Improving length of stay prediction using a hidden Markov model (original) (raw)
Related papers
2012
Intensive care medicine is a large share of the health care budget, and in the last decade there has been an increasing focus on making intensive care medicine more cost-effective by the efficient use of resources while still providing the best outcome for critically-ill patients. One important set of tools to perform this are critical illness severity assessment scores such as Simplified Acute Physiology score (SAPS-I) which help clinicians prioritize resources and determine the appropriate diagnostic/therapeutic plan for each patient. These scores are also used for assessing how medications, care guidelines, surgery, and other interventions impact mortality in Intensive Care Unit (ICU) patients. In an attempt to develop an improved patient-specific prediction of in-hospital mortality, we propose an algorithm based on logistic regression and Hidden-Markov model using vital signs (vitals), laboratory values (labs) and fluid measurements that are commonly available in ICUs. The algor...
Using Clinical Notes with Time Series Data for ICU Management
Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)
Monitoring patients in ICU is a challenging and high-cost task. Hence, predicting the condition of patients during their ICU stay can help provide better acute care and plan the hospital's resources. There has been continuous progress in machine learning research for ICU management, and most of this work has focused on using time series signals recorded by ICU instruments. In our work, we show that adding clinical notes as another modality improves the performance of the model for three benchmark tasks: in-hospital mortality prediction, modeling decompensation, and length of stay forecasting that play an important role in ICU management. While the time-series data is measured at regular intervals, doctor notes are charted at irregular times, making it challenging to model them together. We propose a method to model them jointly, achieving considerable improvement across benchmark tasks over baseline time-series model. Our implementation can be found at https://github. com/kaggarwal/ClinicalNotesICU.
Predicting Prolonged Length of ICU Stay through Machine Learning
Diagnostics, 2021
This study aimed to construct machine learning (ML) models for predicting prolonged length of stay (pLOS) in intensive care units (ICU) among general ICU patients. A multicenter database called eICU (Collaborative Research Database) was used for model derivation and internal validation, and the Medical Information Mart for Intensive Care (MIMIC) III database was used for external validation. We used four different ML methods (random forest, support vector machine, deep learning, and gradient boosting decision tree (GBDT)) to develop prediction models. The prediction performance of the four models were compared with the customized simplified acute physiology score (SAPS) II. The area under the receiver operation characteristic curve (AUROC), area under the precision-recall curve (AUPRC), estimated calibration index (ECI), and Brier score were used to measure performance. In internal validation, the GBDT model achieved the best overall performance (Brier score, 0.164), discrimination ...
Statistical Methods in Medical Research, 2020
Continuous mortality risk monitoring is instrumental to manage a patient's care and to efficiently utilize the limited hospital resources. Due to incompleteness and irregularities of electronic health records (EHR), developing continuous mortality risk prediction using EHR data is a challenge. In this study, we propose a framework to continuously monitor mortality risk, and apply it to the real-world EHR data. The proposed method employs hidden Markov models (temporal technique) that take account of both the previous state of patient's health and the current value of clinical signs. Following the Sepsis-3 definition, we selected 3898 encounters of patients with suspected infection to compare the performance of temporal and non-temporal methods (Decision Tree (DT), Logistic Regression (LR), Naive Bayes (NB), Random Forest (RF), and Support Vector Machine (SVM)). The area under receiver operating characteristics (AUROC) curve, sensitivity, specificity and G-mean were used as performance measures. On the selected data, the AUROC of the proposed temporal framework (0.87) is 9-12% greater than the nontemporal methods (DT: 0.78, NB: 0.79, SVM: 0.79, LR: 0.80 and RF: 0.80). The results also show that our model (G-mean 1 = 4 0.78) provides a better balance between sensitivity and specificity compared to clinically acceptable bedside criteria (G-mean 1 = 4 0.71). The proposed framework leverages the longitudinal data available in EHR and performs better than the non-temporal methods. The proposed method facilitates information related to the time of change of the patient's health that may help practitioners to plan early and develop effective treatment strategies.
