Classification of patients by severity grades during triage in the emergency department using data mining methods

Risk of mortality and cardiopulmonary arrest in critical patients presenting to the emergency department using machine learning and natural language processing

Emergency department triage is the first point in time when a patient's acuity level is determined. The time to assign a priority at triage is short and it is vital to accurately stratify patients at this stage, since under-triage can lead to increased morbidity, mortality and costs. Our aim was to present a model that can assist healthcare professionals in triage decision making, namely in the stratification of patients through the risk prediction of a composite critical outcome-mortality and cardiopulmonary arrest. Our study cohort consisted of 235826 adult patients triaged at a Portuguese Emergency Department from 2012 to 2016. Patients were assigned to emergent, very urgent or urgent priorities of the Manchester Triage System (MTS). Demographics, clinical variables routinely collected at triage and the patients' chief complaint were used. Logistic regression, random forests and extreme gradient boosting were developed using all available variables. The term frequency-inverse document frequency (TF-IDF) natural language processing weighting factor was applied to vectorize the chief complaint. Stratified random sampling was used to split the data into train (70%) and test (30%) data sets. Ten-fold cross validation was performed in train to optimize model hyper-parameters. The performance obtained with the best model was compared against the reference model-a regularized logistic regression trained using only triage priorities. Extreme gradient boosting exhibited good calibration properties and yielded areas under the receiver operating characteristic and precision-recall curves of 0.96 (95% CI 0.95-0.97) and 0.31 (95% CI 0.26-0.36), respectively. The predictors ranked with higher importance by this model were the Glasgow coma score, the patients' age, pulse oximetry and arrival mode. Compared to the reference, the extreme gradient boosting model using clinical variables and the chief complaint presented higher recall for patients assigned MTS-3 and can identify those who are at risk of the composite outcome.

Predicting Intensive Care Unit admission among patients presenting to the emergency department using machine learning and natural language processing

The risk stratification of patients in the emergency department begins at triage. It is vital to stratify patients early based on their severity, since undertriage can lead to increased morbidity, mortality and costs. Our aim was to present a new approach to assist healthcare professionals at triage in the stratification of patients and in identifying those with higher risk of ICU admission. Adult patients assigned Manchester Triage System (MTS) or Emergency Severity Index (ESI) 1 to 3 from a Portuguese and a United States Emergency Departments were analyzed. Variables routinely collected at triage were used and natural language processing was applied to the patient chief complaint. Stratified random sampling was applied to split the data in train (70%) and test (30%) sets and 10-fold cross validation was performed for model training. Logistic regression, random forests, and a random undersampling boosting algorithm were used. We compared the performance obtained with the reference model—using only triage priorities—with the models using additional variables. For both hospitals, a logistic regression model achieved higher overall performance, yielding areas under the receiver operating characteristic and precision-recall curves of 0.91 (95% CI 0.90-0.92) and 0.30 (95% CI 0.27-0.33) for the United States hospital and of 0.85 (95% CI 0.83-0.86) and 0.06 (95% CI 0.05-0.07) for the Portuguese hospital. Heart rate, pulse oximetry, respiratory rate and systolic blood pressure were the most important predictors of ICU admission. Compared to the reference models, the models using clinical variables and the chief complaint presented higher recall for patients assigned MTS/ESI 3 and can identify patients assigned MTS/ESI 3 who are at risk for ICU admission.

Using machine-learning risk prediction models to triage the acuity of undifferentiated patients entering the emergency care system: a systematic review

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.

Accurate and rapid screening model for potential diabetes mellitus

Background

Prediction or early diagnosis of diabetes is crucial for populations with high risk of diabetes.

Methods

In this study, we assessed the ability of five popular classifiers (J48, AdaboostM1, SMO, Bayes Net, and Naïve Bayes) to identify individuals with diabetes based on nine non-invasive and easily obtained clinical features, including age, gender, body mass index (BMI), hypertension, history of cardiovascular disease or stroke, family history of diabetes, physical activity, work stress, and salty food preference. A total of 4205 data entries were obtained from annual physical examination reports for adults in the Shengjing Hospital of China Medical University during January–April 2017. Weka data mining software was used to identify the best algorithm for diabetes classification.

Results

The results indicate that decision tree classifier J48 has the best performance (accuracy = 0.9503, precision = 0.950, recall = 0.950, F-measure = 0.948, and AUC = 0.964). The decision tree structure shows that age is the most significant feature, followed by family history of diabetes, work stress, BMI, salty food preference, physical activity, hypertension, gender, and history of cardiovascular disease or stroke.

Conclusions

Our study shows that decision tree analyses can be applied to screen individuals for early diabetes risk without the need for invasive tests. This procedure will be particularly useful in developing regions with high epidemiological risk and poor socioeconomic status, and enable clinical practitioners to rapidly screen patients for increased risk of diabetes. The key features in the tree structure could further facilitate diabetes prevention through targeted community interventions, which can potentially improve early diabetes diagnosis and reduce burdens on the healthcare system.

