Dataset Information

Diagnosis-specific readmission risk prediction using electronic health data: a retrospective cohort study.

ABSTRACT:

Background

Readmissions after hospital discharge are a common occurrence and are costly for both hospitals and patients. Previous attempts to create universal risk prediction models for readmission have not met with success. In this study we leveraged a comprehensive electronic health record to create readmission-risk models that were institution- and patient- specific in an attempt to improve our ability to predict readmission.

Methods

This is a retrospective cohort study performed at a large midwestern tertiary care medical center. All patients with a primary discharge diagnosis of congestive heart failure, acute myocardial infarction or pneumonia over a two-year time period were included in the analysis.The main outcome was 30-day readmission. Demographic, comorbidity, laboratory, and medication data were collected on all patients from a comprehensive information warehouse. Using multivariable analysis with stepwise removal we created three risk disease-specific risk prediction models and a combined model. These models were then validated on separate cohorts.

Results

3572 patients were included in the derivation cohort. Overall there was a 16.2% readmission rate. The acute myocardial infarction and pneumonia readmission-risk models performed well on a random sample validation cohort (AUC range 0.73 to 0.76) but less well on a historical validation cohort (AUC 0.66 for both). The congestive heart failure model performed poorly on both validation cohorts (AUC 0.63 and 0.64).

Conclusions

The readmission-risk models for acute myocardial infarction and pneumonia validated well on a contemporary cohort, but not as well on a historical cohort, suggesting that models such as these need to be continuously trained and adjusted to respond to local trends. The poor performance of the congestive heart failure model may suggest that for chronic disease conditions social and behavioral variables are of greater importance and improved documentation of these variables within the electronic health record should be encouraged.

SUBMITTER: Hebert C

PROVIDER: S-EPMC4136398 | biostudies-literature | 2014 Aug

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Diagnosis-specific readmission risk prediction using electronic health data: a retrospective cohort study.

Hebert Courtney C Shivade Chaitanya C Foraker Randi R Wasserman Jared J Roth Caryn C Mekhjian Hagop H Lemeshow Stanley S Embi Peter P

BMC medical informatics and decision making 20140804

<h4>Background</h4>Readmissions after hospital discharge are a common occurrence and are costly for both hospitals and patients. Previous attempts to create universal risk prediction models for readmission have not met with success. In this study we leveraged a comprehensive electronic health record to create readmission-risk models that were institution- and patient- specific in an attempt to improve our ability to predict readmission.<h4>Methods</h4>This is a retrospective cohort study perform ...[more]

PMID: 25091637

Similar Datasets

Project description:BackgroundIndividuals from minority ethnic groups in the UK are thought to be at higher risk of developing dementia while facing additional barriers to receiving timely care. However, few studies in the UK have examined if there are ethnic disparities in survival once individuals receive a dementia diagnosis.MethodsWe conducted a retrospective cohort study using electronic health record data of individuals diagnosed with dementia from a large secondary mental healthcare provider in London. Patients from Black African, Black Caribbean, South Asian, White British, and White Irish ethnic backgrounds were followed up for a 10-year period between 01 January 2008 and 31 December 2017. Data were linked to death certificate data from the Office of National Statistics to determine survival from dementia diagnosis. Standardised mortality ratios were calculated to estimate excess deaths in each ethnicity group as compared to the gender- and age-standardised population of England and Wales. We used Cox regression models to compare survival after dementia diagnosis across each ethnicity group.ResultsMortality was elevated at least twofold across all ethnicity groups with dementia compared to the general population in England and Wales. Risk of death was lower in Black Caribbean, Black African, White Irish, and South Asian groups as compared to the White British population, even after adjusting for age, gender, neighbourhood-level deprivation, indicators of mental and physical comorbidities. Risk of death remained lower after additionally accounting for those who emigrated out of the cohort.ConclusionsWhile mortality in dementia is elevated across all ethnic groups as compared to the general population, reasons for longer survival in minority ethnic groups in the UK as compared to the White British group are unclear and merit further exploration. Implications of longer survival, including carer burden and costs, should be considered in policy and planning to ensure adequate support for families and carers of individuals with dementia.

