Article

Predicting Hospital Admissions to Reduce Emergency Department Boarding

Authors:
To read the full-text of this research, you can request a copy directly from the author.

Abstract

Recent research has established that Emergency Department (ED) congestion is often caused by the inability to transition patients into inpatient units within the hospital in a timely fashion. This problem, in which the ED boards inpatients, is common across the U.S. Predicting ED patient admission using demographic and clinical information with only a few admission predictor factors investigated so far. We have developed a prediction model that can be used as a decision support tool and help reduce ED boarding. Using secondary data from the ED of a local hospital, we have examined the importance of eight demographic and clinical determinant factors of ED patients’ admission to the hospital. We have employed Logistic Regression (LR) and Neural Network (NN) modeling techniques and based on our statistical analysis, we have identified encounter reason, age, and radiology exam type as the most significant factors. We have studied patterns between input variables (i.e. age) and output variables (i.e. admitted or not) and have developed a set of rules of thumb for predicting admissions. These unique rules can be used without any modeling or further investigation during operations, therefore providing important information regarding the ultimate status of a patient after ED operations without any time or cost. The study proves that an admission prediction model based on demographic and clinical determinant factors can accurately estimate the likelihood of patient admission, thus decreasing ED boarding and congestion, both significant problems in hospital operations.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the author.

... Patients are experiencing delays in their transfers due to insufficient ward beds or understaffing [16]. Staff availability, staff skillset, as well as timely transferring patients, are other causes of boarding [25]. The lack of efficiency in the continuity of care through step-down and appropriate alternative care provisions also causes bed-blocking, which in turn exacerbates critical care resource shortages [11]. ...
... Moreover, demographic data and clinical information can be used to estimate the likelihood of patient admission accurately. These predictions enable hospital managers to improve their estimation of required resources and the process of assigning beds to patients [25]. Finally, researchers indicate that a frequent assessment of boarders, effective communication and coordination between the ED and other related departments facilitate patient transfers and result in lower patient boarding [25,57]. ...
... These predictions enable hospital managers to improve their estimation of required resources and the process of assigning beds to patients [25]. Finally, researchers indicate that a frequent assessment of boarders, effective communication and coordination between the ED and other related departments facilitate patient transfers and result in lower patient boarding [25,57]. ...
Article
Timely access to health services has become increasingly difficult due to demographic change and aging people growth. These create new heterogeneous challenges for society and healthcare systems. Congestion at acute hospitals has reached unprecedented levels due to the unavailability of acute beds. As a consequence, patients in need of treatment endure prolonged waiting times as a decision whether to admit, transfer, or send them home is made. These long waiting times often result in boarding patients in different places in the hospital. This threatens patient safety and diminishes the service quality while increasing treatment costs. It is argued in the extant literature that improved communication and enhanced patient flow is often more effective than merely increasing hospital capacity. Achieving this effective coordination is challenged by the uncertainties in care demand, the availability of accurate information, the complexity of inter-hospital dynamics and decision times. A hybrid simulation approach is presented in this paper, which aims to offer hospital managers a chance at investigating the patient boarding problem. Integrating ‘System Dynamic’ and ‘Discrete Event Simulation’ enables the user to ease the complexity of patient flow at both macro and micro levels. ‘Design of Experiment’ and ‘Data Envelopment Analysis’ are integrated with the simulation in order to assess the operational impact of various management interventions efficiently. A detailed implementation of the approach is demonstrated on an emergency department (ED) and Acute Medical Unit (AMU) of a large Irish hospital, which serves over 50,000 patients annually. Results indicate that improving transfer rates between hospital units has a significant positive impact. It reduces the number of boarding patients and has the potential to increase access by up to 40% to the case study organization. However, poor communication and coordination, human factors, downstream capacity constraints, shared resources and services between units may affect this access. Furthermore, an increase in staff numbers is required to sustain the acceptable level of service delivery.
... All these studies used any subset of these variables. e patient arrival problems are generally related daily [7,[9][10][11][12][13][14], weekly [4], or monthly [4,15]; however, a few studies focus on hourly arrival rates which include high variation [16][17][18]. e forecasting accuracy worsened as the forecast time intervals became smaller: the best MAPE rate is 2% for a month, 11% for a day, 38% for four-hour period, and 50% for an hour [18]. ...
... e studies were performed using different types of patient data, such as one-year data [4, 5, 9-11, 16, 17], 2-to 6-year data [7,[12][13][14][15]17], 10-year data [19], and the daily number of patient arrivals generally used over a period ranging from 1-to 10-year data (usually 3 years) [8]. Different techniques used for the prediction of patient arrival rate are the ANN [7,9,11,14], the autoregressive integrated moving average (ARIMA) [4,8,[12][13][14], the linear regression (LR) [14,18], the exponential smoothing [14,15,18], the logistic regression [5,10,11,19], the decision tree (DT) [1,10], the gradient boosted machines (GBM) [10], the Poisson regression model [16], the random forest (RF), the AdaBoost (AB), the support vector machine (SVM) [5], and the LSTM model [20]. ...
... e studies were performed using different types of patient data, such as one-year data [4, 5, 9-11, 16, 17], 2-to 6-year data [7,[12][13][14][15]17], 10-year data [19], and the daily number of patient arrivals generally used over a period ranging from 1-to 10-year data (usually 3 years) [8]. Different techniques used for the prediction of patient arrival rate are the ANN [7,9,11,14], the autoregressive integrated moving average (ARIMA) [4,8,[12][13][14], the linear regression (LR) [14,18], the exponential smoothing [14,15,18], the logistic regression [5,10,11,19], the decision tree (DT) [1,10], the gradient boosted machines (GBM) [10], the Poisson regression model [16], the random forest (RF), the AdaBoost (AB), the support vector machine (SVM) [5], and the LSTM model [20]. ...
Full-text available
Article
Emergency departments (EDs) play a vital role in the whole healthcare system as they are the first point of care in hospitals for urgent and critically ill patients. Therefore, effective management of hospital’s ED is crucial in improving the quality of the healthcare service. The effectiveness depends on how efficiently the hospital resources are used, particularly under budget constraints. Simulation modeling is one of the best methods to optimize resources and needs inputs such as patients’ arrival time, patient’s length of stay (LOS), and the route of patients in the ED. This study develops a simulation model to determine the optimum number of beds in an ED by minimizing the patients’ LOS. The hospital data are analyzed, and patients’ LOS and the route of patients in the ED are determined. To determine patients’ arrival times, the features associated with patients’ arrivals at ED are identified. Mean arrival rate is used as a feature in addition to climatic and temporal variables. The exhaustive feature-selection method has been used to determine the best subset of the features, and the mean arrival rate is determined as one of the most significant features. This study is executed using the one-year ED arrival data together with five-year (43.824 study hours) ED arrival data to improve the accuracy of predictions. Furthermore, ten different machine learning (ML) algorithms are used utilizing the same best subset of these features. After a tenfold cross-validation experiment, based on mean absolute percentage error (MAPE), the stateful long short-term memory (LSTM) model performed better than other models with an accuracy of 47%, followed by the decision tree and random forest methods. Using the simulation method, the LOS has been minimized by 7% and the number of beds at the ED has been optimized.
... Emergency department (ED) crowding continues to be a problem in the United States and many other countries [1,2,3,4,5,6], which is associated with increased ambulance diversion and boarding -a practice of holding patients who have already been admitted in the ED until inpatient beds become available. This has profound implications for the quality and safety of the emergency care system: many studies have shown that ED crowding is associated with negative health outcomes in-cluding increased risk of mortality, longer wait time and length of stay, higher hospital costs, and patient dissatisfaction [5,7,8,9]. ...
... This has profound implications for the quality and safety of the emergency care system: many studies have shown that ED crowding is associated with negative health outcomes in-cluding increased risk of mortality, longer wait time and length of stay, higher hospital costs, and patient dissatisfaction [5,7,8,9]. Early identification of patients in need of hospital admission may improve ED efficiency and prevent negative patient outcomes [3]. ...
... Various prediction models have been proposed to identify factors related to hospital admission among ED patients [1,2,3,10,11]. However, previous studies have been limited for several reasons. ...
Full-text available
Article
Objective: To describe and compare logistic regression and neural network modeling strategies to predict hospital admission or transfer following initial presentation to Emergency Department (ED) triage with and without the addition of natural language processing elements. Methods: Using data from the National Hospital Ambulatory Medical Care Survey (NHAMCS), a cross-sectional probability sample of United States EDs from 2012 and 2013 survey years, we developed several predictive models with the outcome being admission to the hospital or transfer vs. discharge home. We included patient characteristics immediately available after the patient has presented to the ED and undergone a triage process. We used this information to construct logistic regression (LR) and multilayer neural network models (MLNN) which included natural language processing (NLP) and principal component analysis from the patient's reason for visit. Ten-fold cross validation was used to test the predictive capacity of each model and receiver operating curves (AUC) were then calculated for each model. Results: Of the 47,200 ED visits from 642 hospitals, 6,335 (13.42%) resulted in hospital admission (or transfer). A total of 48 principal components were extracted by NLP from the reason for visit fields, which explained 75% of the overall variance for hospitalization. In the model including only structured variables, the AUC was 0.824 (95% CI 0.818-0.830) for logistic regression and 0.823 (95% CI 0.817-0.829) for MLNN. Models including only free-text information generated AUC of 0.742 (95% CI 0.731- 0.753) for logistic regression and 0.753 (95% CI 0.742-0.764) for MLNN. When both structured variables and free text variables were included, the AUC reached 0.846 (95% CI 0.839-0.853) for logistic regression and 0.844 (95% CI 0.836-0.852) for MLNN. Conclusions: The predictive accuracy of hospital admission or transfer for patients who presented to ED triage overall was good, and was improved with the inclusion of free text data from a patient's reason for visit regardless of modeling approach. Natural language processing and neural networks that incorporate patient-reported outcome free text may increase predictive accuracy for hospital admission.
... One study examined the generalizability of a logistic regression model across disparate hospitals and retrained the model for each hospital. 14 Other similar studies focused on a single ED, 5,6,13,[15][16][17][18][19] used curated survey data, 10,20 or grouped multiple EDs together for model training. 8 Our study was limited in that only 1 health care system's EDs were represented, although the EDs were substantially variable in size and geography. ...
... Age (13) Temperature (5) Respiratory rate (16) Respiratory rate (7) Diastolic blood pressure (13) Age (20) Feature 10 (weight) Resuscitation status (12) Weakness (5) Diastolic blood pressure (10) Abdominal pain (6) Respiratory rate (13) Chest pain (18) a EKG, electrocardiogram; ESI, emergency severity index. ...
Full-text available
Article
Objective To assess the generalizability of a clinical machine learning algorithm across multiple emergency departments (EDs). Patients and Methods We obtained data on all ED visits at our health care system’s largest ED from May 5, 2018, to December 31, 2019. We also obtained data from 3 satellite EDs and 1 distant-hub ED from May 1, 2018, to December 31, 2018. A gradient-boosted machine model was trained on pooled data from the included EDs. To prevent the effect of differing training set sizes, the data were randomly downsampled to match those of our smallest ED. A second model was trained on this downsampled, pooled data. The model’s performance was compared using area under the receiver operating characteristic (AUC). Finally, site-specific models were trained and tested across all the sites, and the importance of features was examined to understand the reasons for differing generalizability. Results The training data sets contained 1918-64,161 ED visits. The AUC for the pooled model ranged from 0.84 to 0.94 across the sites; the performance decreased slightly when Ns were downsampled to match those of our smallest ED site. When site-specific models were trained and tested across all the sites, the AUCs ranged more widely from 0.71 to 0.93. Within a single ED site, the performance of the 5 site-specific models was most variable for our largest and smallest EDs. Finally, when the importance of features was examined, several features were common to all site-specific models; however, the weight of these features differed. Conclusion A machine learning model for predicting hospital admission from the ED will generalize fairly well within the health care system but will still have significant differences in AUC performance across sites because of site-specific factors.
... Scholars have increasingly focused on using predictive analytics to forecast ED waiting times, due to its superior predictive performance (Islam et al., 2018;Koh et al., 2011). Predictive analytics and data mining encompass various learning techniques aimed at extracting hidden and potentially useful information and patterns from data in large databases, and at making predictions from such data (Golmohammadi, 2016;Roquette et al., 2020;Friedman et al., 2001, chap. 1). ...
... The ANN method aims to simulate the human brain when collecting and processing data for the purpose of learning (Golmohammadi, 2016;Golmohammadi and Radnia, 2016). ...
Article
Emergency Departments (EDs) can better manage activities and resources and anticipate overcrowding through accurate estimations of waiting times. However, the complex nature of EDs imposes a challenge on waiting time prediction. In this paper, we test various machine learning techniques, using predictive analytics, applied to two large datasets from real EDs. We evaluate the predictive ability of Lasso, Random Forest, Support Vector Regression, Artificial Neural Network, and the Ensemble Method, using different error metrics and computational times. To improve the prediction accuracy, new queue-based variables, that capture the current state of the ED, are defined as additional predictors. The results show that the Ensemble Method is the most effective at predicting waiting times. In terms of both accuracy and computational efficiency, Random Forest is a reasonable trade-off. The results have significant practical implications for EDs and hospitals, suggesting that a real-time performance monitoring system that supports operational decision-making is possible.
... Second, it is necessary to estimate patients' remaining length of stay (LoS) within the ED in order to initiate bed allocation with a proper lead time. Fortunately, recent advances in predictive analytics methods and tools (including statistical and machine learning methods exploiting patient data from EHR systems) have led to a growing body of literature on effective future state prediction for ED patients (e.g., Qiu et al. 2014, Golmohammadi 2016 However, despite the growing attention being given to the prediction domain, the operationalization of prediction information has not been fully investigated. In addition, the literature reports widely varying prediction performance, depending on the maturity of the employed HIT systems as well as internal practices in making decisions. ...
... Figure 3 illustrates the ETI strategy for enabling proactive bed allocation by exploiting EHR data. The key assumption is that once a patient enters the ED and begins to undergo triage, testing, and treatment, there is adequate and increasing information about the patient within the EHR, including the patient's health history, to allow more reliable predictions of ED disposition decisions ahead of the actual disposition decisions (Golmohammadi 2016, Araz et al. 2019). This will enable the ED to proactively signal the relevant IU regarding an impending admission and need for a bed. ...
Full-text available
Article
Emergency departments (EDs) across the world are experiencing severe crowding and prolonged patient wait times for hospital admissions (a.k.a. patient “boarding”). Using data from a major healthcare system, we show that EDs suffer from severe boarding not only due to a high level of hospital inpatient bed occupancy but also due to reactive coordination of inpatient bed management activities. To reduce patient boarding, we explore early task initiation for the service network spanning the ED and inpatient units within a hospital. In particular, we investigate the value of predicting ED patient admissions (to be specific, disposition decisions) during the ED caregiving process to proactively initiate downstream tasks for reduced patient boarding. We show that the coordination mechanism can be modeled as a fork–join queueing system. The proposed modeling framework accounts for both imperfect patient disposition predictions and multiple hospital admission sources (in addition to the ED) for inpatient units. We maintain analytical tractability while preserving the complexities of real-world inpatient bed management operations by characterizing the state sets and transition sequences through the Markovian assumption. The proactive inpatient bed allocation scheme can lead to significant reductions in bed allocation delays for ED patients (nearly up to ∼50%) and does not increase delays for other admission sources. The insights from our model should guide hospital managers in embracing proactive coordination and adaptive workflow technologies enabled by modern health information technology systems and predictive analytics.
... Yet, queue-based information inside the process is often unavailable or difficult to extract from the event log (Van der Aalst et al., 2011a; Senderovich et al., 2015). Process mining seems a valuable solution to overcome such limitations (Van der Aalst, 2011b, 2016Senderovich et al., 2015;Senderovich et al.,2016). By undertaking process mining, it is possible to gain insights into the hospital processes, extracting the actual patient-flow and the crowding level of the activities within the ED. ...
... Most of the research focuses solely on data mining techniques for waiting time prediction in ED according to a small group of predictors based on patient characteristics and triage information (Golmohammadi, 2016;Ding et al., 2010). For example, the study of Sun et al. (2012) reports on a quantile regression model to predict individual patient's median and 95th percentile waiting time by using data available at triage (e.g. the end time of triage, the start time of visit, etc.) as predictors. ...
Article
Purpose The purpose of this study is twofold: exploring new queue-based variables enabled by process mining and evaluating their impact on the accuracy of waiting time prediction. Such queue-based predictors that capture the current state of the emergency department (ED) may lead to a significant improvement in the accuracy of the prediction models. Design/methodology/approach Alongside the traditional variables influencing ED waiting time, the authors developed new queue-based predictors exploiting process mining. Process mining techniques allowed the authors to discover the actual patient-flow and derive information about the crowding level of the activities. The proposed predictors were evaluated using linear and nonlinear learning techniques. The authors used real data from an ED. Findings As expected, the main results show that integrating the set of predictors with queue-based variables significantly improves the accuracy of waiting time prediction. Specifically, mean square error values were reduced by about 22 and 23 per cent by applying linear and nonlinear learning techniques, respectively. Practical implications Accurate estimates of waiting time can enable the ED systems to prevent overcrowding e.g. improving the routing of patients in EDs and managing more efficiently the resources. Providing accurate waiting time information also can lead to decreased patients’ dissatisfaction and elopement. Originality/value The novelty of the study relies on the attempt to derive queue-based variables reporting the crowding level of the activities within the ED through process mining techniques. Such information is often unavailable or particularly difficult to extract automatically, due to the characteristics of ED processes.
... Thinks this is a diverticulitis flare up. While multiple studies have described predictive models that use the health information available at the time of triage to make predictions about the eventual admission or discharge of a patient from the ED [4][5][6]8,9], until recently, there was a paucity of techniques that could capture the qualitative information contained within free-text data, such as nursing triage notes [4,5,[15][16][17][18][19][20][21][22]6,[8][9][10][11][12][13][14]. The purpose of this study was to predict inpatient admission to the hospital based on NLP of nursing triage notes in isolation and without the influence of other clinical data such as vital signs, past medical history, or SOAP notes which have been used in the past for this purpose. ...
... Recent advances in NLP have made possible the use of free-text as quantifiable numeric representations of language data for predictive modeling of patient outcomes. These predictive techniques have ranged in complexity from easy-to-use risk scores [10][11][12][13] to analytically complex logistic and Bayesian models [5,6,9,10,[14][15][16][17], random forests [4,[18][19][20], neural networks [4,8,14,15,18,21,22], and support vector techniques [20]. Because the triage process is intended to be fast, the availability of these data can be limited. ...
Article
Background: Nursing triage documentation is the first free-form text data created at the start of an emergency department (ED) visit. These 1-3 unstructured sentences reflect the clinical impression of an experienced nurse and are key in gauging a patient's illness. We aimed to predict final ED disposition using three commonly-employed natural language processing (NLP) techniques of nursing triage notes in isolation from other data. Methods: We constructed a retrospective cohort of all 260,842 consecutive ED encounters in 2015-16, from three clinically heterogeneous academically-affiliated EDs. After exclusion of 3964 encounters based on completeness of triage, and disposition data, we included 256,878 encounters. We defined the outcome as: 1) admission, transfer, or in-ED death [68,092 encounters] vs. 2) discharge, "left without being seen," and "left against medical advice" [188,786 encounters]. The dataset was divided into training and testing subsets. Neural network regression models were trained using bag-of-words, paragraph vectors, and topic distributions to predict disposition and were evaluated using the testing dataset. Results: Area under the curve for disposition using triage notes as bag-of-words, paragraph vectors, and topic distributions were 0.737 (95% CI: 0.734 - 0.740), 0.785 (95% CI: 0.782 - 0.788), and 0.687 (95% CI: 0.684 - 0.690), respectively. Conclusions: Nursing triage notes can be used to predict final ED patient disposition, even when used separately from other clinical information. These findings have substantial implications for future studies, suggesting that free text from medical records may be considered as a critical predictor in research of patient outcomes.
... Because of all these reasons, ED admission process has been addressed by researchers in the literature (Table 3). Golmohammadi (2016) developed a prediction model capable of estimating the likelihood of admission of each ED patient to the hospital (as inpatient) using logistic regression and ANN methods. On conclusion of the study, it was proved that an ED admission forecasting model based on demographic and clinical variables accurately estimated the likelihood of patient admission, thus decreased ED boarding and crowding. ...
... On conclusion of the study, it was proved that an ED admission forecasting model based on demographic and clinical variables accurately estimated the likelihood of patient admission, thus decreased ED boarding and crowding. As in Sun, Heng, Tay, and Seow (2011), Peck et al. (2013) and Golmohammadi (2016) applied logistic regression in their forecasting models. Peck et al. (2013) tested the generalisability for predicting the probabilities that ED patients will be admitted to a hospital inpatient unit. ...
Article
Emergency departments (EDs) provide medical treatment for a broad spectrum of illnesses and injuries to patients who arrive at all hours of the day. The quality and efficient delivery of health care in EDs are associated with a number of factors, such as patient overall length of stay (LOS) and admission, prompt ambulance diversion, quick and accurate triage, nurse and physician assessment, diagnostic and laboratory services, consultations and treatment. One of the most important ways to plan the healthcare delivery efficiently is to make forecasts of ED processes. The aim of this study is thus to provide an exhaustive review for ED stakeholders interested in applying forecasting methods to their ED processes. A categorization, analysis and interpretation of 102 papers is performed for review. This exhaustive review provides an insight for researchers and practitioners about forecasting in EDs in terms of showing current state and potential areas for future attempts.
... With the proposed hybrid ARIMA-LR model, it has been shown to outperform existing models in terms of prediction accuracy. Golmohammadi et al. [15] have developed a prediction model that can be used as a decision support tool and help reduce susceptibility to emergency services, using secondary data from the emergency service of a local hospital, with the help of logistic regression and neural networks. As a result, they stated that their study was able to accurately predict the likelihood of patient admission, thereby reducing emergency room entry and congestion, which are important problems in hospital operations. ...
Full-text available
Preprint
As the demand for medical care increases significantly every day, the issue of managing the volume of patients in hospitals and radiology units becomes more and more important. Due to the radiation emitted by the devices in the radiology unit, the minimum time spent by the patients for radiological imaging in hospitals is of vital importance both for the hospital and the patient. This study aims to estimate the monthly number of images in the hospital radiology unit using deep learning models and statistical-based models, so that it is prepared for the future in a more planned way. While deep learning models such as LSTM, MLP, NNAR and ELM were used for forecasting, statistics- based prediction models such as ARIMA, SES, TBATS, HOLT, and THETAF were also used. In order to evaluate the performance of the models, symmetric mean absolute percentage error (sMAPE), and mean absolute scaled error (MASE), which are very popular recently, were used. Results show that while the LSTM model performed better than the deep learning group in estimating the number of monthly radiological case images, the ARIMA model performed better in the statistical-based group. It is believed that the findings obtained will make important contributions to the future planning of the hospital by increasing both the service quality and patient satisfaction by facilitating the hospital managers in managing the volume of patients coming to the hospital and transferred to the radiology unit more efficiently.
... In this analysis, XGBoost and DNN displayed good AUC values when predicting ED patient hospitalizations. Golmohammadi [15] presented hospitalization predictions using LR, ANN, and a statistical method that patterned the similarity of patient characteristics to predict hospitalization. He showed that the overall accuracy of the three models was greater than 80%. ...
Full-text available
Article
Overcrowding in emergency departments (EDs) has long been a problem worldwide and has serious consequences for patient satisfaction and safety. Typically, overcrowding is caused by delays in the boarding time of ED patients waiting for inpatient beds. If the hospitalization of patients is predicted early enough in EDs, inpatient beds can be prepared in advance and the boarding time can be reduced. We design machine learning-based hospitalization predictive models using data on 27,747 patients and compare the experimental results. Five predictive models are designed: 1) logistic regression, 2) XGBoost, 3) NGBoost, 4) support vector machine, and 5) decision tree models. Based on the predictive results, we estimate the quantitative effects of hospitalization predictions on EDs and wards. Using the data from the ED of a general hospital in South Korea, our experiments show that the ED length of stay of a patient can be reduced by 12.3 minutes on average and the ED can reduce the total length of stay by 333,887 minutes for a year.
... Golmohammadi ve arkadaşları (Golmohammadi, 2016) çalışmaların-da yerel bir hastanenin acil servisinden alınan ikincil verileri kullanarak logistic Regression ve neural networks yardımıyla, karar destek aracı olarak kullanılabilecek ve acil servislere yatkınlığı azaltmaya yardımcı olacak bir tahmin modeli geliştirmişlerdir. Sonuçta, çalışmalarının hasta kabul olasılığını doğru bir şekilde tahmin edebildiğini ve böylece hastane operasyonlarında önemli sorunlar olan acil servise biniş ve tıkanıklığı azalttığını ifade etmişlerdir. ...
Full-text available
Chapter
Bu çalışma kapsamında üstün dielektrik dayanımı ve termal kararlılığı nedeniyle elektriksel yalıtkan olarak kullanılan polimerlerden biri olan PTFE malzemeleri temsilen PTFE (Teflon) bant yüzeyi için temas açısı ölçümleri gerçekleştirilmiştir. Bu bağlamda PTFE dielektrik malzeme yüzeyi için ıslanabilirlik ve buharlaşma hızı incelenmiştir. Elektrik alan ve nem varlığında elektriksel yaşlanma olaylarının başladığı, hızlandığı ve büyüdüğü göz önünde bulundurulduğunda elektriksel yalıtkan malzemelerde hidrofobisitenin önemli bir kavram olduğu aşikardır. PTFE dielektrik malzeme yüzeyi için temas açısı değerleri hidrofobik değerler aldığı gözlemlenmiştir. Damlacığın PTFE malzeme yüzeyine bırakıldıktan sonra durağan bir hal almasıyla birlikte uzunca bir süre ölçülen temas açısı değerlerinin hidrofobik değerlerde olduğu gözlemlenmiştir. Zamanla birlikte buharlaşmanın da etkisiyle damlacık yüksekliğinde ve temas açısında azalmalar gözlemlenmiştir. Damlacık yarıçapında meydana gelen azalmanın yükseklikte meydana gelen azalmanın yanında çok az olduğu gözlemlenmiştir
... Many of these concepts include data insights provided by predictive modeling or ML approaches which relies on data from clinical admissions [15][16][17]. We recognize that providing near-term bed demand forecasts to administrative personnel such as operating room, schedulers, inpatient bed coordinators, and operations managers can increase their ability to assertively maintain efficient levels of occupancy. ...