Chronic Kidney Disease stratification using office visit records: Handling data imbalance via hierarchical meta-classification

Background

Chronic Kidney Disease (CKD) is one of several conditions that affect a growing percentage of the US population; the disease is accompanied by multiple co-morbidities, and is hard to diagnose in-and-of itself. In its advanced forms it carries severe outcomes and can lead to death. It is thus important to detect the disease as early as possible, which can help devise effective intervention and treatment plan.

Here we investigate ways to utilize information available in electronic health records (EHRs) from regular office visits of more than 13,000 patients, in order to distinguish among several stages of the disease. While clinical data stored in EHRs provide valuable information for risk-stratification, one of the major challenges in using them arises from data imbalance. That is, records associated with a more severe condition are typically under-represented compared to those associated with a milder manifestation of the disease. To address imbalance, we propose and develop a sampling-based ensemble approach, hierarchical meta-classification, aiming to stratify CKD patients into severity stages, using simple quantitative non-text features gathered from standard office visit records.

Methods

The proposed hierarchical meta-classification method frames the multiclass classification task as a hierarchy of two subtasks. The first is binary classification, separating records associated with the majority class from those associated with all minority classes combined, using meta-classification. The second subtask separates the records assigned to the combined minority classes into the individual constituent classes.

Results

The proposed method identifies a significant proportion of patients suffering from the more advanced stages of the condition, while also correctly identifying most of the less severe cases, maintaining high sensitivity, specificity and F-measure (≥ 93%). Our results show that the high level of performance attained by our method is preserved even when the size of the training set is significantly reduced, demonstrating the stability and generalizability of our approach.

Conclusion

We present a new approach to perform classification while addressing data imbalance, which is inherent in the biomedical domain. Our model effectively identifies severity stages of CKD patients, using information readily available in office visit records within the realistic context of high data imbalance.

Novel approach for automating medical emergency protocol in military environment

BACKGROUND AND OBJECTIVES

Categorization of the casualties in accordance with medical care priorities is crucial in a military environment. Automation of the triage process is still a challenging task. The goal of the paper is to propose a novel algorithm for automation of medical emergency protocol in the military environment by the creation of classifiers that can provide accurate prioritization of injured soldier cases. It is a part of a complex military telemedicine system that provides continuous monitoring of soldiers' vital data gathered on-site using an unobtrusive set of sensors.

METHODS

After pre-processing the collected raw physiological data and eliminating the outliers using Naïve Bayesian Classifier, the system is capable of calculating the risk level and categorizing the victims based on Markov Decision Process. The NBC has been trained with a dataset that has contained labels and 6 features. Training set has held 8000 randomly chosen samples. Twenty percent of the determined dataset has been used for the validation set.

RESULTS

For algorithm verification, several evaluation scenarios have been created. In each scenario, randomly generated vital sign data describing the hypothetical health condition of soldiers was contemporarily assessed by the system as well as by 50 experienced military medical physicians.

CONCLUSION

The obtained correlation result of the proposed algorithm and medical physicians' classifications is strong evidence that the system can be implemented in warfare emergency medicine.

Inter-labeler and intra-labeler variability of condition severity classification models using active and passive learning methods

BACKGROUND AND OBJECTIVES

Labeling instances by domain experts for classification is often time consuming and expensive. To reduce such labeling efforts, we had proposed the application of active learning (AL) methods, introduced our CAESAR-ALE framework for classifying the severity of clinical conditions, and shown its significant reduction of labeling efforts. The use of any of three AL methods (one well known [SVM-Margin], and two that we introduced [Exploitation and Combination_XA]) significantly reduced (by 48% to 64%) condition labeling efforts, compared to standard passive (random instance-selection) SVM learning. Furthermore, our new AL methods achieved maximal accuracy using 12% fewer labeled cases than the SVM-Margin AL method. However, because labelers have varying levels of expertise, a major issue associated with learning methods, and AL methods in particular, is how to best to use the labeling provided by a committee of labelers. First, we wanted to know, based on the labelers' learning curves, whether using AL methods (versus standard passive learning methods) has an effect on the Intra-labeler variability (within the learning curve of each labeler) and inter-labeler variability (among the learning curves of different labelers). Then, we wanted to examine the effect of learning (either passively or actively) from the labels created by the majority consensus of a group of labelers.

METHODS

We used our CAESAR-ALE framework for classifying the severity of clinical conditions, the three AL methods and the passive learning method, as mentioned above, to induce the classifications models. We used a dataset of 516 clinical conditions and their severity labeling, represented by features aggregated from the medical records of 1.9 million patients treated at Columbia University Medical Center. We analyzed the variance of the classification performance within (intra-labeler), and especially among (inter-labeler) the classification models that were induced by using the labels provided by seven labelers. We also compared the performance of the passive and active learning models when using the consensus label.