The Lancet Digital Health, 2020
Background Many mortality prediction models have been developed for patients in intensive care units (ICUs); most are based on data available at ICU admission. We investigated whether machine learning methods using analyses of time-series data improved mortality prognostication for patients in the ICU by providing real-time predictions of 90-day mortality. In addition, we examined to what extent such a dynamic model could be made interpretable by quantifying and visualising the features that drive the predictions at different timepoints. Methods Based on the Simplified Acute Physiology Score (SAPS) III variables, we trained a machine learning model on longitudinal data from patients admitted to four ICUs in the Capital Region, Denmark, between 2011 and 2016. We included all patients older than 16 years of age, with an ICU stay lasting more than 1 h, and who had a Danish civil registration number to enable 90-day follow-up. We leveraged static data and physiological time-series data from electronic health records and the Danish National Patient Registry. A recurrent neural network was trained with a temporal resolution of 1 h. The model was internally validated using the holdout method with 20% of the training dataset and externally validated using previously unseen data from a fifth hospital in Denmark. Its performance was assessed with the Matthews correlation coefficient (MCC) and area under the receiver operating characteristic curve (AUROC) as metrics, using bootstrapping with 1000 samples with replacement to construct 95% CIs. A Shapley additive explanations algorithm was applied to the prediction model to obtain explanations of the features that drive patient-specific predictions, and the contributions of each of the 44 features in the model were analysed and compared with the variables in the original SAPS III model. Findings From a dataset containing 15 615 ICU admissions of 12 616 patients, we included 14 190 admissions of 11 492 patients in our analysis. Overall, 90-day mortality was 33⋅1% (3802 patients). The deep learning model showed a predictive performance on the holdout testing dataset that improved over the timecourse of an ICU stay: MCC 0⋅29 (95% CI 0⋅25-0⋅33) and AUROC 0⋅73 (0⋅71-0⋅74) at admission, 0⋅43 (0⋅40-0⋅47) and 0⋅82 (0⋅80-0⋅84) after 24 h, 0⋅50 (0⋅46-0⋅53) and 0⋅85 (0⋅84-0⋅87) after 72 h, and 0⋅57 (0⋅54-0⋅60) and 0⋅88 (0⋅87-0⋅89) at the time of discharge. The model exhibited good calibration properties. These results were validated in an external validation cohort of 5827 patients with 6748 admissions: MCC 0⋅29 (95% CI 0⋅27-0⋅32) and AUROC 0⋅75 (0⋅73-0⋅76) at admission, 0⋅41 (0⋅39-0⋅44) and 0⋅80 (0⋅79-0⋅81) after 24 h, 0⋅46 (0⋅43-0⋅48) and 0⋅82 (0⋅81-0⋅83) after 72 h, and 0⋅47 (0⋅44-0⋅49) and 0⋅83 (0⋅82-0⋅84) at the time of discharge. Interpretation The prediction of 90-day mortality improved with 1-h sampling intervals during the ICU stay. The dynamic risk prediction can also be explained for an individual patient, visualising the features contributing to the prediction at any point in time. This explanation allows the clinician to determine whether there are elements in the current patient state and care that are potentially actionable, thus making the model suitable for further validation as a clinical tool. Funding Novo Nordisk Foundation and the Innovation Fund Denmark.
Learning Temporal Rules to Forecast Instability in Intensive Care Patients
Intensive care medicine, 2013
Inductive machine learning, and in particular extraction of association rules from data, has been successfully used in multiple application domains, such as market basket analysis, disease prognosis, fraud detection, and protein sequencing. The appeal of rule extraction techniques stems from their ability to handle intricate problems yet produce models based on rules that can be comprehended by humans, and are therefore more transparent. Human comprehension is a factor that may improve adoption and use of data-driven decision support systems clinically via face validity. In this work, we explore whether we can reliably and informatively forecast cardiorespiratory instability (CRI) in step-down unit (SDU) patients utilizing data from continuous monitoring of physiologic vital sign (VS) measurements. We use a temporal association rule extraction technique in conjunction with a rule fusion protocol to learn how to forecast CRI in continuously monitored patients. We detail our approach and present and discuss encouraging empirical results obtained using continuous multivariate VS data from the bedside monitors of 297 SDU patients spanning 29 346 hours (3.35 patient-years) of observation. We present example rules that have been learned from data to illustrate potential benefits of comprehensibility of the extracted models, and we analyze the empirical utility of each VS as a potential leading indicator of an impending CRI event.
Patient length of stay and mortality prediction: A survey
Health Services Management Research, 2017
Over the past few years, there has been increased interest in data mining and machine learning methods to improve hospital performance, in particular hospitals want to improve their intensive care unit statistics by reducing the number of patients dying inside the intensive care unit. Research has focused on prediction of measurable outcomes, including risk of complications, mortality and length of hospital stay. The length of stay is an important metric both for healthcare providers and patients, influenced by numerous factors. In particular, the length of stay in critical care is of great significance, both to patient experience and the cost of care, and is influenced by factors specific to the highly complex environment of the intensive care unit. The length of stay is often used as a surrogate for other outcomes, where those outcomes cannot be measured; for example as a surrogate for hospital or intensive care unit mortality. The length of stay is also a parameter, which has bee...
