparedness.

SUMMARY

Given its time-sensitive nature, new data systems and analytic methods will be required to examine the impact of emergency care. The linkage of emergency care data systems to outcomes based systems could create an ideal environment to improve patient morbidity and mortality in critical illness and injury.

Research using emergency department-related data sets: current status and future directions

The 2009 Academic Emergency Medicine consensus conference focused on "Public Health in the ED: Surveillance, Screening and Intervention." One conference breakout session discussed the significant research value of health-related data sets. This article represents the proceedings from that session, primarily focusing on emergency department (ED)-related data sets and includes examples of the use of a data set based on ED visits for research purposes. It discusses types of ED-related data sets available, highlights barriers to research use of ED-related data sets, and notes limitations of these data sets. The paper highlights future directions and challenges to using these important sources of data for research, including identification of five main needs related to enhancing the use of ED-related data sets. These are 1) electronic linkage of initial and follow-up ED visits and linkage of information about ED visits to other outcomes, including costs of care, while maintaining de-identification of the data; 2) timely data access with minimal barriers; 3) complete data collection for clinically relevant and/or historical data elements, such as the external cause-of-injury code; 4) easy access to data that can be parsed into smaller jurisdictions (such as states) for policy and/or research purposes, while maintaining confidentiality; and 5) linkages between health survey data and health claims data. ED-related data sets contain much data collected directly from health care facilities, individual patient records, and multiple other sources that have significant potential impact for studying and improving the health of individuals and the population.

Study designs and evaluation models for emergency department public health research

Abstract Public health research requires sound design and thoughtful consideration of potential biases that may influence the validity of results. It also requires careful implementation of protocols and procedures that are likely to translate from the research environment to actual clinical practice. This article is the product of a breakout session from the 2009 Academic Emergency Medicine consensus conference entitled "Public Health in the ED: Screening, Surveillance, and Intervention" and serves to describe in detail aspects of performing emergency department (ED)-based public health research, while serving as a resource for current and future researchers. In doing so, the authors describe methodologic features of study design, participant selection and retention, and measurements and analyses pertinent to public health research. In addition, a number of recommendations related to research methods and future investigations related to public health work in the ED are provided. Public health investigators are poised to make substantial contributions to this important area of research, but this will only be accomplished by employing sound research methodology in the context of rigorous program evaluation.