Full-text available
Article
Background Overcrowding is a serious problem that impacts the ability to provide optimal level of care in a timely manner. High patient volume is known to increase the boarding time at the emergency department (ED), as well as at post-anesthesia care unit (PACU). Furthermore, the same high volume increases inpatient bed transfer times, which causes delays in elective surgeries, increases the probability of near misses, patient safety incidents, and adverse events. Objective The purpose of this study is to develop a Machine Learning (ML) based strategy to predict weekly forecasts of the inpatient bed demand in order to assist the resource planning for the ED and PACU, resulting in a more efficient utilization. Methods The data utilized included all adult inpatient encounters at Geisinger Medical Center (GMC) for the last 5 years. The variables considered were class of inpatient encounter, observation, or surgical overnight recovery (SORU) at the time of their discharge. The ML based strategy is built using the K-means clustering method and the Support Vector Machine Regression technique (K-SVR). Results The performance obtained by the K-SVR strategy in the retrospective cohort amounts to a mean absolute percentage error (MAPE) that ranges between 0.49 and 4.10% based on the test period. Additionally, results present a reduced variability, which translates into more stable forecasting results. Conclusions The results from this study demonstrate the capacity of ML techniques to forecast inpatient bed demand, particularly using K-SVR. It is expected that the implementation of this model in the workflow of bed capacity management will create efficiencies, which will translate in a more reliable, inexpensive and timely care for patients.
... Chonde et al. (2013) developed and compared three models (e.g., artificial neural networks (ANNs), ordinal logistic regression (OLR), and naïve Bayesian networks (NBNs)) to predict the patient's emergency severity index (ESI) at EDs. ESI is a triage algorithm that organizes ED patients into 5 levels that reflect the severity of their symptoms (Tanabe et al., 2004). Golmohammadi (2016) implemented neural networks (NNs) and logistic regression models to identify the relationships among patients' characteristics such as age, radiology images, and the admission decision. Another study developed models to predict early readmissions to hospitals (Futoma et al., 2015). ...
Full-text available
Preprint
This work proposes a framework for optimizing machine learning algorithms. The practicality of the framework is illustrated using an important case study from the healthcare domain, which is predicting the admission status of emergency department (ED) patients (e.g., admitted vs. discharged) using patient data at the time of triage. The proposed framework can mitigate the crowding problem by proactively planning the patient boarding process. A large retrospective dataset of patient records is obtained from the electronic health record database of all ED visits over three years from three major locations of a healthcare provider in the Midwest of the US. Three machine learning algorithms are proposed: T-XGB, T-ADAB, and T-MLP. T-XGB integrates extreme gradient boosting (XGB) and Tabu Search (TS), T-ADAB integrates Adaboost and TS, and T-MLP integrates multi-layer perceptron (MLP) and TS. The proposed algorithms are compared with the traditional algorithms: XGB, ADAB, and MLP, in which their parameters are tunned using grid search. The three proposed algorithms and the original ones are trained and tested using nine data groups that are obtained from different feature selection methods. In other words, 54 models are developed. Performance was evaluated using five measures: Area under the curve (AUC), sensitivity, specificity, F1, and accuracy. The results show that the newly proposed algorithms resulted in high AUC and outperformed the traditional algorithms. The T-ADAB performs the best among the newly developed algorithms. The AUC, sensitivity, specificity, F1, and accuracy of the best model are 95.4%, 99.3%, 91.4%, 95.2%, 97.2%, respectively.
... Die Verwendung von KNN zeigte bereits in anderen, vergleichbaren Studien, z. B. bei der Optimierung der Triage, sehr gute Ergebnisse [6,13]. 3 8 "ROC curve" -KNN mit einer versteckten Schicht und 13 Neuronen, Schwellenwerte 0,3 und 0,5. ROC "receiver operating characteristics", AUROC "area under the ROC", KNN künstliche neuronale Netze somit eine ähnliche Größenordnung auf und der entsprechende F1-Score beträgt 0,67, was bedeutet, dass das KNN, obwohl es unter Nichtberücksichtigung der Unausgewogenheit des Datensatzes trainiert wurde, stationäre Fälle gut identifizieren kann und es nur beschränkt zu falschnegativen und falsch-positiven Klassifizierungen kommt. ...
Full-text available
Article
Zusammenfassung Hintergrund Krankenhäuser generieren einen Teil ihrer stationären Fälle aus ungeplanten Einweisungen über die zentrale Notfallambulanz (ZNA). Die Vorbereitung der Aufnahme benötigt üblicherweise eine ärztliche Entscheidung. Die resultierende Vorbereitungszeit für die Normalstation ist mitunter nicht ausreichend und es entstehen Verzögerungen. Ziel der Arbeit/Fragestellung Anhand der Prognose der Wahrscheinlichkeit einer stationären Aufnahme soll der potenzielle Nutzen des Einsatzes künstlicher neuronaler Netze (KNN) in der ZNA aufgezeigt werden. Dabei stellt sich die Frage, ob Routinedaten, welche in fast jeder ZNA bereits zum Zeitpunkt der Ersteinschätzung zur Verfügung stehen, einen Beitrag zur Reduktion von Verzögerungen bei der stationären Aufnahme leisten können. Material und Methoden Auf Grundlage von beschränkten und anonymisierten Routinedaten aus einem Krankenhausinformationssystem wird für eine ZNA ein KNN entwickelt, das die Vorhersage der stationären Aufnahme ermöglicht. Die Implementierung des KNN erfolgt über die Open-Source-Software R. Ergebnisse Unter Anwendung von Routinedaten erzielt das KNN eine Genauigkeit von 76,64 %. Die Sensitivität, d. h. der Anteil korrekt vorhergesagter Patientenaufnahmen, liegt bei 66,93 % und damit niedriger als die Spezifität (Anteil korrekt vorhergesagter Nichtaufnahmen), die 82,13 % beträgt. Diskussion Bereits unter Verwendung von Routinedaten können KNN einen wertvollen Beitrag für die Ablaufplanung in der ZNA leisten. Es ist zu erwarten, dass zusätzliche Variablen, wie z. B. das Patientenalter, die Prognosegüte steigern.
... Linear methods include Holt-Winters, Multiple Linear Regression (MLR), Exponential Smoothing (ES), Autoregressive Integrated Moving Average (ARIMA), Seasonal Autoregressive Integrated Moving Average (SARIMA), and some other regression-based methods [18,[21][22][23]. Nonlinear methods include the adaptive Neuro-Fuzzy Inference System (ANFIS), ANN, SVM, and LSTM [20,[24][25][26]. In addition to these two groups of methods, some hybrid approaches are developed for this problem to benefit from the advantages of the usage of these methods either individually or integrated, improve accuracy, and decrease modeling errors [12,14,27]. ...
Full-text available
Article
The overall service quality level of Emergency Departments (EDs) can be improved by accurate forecasting of patient visits. Accordingly, this study aims to evaluate the use of three metaheuristic approaches integrated with Artificial Neural Network (ANN) in forecasting daily ED visits. To do this, five performance measures are used for evaluating the accuracy of the proposed approaches, including Bayesian ANN, Genetic Algorithm-based ANN (GA-ANN), and Particle Swarm Optimization algorithm-based ANN (PSO-ANN). The outputs of this study show that the PSO-ANN model provides the most dominant performance in both the training and testing process. The lowest error is obtained with a mean absolute percentage error (MAPE) of 6.3%, Mean Absolute Error (MAE) of 42.797, Mean Squared Error (MSE) of 2499.340, Root Mean Square Error (RMSE) of 49.933, and R-squared (R²) of 0.824 on the training dataset. The lowest error with an MAPE of 6.0%, MAE of 40.888, MSE of 2839.998, RMSE of 53.292, and R² of 0.791 is also obtained on the testing process.
... The MLP consists of multiple parallel layers of nodes connected by weighted links. The input layer contains independent variables, the middle layer (hidden layer) contains processing units, and the output layer contains output variables [19]. The MLP model was designed with one input layer with five inputs, three hidden layers (with 64, 32, and 16 neurons, respectively), and one output layer. ...
Full-text available
Article
Recent developments in machine learning and deep learning have led to the use of multiple algorithms to make better predictions. Surgical units in hospitals allocate their resources for day surgeries based on the number of elective patients, which is mostly disrupted by emergency surgeries. Sixteen different models were constructed for this comparative study, including four simple and twelve hybrid models for predicting the demand for endocrinology, gastroenterology, vascular, urology, and pediatric surgical units. The four simple models used were seasonal autoregressive integrated moving average (SARIMA), support vector regression (SVR), multilayer perceptron (MLP), and long short-term memory (LSTM). The twelve hybrid models used were a combination of any two of the above-mentioned simple models, namely, SARIMA–SVR, SVR–SARIMA, SARIMA–MLP, MLP–SARIMA, SARIMA–LSTM, LSTM–SARIMA, SVR–MLP, MLP–SVR, SVR–LSTM, LSTM–SVR, MLP–LSTM, and LSTM–MLP. Data from the period 2012–2018 were used to build and test the models for each surgical unit. The results indicated that, in some cases, the simple LSTM model outperformed the others while, in other cases, there was a need for hybrid models. This shows that surgical units are unique in nature and need separate models for predicting their corresponding surgical volumes.
... Golmohammadi ve arkadaşları (Golmohammadi, 2016) çalışmaların-da yerel bir hastanenin acil servisinden alınan ikincil verileri kullanarak logistic Regression ve neural networks yardımıyla, karar destek aracı olarak kullanılabilecek ve acil servislere yatkınlığı azaltmaya yardımcı olacak bir tahmin modeli geliştirmişlerdir. Sonuçta, çalışmalarının hasta kabul olasılığını doğru bir şekilde tahmin edebildiğini ve böylece hastane operasyonlarında önemli sorunlar olan acil servise biniş ve tıkanıklığı azalttığını ifade etmişlerdir. ...
... Golmohammadi ve arkadaşları (Golmohammadi, 2016) çalışmaların-da yerel bir hastanenin acil servisinden alınan ikincil verileri kullanarak logistic Regression ve neural networks yardımıyla, karar destek aracı olarak kullanılabilecek ve acil servislere yatkınlığı azaltmaya yardımcı olacak bir tahmin modeli geliştirmişlerdir. Sonuçta, çalışmalarının hasta kabul olasılığını doğru bir şekilde tahmin edebildiğini ve böylece hastane operasyonlarında önemli sorunlar olan acil servise biniş ve tıkanıklığı azalttığını ifade etmişlerdir. ...
... With the assessment provided by the model, ED physicians would be more confident in their decisions regarding patient disposition. For patients who need hospital admission, the process can be initiated earlier according to the prediction given by the model to reduce ED boarding 24,25 . On the other hand, unnecessary examination and observation in the ED could be avoided for those who do not need hospital admission. ...
Full-text available
Article
Timely assessment to accurately prioritize patients is crucial for emergency department (ED) management. Urgent (i.e., level-3, on a 5-level emergency severity index system) patients have become a challenge since under-triage and over-triage often occur. This study was aimed to develop a computational model by artificial intelligence (AI) methodologies to accurately predict urgent patient outcomes using data that are readily available in most ED triage systems. We retrospectively collected data from the ED of a tertiary teaching hospital between January 1, 2015 and December 31, 2019. Eleven variables were used for data analysis and prediction model building, including 1 response, 2 demographic, and 8 clinical variables. A model to predict hospital admission was developed using neural networks and machine learning methodologies. A total of 282,971 samples of urgent (level-3) visits were included in the analysis. Our model achieved a validation area under the curve (AUC) of 0.8004 (95% CI 0.7963–0.8045). The optimal cutoff value identified by Youden's index for determining hospital admission was 0.5517. Using this cutoff value, the sensitivity was 0.6721 (95% CI 0.6624–0.6818), and the specificity was 0.7814 (95% CI 0.7777–0.7851), with a positive predictive value of 0.3660 (95% CI 0.3586–0.3733) and a negative predictive value of 0.9270 (95% CI 0.9244–0.9295). Subgroup analysis revealed that this model performed better in the nontraumatic adult subgroup and achieved a validation AUC of 0.8166 (95% CI 0.8199–0.8212). Our AI model accurately assessed the need for hospitalization for urgent patients, which constituted nearly 70% of ED visits. This model demonstrates the potential for streamlining ED operations using a very limited number of variables that are readily available in most ED triage systems. Subgroup analysis is an important topic for future investigation.
... Rather than replacing the decision-making of the ED triage team, research and implementation efforts are focused on augmenting triage capabilities such as training guidelines 24 and machine learning prediction of diagnosis, mortality, readmission, and length of stay. 25 While predicting hospital admission is also a common application of machine learning in emergency medicine, [26][27][28][29][30][31][32][33] prediction of ICU care need for admitted adult ED patients is a burgeoning area of research. 17,19,[34][35][36][37] The few studies that have been published on this topic either use national survey data, 19,34,35 include all mortality to the composite critical outcome, 17,19,34,36 or only subset adult patients in a narrow range of illness severity. ...
Full-text available
Article
Objective To develop prediction models for intensive care unit (ICU) vs non-ICU level-of-care need within 24 hours of inpatient admission for emergency department (ED) patients using electronic health record data. Materials and Methods Using records of 41 654 ED visits to a tertiary academic center from 2015 to 2019, we tested 4 algorithms—feed-forward neural networks, regularized regression, random forests, and gradient-boosted trees—to predict ICU vs non-ICU level-of-care within 24 hours and at the 24th hour following admission. Simple-feature models included patient demographics, Emergency Severity Index (ESI), and vital sign summary. Complex-feature models added all vital signs, lab results, and counts of diagnosis, imaging, procedures, medications, and lab orders. Results The best-performing model, a gradient-boosted tree using a full feature set, achieved an AUROC of 0.88 (95%CI: 0.87–0.89) and AUPRC of 0.65 (95%CI: 0.63–0.68) for predicting ICU care need within 24 hours of admission. The logistic regression model using ESI achieved an AUROC of 0.67 (95%CI: 0.65–0.70) and AUPRC of 0.37 (95%CI: 0.35–0.40). Using a discrimination threshold, such as 0.6, the positive predictive value, negative predictive value, sensitivity, and specificity were 85%, 89%, 30%, and 99%, respectively. Vital signs were the most important predictors. Discussion and Conclusions Undertriaging admitted ED patients who subsequently require ICU care is common and associated with poorer outcomes. Machine learning models using readily available electronic health record data predict subsequent need for ICU admission with good discrimination, substantially better than the benchmarking ESI system. The results could be used in a multitiered clinical decision-support system to improve ED triage.