Project description:BackgroundSjögren's disease (SD) is an autoimmune disease that is difficult to diagnose early due to its wide spectrum of clinical symptoms and overlap with other autoimmune diseases. SD potentially presents through early oral manifestations prior to showing symptoms of clinically significant dry eyes or dry mouth. We examined the feasibility of utilizing a linked electronic dental record (EDR) and electronic health record (EHR) dataset to identify factors that could be used to improve early diagnosis prediction of SD in a matched case-control study population.MethodsEHR data, including demographics, medical diagnoses, medication history, serological test history, and clinical notes, were retrieved from the Indiana Network for Patient Care database and dental procedure data were retrieved from the Indiana University School of Dentistry EDR. We examined EHR and EDR history in the three years prior to SD diagnosis for SD cases and the corresponding period in matched non-SD controls. Two conditional logistic regression (CLR) models were built using Least Absolute Shrinkage and Selection Operator regression. One used only EHR data and the other used both EHR and EDR data. The ability of these models to predict SD diagnosis was assessed using a concordance index designed for CLR.ResultsWe identified a sample population of 129 cases and 371 controls with linked EDR-EHR data. EHR factors associated with an increased risk of SD diagnosis were the usage of lubricating throat drugs with an odds ratio (OR) of 14.97 (2.70-83.06), dry mouth (OR = 6.19, 2.14-17.89), pain in joints (OR = 2.54, 1.34-4.76), tear film insufficiency (OR = 27.04, 5.37-136.), and rheumatoid factor testing (OR = 6.97, 1.94-25.12). The addition of EDR data slightly improved model concordance compared to the EHR only model (0.834 versus 0.811). Surgical dental procedures (OR = 2.33, 1.14-4.78) were found to be associated with an increased risk of SD diagnosis while dental diagnostic procedures (OR = 0.45, 0.20-1.01) were associated with decreased risk.ConclusionUtilizing EDR data alongside EHR data has the potential to improve prediction models for SD. This could improve the early diagnosis of SD, which is beneficial to slowing or preventing complications of SD.

Project description:BackgroundHip fracture is associated with immobility, morbidity, mortality, and high medical cost. Due to limited availability of dual-energy X-ray absorptiometry (DXA), hip fracture prediction models without using bone mineral density (BMD) data are essential. We aimed to develop and validate 10-year sex-specific hip fracture prediction models using electronic health records (EHR) without BMD.MethodsIn this retrospective, population-based cohort study, anonymized medical records were retrieved from the Clinical Data Analysis and Reporting System for public healthcare service users in Hong Kong aged ≥60 years as of 31 December 2005. A total of 161,051 individuals (91,926 female; 69,125 male) with complete follow-up from 1 January 2006 till the study end date on 31 December 2015 were included in the derivation cohort. The sex-stratified derivation cohort was randomly divided into 80% training and 20% internal testing datasets. An independent validation cohort comprised 3046 community-dwelling participants aged ≥60 years as of 31 December 2005 from the Hong Kong Osteoporosis Study, a prospective cohort which recruited participants between 1995 and 2010. With 395 potential predictors (age, diagnosis, and drug prescription records from EHR), 10-year sex-specific hip fracture prediction models were developed using stepwise selection by logistic regression (LR) and four machine learning (ML) algorithms (gradient boosting machine, random forest, eXtreme gradient boosting, and single-layer neural networks) in the training cohort. Model performance was evaluated in both internal and independent validation cohorts.FindingsIn female, the LR model had the highest AUC (0.815; 95% Confidence Interval [CI]: 0.805-0.825) and adequate calibration in internal validation. Reclassification metrics showed the LR model had better discrimination and classification performance than the ML algorithms. Similar performance was attained by the LR model in independent validation, with high AUC (0.841; 95% CI: 0.807-0.87) comparable to other ML algorithms. In internal validation for male, LR model had high AUC (0.818; 95% CI: 0.801-0.834) and it outperformed all ML models as indicated by reclassification metrics, with adequate calibration. In independent validation, the LR model had high AUC (0.898; 95% CI: 0.857-0.939) comparable to ML algorithms. Reclassification metrics demonstrated that LR model had the best discrimination performance.InterpretationEven without using BMD data, the 10-year hip fracture prediction models developed by conventional LR had better discrimination performance than the models developed by ML algorithms. Upon further validation in independent cohorts, the LR models could be integrated into the routine clinical workflow, aiding the identification of people at high risk for DXA scan.FundingHealth and Medical Research Fund, Health Bureau, Hong Kong SAR Government (reference: 17181381).

Dataset Information

Diagnosis-specific readmission risk prediction using electronic health data: a retrospective cohort study.

Background

Methods

Results

Conclusions

Publications

Diagnosis-specific readmission risk prediction using electronic health data: a retrospective cohort study.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets