Readmission Risk Prediction Research Articles

Identifying patients at highest-risk: the best timing to apply a readmission predictive model

BackgroundMost of readmission prediction models are implemented at the time of patient discharge. However, interventions which include an early in-hospital component are critical in reducing readmissions and improving patient outcomes. Thus, at-discharge high-risk identification may be too late for effective intervention. Nonetheless, the tradeoff between early versus at-discharge prediction and the optimal timing of the risk prediction model application remains to be determined. We examined a high-risk patient selection process with readmission prediction models using data available at two time points: at admission and at the time of hospital discharge.MethodsAn historical prospective study of hospitalized adults (≥65 years) discharged alive from internal medicine units in Clalit’s (the largest integrated payer-provider health fund in Israel) general hospitals in 2015. The outcome was all-cause 30-day emergency readmissions to any internal medicine ward at any hospital. We used the previously validated Preadmission Readmission Detection Model (PREADM) and developed a new model incorporating PREADM with hospital data (PREADM-H). We compared the percentage of overlap between the models and calculated the positive predictive value (PPV) for the subgroups identified by each model separately and by both models.ResultsThe final cohort included 35,156 index hospital admissions. The PREADM-H model included 17 variables with a C-statistic of 0.68 (95% CI: 0.67–0.70) and PPV of 43.0% in the highest-risk categories. Of patients categorized by the PREADM-H in the highest-risk decile, 78% were classified similarly by the PREADM. The 22% (n = 229) classified by the PREADM-H at the highest decile, but not by the PREADM, had a PPV of 37%. Conversely, those classified by the PREADM into the highest decile but not by the PREADM-H (n = 218) had a PPV of 31%.ConclusionsThe timing of readmission risk prediction makes a difference in terms of the population identified at each prediction time point – at-admission or at-discharge. Our findings suggest that readmission risk identification should incorporate a two time-point approach in which preadmission data is used to identify high-risk patients as early as possible during the index admission and an “all-hospital” model is applied at discharge to identify those that incur risk during the hospital stay.

Can We Do More With Less While Building Predictive Models? A Study in Parsimony of Risk Models for Predicting Heart Failure Readmissions.

Hospital readmission due to heart failure is a topic of concern for patients and hospitals alike: it is both the most frequent and expensive diagnosis for hospitalization. Therefore, accurate prediction of readmission risk while patients are still in the hospital helps to guide appropriate postdischarge interventions. As our understanding of the disease and the volume of electronic health record data both increase, the number of predictors and model-building time for predicting risk grow rapidly. This suggests a need to use methods for reducing the number of predictors without losing predictive performance. We explored and described three such methods and demonstrated their use by applying them to a real-world dataset consisting of 57 variables from health data of 1210 patients from one hospital system. We compared all models generated from predictor reduction methods against the full, 57-predictor model for predicting risk of 30-day readmissions for patients with heart failure. Our predictive performance, measured by the C-statistic, ranged from 0.630 to 0.840, while model-building time ranged from 10 minutes to 10 hours. Our final model achieved a C-statistic (0.832) comparable to the full model (0.840) in the validation cohort while using only 16 predictors and providing a 66-fold improvement in model-building time.

Hospital readmission risk prediction based on claims data available at admission: a pilot study in Switzerland.