... Apart from sex these were also reported as some of the most influential for predicting admission, particularly at an early stage. To further increase model performance numerous researchers included significant predictors such as vitals (LaMantia et al., 2010;Goto et al., 2019), pain scores (Barak-Corren et al., 2017b), anthropometrics (Barak-Corren et al., 2017a;Patel et al., 2018), medication (Barak-Corren et al., 2017a,b), radiology (Golmohammadi, 2016), and laboratory (Kim et al., 2014;Barak-Corren et al., 2017b) tests ordered. For one paediatric study that created models after 0, 10, 30 and 60 min, the inclusion of these types of predictors resulted in an Area Under the Curve (AUC) of 0.789 for 0 min up to an outstanding discrimination value of 0.913 at 60 min upon evaluation (Barak-Corren et al., 2017a). ...
Full-text available
Article
Introduction: Patients boarding in the Emergency Department can contribute to overcrowding, leading to longer waiting times and patients leaving without being seen or completing their treatment. The early identification of potential admissions could act as an additional decision support tool to alert clinicians that a patient needs to be reviewed for admission and would also be of benefit to bed managers in advance bed planning for the patient. We aim to create a low-dimensional model predicting admissions early from the paediatric Emergency Department. Methods and Analysis: The methodology Cross Industry Standard Process for Data Mining (CRISP-DM) will be followed. The dataset will comprise of 2 years of data, ~76,000 records. Potential predictors were identified from previous research, comprising of demographics, registration details, triage assessment, hospital usage and past medical history. Fifteen models will be developed comprised of 3 machine learning algorithms (Logistic regression, naïve Bayes and gradient boosting machine) and 5 sampling methods, 4 of which are aimed at addressing class imbalance (undersampling, oversampling, and synthetic oversampling techniques). The variables of importance will then be identified from the optimal model (selected based on the highest Area under the curve) and used to develop an additional low-dimensional model for deployment. Discussion: A low-dimensional model comprised of routinely collected data, captured up to post triage assessment would benefit many hospitals without data rich platforms for the development of models with a high number of predictors. Novel to the planned study is the use of data from the Republic of Ireland and the application of sampling techniques aimed at improving model performance impacted by an imbalance between admissions and discharges in the outcome variable.
... This limits the benefit of grouping high vs low risk of bias studies. Most studies had low applicability concern, except for six studies [26,30,38,41,46,49]. ...
Full-text available
Article
Background: The primary objective of this review is to assess the accuracy of machine learning methods in their application of triaging the acuity of patients presenting in the Emergency Care System (ECS). The population are patients that have contacted the ambulance service or turned up at the Emergency Department. The index test is a machine-learning algorithm that aims to stratify the acuity of incoming patients at initial triage. This is in comparison to either an existing decision support tool, clinical opinion or in the absence of these, no comparator. The outcome of this review is the calibration, discrimination and classification statistics. Methods: Only derivation studies (with or without internal validation) were included. MEDLINE, CINAHL, PubMed and the grey literature were searched on the 14th December 2019. Risk of bias was assessed using the PROBAST tool and data was extracted using the CHARMS checklist. Discrimination (C-statistic) was a commonly reported model performance measure and therefore these statistics were represented as a range within each machine learning method. The majority of studies had poorly reported outcomes and thus a narrative synthesis of results was performed. Results: There was a total of 92 models (from 25 studies) included in the review. There were two main triage outcomes: hospitalisation (56 models), and critical care need (25 models). For hospitalisation, neural networks and tree-based methods both had a median C-statistic of 0.81 (IQR 0.80-0.84, 0.79-0.82). Logistic regression had a median C-statistic of 0.80 (0.74-0.83). For critical care need, neural networks had a median C-statistic of 0.89 (0.86-0.91), tree based 0.85 (0.84-0.88), and logistic regression 0.83 (0.79-0.84). Conclusions: Machine-learning methods appear accurate in triaging undifferentiated patients entering the Emergency Care System. There was no clear benefit of using one technique over another; however, models derived by logistic regression were more transparent in reporting model performance. Future studies should adhere to reporting guidelines and use these at the protocol design stage. Registration and funding: This systematic review is registered on the International prospective register of systematic reviews (PROSPERO) and can be accessed online at the following URL: https://www.crd.york.ac.uk/PROSPERO/display_record.php?ID=CRD42020168696This study was funded by the NIHR as part of a Clinical Doctoral Research Fellowship.
... Their simulated results showed a significant improvement in reducing the waiting time. Other researchers have proposed to apply industrial engineering and operation research tools to reduce waiting time [42], and developing forecasting models to predict the number of patients in order to adapt hospital resources [43]. ...
Article
The objective of this paper is to examine how hospital logistics influence patient satisfaction. Based on the literature, we proposed that hospital logistics has four factors: physical accessibility, waiting time, hospital hotel services and administrative procedures. A PLS-SEM model was developed and evaluated using data from a survey of 21 6 patients conducted in three public hospitals in Fez-Morocco. It was found that hospital logistics positively impacts patient satisfaction, which proves its role as a key driver of patient satisfaction. Consequently, healthcare decision-makers are called upon to pay more attention to logistics activities at the hospital to, among other things, improve patient satisfaction.
... Indeed, Gupta and Denton [45] proposed to apply industrial engineering and operational research tools to optimize appointment planning. Golmohammadi [46] developed statistical and mathematical models capable of predicting the number of patients coming from the emergency department in order to prepare for their reception in the care units. The objective was to reduce the waiting time for preparing patients' reception. ...
Article
The purpose of this qualitative research was to explore the influence of hospital logistics on quality of care and patient satisfaction. Hospital logistics was assessed by considering five factors in the patient pathway, namely: physical accessibility of care, waiting time, consultation time, administrative procedures and hospital hotel services. Semi-structured interviews were conducted, following an interview guide, with two categories of participants: patients and healthcare professionals. The interviews were transcribed and then a thematic analysis method was conducted using QSR NVivo 10 software. Results showed that hospital logistics has a direct impact on quality of care and patient satisfaction. All the participants' testimonies highlighted the critical and crucial role that hospital logistics plays in the perception of quality of care and patient satisfaction. It is recommended to pay great attention to hospital logistics activities in order to improve quality and satisfaction.
... These initial findings provide the basis to evaluate this model of There are a number of important implications for these findings. Firstly, although there are numerous analytics and risk prediction tools in use clinically, [12][13][14] and several that have reported admission prediction models, [15][16][17][18] this is the first initiative in the Australian context to demonstrate performance improvements associated with real-time use of a clinical analytics tool. It should be noted that the tool itself was not designed to directly alter patient flow. ...
Full-text available
Article
Background: The Sydney Triage to Admission Risk Tool (START) is a validated clinical analytics tool designed to estimate the probability of in-patient admission based on Emergency Department triage characteristics. Methods: This was a single centre pilot implementation study using a matched case control sample of patients assessed at ED triage. Patients in the intervention group were identified at triage by the START tool as likely requiring in-patient admission and briefly assessed by an ED Consultant. Bed management were notified of these patients and their likely admitting team based on senior early assessment. Matched controls were identified on the same day of presentation if they were admitted to the same in-patient teams as patients in the intervention group and same START score category. Outcomes were ED length of stay and proportion of patients correctly classified as an in-patient admission by the START tool. Results: One hundred and thirteen patients were assessed using the START-based model of care. When compared with matched control patients, this intervention model of care was associated with a significant reduction in ED length of stay [301 min (IQR 225-397) versus 423 min (IQR 297-587) p < 0.001] and proportion of patients meeting 4 h length of stay thresholds increased from 24 to 45% (p < 0.001). Conclusion: In this small pilot implementation study, the START tool, when used in conjunction with senior early assessment was associated with a reduction in ED length of stay. Further controlled studies are now underway to further examine its utility across other ED settings.
... ANNs are mathematical models that mimic the process of learning in the human brain by collecting and processing data. ANNs are able to find complex input-output relationships without assumptions of normality, linearity and variable independents (Golmohammadi, 2016;Thomaidis and Dounias, 2012). ...
Article
Purpose This study aims to investigate the factors affecting daily demand in an emergency department (ED) and to provide a forecasting tool in a public hospital for horizons of up to seven days. Design/methodology/approach In this study, first, the important factors to influence the demand in EDs were extracted from literature then the relevant factors to the study are selected. Then, a deep neural network is applied to constructing a reliable predictor. Findings Although many statistical approaches have been proposed for tackling this issue, better forecasts are viable by using the abilities of machine learning algorithms. Results indicate that the proposed approach outperforms statistical alternatives available in the literature such as multiple linear regression, autoregressive integrated moving average, support vector regression, generalized linear models, generalized estimating equations, seasonal ARIMA and combined ARIMA and linear regression. Research limitations/implications The authors applied this study in a single ED to forecast patient visits. Applying the same method in different EDs may give a better understanding of the performance of the model to the authors. The same approach can be applied in any other demand forecasting after some minor modifications. Originality/value To the best of the knowledge, this is the first study to propose the use of long short-term memory for constructing a predictor of the number of patient visits in EDs.
... In the literature, there have been many studies which used different functions of data mining such as for clustering the patients, [3][4][5] classifying them, 6 or generating predictions. [7][8][9] However, to the best of the knowledge, use of association analysis or association rule mining (ARM) is very rare in ED context. ...
Full-text available
Article
Diagnostic tests are widely used in emergency departments to make detailed investigations on diagnosis and treat patients correctly. However, since these tests are expensive and time-consuming, ordering correct tests for patients is crucial for efficient use of hospital resources. Thus, understanding the relation between diagnosis and diagnostic test requirement becomes an important issue in emergency departments. Association rule mining was used to extract hidden patterns and relation between diagnosis and diagnostic test requirement in real-life medical data received from an emergency department. Apriori was used as an association rule mining algorithm. Diagnosis was grouped into 21 categories based on International Classification of Disease, and laboratory tests were grouped into four main categories (hemogram, biochemistry, cardiac enzyme, urine and human excrement related). Both positive and negative rules were discovered. Since the nature of the data had the dominance of negative values, higher number of negative rules with higher confidences were discovered compared to positive ones. The extracted rules were validated by emergency department experts and practitioners. It was concluded that understanding the association between patient’s diagnosis and diagnostic test requirement can improve decision-making and efficient use of resources in emergency departments. Association rules can also be used for supporting physicians to treat patients.
... ANNs are mathematical models that mimic the process of learning in the human brain by collecting and processing data. ANNs are able to find complex input-output relationships without assumptions of normality, linearity and variable independents (Golmohammadi, 2016;Thomaidis and Dounias, 2012). ...
Article
Purpose The complexity and interdisciplinarity of healthcare industry problems make this industry one of the attention centers of computer-based simulation studies to provide a proper tool for interaction between decision-makers and experts. The purpose of this study is to present a metamodel-based simulation optimization in an emergency department (ED) to allocate human resources in the best way to minimize door to doctor time subject to the problem constraints which are capacity and budget. Design/methodology/approach To obtain the objective of this research, first the data are collected from a public hospital ED in Brazil, and then an agent-based simulation is designed and constructed. Afterwards, three machine-learning approaches, namely, adaptive neuro-fuzzy inference system (ANFIS), feed forward neural network (FNN) and recurrent neural network (RNN), are used to build an ensemble metamodel through adaptive boosting. Finally, the results from the metamodel are applied in a discrete imperialist competitive algorithm (ICA) for optimization. Findings Analyzing the results shows that the yellow zone section is considered as a potential bottleneck of the ED. After 100 executions of the algorithm, the results show a reduction of 24.82 per cent in the door to doctor time with a success rate of 59 per cent. Originality/value This study fulfils an identified need to optimize human resources in an ED with less computational time.
... The performance of neural network models has been found superior to traditional linear statistical models, such as MLR and structural equations modeling (Chong et al., 2013;Sharma et al., 2015). Furthermore, the neural network model can overcome the mandatory assumptions, such as linearity, normality and independence of predictors used in traditional statistical models (Golmohammadi, 2016). The back propagation neural network is a commonly used model in business research. ...
Article
Purpose The purpose of this study is to examine the predictabilities of five intra-personal factors to predict pro-environmental consumer behavior (PECB) and the moderating role of religiosity in Oman. Design/methodology/approach The study uses neural network to analyze the antecedents/antecedents × religiosity → PECB relationships by using a sample of 306 consumers from Oman. Findings This study finds that the most important predictors of PECB, according to the order of importance, are attitude × religiosity, knowledge, concern × religiosity, knowledge × religiosity, value, religiosity, attitude, concern and value × religiosity. Research limitations/implications The convenience sample from a single Islamic country limits the generalizability of the findings. Future studies should use probabilistic sampling techniques and multiple Islamic countries located in different geographical regions. Practical implications To promote PECB, businesses and policymakers should provide environmental education to expand knowledge and value, leverage ecological religious values in integrated marketing communications, make positive inducements to change attitude and concern enhancing interventions. Social implications As religiosity enhances PECB by moderating the impacts of environmental intra-personal factors on PECB, businesses and policymakers should find ways to use faith-based ecological messages in Islamic countries. Originality/value Determining the predictabilities of psychological factors and their interactions with religiosity to predict PECB in Islamic countries is necessary for promoting environmentally friendly products in Islamic countries and for reducing the ecological damage to the environment.