RESULTS

The AL methods: produced, for the models induced from each labeler, smoother Intra-labeler learning curves during the training phase, compared to the models produced when using the passive learning method. The mean standard deviation of the learning curves of the three AL methods over all labelers (mean: 0.0379; range: [0.0182 to 0.0496]), was significantly lower (p=0.049) than the Intra-labeler standard deviation when using the passive learning method (mean: 0.0484; range: [0.0275-0.0724). Using the AL methods resulted in a lower mean Inter-labeler AUC standard deviation among the AUC values of the labelers' different models during the training phase, compared to the variance of the induced models' AUC values when using passive learning. The Inter-labeler AUC standard deviation, using the passive learning method (0.039), was almost twice as high as the Inter-labeler standard deviation using our two new AL methods (0.02 and 0.019, respectively). The SVM-Margin AL method resulted in an Inter-labeler standard deviation (0.029) that was higher by almost 50% than that of our two AL methods The difference in the inter-labeler standard deviation between the passive learning method and the SVM-Margin learning method was significant (p=0.042). The difference between the SVM-Margin and Exploitation method was insignificant (p=0.29), as was the difference between the Combination_XA and Exploitation methods (p=0.67). Finally, using the consensus label led to a learning curve that had a higher mean intra-labeler variance, but resulted eventually in an AUC that was at least as high as the AUC achieved using the gold standard label and that was always higher than the expected mean AUC of a randomly selected labeler, regardless of the choice of learning method (including a passive learning method). Using a paired t-test, the difference between the intra-labeler AUC standard deviation when using the consensus label, versus that value when using the other two labeling strategies, was significant only when using the passive learning method (p=0.014), but not when using any of the three AL methods.

CONCLUSIONS

The use of AL methods, (a) reduces intra-labeler variability in the performance of the induced models during the training phase, and thus reduces the risk of halting the process at a local minimum that is significantly different in performance from the rest of the learned models; and (b) reduces Inter-labeler performance variance, and thus reduces the dependence on the use of a particular labeler. In addition, the use of a consensus label, agreed upon by a rather uneven group of labelers, might be at least as good as using the gold standard labeler, who might not be available, and certainly better than randomly selecting one of the group's individual labelers. Finally, using the AL methods: when provided by the consensus label reduced the intra-labeler AUC variance during the learning phase, compared to using passive learning.

Text mining approach to predict hospital admissions using early medical records from the emergency department

OBJECTIVE

Emergency department (ED) overcrowding is a serious issue for hospitals. Early information on short-term inward bed demand from patients receiving care at the ED may reduce the overcrowding problem, and optimize the use of hospital resources. In this study, we use text mining methods to process data from early ED patient records using the SOAP framework, and predict future hospitalizations and discharges.

DESIGN

We try different approaches for pre-processing of text records and to predict hospitalization. Sets-of-words are obtained via binary representation, term frequency, and term frequency-inverse document frequency. Unigrams, bigrams and trigrams are tested for feature formation. Feature selection is based on χ² and F-score metrics. In the prediction module, eight text mining methods are tested: Decision Tree, Random Forest, Extremely Randomized Tree, AdaBoost, Logistic Regression, Multinomial Naïve Bayes, Support Vector Machine (Kernel linear) and Nu-Support Vector Machine (Kernel linear).

MEASUREMENTS

Prediction performance is evaluated by F1-scores. Precision and Recall values are also informed for all text mining methods tested.

RESULTS

Nu-Support Vector Machine was the text mining method with the best overall performance. Its average F1-score in predicting hospitalization was 77.70%, with a standard deviation (SD) of 0.66%.

CONCLUSIONS

The method could be used to manage daily routines in EDs such as capacity planning and resource allocation. Text mining could provide valuable information and facilitate decision-making by inward bed management teams.

Using machine learning classifiers to assist healthcare-related decisions: classification of electronic patient records

Surveillance Levels (SLs) are categories for medical patients (used in Brazil) that represent different types of medical recommendations. SLs are defined according to risk factors and the medical and developmental history of patients. Each SL is associated with specific educational and clinical measures. The objective of the present paper was to verify computer-aided, automatic assignment of SLs. The present paper proposes a computer-aided approach for automatic recommendation of SLs. The approach is based on the classification of information from patient electronic records. For this purpose, a software architecture composed of three layers was developed. The architecture is formed by a classification layer that includes a linguistic module and machine learning classification modules. The classification layer allows for the use of different classification methods, including the use of preprocessed, normalized language data drawn from the linguistic module. We report the verification and validation of the software architecture in a Brazilian pediatric healthcare institution. The results indicate that selection of attributes can have a great effect on the performance of the system. Nonetheless, our automatic recommendation of surveillance level can still benefit from improvements in processing procedures when the linguistic module is applied prior to classification. Results from our efforts can be applied to different types of medical systems. The results of systems supported by the framework presented in this paper may be used by healthcare and governmental institutions to improve healthcare services in terms of establishing preventive measures and alerting authorities about the possibility of an epidemic.