ObjectivesEvaluating whether future studies to develop prediction models for early readmissions based on health insurance claims data available at the time of a hospitalisation are worthwhile.DesignRetrospective cohort study of hospital admissions with discharge dates between 1 January 2014 and 31 December 2016.SettingAll-cause acute care hospital admissions in the general population of Switzerland, enrolled in the Helsana Group, a large provider of Swiss mandatory health insurance.ParticipantsThe mean age of 138 222 hospitalised adults included in the study was 60.5 years. Patients were included only with their first index hospitalisation. Patients who deceased during the follow-up period were excluded, as well as patients admitted from and/or discharged to nursing homes or rehabilitation clinics.MeasuresThe primary outcome was 30-day readmission rate. Area under the receiver operating characteristic curve (AUC) was used to measure the discrimination of the developed logistic regression prediction model. Candidate variables were theory based and derived from a systematic literature search.ResultsWe observed a 30-day readmission rate of 7.5%. Fifty-five candidate variables were identified. The final model included pharmacy-based cost group (PCG) cancer, PCG cardiac disease, PCG pain, emergency index admission, number of emergency visits, costs specialists, costs hospital outpatient, costs laboratory, costs therapeutic devices, costs physiotherapy, number of outpatient visits, sex, age group and geographical region as predictors. The prediction model achieved an AUC of 0.60 (95% CI 0.60 to 0.61).ConclusionsBased on the results of our study, it is not promising to invest resources in large-scale studies for the development of prediction tools for hospital readmissions based on health insurance claims data available at admission. The data proved appropriate to investigate the occurrence of hospitalisations and subsequent readmissions, but we did not find evidence for the potential of a clinically helpful prediction tool based on patient-sided variables alone.

Derivation and external validation of a simple risk tool to predict 30-day hospital readmissions after transcatheter aortic valve replacement.

Patients undergoing transcatheter aortic valve replacement (TAVR) possess a higher risk of recurrent healthcare resource utilisation due to multiple comorbidities, frailty, and advanced age. We sought to devise a simple tool to identify TAVR patients at increased risk of 30-day readmission. We used the Nationwide Readmissions Database from January 2013 to September 2015. Complex survey methods and hierarchical regression in R were implemented to create a prediction tool to determine probability of 30-day readmission. Boot-strapped internal validation and cross-validation were performed to assess model accuracy. External validation was performed using a single-centre data set. Of 39,305 patients who underwent endovascular TAVR, 6,380 (16.2%) were readmitted within 30 days. The final 30-day readmission risk prediction tool included the following variables: chronic kidney disease, end-stage renal disease on dialysis (ESRD), anaemia, chronic lung disease, chronic liver disease, atrial fibrillation, length of stay, acute kidney injury, and discharge disposition. ESRD (OR 2.11, 95% CI: 1.7-2.63), length of stay ≥5 days (OR 1.64, 95% CI: 1.50-1.79), and short-term hospital discharge disposition (OR 1.81, 95% CI: 1.2-2.7) were the strongest predictors. The c-statistic of the prediction model was 0.63. The c-statistic in the external validation cohort was 0.69. On internal calibration, the tool was extremely accurate in predicting readmissions up to 25%. A simple and easy-to-use risk prediction tool utilising standard clinical parameters identifies TAVR patients at increased risk of 30-day readmission. The tool may consequently inform hospital discharge planning, optimise transitions of care, and reduce resource utilisation.

The Impact of Pharmacy-specific Predictors on the Performance of 30-Day Readmission Risk Prediction Models.

Pharmacists are an expensive and limited resource in the hospital and outpatient setting. A pharmacist can spend up to 25% of their day planning. Time spent planning is time not spent delivering an intervention. A readmission risk adjustment model has potential to be used as a universal outcome-based prioritization tool to help pharmacists plan their interventions more efficiently. Pharmacy-specific predictors have not been used in the constructs of current readmission risk models. We assessed the impact of adding pharmacy-specific predictors on performance of readmission risk prediction models. We used an observational retrospective cohort study design to assess whether pharmacy-specific predictors such as an aggregate pharmacy score and drug classes would improve the prediction of 30-day readmission. A model of age, sex, length of stay, and admission category predictors was used as the reference model. We added predictor variables in sequential models to evaluate the incremental effect of additional predictors on the performance of the reference. We used logistic regression to regress the outcomes on predictors in our derivation dataset. We derived and internally validated our models through a 50:50 split validation of our dataset. Our study population (n=350,810) was of adult admissions at hospitals in a large integrated health care delivery system. Individually, the aggregate pharmacy score and drug classes caused a nearly identical but moderate increase in model performance over the reference. As a single predictor, the comorbidity burden score caused the greatest increase in model performance when added to the reference. Adding the severity of illness score, comorbidity burden score and the aggregate pharmacy score to the reference caused a cumulative increase in model performance with good discrimination (c statistic, 0.712; Nagelkerke R, 0.112). The best performing model included all predictors: severity of illness score, comorbidity burden score, aggregate pharmacy score, diagnosis groupings, and drug subgroups. Adding the aggregate pharmacy score to the reference model significantly increased the c statistic but was out-performed by the comorbidity burden score model in predicting readmission. The need for a universal prioritization tool for pharmacists may therefore be potentially met with the comorbidity burden score model. However, the aggregate pharmacy score and drug class models still out-performed current Medicare readmission risk adjustment models. Pharmacists have a great role in preventing readmission, and therefore can potentially use one of our models: comorbidity burden score model, aggregate pharmacy score model, drug class model or complex model (a combination of all 5 major predictors) to prioritize their interventions while exceeding Medicare performance measures on readmission. The choice of model to use should be based on the availability of these predictors in the health care system.

Abstract 161: Adding Social and Functional Determinants of Health Improves Readmission Risk Prediction in Urban Underserved Adults Hospitalized for Acute Myocardial Infarction

Background: Although social and functional determinants of health (SFDoH) are increasingly recognized to contribute to increased 30-day readmission risk for cardiovascular illnesses, most health systems lack individual-level data to enable individual risk stratification and targeted interventions to address these needs. We sought to determine whether electronic health record (EHR) data in SFDoH domains collected as a part of routine inpatient care from a safety-net hospital could be used to enhance prediction of 30-day readmissions beyond traditional clinical risk factors alone among underserved adults hospitalized with acute myocardial infarction (AMI). Methods: We used Epic EHR data on AMI hospitalizations from 2014-2017 in an urban safety-net hospital in Dallas to predict non-elective 30-day readmissions to any of 75 hospitals in north Texas, using a regional all-payer hospitalization database. We started with the AMI READMITS score, our previously validated model consisting of 7 clinical risk factors. Next, we derived and validated an enhanced model incorporating SFDoH predictors. EHR data for candidate SFDoH predictors included structured, semi-structured, and short free-text fields from demographics, encounter types, note types, and case manager, social worker, and nursing inpatient flowsheets. We selected novel predictors using stepwise backward selection with a threshold of p≤0.20, forcing in the composite AMI READMITS score. We conducted sensitivity analyses using forward selection, lasso regression, and random forest which resulted in similar models. We internally validated our final model using 5-fold cross-validation and compared model performance to the AMI READMITS score. Results: Of 1,071 hospitalizations, 154 (14.5%) were readmitted within 30 days. In addition to the composite AMI READMITS score, our final model included 9 SFDoH risk factors, including neighborhood poverty (aOR 1.39, 95% CI 0.91-2.14), address change in the past year (aOR 1.53, 95% CI 0.96-2.46), any opioid use in the past year (aOR 2.35, 95% CI 1.36-4.05), history of durable medical equipment use in the past year (aOR 1.70, 95% CI 1.02-2.82), and ADL impairments in feeding, bathing, and toileting in the past year (aOR 1.81 for all three impairments, 95% CI 0.99-3.29). The AMI READMITS+SFDoH model had better discrimination than AMI READMITS alone (optimism-corrected C-statistic 0.69 vs 0.62, p<0.001), was able to predict a broader range of probabilities for readmission risk (3.6-72.7% vs. 8.2-37.2%), and improved net reclassification (categorical NRI of highest 2 risk quintiles vs. lowest 3 of 0.09, 95% CI 0.00-0.18). Conclusions: Incorporating existing EHR data on social and functional determinants of health improved readmission risk prediction after an AMI hospitalization in an urban underserved population.

Modeling Biobehavioral Rhythms with Passive Sensing in the Wild

Biobehavioral rhythms are associated with numerous health and life outcomes. We study the feasibility of detecting rhythms in data that is passively collected from Fitbit devices and using the obtained model parameters to predict readmission risk after pancreatic surgery. We analyze data from 49 patients who were tracked before surgery, in hospital, and after discharge. Our analysis produces a model of individual patients' rhythms for each stage of treatment that is predictive of readmission. All of the rhythm-based models outperform the traditional approaches to readmission risk stratification that uses administrative data.

Predictive Modeling of the Hospital Readmission Risk from Patients\u2019 Claims Data Using Machine Learning: A Case Study on COPD

Chronic Obstructive Pulmonary Disease (COPD) is a prevalent chronic pulmonary condition that affects hundreds of millions of people all over the world. Many COPD patients got readmitted to hospital within 30 days after discharge due to various reasons. Such readmission can usually be avoided if additional attention is paid to patients with high readmission risk and appropriate actions are taken. This makes early prediction of the hospital readmission risk an important problem. The goal of this paper is to conduct a systematic study on developing different types of machine learning models, including both deep and non-deep ones, for predicting the readmission risk of COPD patients. We evaluate those different approaches on a real world database containing the medical claims of 111,992 patients from the Geisinger Health System from January 2004 to September 2015. The patient features we build the machine learning models upon include both knowledge-driven ones, which are the features extracted according to clinical knowledge potentially related to COPD readmission, and data-driven features, which are extracted from the patient data themselves. Our analysis showed that the prediction performance in terms of Area Under the receiver operating characteristic (ROC) Curve (AUC) can be improved from around 0.60 using knowledge-driven features, to 0.653 by combining both knowledge-driven and data-driven features, based on the one-year claims history before discharge. Moreover, we also demonstrate that the complex deep learning models in this case cannot really improve the prediction performance, with the best AUC around 0.65.

Risk factors associated with paediatric unplanned hospital readmissions: a systematic review

ObjectiveTo synthesise evidence on risk factors associated with paediatric unplanned hospital readmissions (UHRs).DesignSystematic review.Data sourceCINAHL, EMBASE (Ovid) and MEDLINE from 2000 to 2017.Eligibility criteriaStudies published in English with full-text access...

Pre- and within hospitalization risk factors for readmission of older adults

Introduction: Unplanned readmission is indicative of poor quality of care at the hospital-community interface. Yet, the unique contribution of in-hospital care processes to readmission risk is unknown. We aimed to test whether information on processes of hospitalization, such as mobility, functioning, and nutritional intake, contribute to the ability to accurately identify older adults at high-risk of readmission beyond their at-admission risk. Methods: Prospective cohort study included 559 older adults (aged≥70) admitted to internal medicine wards for non-disabling conditions and discharged to their homes. Data on any urgent readmission up to 30 days after discharge were retrieved from an electronic health record (EHR) database. Information on pre-admission (e.g., chronic conditions and healthcare utilization in the prior year), at-admission (e.g., functional and nutritional status) risk factors was collected from EHRs and survey. Data on processes of hospitalization: mobility, incontinence care, nutritional intake and functional decline, were collected using validated questionnaires. A multiple logistic regression model for predicting readmission was developed. Results: Readmission occurred in 15.2% of the cases. At-admission high and medium (vs. low) risk of malnutrition (Odds ratio (OR)=2.89, 95% Confidence interval (CI)=1.44-5.80; OR=2.35, 95% CI=1.13-4.91, respectively), higher serum albumin levels (OR for each mg/dL increase=0.45, 95% CI=0.28-0.74), prior hospitalizations (OR for each additional hospitalization=1.30, 95% CI=1.13-1.50), diagnoses of malignancy (OR=2.25, 95% CI=1.31-3.86), and/or chronic renal failure (OR=2.90, 95% CI=1.71-4.89), yielded good discrimination (c-statistic=0.79). Among the processes of hospitalization, only admission to discharge functional decline was significantly associated with readmission (OR for each 10-point decrease in the Barthel Index = 1.32, 95% CI = 1.02–1.72), but did not contribute to the overall discrimination of the model as compared with the at-admission data. Classification of patients into risk categories by at-admission combined with the in-hospital data shows that identifying patients in the top-tier category (10th highest percentile) according to the at-admission model misses 7/55 (12.7%) of patients who would have been categorized as highest-risk if risk identification were postponed to the discharge date. Discussion: Information on processes of hospitalization does not contribute to the accuracy of a readmission risk prediction model that utilizes data on prior hospitalizations, baseline nutritional and functional status and chronic morbidity (CRF and malignancy). Conclusion: Older adults can be identified as high-risk for readmission early during the index admission. Early evaluation of the readmission risk is likely to allow timely intervention to improve integration of care transitions in the target population. Further at-discharge functional assessment can detect additional patients whose readmission risk changes during the index admission and who should be considered for inclusion in readmission-prevention interventions. Lessons learned: At-admission clinical and functional status combined with at-discharge additional functional assessment can be used to target older adults at high-risk for readmission. Limitations: A convenient sample of a relatively high functioning group of patients. These inclusion criteria may have affected the readmission rates. Suggestions for future research: Future studies should examine the significance of in-hospital functioning in other older adult populations, such as greater mix of baseline functioning and morbidity.

Predicting ICU readmission using grouped physiological and medication trends.

Patients who are readmitted to an intensive care unit (ICU) usually have a high risk of mortality and an increased length of stay. ICU readmission risk prediction may help physicians to re-evaluate the patient's physical conditions before patients are discharged and avoid preventable readmissions. ICU readmission prediction models are often built based on physiological variables. Intuitively, snapshot measurements, especially the last measurements, are effective predictors that are widely used by researchers. However, methods that only use snapshot measurements neglect predictive information contained in the trends of physiological and medication variables. Mean, maximum or minimum values take multiple time points into account and capture their summary statistics, however, these statistics are not able to catch the detailed picture of temporal trends. In this study, we find strong predictors with ability of capturing detailed temporal trends of variables for 30-day readmission risk and build prediction models with high accuracy. We study physiological measurements and medications from the Multiparameter Intelligent Monitoring in Intensive Care II (MIMIC-II) clinical dataset. Time series of each variable are converted into trend graphs with nodes being discretized measurements of each variable. Then we extract important temporal trends by applying frequent subgraph mining on the trend graphs. The frequency of a subgraph is a good cue to find important temporal trends since similar patients often share similar trends regarding their pathophysiological evolution under medical interventions. Important temporal trends are then grouped automatically by non-negative matrix factorization. The grouped trends could be considered as an approximate representation of patients' pathophysiological states and medication profiles. We train a logistic regression model to predict 30-day ICU readmission risk based on snapshot measurements, grouped physiological trends and medication trends. Our dataset consists of 1170 patients who are alive 30 days after discharge from ICU and have at least 12 h of data. In the dataset, 860 patients were not readmitted and 310 were readmitted, within 30 days after discharge. Our model outperforms all comparison models, and shows an improvement in the area under the receiver operating characteristic curve (AUC) of almost 4% from the best comparison model. Grouped physiological and medication trends carry predictive information for ICU readmission risk. In order to build predictive models with higher accuracy, we should add grouped physiological and medication trends as complementary features to snapshot measurements.

Analysing repeated hospital readmissions using data mining techniques

Few studies have examined how to identify future readmission of patients with a large number of repeat emergency department (ED) visits. We explore 30-day readmission risk prediction using Microsoft’s AZURE machine learning software and compare five classification methods: Logistic Regression, Boosted Decision Trees (BDTs), Support Vector Machine (SVM), Bayes Point Machine (BPM), and Two-Class Neural Network (TCNN). We predict the last readmission visit of frequent ED patients extracted from the electronic health records of their 8455 penultimate visits. The methods show differential improvement, with the BDT indicating marginally better AUC (area under the ROC curve) than logistic regression and BPM, followed by the TCNN and SVM. A comparison of BDT and Logistic Regression results for correct and incorrect classification highlights the similarities and differences in the significant predictors identified by each method. Future research may incorporate time-varying covariates to identify other longitudinal factors that can lead to readmission risk reduction.

An Analytical Framework for TJR Readmission Prediction and Cost-Effective Intervention

This paper introduces an analytical framework for assessing the cost-effectiveness of intervention strategies to reduce total joint replacement (TJR) readmissions. In such a framework, a machine learning-based readmission risk prediction model is developed to predict an individual TJR patient's risk of hospital readmission within 90 days post-discharge. Specifically, through data sampling and boosting techniques, we overcome the class imbalance problem by iteratively building an ensemble of models. Then, utilizing the results of the predictive model, and by taking into account the imbalanced misclassification costs between readmitted and nonreadmitted patients, a cost analysis framework is introduced to support decision making in selecting cost-effective intervention policies. Finally, using this framework, a case study at a community hospital is presented to demonstrate the applicability of the analysis.

Validation of the Registered Nurse Assessment of Readiness for Hospital Discharge Scale.

Statistical models for predicting readmissions have been published for high-risk patient populations but typically focus on patient characteristics; nurse judgment is rarely considered in a formalized way to supplement prediction models. The purpose of this study was to determine psychometric properties of long and short forms of the Registered Nurse Readiness for Hospital Discharge Scale (RN-RHDS), including reliability, factor structure, and predictive validity. Data were aggregated from two studies conducted at four hospitals in the Midwestern United States. The RN-RHDS was completed within 4 hours before hospital discharge by the discharging nurse. Data on readmissions and emergency department visits within 30 days were extracted from electronic medical records. The RN-RHDS, both long and short forms, demonstrate acceptable reliability (Cronbach's alphas of .90 and .73, respectively). Confirmatory factor analysis demonstrated less than adequate fit with the same four-factor structure observed in the patient version. Exploratory factor analysis identified three factors, explaining 60.2% of the variance. When nurses rate patients as less ready to go home (<7 out of 10), patients are 6.4-9.3 times more likely to return to the hospital within 30 days, in adjusted models. The RN-RHDS, long and short forms, can be used to identify medical-surgical patients at risk for potential unplanned return to hospital within 30 days, allowing nurses to use their clinical judgment to implement interventions prior to discharge. Use of the RN-RHDS could enhance current readmission risk prediction models.

A machine learning model to predict the risk of 30-day readmissions in patients with heart failure: a retrospective analysis of electronic medical records data

BackgroundHeart failure is one of the leading causes of hospitalization in the United States. Advances in big data solutions allow for storage, management, and mining of large volumes of structured and semi-structured data, such as complex healthcare data. Applying these advances to complex healthcare data has led to the development of risk prediction models to help identify patients who would benefit most from disease management programs in an effort to reduce readmissions and healthcare cost, but the results of these efforts have been varied. The primary aim of this study was to develop a 30-day readmission risk prediction model for heart failure patients discharged from a hospital admission.MethodsWe used longitudinal electronic medical record data of heart failure patients admitted within a large healthcare system. Feature vectors included structured demographic, utilization, and clinical data, as well as selected extracts of un-structured data from clinician-authored notes. The risk prediction model was developed using deep unified networks (DUNs), a new mesh-like network structure of deep learning designed to avoid over-fitting. The model was validated with 10-fold cross-validation and results compared to models based on logistic regression, gradient boosting, and maxout networks. Overall model performance was assessed using concordance statistic. We also selected a discrimination threshold based on maximum projected cost saving to the Partners Healthcare system.ResultsData from 11,510 patients with 27,334 admissions and 6369 30-day readmissions were used to train the model. After data processing, the final model included 3512 variables. The DUNs model had the best performance after 10-fold cross-validation. AUCs for prediction models were 0.664 ± 0.015, 0.650 ± 0.011, 0.695 ± 0.016 and 0.705 ± 0.015 for logistic regression, gradient boosting, maxout networks, and DUNs respectively. The DUNs model had an accuracy of 76.4% at the classification threshold that corresponded with maximum cost saving to the hospital.ConclusionsDeep learning techniques performed better than other traditional techniques in developing this EMR-based prediction model for 30-day readmissions in heart failure patients. Such models can be used to identify heart failure patients with impending hospitalization, enabling care teams to target interventions at their most high-risk patients and improving overall clinical outcomes.

Intravenous Acetaminophen May Be Associated with Reduced Odds of 30-Day Readmission after Total Knee Arthroplasty.

The purpose of this study was (1) to evaluate 30-day readmission rates in total knee arthroplasty (TKA) patients who either received intravenous (IV) or oral (PO) acetaminophen (APAP) perioperatively and (2) to extrapolate the potential annual cost savings on the national level. This was a review of 190,691 TKA recipients between the years 2012 and 2015 who received either IV (n = 56,475) or PO APAP (n = 134,216). All-cause readmissions that occurred between patient discharge and 30 days postdischarge were recorded. Continuous and categorical variables were evaluated using t-test and chi-square test, respectively. A logistic regression analysis was conducted to assess the effect of IV APAP on 30-day readmission. We also performed a literature review on 30-day readmission rates and risk prediction tools for TKA and correlated these with our findings. In addition, we extrapolated potential cost savings on the national level. The readmission rate was 0.04% in the IV and 0.14% in the PO APAP cohort (69% decreased risk; odds ratio = 0.31; 95% confidence interval = 0.20-0.47; p < 0.001). The readmission rate in this patient population appears to be markedly lower, when compared with previous reports. This reduction in readmissions may potentially result in $160 million savings per year. The use of IV APAP in TKA patients resulted in lower readmission rates, which may be valuable in clinical decision making by surgeons and health care administrators looking to lower costs of care.

Predicting 30-Day Hospital Readmissions in Acute Myocardial Infarction: The AMI "READMITS" (Renal Function, Elevated Brain Natriuretic Peptide, Age, Diabetes Mellitus, Nonmale Sex, Intervention with Timely Percutaneous Coronary Intervention, and Low Systolic Blood Pressure) Score.

BackgroundReadmissions after hospitalization for acute myocardial infarction (AMI) are common. However, the few currently available AMI readmission risk prediction models have poor‐to‐modest predictive ability and are not readily actionable in real time. We sought to develop an actionable and accurate AMI readmission risk prediction model to identify high‐risk patients as early as possible during hospitalization.Methods and ResultsWe used electronic health record data from consecutive AMI hospitalizations from 6 hospitals in north Texas from 2009 to 2010 to derive and validate models predicting all‐cause nonelective 30‐day readmissions, using stepwise backward selection and 5‐fold cross‐validation. Of 826 patients hospitalized with AMI, 13% had a 30‐day readmission. The first‐day AMI model (the AMI “READMITS” score) included 7 predictors: renal function, elevated brain natriuretic peptide, age, diabetes mellitus, nonmale sex, intervention with timely percutaneous coronary intervention, and low systolic blood pressure, had an optimism‐corrected C‐statistic of 0.73 (95% confidence interval, 0.71–0.74) and was well calibrated. The full‐stay AMI model, which included 3 additional predictors (use of intravenous diuretics, anemia on discharge, and discharge to postacute care), had an optimism‐corrected C‐statistic of 0.75 (95% confidence interval, 0.74–0.76) with minimally improved net reclassification and calibration. Both AMI models outperformed corresponding multicondition readmission models.ConclusionsThe parsimonious AMI READMITS score enables early prospective identification of high‐risk AMI patients for targeted readmissions reduction interventions within the first 24 hours of hospitalization. A full‐stay AMI readmission model only modestly outperformed the AMI READMITS score in terms of discrimination, but surprisingly did not meaningfully improve reclassification.

Evaluating Patient Readmission Risk: A Predictive Analytics Approach

With the emergence of the Hospital Readmission Reduction Program of the Center for Medicare and Medicaid Services on October 1, 2012, forecasting unplanned patient readmission risk became crucial to the healthcare domain. There are tangible works in the literature emphasizing on developing readmission risk prediction models; However, the models are not accurate enough to be deployed in an actual clinical setting. Our study considers patient readmission risk as the objective for optimization and develops a useful risk prediction model to address unplanned readmissions. Furthermore, Genetic Algorithm and Greedy Ensemble is used to optimize the developed model constraints.

Readmission Prediction for Patients with Heterogeneous Hazard: A Trajectory-Based Deep Learning Approach

Hospital readmission refers to the situation where a patient is re-hospitalized with the same primary diagnosis within a specific time interval after discharge. Hospital readmission causes $26 billion preventable expenses to the U.S. health systems annually and often indicates suboptimal patient care. To alleviate those severe financial and health consequences, it is crucial to proactively predict patients’ readmission risk. Such prediction is challenging because the evolution of patients’ medical history is dynamic and complex. The state-of-the-art studies apply statistical models which use static predictors in a period, failing to consider patients’ heterogeneous medical history. Our approach – Trajectory-BAsed DEep Learning (TADEL) – is motivated to tackle the deficiencies of the existing approaches by capturing dynamic medical history. We evaluate TADEL on a five-year national Medicare claims dataset reaching an F1 score of 87.3%. Our approach significantly outperforms all the state-of-the-art methods. Our findings suggest that health status factors and insurance coverage are important predictors for readmission. This study contributes to IS literature and analytical methodology by formulating the trajectory-based readmission prediction problem and developing a novel deep-learning-based readmission risk prediction framework. We also deliver implementable methods to assess patients’ readmission risk and take early interventions to avoid potential negative consequences.

Predicting readmission risk shortly after admission for CABG surgery.

Reducing preventable hospital readmissions after coronary artery bypass graft (CABG) surgery has become a national priority. Predictive models can be used to identify patients at high risk for readmission. However, the majority of the existing models are based on data available at discharge. We sought to develop a model to predict hospital readmission using data available soon after admission for isolated CABG surgery. Fifty risk factors were included in a bivariate analysis, 16 of which were significantly associated (P < 0.05) with readmissions and were entered into a multivariate logistic regression and removed stepwise, using backward elimination procedures. The derived model was then validated on 896 prospective isolated CABG cases. Of 2589 isolated CABG patients identified between December 1, 2010, and June 30, 2014, 237(9.15%) were readmitted within 30 days. Five risk factors were predictive of 30-day all-cause readmission: age (odds ratio [OR] = 1.03; 95% confidence interval [CI]: 1.01-1.05; P = 0.004), prior heart failure (OR = 1.55; 95%CI: 1.07-2.24; P = 0.020), total albumin prior to surgery (OR = 0.68; 95%CI: 0.05-0.94; P = 0.021), previous myocardial infarction (OR = 1.44; 95%CI: 1.00-2.08; P = 0.50), and history of diabetes (OR = 1.54; 95%CI: 1.09-2.19; P = 0.015). The area under the curve c-statistic was 0.63 in the derivation sample and 0.65 in the validation sample showing good discrimination. A 30-day all-cause readmission among isolated CABG patients can be predicted soon after admission with a small number of risk factors.