... Avec ces liens, il peut atteindre le niveau de la projection en prévoyant le nombre de transferts possible dans les prochains jours. Il est important de mentionner que certains éléments comme le nombre futur de lits occupés ou d'admissions peuvent être prédits par des approches statistiques traditionnelles [2,15]. À la suite de l'analyse de la tàche dirigée par les buts, un schéma du système causal de l'USIP a été réalisé. ...
Full-text available
Conference Paper
Le tableau de bord de gestion (TBG) est un outil visuel utilisé par les gestionnaires pour réaliser le suivi de la performance organisationnelle. Il consiste à afficher de l’information portant sur l’état de l’organisation afin de soutenir la prise de décision. Le TBG doit ainsi permettre au gestionnaire d’obtenir une représentation mentale véridique et complète de l’état de la situation. Le modèle théorique de la conscience de la situation (CS) caractérise cette représentation mentale en trois niveaux : perception, compréhension et projection. La prise de décision s’appuie en grande partie sur la capacité à anticiper l’état futur de l’environnement qui réfère au niveau de la projection de la CS. Cet article vise donc à proposer une démarche de conception de TBG pour soutenir non seulement la perception et la compréhension, mais aussi la projection.
Article
Overcrowding in emergency department (ED) causes lengthy waiting times, reduces adequate emergency care and increases rate of mortality. Accurate prediction of daily ED visits and allocating resources in advance is one of the solutions to ED overcrowding problem. In this paper, a deep stacked architecture is being proposed and applied to the daily ED visits prediction problem with deep components such as Long Short Term Memory (LSTM), Gated Recurrent Units (GRU) and simple Recurrent Neural Network (RNN). The proposed architecture achieves very high mean accuracy level (94.28–94.59%) in daily ED visits predictions. We have also compared the performance of this architecture with non-stacked deep models and traditional prediction models. The results indicate that deep stacked models outperform (4–7%) the traditional prediction models and other non-stacked deep learning models (1–2%) in our prediction tasks. The application of deep neural network in ED visits prediction is novel as this is one of the first studies to apply a deep stacked architecture in this field. Importantly, our models have achieved better prediction accuracy (in one case comparable) than the state-of-the-art in the literature.
Article
This work proposes a framework for optimizing machine learning algorithms. The practicality of the framework is illustrated using an important case study from the healthcare domain, which is predicting the admission status of emergency department (ED) patients (e.g., admitted vs. discharged) using patient data at the time of triage. The proposed framework can mitigate the crowding problem by proactively planning the patient boarding process. A large retrospective dataset of patient records is obtained from the electronic health record database of all ED visits over three years from three major locations of a healthcare provider in the Midwest of the US. Three machine learning algorithms are proposed: T-XGB, T-ADAB, and T-MLP. T-XGB integrates extreme gradient boosting (XGB) and Tabu Search (TS), T-ADAB integrates Adaboost and TS, and T-MLP integrates multi-layer perceptron (MLP) and TS. The proposed algorithms are compared with the traditional algorithms: XGB, ADAB, and MLP, in which their parameters are tunned using grid search. The three proposed algorithms and the original ones are trained and tested using nine data groups that are obtained from different feature selection methods. In other words, 54 models are developed. Performance was evaluated using five measures: Area under the curve (AUC), sensitivity, specificity, F1, and accuracy. The results show that the newly proposed algorithms resulted in high AUC and outperformed the traditional algorithms. The T-ADAB performs the best among the newly developed algorithms. The AUC, sensitivity, specificity, F1, and accuracy of the best model are 95.4%, 99.3%, 91.4%, 95.2%, 97.2%, respectively.
Article
The volume of daily patient arrivals at Emergency Departments (EDs) is unpredictable and is a significant reason of ED crowding in hospitals worldwide. Timely forecast of patients arriving at ED can help the hospital management in early planning and avoiding of overcrowding. Many different ED patient arrivals forecasting models have previously been proposed by using time series analysis methods. Even though the time series methods such as Linear and Logistic Regression, Autoregressive Integrated Moving Average (ARIMA), Seasonal ARIMA (SARIMA), Exponential Smoothing (ES), and Artificial Neural Network (ANN) have been explored extensively for the ED forecasting model development, the few significant limitations of these methods associated in the analysis of time series data make the models inadequate in many practical situations. Therefore, in this paper, ML-based Random Forest (RF) regressor, and DNN-based Long Short-term Memory (LSTM) and Convolutional Neural network (CNN) methods, which have not been explored to the same extent as the other time series techniques, are implemented by incorporating meteorological and calendar parameters for the development of forecasting models. The performances of developed three models in forecasting the ED patient arrivals are evaluated. Among the three models, CNN outperformed for short-term (3 days in advance) patient arrivals prediction with MAPE of 9.24% and LSTM performed better for moderate-term (7 days in advance) patient arrivals prediction with MAPE of 8.91% using weather forecast information. Whereas, LSTM model outperformed with MAPE of 8.04% compared to 9.53% by CNN and 10.10% by RF model for current day prediction of patient arrivals using 3 days past weather information. Thus, for short-term ED patient arrivals forecasting, DNN-based model performed better compared to RF regressor ML-based model.
Full-text available
Conference Paper
Fluctuating patient arrivals and bed management in Inpatient departments of hospitals necessitated hospital management to understand the dynamics of patient flow so as to optimally engage in resource allocation and strategic planning. Within such a framework, the ability to analyze patient admissions, length of stay and predict demand has significant implications in decision making. In this study, data of 908 patients with respect to age, gender, date and time of admissions and discharge from inpatient department over a period of three months has been recorded from a well- established multi-speciality level II referral hospital in Bengaluru city with 52 inpatient beds. Details were recorded in a spreadsheet. The hospital operates in three working shifts, each with a duration of eight hours. Admission distribution by day of the week and length of stay has been analyzed. The computed bed occupancy rate provides an understanding on utilization and planning deployment of beds, which is a very important resource in a hospital setting. Data was divided into training and test dataset. Time series models are used to model patient admissions and discharge in the inpatient setting using MS Excel and R-studio. This kind of an insight will enable health care units to be better prepared to handle the demand and also plan suitable deployment of resources to enable better planning, thereby improving their operational efficiency.
Article
Background Since providing timely care is the primary concern of emergency departments (EDs), long waiting times increase patient dissatisfaction and adverse outcomes. Especially in overcrowded ED environments, emergency care quality can be significantly improved by developing predictive models of patients' waiting and treatment times to use in ED operations planning. Methods Retrospective data on 37,711 patients arriving at the ED of a large urban hospital were examined. Ordinal logistic regression models were proposed to identify factors causing increased waiting and treatment times and classify patients with longer waiting and treatment times. Results According to the proposed ordinal logistic regression model for waiting time prediction, age, arrival mode, and ICD-10 encoded diagnoses are all significant predictors. The model had 52.247% accuracy. The model for treatment time showed that in addition to age, arrival mode, and diagnosis, triage level was also a significant predictor. The model had 66.365% accuracy. The model coefficients had negative signs in the corresponding models, indicating that waiting times are negatively related to treatment times. Conclusion By predicting patients' waiting and treatment times, ED workloads can be assessed instantly. This enables ED personnel to be scheduled to better manage demand supply deficiencies, increase patient satisfaction by informing patients and relatives about expected waiting times, and evaluate performances to improve ED operations and emergency care quality.
Full-text available
Article
Reducing patient waits is vital to respond for the increasing demand for quality health services. Toward this endeavour, the aim of this research is to explore the current situation of application of lean six sigma (LSS) techniques in the healthcare sector as a strategy to address negative effects of long waiting of patients. The study reviews the outcomes of application LSS techniques to increase patient's satisfaction with healthcare. A comprehensive review of the literature dedicated to apply LSS in healthcare is used to generate a synthesis of the literature around the chosen research aim. The review focuses on research addressed various types of issues, which can cause increasing waiting time at healthcare centres. The study confirms the significant of LSS application in reducing waiting time of patients at healthcare centres.
Article
Background Emergency departments (ED) are a portal of entry into the hospital and are uniquely positioned to influence the health care trajectories of older adults seeking medical attention. Older adults present to the ED with distinct needs and complex medical histories, which can make disposition planning more challenging. Machine Learning (ML) approaches have been previously used to inform decision-making surrounding ED disposition in the general population. However, little is known about the performance and utility of ML methods in predicting hospital admission among older ED patients. We applied a series of ML algorithms to predict ED admission among older adults and discuss their clinical and policy implications. Materials and Methods We analyzed the Canadian data from the interRAI multinational ED study, the largest prospective cohort study of older ED patients to date. The data included 2,274 ED patients 75 years of age and older from eight ED sites across Canada between November 2009 and April 2012. Data were extracted from the interRAI ED Contact Assessment, with predictors including a series of geriatric syndromes, functional assessments, and baseline care needs. We applied a total of five ML algorithms. Models were trained, assessed, and analyzed using 10-fold cross-validation. The performance of predictive models was measured using the area under the receiver operating characteristic curve (AUC). We also report the accuracy, sensitivity, and specificity of each model to supplement performance interpretation. Results Gradient boosted trees was the most accurate model to predict older ED patients who would require hospitalization (AUC = 0.80). The five most informative features include home intravenous therapy, time of ED presentation, a requirement for formal support services, independence in walking, and the presence of an unstable medical condition. Conclusion To the best of our knowledge, this is the first study to predict hospital admission in older ED patients using a series of geriatric syndromes and functional assessments. We were able to predict hospital admission in older ED patients with good accuracy using the items available in the interRAI ED Contact Assessment. This information can be used to inform decision-making about ED disposition and may expedite admission processes and proactive discharge planning.
Article
Background Emergency departments (EDs) play an important role in health systems since they are the front line for patients with emergency medical conditions who frequently require diagnostic tests and timely treatment. Objective To improve decision-making and accelerate processes in EDs, this study proposes predictive models for classifying patients according to whether or not they are likely to require a diagnostic test based on referral diagnosis, age, gender, triage category and type of arrival. Method Retrospective data were categorised into four output patient groups: not requiring any diagnostic test (group A); requiring a radiology test (group B); requiring a laboratory test (group C); requiring both tests (group D). Multivariable logistic regression models were used, with the outcome classifications represented as a series of binary variables: test (1) or no test (0); in the case of group A, no test (1) or test (0). Results For all models, age, triage category, type of arrival and referral diagnosis were significant predictors whereas gender was not. The main referral diagnosis with high model coefficients varied by designed output groups (groups A, B, C and D). The overall accuracies of the logistic regression models for groups A, B, C and D were, respectively, 74.11%, 73.07%, 82.47% and 85.79%. Specificity metrics were higher than the sensitivities for groups B, C and D, meaning that these models were better able to predict negative outcomes. Implications These results provide guidance for ED triage staff, researchers and practitioners in making rapid decisions regarding patients’ diagnostic test requirements based on specified variables in the predictive models. This is critical in ED operations planning as it potentially decreases waiting times, while increasing patient satisfaction and operational performance.
Chapter
In this study, linear regression and neural network-based hybrid models are developed for modelling the daily ED visits. Month and week of the year, day of the week, and period of the day, are used as input variables of the linear regression model. Generated forecasts and the residuals are further processed through a multilayer perceptron model to improve the performance of forecasting. To obtain forecasts for daily number of patient visits, aggregation is used where the obtained periodical forecasts are summed up. By comparing the performances of models in generating periodical and daily forecasts, this chapter not only shows that hybrid model improves the forecasting performance significantly, but also aggregation fits well in practice.
Article
Balancing bed allocation is a critical but cumbersome decision-making process in hospitals due to limited capacity, fluctuations in the rate of patient arrival and service interactions among various units; typically, this will cause blockages in multi-stage healthcare services. Accurately estimating the blocking probability is an important task in order to improve the performance of healthcare systems. Early studies assumed either unlimited bed capacity or no service interaction among units. In this study, we consider the correlation between the blockage and service time of the subsequent stage and apply a multi-stage tandem-queuing model with limited bed capacity and service interactions to model healthcare systems. We develop two effective heuristics to estimate the patient-blocking probability, which are then used to develop an integrated mathematical model for bed allocation. We collect real-world data from a tertiary hospital in China to delineate the effect of service interactions while estimating the blocking probability and use non-parametric rank-sum tests to verify and compare the relative performances of the proposed model against two popular heuristics. Our comparative results illustrate that the proposed model is as accurate as simulations. We also observe that increasing the number of beds during the first stage is more effective in reducing blockage than doing so later in case of a limited number of beds.
Article
Emergency departments (EDs) are the largest departments of hospitals which encounter high variety of cases as well as high level of patient volumes. Thus, an efficient classification of those patients at the time of their registration is very important for the operations planning and management. Using secondary data from the ED of an urban hospital, we examine the significance of factors while classifying patients according to their length of stay. Random Forest, Classification and Regression Tree, Logistic Regression (LR), and Multilayer Perceptron (MLP) were adopted in the data set of July 2016, and these algorithms were tested in data set of August 2016. Besides adopting and testing the algorithms on the whole data set, patients in these sets were grouped into 21 based on the similarities in their diagnoses and the algorithms were also performed in these subgroups. Performances of the classifiers were evaluated based on the sensitivity, specificity, and accuracy. It was observed that sensitivity, specificity, and accuracy values of the classifiers were similar, where LR and MLP had somehow higher values. In addition, the average performance of the classifying patients within the subgroups outperformed the classifying based on the whole data set for each of the classifiers.
Conference Paper
A management dashboard is a visual tool used by managers to monitor organizational performance. It involves displaying information about the state of the organization to decision-making. A management dashboard must thus enable the manager to obtain a veracious and complete mental representation of the state of a situation. The theoretical model of situational awareness (SA) characterizes this mental representation in three levels: perception, comprehension and projection. Decision-making relies heavily on the ability to anticipate the future state of the environment, which refers to the SA level of projection. This article aims to propose a management dashboard design approach to support not only the levels of perception and comprehension, but also projection.
Article
Objective: The present study aims to prospectively validate the Sydney Triage to Admission Risk Tool (START) to predict ED disposition. Methods: This was a prospective validation study at two metropolitan EDs in Sydney, Australia. Consecutive triage encounters were observed by a trained researcher and START scores calculated. The primary outcome was patient disposition (discharge or inpatient admission) from the ED. Multivariable logistic regression was used to estimate area under curve of receiver operator characteristic (AUC ROC) for START scores as well as START score in combination with other variables such as frailty, general practitioner referral, overcrowding and major medical comorbidities. Results: There were 894 patients analysed during the study period. The START score when applied to the data had AUC ROC of 0.80 (95% CI 0.77-0.83). The inclusion of other clinical variables identified at triage did not improve the overall performance of the model with an AUC ROC of 0.81 (95% CI 0.78-0.84) in the present study. Conclusion: The overall performance of the START tool with respect to model discrimination and accuracy has been prospectively validated. Further clinical trials are required to test the clinical effectiveness of the tool in improving patient flow and overall ED performance.
Full-text available
Article
An efficient decision-making model was developed to select suppliers using multi-layer feed forward neural networks. A set of input functions for supplier selection criteria was defined to create input data for training the model. Both types of criteria, qualitative and quantitative, were considered in the model. Fuzzy techniques were applied to convert qualitative data to quantitative data. Pairwise comparisons matrices were applied for output values and weight assignment. The neural network model structure was designed and tested based on backpropagation. The results of the neural network model indicated that the proper structure of the model had a crucial effect on its performance. The selection of appropriate initial weights, learning rate and momentum were critical in improving the model performance. To prove the capability of the proposed model, suppliers of three products were ranked based on the proposed model and the results were compared with the managers’ ranking. The proposed neural network model can use historical data of suppliers to evaluate their performance in the vendor supplier selection decision. The vendor can update the suppliers’ database information over time for future decisions.
Full-text available
Article
This paper, using detailed time measurements of patients complemented by interviews with hospital management and staff, examines three facets of an emergency room's (ER) operational performance: (1) effectiveness of the triage system in rationing patient treatment; (2) factors influencing ER's operational performance in general and the trade-offs in flow times, inventory levels (that is the number of patients waiting in the system), and resource utilization; (3) the impacts of potential process and staffing changes to improve the ER's performance. Specifically, the paper discusses four proposals for streamlining the patient flow: establishing designated tracks ("fast track," "diagnostic track"), creating a "holding" area for certain type of patients, introducing a protocol that would reduce the load on physicians by allowing a registered nurse to order testing and treatment for some patients, and potentially and in the longer term, moving from non-ER specialist physicians to ER specialists. The paper's findings are based on analyzing the paths and flow times of close to two thousand patients in the emergency room of the Medical Center of Leeuwarden (MCL), The Netherlands. Using exploratory data analysis the paper presents generalizable findings about the impacts of various factors on ER's lead-time performance and shows how the proposals fit with well-documented process improvement theories.
Full-text available
Article
In this paper, a fuzzy multi-criteria decision making model is presented based on a feed forward artificial neural network. This model is used to capture and represent the decision makers' preferences. The topology of the neural network model is developed to train the model. The proposed model can use historical data and update the database information for alternatives over time for future decisions. Basically, multi-criteria decision making problems are formulated, and neural network is used to learn the relation among criteria and alternatives and rank the alternatives. We do not use any utility function for the modeling; however, a unique method is proposed for eliciting the information from decision makers. The proposed model is applicable for a wide variety of multi-attribute decision making problems and can be used for future ranking or selection without managers' judgment effort. Simulation of the managers' decisions is demonstrated in detail and the design and implementation of the model are illustrated by a case study.
Full-text available
Article
The problem of emergency department (ED) overcrowding has reached crisis proportions in the last decade. In 2005, the National Academy of Engineering and the Institute of Medicine reported on the important role of simulation as a systems analysis tool that can have an impact on care processes at the care-team, organizational, and environmental levels. Simulation has been widely used to understand causes of ED overcrowding and to test interventions to alleviate its effects. In this paper, we present a systematic review of ED simulation literature from 1970 to 2006 from healthcare, systems engineering, operations research and computer science publication venues. The goals of this review are to highlight the contributions of these simulation studies to our understanding of ED overcrowding and to discuss how simulation can be better used as a tool to address this problem. We found that simulation studies provide important insights into ED overcrowding but they also had major limitations that must be addressed.
Full-text available
Article
Access block affecting the emergency department (ED), also known as boarding in the United States and Canada, can be described as a phenomenon comprising almost all the challenges in the world of modern EDs. We use the analogy of parallel universes to illustrate both the complexity and the severity of the problem. In the world of physics, many attempts have been made to create a mathematical solution that can answer the more basic questions about physical phenomena in the universe. This has been known as ‘Theory of Everything’. Albert Einstein spent 30 years of his life trying to solve this ‘Theory of Everything’, but failed [1].
Full-text available
Article
Access block refers to the situation where patients in the emergency department (ED) requiring inpatient care are unable to gain access to appropriate hospital beds within a reasonable time frame. We systematically evaluated the relationship between access block, ED overcrowding, ambulance diversion, and ED activity. This was a retrospective analysis of data from the Emergency Department Information System for the three major central metropolitan EDs in Perth, Western Australia, for the calendar years 2001-2. Bivariate analyses were performed in order to study the relationship between a range of emergency department workload variables, including access block (>8 hour total ED stay for admitted patients), ambulance diversion, ED overcrowding, and ED waiting times. We studied 259,580 ED attendances. Total diversion hours increased 74% from 3.39 hours/day in 2001 to 5.90 hours/day in 2002. ED overcrowding (r = 0.96; 95% confidence interval (CI) 0.91 to 0.98), ambulance diversion (r = 0.75; 95% CI 0.49 to 0.88), and ED waiting times for care (r = 0.83; 95% CI 0.65 to 0.93) were strongly correlated with high levels of ED occupancy by access blocked patients. Total attendances, admissions, discharges, and low acuity patient attendances were not associated with ambulance diversion. Reducing access block should be the highest priority in allocating resources to reduce ED overcrowding. This would result in reduced overcrowding, reduced ambulance diversion, and improved ED waiting times. Improving hospital inpatient flow, which would directly reduce access block, is most likely to achieve this.
Full-text available
Article
Five-point triage assessment scales currently used in many emergency departments (EDs) across the country have been shown to be accurate and reliable. We have found the system to be highly predictive of outcome (hospital admission, intensive care unit/operating room admission, or death) at either extreme of the scale but much less predictive in the middle triage group. This is problematic because the middle triage acuity group is the largest, in our experience comprising almost half of all patients. Patients triaged to the 2 highest acuity categories (A and B) have admission/ED death rates of 76% and 43%, respectively. In contrast, the 2 lowest acuity categories (D and E) have admission/ED death rates of 1% or less. The middle category (C), however, has an overall admission/ED death rate of 10%, too high to be comfortable with prolonged delays in the ED evaluation of these patients. We studied this group to determine if easily obtainable clinical factors could identify higher-risk patients in this heterogeneous category. Data were obtained from a retrospective, cross-sectional study of all patients seen in 2001 at an urban academic hospital ED. The main outcome measure for multivariate logistic regression models was hospital admission among patients triaged as acuity C. Acuity C patients who were 65 years or older, presenting with weakness or dizziness, shortness of breath, abdominal pain, or a final diagnosis related group diagnosis of psychosis, were more likely to be admitted than patients originally triaged in category B. These findings suggest that a few easily obtainable clinical factors may significantly improve the accuracy of triage and resource allocation among patients assigned with a middle-acuity score.
Article
Study objective: Using Internet data to forecast emergency department (ED) visits might enable a model that reflects behavioral trends and thereby be a valid tool for health care providers with which to allocate resources and prevent crowding. The aim of this study is to investigate whether Web site visits to a regional medical Web site, the Stockholm Health Care Guide, a proxy for the general public's concern of their health, could be used to predict the ED attendance for the coming day. Methods: In a retrospective, observational, cross-sectional study, a model for forecasting the daily number of ED visits was derived and validated. The model was derived through regression analysis, using visits to the Stockholm Health Care Guide Web site between 6 pm and midnight and day of the week as independent variables. Web site visits were measured with Google Analytics. The number of visits to the ED within the region was retrieved from the Stockholm County Council administrative database. All types of ED visits (including adult, pediatric, and gynecologic) were included. The period of August 13, 2011, to August 12, 2012, was used as a training set for the model. The hourly variation of visits was analyzed for both Web site and the ED visits to determine the interval of hours to be used for the prediction. The model was validated with mean absolute percentage error for August 13, 2012, to October 31, 2012. Results: The correlation between the number of Web site visits between 6 pm and midnight and ED visits the coming day was significant (r=0.77; P<.001). The best forecasting results for ED visits were achieved for the entire county, with a mean absolute percentage error of 4.8%. The result for the individual hospitals ranged between mean absolute percentage error 5.2% and 13.1%. Conclusion: Web site visits may be used in this fashion to predict attendance to the ED. The model works both for the entire region and for individual hospitals. The possibility of using Internet data to predict ED visits is promising.
Article
We apply discrete event simulation to characterize the patient flow affects of using admission predictions and current state information, generated in an Emergency Department (ED), to influence the prioritization of inpatient unit (IU) physicians between treating and discharging IU patients. Shared information includes crowding levels and total expected bed need (based on the sum of individual patients’ imperfect admission predictions and perfect admission predictions). It is found that sharing prediction and crowding information to influence inpatient staff priorities, using specific information sensitivity schedules, can result in statistically significant (p ≪ 0.05) reductions in boarding time (between 11.69% and 18.38% compared to baseline performance). The range of improvement is dependent on varying simulated hospital configurations.
Article
Background Emergency departments (ED) continue to evolve models of care and streaming as interventions to tackle the effects of access block and overcrowding. Tertiary ED may be able to design patient-flow based on predicted dispositions in the department. Segregating discharge-stream patients may help develop patient-flows within the department, which is less affected by availability of beds in a hospital. We aim to determine if triage nurses and ED doctors can predict disposition outcomes early in the patient journey and thus lead to successful streaming of patients in the ED. Methods During this study, triage nurses and ED doctors anonymously predicted disposition outcomes for patients presenting to triage after their brief assessments. Patient disposition at the 24-h post ED presentation was considered as the actual outcome and compared against predicted outcomes. Results Triage nurses were able to predict actual discharges of 445 patients out of 490 patients with a positive predictive value (PPV) of 90.8% (95% CI 87.8–93.2%). ED registrars were able to predict actual discharges of 85 patients out of 93 patients with PPV of 91.4% (95% CI 83.3–95.9%). ED consultants were able to predict actual discharges of 111 patients out of 118 patients with PPV 94.1% (95% CI 87.7–97.4%). PPVs for admission among ED consultants, ED registrars and Triage nurses were 59.7%, 54.4% and 48.5% respectively. Conclusions Triage nurses, ED consultants and ED registrars are able to predict a patient's discharge disposition at triage with high levels of confidence. Triage nurses, ED consultants, and ED registrars can predict patients who are likely to be admitted with equal ability. This data may be used to develop specific admission and discharge streams based on early decision-making in EDs by triage nurses, ED registrars or ED consultants.
Article
Artificial neural networks are appearing as useful alternatives to traditional statistical modelling techniques in many scientific disciplines. This paper presents a general introduction and discussion of recent applications of the multilayer perceptron, one type of artificial neural network, in the atmospheric sciences.
Article
Ambulance diversion (AD) is used by emergency departments (EDs) to relieve congestion by requesting ambulances to bypass the ED and transport patients to another facility. We study optimal AD control policies using a Markov Decision Process (MDP) formulation that minimizes the average time that patients wait beyond their recommended safety time threshold. The model assumes that patients can be treated in one of two treatment areas and that the distribution of the time to start treatment at the neighboring facility is known. Assuming Poisson arrivals and exponential times for the length of stay in the ED, we show that the optimal AD policy follows a threshold structure, and explore the behavior of optimal policies under different scenarios. We analyze the value of information on the time to start treatment in the neighboring hospital, and show that optimal policies depend strongly on the congestion experienced by the other facility. Simulation is used to compare the performance of the proposed MDP model to that of simple heuristics under more realistic assumptions. Results indicate that the MDP model performs significantly better than the tested heuristics under most cases. Finally, we discuss practical issues related to the implementation of the policies prescribed by the MDP.
Article
Our objective was to apply neural network methodology to determine whether adding coded chief complaint (CCC) data to triage information would result in an improved hospital admission prediction model than one without CCC data. We carried out a retrospective derivation and validation cohort study of all adult emergency department visits to a single center. We downloaded triage, chief complaint, and admission/discharge data on each included visit. Using a CCC algorithm and the Levenberg-Marquardt back-propagation learning method, we derived hospital admission prediction models without and with CCC data and applied these to the validation cohort, reporting the prediction models' characteristics. A total of 74 056 emergency department visits were included in the derivation cohort, 85 144 in the validation cohort with 213 CCC categories. The sensitivity/specificity of the derivation cohort models without and with CCC data were 64.0% [95% confidence interval (CI): 63.7-64.3], 87.7% (95% CI: 87.4-88.0), 59.8% (95% CI: 59.5-60.3%), and 91.7% (95% CI: 91.4-92.0) respectively. The sensitivity/specificity of the derived models without and with CCC data applied to the validation cohort were 60.7% (95% CI: 60.4-61.0), 87.7% (95% CI: 87.4-88.0), 59.8% (95% CI: 59.5-60.3), and 90.6% (95% CI: 90.3-90.9) respectively. The area under the curve in the validation cohort for the derived models without and with CCC data were 0.840 (95% CI: 0.838-0.842) and 0.860 (95% CI: 0.858-0.862). Net reclassification index (0.156; 95% CI: 0.148-0.163) and integrated discrimination improvement (0.060; 95% CI: 0.058-0.061) in the CCC model were significant. Neural net methodology application resulted in the derivation and validation of a modestly stronger hospital admission prediction model after the addition of CCC data.
Article
Ambulance offload delays are a growing concern for health care providers in many countries. Offload delays occur when ambulance paramedics arriving at a hospital Emergency Department (ED) cannot transfer patient care to staff in the ED immediately. This is typically caused by overcrowding in the ED. Using queueing theory, we model the interface between a regional Emergency Medical Services (EMS) provider and multiple EDs that serve both ambulance and walk-in patients. We introduce Markov chain models for the system and solve for the steady state probability distributions of queue lengths and waiting times using matrix-analytic methods. We develop several algorithms for computing performance measures for the system, particularly the offload delays for ambulance patients. Using these algorithms, we analyze several three-hospital systems and assess the impact of system resources on offload delays. In addition, simulation is used to validate model assumptions.
Article
Boarding of admitted patients in the emergency department (ED) is a major cause of crowding. One alternative to boarding in the ED, a full-capacity protocol where boarded patients are redeployed to inpatient units, can reduce crowding and improve overall flow. Our aim was to compare patient satisfaction with boarding in the ED vs. inpatient hallways. We performed a structured telephone survey regarding patient experiences and preferences for boarding among admitted ED patients who experienced boarding in the ED hallway and then were subsequently transferred to inpatient hallways. Demographic and clinical characteristics, as well as patient preferences, including items related to patient comfort and safety using a 5-point scale, were recorded and descriptive statistics were used to summarize the data. Of 110 patients contacted, 105 consented to participate. Mean age was 57 ± 16 years and 52% were female. All patients were initially boarded in the ED in a hallway before their transfer to an inpatient hallway bed. The overall preferred location after admission was the inpatient hallway in 85% (95% confidence interval 75-90) of respondents. In comparing ED vs. inpatient hallway boarding, the following percentages of respondents preferred inpatient boarding with regard to the following 8 items: rest, 85%; safety, 83%; confidentiality, 82%; treatment, 78%; comfort, 79%; quiet, 84%; staff availability, 84%; and privacy, 84%. For no item was there a preference for boarding in the ED. Patients overwhelmingly preferred the inpatient hallway rather than the ED hallway when admitted to the hospital.
Article
To evaluate an automatic forecasting algorithm in order to predict the number of monthly emergency department (ED) visits one year ahead. We collected retrospective data of the number of monthly visiting patients for a 6-year period (2005-2011) from 4 Belgian Hospitals. We used an automated exponential smoothing approach to predict monthly visits during the year 2011 based on the first 5years of the dataset. Several in- and post-sample forecasting accuracy measures were calculated. The automatic forecasting algorithm was able to predict monthly visits with a mean absolute percentage error ranging from 2.64% to 4.8%, indicating an accurate prediction. The mean absolute scaled error ranged from 0.53 to 0.68 indicating that, on average, the forecast was better compared with in-sample one-step forecast from the naïve method. The applied automated exponential smoothing approach provided useful predictions of the number of monthly visits a year in advance.
Article
An integral part of econometric practice is to test the adequacy of model specifications. If a model is adequately specified, it should not leave interesting features of the data-generating process in the errors. Despite the common tradition, the importance of diagnostic checking as a safeguard against mis-specification has only recently been recognized by neural network (NN) practitioners, possibly because this type of semi-parametric methodology was not originally designed for economic and financial applications. The purpose of this paper is to compare a number of analytical statistical testing procedures suitable to diagnostic checking on a neural network regression model. We present the standard Lagrange multiplier (LM) testing framework designed under the assumption of identically distributed disturbances and also examine two modifications that are robust to heteroskedasticity in errors. One modification also gives the researcher an opportunity to incorporate information concerning the volatility structure of the data-generating process in the testing procedure. By means of a Monte Carlo simulation, we investigate the performance of these tests under GARCH-type heteroskedasticity in errors and various distributional assumptions. The results show that although the primary concern of the researcher may be to design a regression model that accurately captures relations in the mean of the conditional distribution, developing a good approximation of the underlying volatility structure generally increases the efficiency of tests in detecting non-adequacy of a NN model. †http://fidelity.fme.ae gean.gr/decision
Article
b>Background: Streamlining emergency department (ED) care of patients with chronic obstructive pulmonary disease (COPD) may be an important strategy in managing the increasing burden of this disease. Study objectives: The aim of this study was to identify factors predictive of hospital admission in ED patients with COPD, specifically factors that can be used early in the ED episode of care. Methods: Using retrospective regression analysis, case data from 321 randomly selected medical records from five Australian EDs were analysed. Patient characteristics, triage and ED system features, physiological status, and ED treatment during the first four hours of ED care were compared between discharged and admitted patients. Results: Factors available on ED arrival associated with increased likelihood of admission were: age (OR = 1.04, p = 0.008) respiratory symptoms affecting activities of daily living (OR = 1.8, p = 0.043) and signs of respiratory dysfunction (OR = 2.5, p = 0.005). Factors available from the first four hours of ED care associated with increased likelihood of admission were: age (OR = 1.04, p = 0.021), oxygen use at four hours (OR = 3.5, p = 0.002) and IV antibiotic administration (OR = 2.6, p = 0.026). There were conflicting findings regarding the association between ambulance transport and admission.<br /
Article
This research examines the adoption of an interorganizational system standard and its benefits by using RosettaNet as a case study. A comprehensive research framework derived from institutional theory and a Technology–Organization–Environment model was developed for this research. Data were collected from a sample of 212 Malaysian manufacturing firms. A multi-state analytic approach was proposed whereby the research model was tested using structural equation modeling (SEM), and the results from SEM were used as inputs for a neural network model for predicting RosettaNet adoption. The results showed that factors related to the environment, the Interorganizational Relationship (IOR) and an information sharing culture have a positive influence on the adoption of RosettaNet. In terms of organizational factors, top management support was found to have a positive and significant relationship with RosettaNet adoption. The results also showed that RosettaNet adoption has a significant and positive relationship with organizational performance. The research findings can also assist managerial decision making for those organizations planning to adopt RosettaNet. This research reduces the previous research gap by advancing understanding on the relationship of adoption factors and RosettaNet adoption, and extends previous approaches on RosettaNet adoption by investigating the relationships between RosettaNet adoption and organizational performance. Improved existing technology adoption methodology was achieved by integrating both SEM and neural network for examining the adoptions of RosettaNet.
Article
Objectives: The objectives were to evaluate three models that use information gathered during triage to predict, in real time, the number of emergency department (ED) patients who subsequently will be admitted to a hospital inpatient unit (IU) and to introduce a new methodology for implementing these predictions in the hospital setting. Methods: Three simple methods were compared for predicting hospital admission at ED triage: expert opinion, naïve Bayes conditional probability, and a generalized linear regression model with a logit link function (logit-linear). Two months of data were gathered from the Boston VA Healthcare System's 13-bed ED, which receives approximately 1,100 patients per month. Triage nurses were asked to estimate the likelihood that each of 767 triaged patients from that 2-month period would be admitted after their ED treatment, by placing them into one of six categories ranging from low to high likelihood. Logit-linear regression and naïve Bayes models also were developed using retrospective data and used to estimate admission probabilities for each patient who entered the ED within a 2-month time frame, during triage hours (1,160 patients). Predictors considered included patient age, primary complaint, provider, designation (ED or fast track), arrival mode, and urgency level (emergency severity index assigned at triage). Results: Of the three methods considered, logit-linear regression performed the best in predicting total bed need, with a receiver operating characteristic (ROC) area under the curve (AUC) of 0.887, an R(2) of 0.58, an average estimation error of 0.19 beds per day, and on average roughly 3.5 hours before peak demand occurred. Significant predictors were patient age, primary complaint, bed type designation, and arrival mode (p < 0.0001 for all factors). The naïve Bayesian model had similar positive predictive value, with an AUC of 0.841 and an R(2) of 0.58, but with average difference in total bed need of approximately 2.08 per day. Triage nurse expert opinion also had some predictive capability, with an R(2) of 0.52 and an average difference in total bed need of 1.87 per day. Conclusions: Simple probability models can reasonably predict ED-to-IU patient volumes based on basic data gathered at triage. This predictive information could be used for improved real-time bed management, patient flow, and discharge processes. Both statistical models were reasonably accurate, using only a minimal number of readily available independent variables.
Article
The practice of keeping admitted patients on stretchers in hospital emergency department hallways for hours or days, called "boarding," causes emergency department crowding and can be harmful to patients. Boarding increases patients' morbidity, lengths of hospital stay, and mortality. Strategies that optimize bed management reduce boarding by improving the efficiency of hospital patient flow, but these strategies are grossly underused. Convincing hospital leaders of the value of such solutions, and educating patients to advocate for such changes, may promote improvements. If these strategies do not work, legislation may be required to effect meaningful change.
Article
Much attention has been paid to lengthy wait times in emergency departments (EDs) and much research has sought to improve ED performance. However, ED congestion is often caused by the inability to move patients into the wards while the wards in turn are often congested primarily due to patients waiting for a bed in a long-term care (LTC) facility. The scheduling of clients to LTC is a complex problem that is compounded by the variety of LTC beds (different facilities and room accommodations), the presence of client choice and the competing demands of the hospital and community populations. We present a Markov decision process (MDP) model that determines the required access in order for the census of patients waiting for LTC in the hospitals to remain below a given threshold. We further present a simulation model that incorporates both hospital and community demand for LTC in order to predict the impact of implementing the policy derived from the MDP on the community client wait times and to aid in capacity planning for the future. We test the MDP policy vs. current practice as well as against a number of other proposed policy changes.
Article
Background: Boarding of inpatients in the Emergency Department (ED) has been widely recognized as a major contributor to ED crowding and a cause of adverse outcomes. We hypothesize that these deleterious effects extend to those patients who are discharged from the ED by increasing their length of stay (LOS). Study objectives: This study investigates the impact of boarding inpatients on the ED LOS of discharged patients. Methods: This retrospective, observational, cohort study investigated the association between ED boarder burden and discharged patient LOS over a 3-year period in an urban, academic tertiary care ED. Median ED LOS of 179,840 discharged patients was calculated for each quartile of the boarder burden at time of arrival, and Spearman correlation coefficients were used to summarize the relationship. Subgroup analyses were conducted, stratified by patient acuity defined by triage designation, and hour of arrival. Results: Overall median discharged patient ED LOS increased by boarder burden quartile (205 [95% confidence interval (CI) 203-207], 215 [95% CI 214-217], 221 [95% CI 219-223], and 221 [95% CI 219-223] min, respectively), with a Spearman correlation of 0.25 between daily total boarder burden hours and median LOS. When stratified by patient acuity and hour of arrival (11:00 a.m.-11:00 p.m.), LOS of medium-acuity patients increased significantly by boarder burden quartile (252 [95% CI 247-255], 271 [95% CI 267-275], 285 [95% CI 95% CI 278-289], and 309 [95% CI 305-315] min, respectively) with a Spearman correlation of 0.18. Conclusion: In this retrospective study, increasing boarder burden was associated with increasing LOS of patients discharged from the ED, with the greatest effect between 11:00 a.m. and 11:00 p.m. on medium-acuity patients. This relationship between LOS and ED capacity limitation by inpatient boarders has important implications, as ED and hospital leadership increasingly focus on ED LOS as a measure of efficiency and throughput.
Article
Specialty consultations and waiting for admission to a hospital bed are major contributors to increased length of stay and overcrowding in the emergency department. We implemented a computerized short messaging service to inform care providers of patient delay in order to reduce length of stay. The purpose of this study was to evaluate the effects of this strategy on length of stay in the emergency department. This was a before-and-after observational study. Prior to this study, we registered the mobile phone numbers of all board certified specialists into a computerized physician order entry database and developed an auto-sending short messaging program linked to consultation orders. The short message was transmitted at 2 and 4h after consultation, when a disposition was not yet established, and at 8h after the admission order if the patient was still waiting. The length of stay of consulted patients and intervals such as consultation time (registration-consultation), disposition time (consultation-admission decision), and boarding time (admission decision-hospitalization) of admitted patients were compared between the pre-implementation (September 2009) and post-implementation period (November 2009). Subgroup analyses of disposition time were performed according to time of consultation and the number of consultations. A total of 7518 patients visited the emergency department during the pre-periods and post-periods. Among them, 3335 patients required specialty consultations. The median length of stay of consulted patients decreased significantly after implementation of the messaging system (pre-207 min vs. post-193 min, p<0.001). Among admitted patients, the median length of stay decreased by 36 min from 294 min to 258 min (p<0.001). In the subgroup analysis, times for establishing patient dispositions decreased significantly when the consultation was performed at night and when there was only one department consulted. The numbers of patients with disposition times within 2 and 4h and boarding times within 8h were all increased after implementation of the short message service program. This study suggested that the computerized physician order entry-based short messaging service program, used to inform decision-makers of patient delay, could reduce the length of stay for consulted patients in the emergency department.
Conference Paper
With the rapid outstripping of healthcare resources by the demands on hospital care, it is important to find more effective and efficient ways for managing care. This research is aimed at developing new admission prediction models using various pre-hospital variables to help hospital estimate the patients to be admitted. We developed a framework of hospital admission prediction and proposed two novel approaches to capture semantics of chief complaints to enhance prediction. Our experiments on a hospital dataset demonstrated that our proposed models outperformed several benchmark methods.
Article
The global economic crisis has a significant impact on healthcare resource provision worldwide. The management of limited healthcare resources is further challenged by the high level of uncertainty in demand, which can lead to unbalanced utilization of the available resources and a potential deterioration of patient satisfaction in terms of longer waiting times and perceived reduced quality of services. Therefore, healthcare managers require timely and accurate tools to optimize resource utility in a complex and ever-changing patient care process. An interactive simulation-based decision support framework is presented in this paper for healthcare process improvement. Complexity and different levels of variability within the process are incorporated into the process modeling phase, followed by developing a simulation model to examine the impact of potential alternatives. As a performance management tool, balanced scorecard (BSC) is incorporated within the framework to support continual and sustainable improvement by using strategic-linked performance measures and actions. These actions are evaluated by the simulation model developed, whilst the trade-off between objectives, though somewhat conflicting, is analysed by a preference model. The preference model is designed in an interactive and iterative process considering decision makers preferences regarding the selected key performance indicators (KPIs). A detailed implementation of the framework is demonstrated on an emergency department (ED) of an adult teaching hospital in north Dublin, Ireland. The results show that the unblocking of ED outflows by in-patient bed management is more effective than increasing only the ED physical capacity or the ED workforce.
Article
To be able to predict, at the time of triage, whether a need for hospital admission exists for emergency department (ED) patients may constitute useful information that could contribute to systemwide hospital changes designed to improve ED throughput. The objective of this study was to develop and validate a predictive model to assess whether a patient is likely to require inpatient admission at the time of ED triage, using routine hospital administrative data. Data collected at the time of triage by nurses from patients who visited the ED in 2007 and 2008 were extracted from hospital administrative databases. Variables included were demographics (age, sex, and ethnic group), ED visit or hospital admission in the preceding 3 months, arrival mode, patient acuity category (PAC) of the ED visit, and coexisting chronic diseases (diabetes, hypertension, and dyslipidemia). Chi-square tests were used to study the association between the selected possible risk factors and the need for hospital admission. Logistic regression was applied to develop the prediction model. Data were split for derivation (60%) and validation (40%). Receiver operating characteristic curves and goodness-of-fit tests were applied to the validation data set to evaluate the model. Of 317,581 ED patient visits, 30.2% resulted in immediate hospital admission. In the developed predictive model, age, PAC status, and arrival mode were most predictive of the need for immediate hospital inpatient admission. The c-statistic of the receiver operating characteristic (ROC) curve was 0.849 (95% confidence interval [CI] = 0.847 to 0.851). The goodness-of-fit test showed that the predicted patients' admission risks fit the patients' actual admission status well. A model for predicting the risk of immediate hospital admission at triage for all-cause ED patients was developed and validated using routinely collected hospital data. Early prediction of the need for hospital admission at the time of triage may help identify patients deserving of early admission planning and resource allocation and thus potentially reduce ED overcrowding.
Article
Although the provision of inpatient care is not typically associated with emergency nursing, it is the new reality in many departments. Given the number of admitted patients boarded in the emergency department for part or all of their hospital stay, it is important to know who these patients are. The purpose of this analysis was to determine whether the occurrence of ED boarding could be predicted by factors specific to the type and timing of the ED visit or whether patient characteristics also affected these decisions. A retrospective review of administrative data for a 1-year period was conducted. Chi-square and logistic regression analyses were used to determine whether the likelihood of being boarded for more than 2 hours could be predicted by factors specific to the type of visit (ie, triage level and admission type) and timing of the visit (ie, time of day and day of week) or whether patient characteristics (ie, sex and age group) also played a role. Slightly more than half of patients remained in the emergency department for more than 2 hours following receipt of an admission order. Results suggest the likelihood of boarding was highest for those who were medical admissions and admitted on a weekday or during the night shift. Even after accounting for these factors, patient characteristics improved the ability to predict ED boarding. Female patients and those 65 years of age or older were more likely to be boarded. Findings suggest that in addition to their usual responsibilities, emergency nurses are providing care to a group of inpatients who tend to have high medical and nursing care needs.
Article
Patient crowding and boarding in Emergency Departments (EDs) impair the quality of care as well as patient safety and satisfaction. Improved timing of inpatient discharges could positively affect ED boarding, and this hypothesis can be tested with computer modeling. Modeling enables analysis of the impact of inpatient discharge timing on ED boarding. Three policies were tested: a sensitivity analysis on shifting the timing of current discharge practices earlier; discharging 75% of inpatients by 12:00 noon; and discharging all inpatients between 8:00 a.m. and 4:00 p.m. A cross-sectional computer modeling analysis was conducted of inpatient admissions and discharges on weekdays in September 2007. A model of patient flow streams into and out of inpatient beds with an output of ED admitted patient boarding hours was created to analyze the three policies. A mean of 38.8 ED patients, 22.7 surgical patients, and 19.5 intensive care unit transfers were admitted to inpatient beds, and 81.1 inpatients were discharged daily on September 2007 weekdays: 70.5%, 85.6%, 82.8%, and 88.0%, respectively, occurred between noon and midnight. In the model base case, total daily admitted patient boarding hours were 77.0 per day; the sensitivity analysis showed that shifting the peak inpatient discharge time 4h earlier eliminated ED boarding, and discharging 75% of inpatients by noon and discharging all inpatients between 8:00 a.m. and 4:00 p.m. both decreased boarding hours to 3.0. Timing of inpatient discharges had an impact on the need to board admitted patients. This model demonstrates the potential to reduce or eliminate ED boarding by improving inpatient discharge timing in anticipation of the daily surge in ED demand for inpatient beds.
Article
Prospective and retrospective access block hospital intervention studies from 1998 to 2008 were reviewed to assess the evidence for interventions around access block and ED overcrowding, including over 220 documents reported in Medline and data extracted from The State of our Public Hospitals Reports. There is an estimated 20-30% increased mortality rate due to access block and ED overcrowding. The main causes are major increases in hospital admissions and ED presentations, with almost no increase in the capacity of hospitals to meet this demand. The rate of available beds in Australia reduced from 2.6 beds per 1000 (1998-1999) to 2.4 beds per 1000 (2002-2007) in 2002, and has remained steady at between 2.5-2.6 beds per 1000. In the same period, the number of ED visits increased over 77% from 3.8 million to 6.74 million. Similarly, the number of public hospital admissions increased at an average rate of 3.4% per year from 3.7 to 4.7 million. Compared with 1998-1999 rates, the number of available beds in 2006-2007 is thus similar (2.65 vs 2.6 beds per 1000), but the number of ED presentations has almost doubled. All patient groups are affected by access block. Access block interventions may temporarily reduce some of the symptoms of access block, but many measures are not sustainable. The root cause of the problem will remain unless hospital capacity is addressed in an integrated approach at both national and state levels.
Article
In this paper, a decision-making model was developed to select suppliers using neural networks (NNs). This model used historical supplier performance data for selection of vendor suppliers. Input and output were designed in a unique manner for training purposes. The managers' judgments about suppliers were simulated by using a pairwise comparisons matrix for output estimation in the NN. To obtain the benefit of a search technique for model structure and training, genetic algorithm (GA) was applied for the initial weights and architecture of the network. The suppliers' database information (input) can be updated over time to change the suppliers' score estimation based on their performance. The case study illustrated shows how the model can be applied for suppliers' selection.
Article
Overcrowding in hospitals, especially in EDs, is a serious problem in the United States, Europe, and Taiwan. However, the association between prolonged ED boarding stay and mortality in patients with necrotizing fasciitis remains underinvestigated. This was a retrospective study. A total of 195 patients were enrolled and analyzed. The sample was divided into 2 groups: nonmortality and mortality. A stepwise logistic regression model was developed to investigate 3 factors of clinical relevance predicting patient mortality. The results of the stepwise logistic regression analysis revealed that hypotension (odds ratio [OR], 32.9; 95% confidence interval [CI], 6.9-156.0) and prolonged ED boarding stay (OR, 3.4; 95% CI 1.3-8.6) were both associated with higher mortality. Early operation (OR: 0.16; 95% CI: 0.06-0.45) was associated with lower mortality. Prolonged ED boarding stay was associated with increased mortality in patients with necrotizing fasciitis. Early operation (within 24 hours of ED arrival) was associated with decreased mortality.
Article
We describe the frequency of undesirable events among patients boarding at a single, urban, tertiary, teaching emergency department (ED) through retrospective chart abstraction. This was a chart review of all patients admitted during 3 randomly selected days in 2003 (n=162) to track the frequency of undesirable events such as missed relevant home medications, missed laboratory test results, arrhythmias, or other adverse events. One hundred fifty-one charts were abstracted (93.2%); 27.8% had an undesirable event, 17.9% missed a relevant home medication, and 3.3% had a preventable adverse event. There was a higher frequency of undesirable events among older patients (35.9%, aged >50 years; 7.3%, aged 20 to 49 years; 28.6%, aged 0 to 19 years) and those with more comorbidities (44.4% among Charlson score >or=3; 30.8% score 2; 36.1% score 1; 14.5% score 0). A substantial frequency of undesirable events occurs while patients board in the ED. These events are more frequent in older patients or those with more comorbidities. Future studies need to compare the rates of undesirable events among patients boarding in the ED versus inpatient units.
Article
The objective was to study the association between factors related to emergency department (ED) crowding and patient satisfaction. The authors performed a retrospective cohort study of all patients admitted through the ED who completed Press-Ganey patient satisfaction surveys over a 2-year period at a single academic center. Ordinal and binary logistic regression was used to study the association between validated ED crowding factors (such as hallway placement, waiting times, and boarding times) and patient satisfaction with both ED care and assessment of satisfaction with the overall hospitalization. A total of 1,501 hospitalizations for 1,469 patients were studied. ED hallway use was broadly predictive of a lower likelihood of recommending the ED to others, lower overall ED satisfaction, and lower overall satisfaction with the hospitalization (p < 0.05). Prolonged ED boarding times and prolonged treatment times were also predictive of lower ED satisfaction and lower satisfaction with the overall hospitalization (p < 0.05). Measures of ED crowding and ED waiting times predicted ED satisfaction (p < 0.05), but were not predictive of satisfaction with the overall hospitalization. A poor ED service experience as measured by ED hallway use and prolonged boarding time after admission are adversely associated with ED satisfaction and predict lower satisfaction with the entire hospitalization. Efforts to decrease ED boarding and crowding might improve patient satisfaction.
Article
Emergency department (ED) crowding has become a major barrier to receiving timely emergency care in the United States. Despite widespread recognition of the problem, the research and policy agendas needed to understand and address ED crowding are just beginning to unfold. We present a conceptual model of ED crowding to help researchers, administrators, and policymakers understand its causes and develop potential solutions. The conceptual model partitions ED crowding into 3 interdependent components: input, throughput, and output. These components exist within an acute care system that is characterized by the delivery of unscheduled care. The goal of the conceptual model is to provide a practical framework on which an organized research, policy, and operations management agenda can be based to alleviate ED crowding.
Article
To compare triage decisions of an automated emergency department triage system with decisions made by an emergency specialist. In a retrospective setting, data extracted from charts of 90 patients with chief complaint of non-traumatic abdominal pain were used as input for triage system and emergency medicine specialist. The final disposition and diagnoses of the physicians who visited the patient in Emergency Department (ED) as reflected in the medical records were considered as control. Results were compared by chi(2)-test and a binary logistic regression model. Compared to emergency medicine specialist, triage system had higher sensitivity (90% versus 64%) and lower specificity (25% versus 48%) for patients who required hospitalization. The triage system successfully predicted the Admit decisions made in the ED whereas the emergency medicine specialist decisions could not predict the ED disposition. Both triage system and emergency medicine specialist properly disposed 56% of cases, however, the emergency medicine specialist in this study under-disposed more patients than the triage system considering Admit disposition (p=0.004) while he appropriately discharged more patients compared to the triage system (p=0.017). The triage system studied here shows promise as a triage decision support tool to be used for telephone triage and triage in the emergency departments. This technology may also be useful to the patients as a self-triage tool. However, the efficiency of this particular application of this technology is unclear.
Article
Emergency Department (ED) crowding and ambulance diversion have been increasingly significant national problems for more than a decade. Surveys of hospital directors have reported overcrowding in almost every state and 91% of hospital ED directors report overcrowding as a problem. The problem has developed because of multiple factors in the past 20 years, including a steady downsizing in hospital capacity, closures of a significant number of EDs, increased ED volume, growing numbers of uninsured, and deceased reimbursement for uncompensated care. Initial position statements from major organizations, including JCAHO and the General Accounting Office, suggested the problem of overcrowding was due to inappropriate use of emergency services by those with no urgent conditions, probably cyclical, and needed no specific policy response. More recently, the same and other organizations have more forcefully highlighted the problem of overcrowding and focused on the inability to transfer emergency patients to inpatient beds as the single most important factor contributing to ED overcrowding. This point has been further solidified by initial overcrowding research. This article will review how overcrowding occurred with a focus on the significance and potential remedies of extended boarding of admitted patients in the Emergency Department.