Predicting Patient-ventilator Asynchronies with Hidden Markov Models
Scientific Reports
In mechanical ventilation, it is paramount to ensure the patient's ventilatory demand is met while minimizing asynchronies. We aimed to develop a model to predict the likelihood of asynchronies occurring. We analyzed 10,409,357 breaths from 51 critically ill patients who underwent mechanical ventilation >24 h. Patients were continuously monitored and common asynchronies were identified and regularly indexed. Based on discrete time-series data representing the total count of asynchronies, we defined four states or levels of risk of asynchronies, z1 (very-low-risk)-z4 (very-high-risk). A Poisson hidden Markov model was used to predict the probability of each level of risk occurring in the next period. Long periods with very few asynchronous events, and consequently very-low-risk, were more likely than periods with many events (state z4). States were persistent; large shifts of states were uncommon and most switches were to neighbouring states. Thus, patients entering states with a high number of asynchronies were very likely to continue in that state, which may have serious implications. This novel approach to dealing with patient-ventilator asynchrony is a first step in developing smart alarms to alert professionals to patients entering high-risk states so they can consider actions to improve patient-ventilator interaction. Patients in intensive care units (ICU) sometimes need mechanical ventilation to improve alveolar ventilation and oxygenation while decreasing the load on the respiratory muscles. Patients may undergo mechanical ventilation for several days until their condition improves. Although mechanical ventilation is a life-saving intervention, numerous complications can develop. Ventilator cycles must match the patient's own rhythm of breathing; however, mismatching is common, resulting in poor patient-ventilator interaction with deleterious consequences 1-5. When patient-ventilator asynchronies occur, breathing becomes more difficult and the patient's condition can worsen. Asynchronies are more dangerous when their frequency is relatively high 4. Asynchronies can prolong mechanical ventilation and ICU stays 4,6 , increase the probability of respiratory muscle and lung injury 7,8 , increase mortality 3,9 , and lead to other complications 10,11. Personalized or precision medicine is an emerging concept that will change clinical practice in ICUs in the short-to-mid term, helping physicians choose the right therapy at the right time 12,13. ICU patients are intensely and continuously monitored, generating extremely large datasets. All this data is readily available and can be exploited with big data tools and automated learning systems, providing a unique opportunity to improve decision-making in this demanding environment. Based on their understanding of the physiological principles involved and evidence from clinical studies, physicians manage mechanical ventilation by assessing waveforms on bedside monitors. However, most perform poorly at managing patient-ventilator interactions, often failing to identify common forms of asynchronies 14,15. Moreover, patients take several thousands of breaths each day, and busy professionals can observe only a small proportion of these. Early detection of an increased frequency of asynchronies could alert physicians to
JAMIA Open, 2024
Objective: This study aimed to investigate the predictive capabilities of historical patient records to predict patient adverse outcomes such as mortality, readmission, and prolonged length of stay (PLOS). Methods: Leveraging a de-identified dataset from a tertiary care university hospital, we developed an eXplainable Artificial Intelligence (XAI) framework combining tree-based and traditional machine learning (ML) models with interpretations and statistical analysis of predictors of mortality, readmission, and PLOS. Results: Our framework demonstrated exceptional predictive performance with a notable area under the receiver operating characteristic (AUROC) of 0.9625 and an area under the precision-recall curve (AUPRC) of 0.8575 for 30-day mortality at discharge and an AUROC of 0.9545 and AUPRC of 0.8419 at admission. For the readmission and PLOS risk, the highest AUROC achieved were 0.8198 and 0.9797, respectively. The tree-based models consistently outperformed the traditional ML models in all 4 prediction tasks. The key predictors were age, derived temporal features, routine laboratory tests, and diagnostic and procedural codes. Conclusion: The study underscores the potential of leveraging medical history for enhanced hospital predictive analytics. We present an accurate and intuitive framework for early warning models that can be easily implemented in the current and developing digital health platforms to predict adverse outcomes accurately.
Machine Learning and Medical Data: Predicting ICU Mortality and Re-admission Risks
Intensive care units (ICUs) are divisions where critically ill patients are treated by medical experts. The unmet and vital need for automated clinical decision-making mechanisms is critical to maneuvering the large influx of patients. This became more apparent after the COVID-19 pandemic. Existing studies focus on determining the probability of patients dying in the ICUs and prioritizing patients in dire need. Only a few studies have calculated the patient’s probability of returning to the ICUs after discharge. These studies reduce the problem into a binary task of predicting mortality or re-admission only. However, this is unrealistic since both outcomes are highly possible for each patient. In this interdisciplinary study, two main contributions are proposed for the automated clinical decision-making stateof-the-art: (1) using the real-life data collected from thousands of ICU patients by healthcare professionals, three possibilities (recovery, mortality, and returning to the intensive care unit within 30 days) are predicted for patients in intensive care instead of just one possibility. (2) A novel feature extraction approach is proposed by the biomedical expert in our team. Four machine learning algorithms are applied to the finalized feature set to understand the difference between the binary and the multi-class classification problems. Obtained results reach 78% success, proving the possibility of developing better clinical decision-making mechanisms for ICUs.