Dataset Information

Using age, triage score, and disposition data from emergency department electronic records to improve Influenza-like illness surveillance.

ABSTRACT:

Objective

Markers of illness severity are increasingly captured in emergency department (ED) electronic systems, but their value for surveillance is not known. We assessed the value of age, triage score, and disposition data from ED electronic records for predicting influenza-related hospitalizations.

Materials and methods

From June 2006 to January 2011, weekly counts of pneumonia and influenza (P&I) hospitalizations from five Montreal hospitals were modeled using negative binomial regression. Over lead times of 0-5 weeks, we assessed the predictive ability of weekly counts of 1) total ED visits, 2) ED visits with influenza-like illness (ILI), and 3) ED visits with ILI stratified by age, triage score, or disposition. Models were adjusted for secular trends, seasonality, and autocorrelation. Model fit was assessed using Akaike information criterion, and predictive accuracy using the mean absolute scaled error (MASE).

Results

Predictive accuracy for P&I hospitalizations during non-pandemic years was improved when models included visits from patients ?65 years old and visits resulting in admission/transfer/death (MASE of 0.64, 95% confidence interval (95% CI) 0.54-0.80) compared to overall ILI visits (0.89, 95% CI 0.69-1.10). During the H1N1 pandemic year, including visits from patients <18 years old, visits with high priority triage scores, or visits resulting in admission/transfer/death resulted in the best model fit.

Discussion

Age and disposition data improved model fit and moderately reduced the prediction error for P&I hospitalizations; triage score improved model fit only during the pandemic year.

Conclusion

Incorporation of age and severity measures available in ED records can improve ILI surveillance algorithms.

SUBMITTER: Savard N

PROVIDER: S-EPMC7784300 | biostudies-literature | 2015 May

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Using age, triage score, and disposition data from emergency department electronic records to improve Influenza-like illness surveillance.

Savard Noémie N Bédard Lucie L Allard Robert R Buckeridge David L DL

Journal of the American Medical Informatics Association : JAMIA 20150226 3

<h4>Objective</h4>Markers of illness severity are increasingly captured in emergency department (ED) electronic systems, but their value for surveillance is not known. We assessed the value of age, triage score, and disposition data from ED electronic records for predicting influenza-related hospitalizations.<h4>Materials and methods</h4>From June 2006 to January 2011, weekly counts of pneumonia and influenza (P&I) hospitalizations from five Montreal hospitals were modeled using negative binomia ...[more]

PMID: 25725005

Similar Datasets

Project description:BackgroundEmergency departments (ED) are becoming increasingly overwhelmed, increasing poor outcomes. Triage scores aim to optimize the waiting time and prioritize the resource usage. Artificial intelligence (AI) algorithms offer advantages for creating predictive clinical applications.ObjectiveEvaluate a state-of-the-art machine learning model for predicting mortality at the triage level and, by validating this automatic tool, improve the categorization of patients in the ED.DesignAn institutional review board (IRB) approval was granted for this retrospective study. Information of consecutive adult patients (ages 18-100) admitted at the emergency department (ED) of one hospital were retrieved (January 1, 2012-December 31, 2018). Features included the following: demographics, admission date, arrival mode, referral code, chief complaint, previous ED visits, previous hospitalizations, comorbidities, home medications, vital signs, and Emergency Severity Index (ESI). The following outcomes were evaluated: early mortality (up to 2 days post ED registration) and short-term mortality (2-30 days post ED registration). A gradient boosting model was trained on data from years 2012-2017 and examined on data from the final year (2018). The area under the curve (AUC) for mortality prediction was used as an outcome metric. Single-variable analysis was conducted to develop a nine-point triage score for early mortality.Key resultsOverall, 799,522 ED visits were available for analysis. The early and short-term mortality rates were 0.6% and 2.5%, respectively. Models trained on the full set of features yielded an AUC of 0.962 for early mortality and 0.923 for short-term mortality. A model that utilized the nine features with the highest single-variable AUC scores (age, arrival mode, chief complaint, five primary vital signs, and ESI) yielded an AUC of 0.962 for early mortality.ConclusionThe gradient boosting model shows high predictive ability for screening patients at risk of early mortality utilizing data available at the time of triage in the ED.

Project description:BackgroundSpecialty consultation is a critical aspect of emergency department (ED) practice, and a delay in providing consultation might have a significant clinical effect and worsen ED overcrowding. Although mobile electronic medical records (EMR) are being increasingly used and are known to improve the workflow of health care providers, limited studies have evaluated their effectiveness in real-life clinical scenarios.ObjectiveFor this study, we aimed to determine the association between response duration to an ED specialty consultation request and the frequency of mobile EMR use.MethodsThis retrospective study was conducted in an academic ED in Seoul, South Korea. We analyzed EMR and mobile EMR data from May 2018 to December 2018. Timestamps of ED consultation requests were retrieved from a PC-based EMR, and the response interval was calculated. Doctors' log frequencies were obtained from the mobile EMR, and we merged data using doctors' deidentification numbers. Pearson's product-moment correlation was performed to identify this association. The primary outcome was the relationship between the frequency of mobile EMR usage and the time interval from ED request to consultation completion by specialty doctors. The secondary outcome was the relationship between the frequency of specialty doctors' mobile EMR usage and the response time to consultation requests.ResultsA total of 25,454 consultations requests were made for 15,555 patients, and 252 specialty doctors provided ED specialty consultations. Of the 742 doctors who used the mobile EMR, 208 doctors used it for the specialty consultation process. After excluding the cases lacking essential information, 21,885 consultations with 208 doctors were included for analysis. According to the mobile EMR usage pattern, the average usage frequency of all users was 13.3 logs/day, and the average duration of the completion of the specialty consultation was 51.7 minutes. There was a significant inverse relationship between the frequency of mobile EMR usage and time interval from ED request to consultation completion by specialty doctors (coefficient=-0.19; 95% CI -0.32 to -0.06; P=.005). Secondary analysis with the response time was done. There was also a significant inverse relationship between the frequency of specialty doctors' mobile EMR usage and the response time to consultation requests (coefficient=-0.18; 95% CI -0.30 to -0.04; P=.009).ConclusionsOur findings suggest that frequent mobile EMR usage is associated with quicker response time to ED consultation requests.

Project description:Post-market medical device surveillance is a challenge facing manufacturers, regulatory agencies, and health care providers. Electronic health records are valuable sources of real-world evidence for assessing device safety and tracking device-related patient outcomes over time. However, distilling this evidence remains challenging, as information is fractured across clinical notes and structured records. Modern machine learning methods for machine reading promise to unlock increasingly complex information from text, but face barriers due to their reliance on large and expensive hand-labeled training sets. To address these challenges, we developed and validated state-of-the-art deep learning methods that identify patient outcomes from clinical notes without requiring hand-labeled training data. Using hip replacements-one of the most common implantable devices-as a test case, our methods accurately extracted implant details and reports of complications and pain from electronic health records with up to 96.3% precision, 98.5% recall, and 97.4% F1, improved classification performance by 12.8-53.9% over rule-based methods, and detected over six times as many complication events compared to using structured data alone. Using these additional events to assess complication-free survivorship of different implant systems, we found significant variation between implants, including for risk of revision surgery, which could not be detected using coded data alone. Patients with revision surgeries had more hip pain mentions in the post-hip replacement, pre-revision period compared to patients with no evidence of revision surgery (mean hip pain mentions 4.97 vs. 3.23; t?=?5.14; p?<?0.001). Some implant models were associated with higher or lower rates of hip pain mentions. Our methods complement existing surveillance mechanisms by requiring orders of magnitude less hand-labeled training data, offering a scalable solution for national medical device surveillance using electronic health records.

Project description:ObjectiveTo predict hospital admission at the time of ED triage using patient history in addition to information collected at triage.MethodsThis retrospective study included all adult ED visits between March 2014 and July 2017 from one academic and two community emergency rooms that resulted in either admission or discharge. A total of 972 variables were extracted per patient visit. Samples were randomly partitioned into training (80%), validation (10%), and test (10%) sets. We trained a series of nine binary classifiers using logistic regression (LR), gradient boosting (XGBoost), and deep neural networks (DNN) on three dataset types: one using only triage information, one using only patient history, and one using the full set of variables. Next, we tested the potential benefit of additional training samples by training models on increasing fractions of our data. Lastly, variables of importance were identified using information gain as a metric to create a low-dimensional model.ResultsA total of 560,486 patient visits were included in the study, with an overall admission risk of 29.7%. Models trained on triage information yielded a test AUC of 0.87 for LR (95% CI 0.86-0.87), 0.87 for XGBoost (95% CI 0.87-0.88) and 0.87 for DNN (95% CI 0.87-0.88). Models trained on patient history yielded an AUC of 0.86 for LR (95% CI 0.86-0.87), 0.87 for XGBoost (95% CI 0.87-0.87) and 0.87 for DNN (95% CI 0.87-0.88). Models trained on the full set of variables yielded an AUC of 0.91 for LR (95% CI 0.91-0.91), 0.92 for XGBoost (95% CI 0.92-0.93) and 0.92 for DNN (95% CI 0.92-0.92). All algorithms reached maximum performance at 50% of the training set or less. A low-dimensional XGBoost model built on ESI level, outpatient medication counts, demographics, and hospital usage statistics yielded an AUC of 0.91 (95% CI 0.91-0.91).ConclusionMachine learning can robustly predict hospital admission using triage information and patient history. The addition of historical information improves predictive performance significantly compared to using triage information alone, highlighting the need to incorporate these variables into prediction models.

Dataset Information

Using age, triage score, and disposition data from emergency department electronic records to improve Influenza-like illness surveillance.

Objective

Materials and methods

Results

Discussion

Conclusion

Publications

Using age, triage score, and disposition data from emergency department electronic records to improve Influenza-like illness surveillance.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure