Dataset Information

Real-time forecasting of emergency department arrivals using prehospital data.

ABSTRACT:

Background

Crowding in emergency departments (EDs) is a challenge globally. To counteract crowding in day-to-day operations, better tools to improve monitoring of the patient flow in the ED is needed. The objective of this study was the development of a continuously updated monitoring system to forecast emergency department (ED) arrivals on a short time-horizon incorporating data from prehospital services.

Methods

Time of notification and ED arrival was obtained for all 191,939 arrivals at the ED of a Norwegian university hospital from 2010 to 2018. An arrival notification was an automatically captured time stamp which indicated the first time the ED was notified of an arriving patient, typically by a call from an ambulance to the emergency service communication center. A Poisson time-series regression model for forecasting the number of arrivals on a 1-, 2- and 3-h horizon with continuous weekly and yearly cyclic effects was implemented. We incorporated time of arrival notification by modelling time to arrival as a time varying hazard function. We validated the model on the last full year of data.

Results

In our data, 20% of the arrivals had been notified more than 1 hour prior to arrival. By incorporating time of notification into the forecasting model, we saw a substantial improvement in forecasting accuracy, especially on a one-hour horizon. In terms of mean absolute prediction error, we observed around a six percentage-point decrease compared to a simplified prediction model. The increase in accuracy was particularly large for periods with large inflow.

Conclusions

The proposed model shows increased predictability in ED patient inflow when incorporating data on patient notifications. This approach to forecasting arrivals can be a valuable tool for logistic, decision making and ED resource management.

SUBMITTER: Asheim A

PROVIDER: S-EPMC6683581 | biostudies-literature | 2019 Aug

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Real-time forecasting of emergency department arrivals using prehospital data.

Asheim Andreas A Bache-Wiig Bjørnsen Lars P LP Næss-Pleym Lars E LE Uleberg Oddvar O Dale Jostein J Nilsen Sara M SM

BMC emergency medicine 20190805 1

<h4>Background</h4>Crowding in emergency departments (EDs) is a challenge globally. To counteract crowding in day-to-day operations, better tools to improve monitoring of the patient flow in the ED is needed. The objective of this study was the development of a continuously updated monitoring system to forecast emergency department (ED) arrivals on a short time-horizon incorporating data from prehospital services.<h4>Methods</h4>Time of notification and ED arrival was obtained for all 191,939 ar ...[more]

PMID: 31382882

Similar Datasets

Project description:RationaleEarly antibiotics improve outcomes for patients with sepsis. Factors influencing antibiotic timing in emergency department (ED) sepsis remain unclear.ObjectivesDetermine the relationship between prehospital level of care of patients with sepsis and ED door-to-antibiotic time.MethodsThis retrospective cohort study comprised patients admitted from the community to an academic ED June 2009 to February 2015 with fluid-refractory sepsis or septic shock. Transfer patients and those whose antibiotics began before ED arrival or after ED discharge were excluded. We used multivariable regression to evaluate the association between the time from ED arrival to antibiotic initiation and prehospital level of care, defined as the highest level of emergency medical services received: none, basic life support (BLS) ambulance, or advanced life support (ALS) ambulance. We measured variation in this association when hypotension was or was not present by ED arrival.ResultsAmong 361 community-dwelling patients with sepsis, the level of prehospital care correlated with illness severity. ALS-treated patients received antibiotics faster than patients who did not receive prehospital care (median, 103 [interquartile range, 75 to 135] vs. 144 [98 to 251] minutes, respectively) or BLS-only patients (168 [100-250] minutes; P < 0.001 for each pairwise comparison with ALS). This pattern persisted after multivariable adjustment, where ALS care (-43 min; 95% confidence interval [CI], -84 to -2; P = 0.033) but not BLS-only care (-4 min; 95% CI, -41 to +34; P = 0.97) was associated with less antibiotic delay compared with no prehospital care. ALS-treated patients more frequently received antibiotics within 3 hours of ED arrival (91%) compared with walk-in patients (62%; adjusted odds ratio, 3.11; 95% CI, 1.20 to 8.03; P = 0.015) or BLS-treated patients (56%; adjusted odds ratio, 4.51; 95% CI, 1.89 to 11.35; P < 0.001). ALS-treated patients started antibiotics faster than walk-in patients in the absence of hypotension by ED arrival (-41 min; 95% CI, -110 to -13; P = 0.009) but not when hypotension was present (+25 min; 95% CI, -43 to +92; P = 0.66).ConclusionsPrehospital ALS but not BLS-only care was associated with faster antibiotic initiation for patients with sepsis without hypotension. Process redesign for non-ALS patients may improve antibiotic timeliness for ED sepsis.

Project description:BackgroundAlthough daily emergency department (ED) data is a source of information that often includes residence, its potential for space-time analyses at the individual level has not been fully explored. We propose that ED data collected for surveillance purposes can also be used to inform spatial and temporal patterns of disease using generalized additive models (GAMs). This paper describes the methods for adapting GAMs so they can be applied to ED data.MethodsGAMs are an effective approach for modeling spatial and temporal distributions of point-wise data, producing smoothed surfaces of continuous risk while adjusting for confounders. In addition to disease mapping, the method allows for global and pointwise hypothesis testing and selection of statistically optimum degree of smoothing using standard statistical software. We applied a two-dimensional GAM for location to ED data of overlapping calendar time using a locally-weighted regression smoother. To illustrate our methods, we investigated the association between participants' address and the risk of gastrointestinal illness in Cape Cod, Massachusetts over time.ResultsThe GAM space-time analyses simultaneously smooth in units of distance and time by using the optimum degree of smoothing to create data frames of overlapping time periods and then spatially analyzing each data frame. When resulting maps are viewed in series, each data frame contributes a movie frame, allowing us to visualize changes in magnitude, geographic size, and location of elevated risk smoothed over space and time. In our example data, we observed an underlying geographic pattern of gastrointestinal illness with risks consistently higher in the eastern part of our study area over time and intermittent variations of increased risk during brief periods.ConclusionsSpatial-temporal analysis of emergency department data with GAMs can be used to map underlying disease risk at the individual-level and view changes in geographic patterns of disease over time while accounting for multiple confounders. Despite the advantages of GAMs, analyses should be considered exploratory in nature. It is possible that even with a conservative cutoff for statistical significance, results of hypothesis testing may be due to chance. This paper illustrates that GAMs can be adapted to measure geographic trends in public health over time using ED data.

Project description:BackgroundHyperglycemic crises are associated with high morbidity and mortality. Previous studies have proposed methods to predict adverse outcomes of patients in hyperglycemic crises; however, artificial intelligence (AI) has never been used to predict adverse outcomes. We implemented an AI model integrated with the hospital information system (HIS) to clarify whether AI could predict adverse outcomes.MethodsWe included 2,666 patients with hyperglycemic crises from emergency departments (ED) between 2009 and 2018. The patients were randomized into a 70%/30% split for AI model training and testing. Twenty-two feature variables from the electronic medical records were collected. The performance of the multilayer perceptron (MLP), logistic regression, random forest, Light Gradient Boosting Machine (LightGBM), support vector machine (SVM), and K-nearest neighbor (KNN) algorithms was compared. We selected the best algorithm to construct an AI model to predict sepsis or septic shock, intensive care unit (ICU) admission, and all-cause mortality within 1 month. The outcomes between the non-AI and AI groups were compared after implementing the HIS and predicting the hyperglycemic crisis death (PHD) score.ResultsThe MLP had the best performance in predicting the three adverse outcomes, compared with the random forest, logistic regression, SVM, KNN, and LightGBM models. The areas under the curves (AUCs) using the MLP model were 0.852 for sepsis or septic shock, 0.743 for ICU admission, and 0.796 for all-cause mortality. Furthermore, we integrated the AI predictive model with the HIS to assist decision making in real time. No significant differences in ICU admission or all-cause mortality were detected between the non-AI and AI groups. The AI model performed better than the PHD score for predicting all-cause mortality (AUC 0.796 vs. 0.693).ConclusionsA real-time AI predictive model is a promising method for predicting adverse outcomes in ED patients with hyperglycemic crises. Further studies recruiting more patients are warranted.

Project description:IntroductionBenefits of post-simulation debriefings as an educational and feedback tool have been widely accepted for nearly a decade. Real-time, non-critical incident debriefing is similar to post-simulation debriefing; however, data on its practice in academic emergency departments (ED), is limited. Although tools such as TeamSTEPPS® (Team Strategies and Tools to Enhance Performance and Patient Safety) suggest debriefing after complicated medical situations, they do not teach debriefing skills suited to this purpose. Anecdotal evidence suggests that real-time debriefings (or non-critical incident debriefings) do in fact occur in academic EDs;, however, limited research has been performed on this subject. The objective of this study was to characterize real-time, non-critical incident debriefing practices in emergency medicine (EM).MethodsWe conducted this multicenter cross-sectional study of EM attendings and residents at four large, high-volume, academic EM residency programs in New York City. Questionnaire design was based on a Delphi panel and pilot testing with expert panel. We sought a convenience sample from a potential pool of approximately 300 physicians across the four sites with the goal of obtaining >100 responses. The survey was sent electronically to the four residency list-serves with a total of six monthly completion reminder emails. We collected all data electronically and anonymously using SurveyMonkey.com; the data were then entered into and analyzed with Microsoft Excel.ResultsThe data elucidate various characteristics of current real-time debriefing trends in EM, including its definition, perceived benefits and barriers, as well as the variety of formats of debriefings currently being conducted.ConclusionThis survey regarding the practice of real-time, non-critical incident debriefings in four major academic EM programs within New York City sheds light on three major, pertinent points: 1) real-time, non-critical incident debriefing definitely occurs in academic emergency practice; 2) in general, real-time debriefing is perceived to be of some value with respect to education, systems and performance improvement; 3) although it is practiced by clinicians, most report no formal training in actual debriefing techniques. Further study is needed to clarify actual benefits of real-time/non-critical incident debriefing as well as details on potential pitfalls of this practice and recommendations for best practices for use.

Dataset Information

Real-time forecasting of emergency department arrivals using prehospital data.

Background

Methods

Results

Conclusions

Publications

Real-time forecasting of emergency department arrivals using prehospital data.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets