Electronic healthcare records and external outcome data for hospitalized patients with heart failure.
Ontology highlight
ABSTRACT:
SUBMITTER: Zhang Z
PROVIDER: S-EPMC7865067 | biostudies-literature | 2021 Feb
REPOSITORIES: biostudies-literature
Publications
Electronic healthcare records and external outcome data for hospitalized patients with heart failure.
Zhang Zhongheng Z Cao Linghong L Chen Rangui R Zhao Yan Y Lv Lukai L Xu Ziyin Z Xu Ping P
Scientific data 20210205 1
Heart failure is one of the most important reasons for hospitalization among elderly individuals and is associated with significant mortality and morbidity. Epidemiological studies require the establishment of high-quality databases. Several datasets that primarily involve heart failure populations have been established in Western countries and have generated many high-quality studies. However, no such dataset is available from China. Due to differences in genetic background and healthcare syste ...[more]
Similar Datasets
Project description:BackgroundWith the prevalence of cardiovascular diseases increasing worldwide, early prediction and accurate assessment of heart failure (HF) risk are crucial to meet the clinical demand.ObjectiveOur study objective was to develop machine learning (ML) models based on real-world electronic health records to predict 1-year in-hospital mortality, use of positive inotropic agents, and 1-year all-cause readmission rate.MethodsFor this single-center study, we recruited patients with newly diagnosed HF hospitalized between December 2010 and August 2018 at the First Affiliated Hospital of Dalian Medical University (Liaoning Province, China). The models were constructed for a population set (90:10 split of data set into training and test sets) using 79 variables during the first hospitalization. Logistic regression, support vector machine, artificial neural network, random forest, and extreme gradient boosting models were investigated for outcome predictions.ResultsOf the 13,602 patients with HF enrolled in the study, 537 (3.95%) died within 1 year and 2779 patients (20.43%) had a history of use of positive inotropic agents. ML algorithms improved the performance of predictive models for 1-year in-hospital mortality (areas under the curve [AUCs] 0.92-1.00), use of positive inotropic medication (AUCs 0.85-0.96), and 1-year readmission rates (AUCs 0.63-0.96). A decision tree of mortality risk was created and stratified by single variables at levels of high-sensitivity cardiac troponin I (<0.068 μg/L), followed by percentage of lymphocytes (<14.688%) and neutrophil count (4.870×109/L).ConclusionsML techniques based on a large scale of clinical variables can improve outcome predictions for patients with HF. The mortality decision tree may contribute to guiding better clinical risk assessment and decision making.
| S-EPMC8094022 | biostudies-literature
Project description:AimsThe objective of this study was to estimate the nationwide annual incidence, healthcare utilization, and mortality among hospitalized heart failure (HF) patients in Taiwan.Methods and resultsPeople aged 20 years or older and having been newly admitted for HF between 2010 and 2015 were identified from Taiwan's National Health Insurance Research Database. For 124 816 patients with incident HF hospitalizations between 2010 and 2012, we further analysed their treatment patterns, healthcare utilizations, and mortality during index hospitalization and within 3 years following discharge from the index hospitalization. The age-stratified incidences were declined by 10-20% in people aged 55 years or older, but increased by ~4% among people younger than 44 years old between 2010 and 2015. For all incident hospitalized HF patients, the percentages of patients visiting the emergency room, were rehospitalized, and treated with guideline-directed medical therapy were highest in the first year. Approximately two-thirds of subsequent hospitalizations were due to non-HF and non-cardiovascular causes. The all-cause mortality rate during index hospitalization was 8.5%, whereas the mortality rates at 30 days, 90 days, 180 days, 1 year, 2 years, and 3 years following discharge were 3.5%, 8.9%, 14.4%, 22.5%, 33.9%, and 42.8%, respectively, for those surviving index HF hospitalization. Non-cardiovascular disease-related deaths accounted for nearly 60% of all deaths.ConclusionsOur study reveals that, in contemporary Taiwan, the >10% annual mortality following the first year of hospitalization, 30% deaths occurring outside the hospital, and 60% non-cardiovascular-related deaths, along with the decreasing use of guideline-directed medical therapy, highlight sectors requiring more attention.
| S-EPMC7755001 | biostudies-literature
Project description:BackgroundLinkage of electronic healthcare records is becoming increasingly important for research purposes. However, linkage error due to mis-recorded or missing identifiers can lead to biased results. We evaluated the impact of linkage error on estimated infection rates using two different methods for classifying links: highest-weight (HW) classification using probabilistic match weights and prior-informed imputation (PII) using match probabilities.MethodsA gold-standard dataset was created through deterministic linkage of unique identifiers in admission data from two hospitals and infection data recorded at the hospital laboratories (original data). Unique identifiers were then removed and data were re-linked by date of birth, sex and Soundex using two classification methods: i) HW classification - accepting the candidate record with the highest weight exceeding a threshold and ii) PII-imputing values from a match probability distribution. To evaluate methods for linking data with different error rates, non-random error and different match rates, we generated simulation data. Each set of simulated files was linked using both classification methods. Infection rates in the linked data were compared with those in the gold-standard data.ResultsIn the original gold-standard data, 1496/20924 admissions linked to an infection. In the linked original data, PII provided least biased results: 1481 and 1457 infections (upper/lower thresholds) compared with 1316 and 1287 (HW upper/lower thresholds). In the simulated data, substantial bias (up to 112%) was introduced when linkage error varied by hospital. Bias was also greater when the match rate was low or the identifier error rate was high and in these cases, PII performed better than HW classification at reducing bias due to false-matches.ConclusionsThis study highlights the importance of evaluating the potential impact of linkage error on results. PII can help incorporate linkage uncertainty into analysis and reduce bias due to linkage error, without requiring identifiers.
| S-EPMC4015706 | biostudies-literature
Project description:ObjectivesTo characterise the comorbidities of heart failure (HF) in men and women, to explore their clustering into multimorbidity patterns, and to measure the impact of such patterns on the risk of hospitalisation and mortality.DesignObservational retrospective population study based on electronic health records.SettingEpiChron Cohort (Aragón, Spain).ParticipantsAll the primary and hospital care patients of the EpiChron Cohort with a diagnosis of HF on 1 January 2011 (ie, 8488 women and 6182 men). We analysed all the chronic diseases registered in patients' electronic health records until 31 December 2011.Primary outcomeWe performed an exploratory factor analysis to identify the multimorbidity patterns in men and women, and logistic and Cox proportional-hazards regressions to investigate the association between the patterns and the risk of hospitalisation in 2012, and of 3-year mortality.ResultsAlmost all HF patients (98%) had multimorbidity, with an average of 7.8 chronic diseases per patient. We identified six different multimorbidity patterns, named cardiovascular, neurovascular, coronary, metabolic, degenerative and respiratory. The most prevalent were the degenerative (64.0%) and cardiovascular (29.9%) patterns in women, and the metabolic (49.3%) and cardiovascular (43.2%) patterns in men. Every pattern was associated with higher hospitalisation risks; and the cardiovascular, neurovascular and respiratory patterns significantly increased the likelihood of 3-year mortality.ConclusionsMultimorbidity is the norm rather than the exception in patients with heart failure, whose comorbidities tend to cluster together beyond simple chance in the form of multimorbidity patterns that have different impact on health outcomes. This knowledge could be useful to better understand common pathophysiological pathways underlying this condition and its comorbidities, and the factors influencing the prognosis of men and women with HF. Further large scale longitudinal studies are encouraged to confirm the existence of these patterns as well as their differential impact on health outcomes.
| S-EPMC7008407 | biostudies-literature
Project description:BackgroundUsing electronic health records data to predict events and onset of diseases is increasingly common. Relatively little is known, although, about the tradeoffs between data requirements and model utility.Methods and resultsWe examined the performance of machine learning models trained to detect prediagnostic heart failure in primary care patients using longitudinal electronic health records data. Model performance was assessed in relation to data requirements defined by the prediction window length (time before clinical diagnosis), the observation window length (duration of observation before prediction window), the number of different data domains (data diversity), the number of patient records in the training data set (data quantity), and the density of patient encounters (data density). A total of 1684 incident heart failure cases and 13 525 sex, age-category, and clinic matched controls were used for modeling. Model performance improved as (1) the prediction window length decreases, especially when <2 years; (2) the observation window length increases but then levels off after 2 years; (3) the training data set size increases but then levels off after 4000 patients; (4) more diverse data types are used, but, in order, the combination of diagnosis, medication order, and hospitalization data was most important; and (5) data were confined to patients who had ≥10 phone or face-to-face encounters in 2 years.ConclusionsThese empirical findings suggest possible guidelines for the minimum amount and type of data needed to train effective disease onset predictive models using longitudinal electronic health records data.
| S-EPMC5341145 | biostudies-literature
Project description:BACKGROUND:Heart 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. METHODS:We 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. RESULTS:Data 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. CONCLUSIONS:Deep 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.
| S-EPMC6013959 | biostudies-literature
Project description:BackgroundOlder patients who are hospitalized for acute decompensated heart failure have high rates of physical frailty, poor quality of life, delayed recovery, and frequent rehospitalizations. Interventions to address physical frailty in this population are not well established.MethodsWe conducted a multicenter, randomized, controlled trial to evaluate a transitional, tailored, progressive rehabilitation intervention that included four physical-function domains (strength, balance, mobility, and endurance). The intervention was initiated during, or early after, hospitalization for heart failure and was continued after discharge for 36 outpatient sessions. The primary outcome was the score on the Short Physical Performance Battery (total scores range from 0 to 12, with lower scores indicating more severe physical dysfunction) at 3 months. The secondary outcome was the 6-month rate of rehospitalization for any cause.ResultsA total of 349 patients underwent randomization; 175 were assigned to the rehabilitation intervention and 174 to usual care (control). At baseline, patients in each group had markedly impaired physical function, and 97% were frail or prefrail; the mean number of coexisting conditions was five in each group. Patient retention in the intervention group was 82%, and adherence to the intervention sessions was 67%. After adjustment for baseline Short Physical Performance Battery score and other baseline characteristics, the least-squares mean (±SE) score on the Short Physical Performance Battery at 3 months was 8.3±0.2 in the intervention group and 6.9±0.2 in the control group (mean between-group difference, 1.5; 95% confidence interval [CI], 0.9 to 2.0; P<0.001). At 6 months, the rates of rehospitalization for any cause were 1.18 in the intervention group and 1.28 in the control group (rate ratio, 0.93; 95% CI, 0.66 to 1.19). There were 21 deaths (15 from cardiovascular causes) in the intervention group and 16 deaths (8 from cardiovascular causes) in the control group. The rates of death from any cause were 0.13 and 0.10, respectively (rate ratio, 1.17; 95% CI, 0.61 to 2.27).ConclusionsIn a diverse population of older patients who were hospitalized for acute decompensated heart failure, an early, transitional, tailored, progressive rehabilitation intervention that included multiple physical-function domains resulted in greater improvement in physical function than usual care. (Funded by the National Institutes of Health and others; REHAB-HF ClinicalTrials.gov number, NCT02196038.).
| S-EPMC8353658 | biostudies-literature
Project description:Hospitalized patients with heart failure (HF) undergoing bedside glucose monitoring with subcutaneous insulin orders were retrospectively identified over 2 years. Hypoglycaemia was defined as any glucose value <3.9 mmol/L (70 mg/dL) within 24 hours of admission (Hypo1day ) or throughout the hospitalization (HypoT ) or any glucose value <2.2 mmol/L (40 mg/dL) throughout the hospitalization (HypoSevere ). A total of 13 424 patients were included, of whom 2484 had HF. Patients with HF were more likely to have Hypo1day (9.1% vs 7.0%, P = .0003), HypoT (28% vs 18.5%, P < .0001), or Hypo Severe (3.4% vs 2.1%, P = .0001). After controlling for other variables, the odds of Hypo1day were similar between the HF and non-HF groups (odds ratio [OR] 1.14, 95% CI 0.94-1.39, P = .18, fully adjusted model), slightly lower for HypoT (OR 0.85, 95% CI 0.73-0.99, P = .03, fully adjusted model), and similar for HypoSevere (OR 1.25, 95% CI 0.91-1.70, P = .17). Hypo1day , HypoT and HypoSevere were all associated with increased mortality; there was no evidence of an interaction by HF status. Hypoglycaemia occurred at a similar or lower frequency in hospitalized patients with HF compared to those without HF. Hypoglycaemia was associated with increased hospital mortality, regardless of HF status.
| S-EPMC6625647 | biostudies-literature
Project description:Background Our aim was to calibrate and externally revalidate the ELAN-HF (European Collaboration on Acute Decompensated Heart Failure) score, to confirm and improve on a previous external validation of the risk score. Methods and Results The ELAN-HF score predicts 6-month all-cause mortality in patients hospitalized for acute decompensated heart failure using absolute and percentage change of NT-proBNP (N-terminal pro-B-type natriuretic peptide) levels in addition to clinical variables. For the external validation, we used the PRIMA II (Can NT-proBNP-Guided Therapy During Hospital Admission for Acute Decompensated Heart Failure Reduce Mortality and Readmissions?) trial. For both data sets, observed versus predicted mortality was compared for the 4 risk categories; and the mean predicted mortality was plotted against the observed mortality with calculation of a correlation coefficient and SEE. The model discriminant ability was determined by comparing the C-statistics for both data sets. The predicted versus actual 6-month mortality values in the derivation cohort were 3.7% versus 3.6% for the low-risk category, 9.4% versus 9.2% for the intermediate-risk category, 24.2% versus 23.5% for the high-risk category, and 54.2% versus 51.1% for the very-high-risk category. The correlation between predicted and observed mortality by deciles was 0.92, with an SEE of ±4%. In the validation cohort, predicted versus actual 6-month mortality values were 3.0% versus 2.2% for the low-risk category, 9.4% versus 8.2% for the intermediate-risk category, 25.0% versus 22.9% for the high-risk category, and 56.8% versus 53.6% for the very-high-risk category. The correlation between predicted and actual mortality by quintiles was 0.99, with an SEE of ±2%. There was no significant difference in C-statistic between the derivation cohort (0.78; 95% CI, 0.74-0.82) and the validation cohort (0.77; 95% CI, 0.69-0.84; P=0.693). Conclusions Our study confirms that the ELAN-HF score predicts accurately 6-month mortality in patients hospitalized for acute decompensated heart failure with the use of easily obtained characteristics.
| S-EPMC6662127 | biostudies-literature
Project description:BACKGROUND:In order to examine the potential clinical value of integrating family history information directly from the electronic health records of patients' family members, the electronic health records of individuals in Clalit Health Services, the largest payer/provider in Israel, were linked with the records of their parents. METHODS:We describe the results of a novel approach for creating data-derived family history information for 2 599 575 individuals, focusing on three chronic diseases: asthma, cardiovascular disease (CVD) and diabetes. RESULTS:In our cohort, there were 256 598 patients with asthma, 55 309 patients with CVD and 66 324 patients with diabetes. Of the people with asthma, CVD or diabetes, the percentage that also had a family history of the same disease was 22.0%, 70.8% and 70.5%, respectively. CONCLUSIONS:Linking individuals' health records with their data-derived family history has untapped potential for supporting diagnostic and clinical decision-making.
| S-EPMC7183366 | biostudies-literature