Dataset Information

Recurrent Neural Networks for Early Detection of Heart Failure From Longitudinal Electronic Health Record Data: Implications for Temporal Modeling With Respect to Time Before Diagnosis, Data Density, Data Quantity, and Data Type.

ABSTRACT: BACKGROUND:We determined the impact of data volume and diversity and training conditions on recurrent neural network methods compared with traditional machine learning methods. METHODS AND RESULTS:Using longitudinal electronic health record data, we assessed the relative performance of machine learning models trained to detect a future diagnosis of heart failure in primary care patients. Model performance was assessed in relation to data parameters defined by the combination of different data domains (data diversity), the number of patient records in the training data set (data quantity), the number of encounters per patient (data density), the prediction window length, and the observation window length (ie, the time period before the prediction window that is the source of features for prediction). Data on 4370 incident heart failure cases and 30?132 group-matched controls were used. Recurrent neural network model performance was superior under a variety of conditions that included (1) when data were less diverse (eg, a single data domain like medication or vital signs) given the same training size; (2) as data quantity increased; (3) as density increased; (4) as the observation window length increased; and (5) as the prediction window length decreased. When all data domains were used, the performance of recurrent neural network models increased in relation to the quantity of data used (ie, up to 100% of the data). When data are sparse (ie, fewer features or low dimension), model performance is lower, but a much smaller training set size is required to achieve optimal performance compared with conditions where data are more diverse and includes more features. CONCLUSIONS:Recurrent neural networks are effective for predicting a future diagnosis of heart failure given sufficient training set size. Model performance appears to continue to improve in direct relation to training set size.

SUBMITTER: Chen R

PROVIDER: S-EPMC6814386 | biostudies-literature | 2019 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Recurrent Neural Networks for Early Detection of Heart Failure From Longitudinal Electronic Health Record Data: Implications for Temporal Modeling With Respect to Time Before Diagnosis, Data Density, Data Quantity, and Data Type.

Chen Robert R Stewart Walter F WF Sun Jimeng J Ng Kenney K Yan Xiaowei X

Circulation. Cardiovascular quality and outcomes 20191015 10

<h4>Background</h4>We determined the impact of data volume and diversity and training conditions on recurrent neural network methods compared with traditional machine learning methods.<h4>Methods and results</h4>Using longitudinal electronic health record data, we assessed the relative performance of machine learning models trained to detect a future diagnosis of heart failure in primary care patients. Model performance was assessed in relation to data parameters defined by the combination of di ...[more]

PMID: 31610714

Similar Datasets

Project description:ImportanceInterventions to improve judicious prescribing of opioid analgesics for acute pain are needed owing to the risks of diversion, misuse, and overdose.ObjectiveTo assess the effect of modifying opioid analgesic prescribing defaults in the electronic health record (EHR) on prescribing and health service use.Design, setting, and participantsA cluster randomized clinical trial with 2 parallel arms was conducted between June 13, 2016, and June 13, 2018, in a large urban health care system comprising 32 primary care and 4 emergency department (ED) sites in the Bronx, New York. Data were analyzed using a difference-in-differences method from 6 months before implementation through 18 months after implementation. Data were analyzed from January 2019 to February 2020.InterventionsA default dispense quantity for new opioid analgesic prescriptions of 10 tablets (intervention) vs no change (control) in the EHR.Main outcomes and measuresThe primary outcome was the quantity of opioid analgesics prescribed with the new default prescription. Secondary outcomes were opioid analgesic reorders and health service use within 30 days after the new prescription. Intention-to-treat analysis was conducted.ResultsOverall, 21 331 patients received a new opioid analgesic prescription from 490 prescribers. Comparing the intervention and control arms, site, prescriber, and patient characteristics were similar. For the new prescription, compared with the control arm, patients in the intervention arm had significantly more prescriptions for 10 tablets or fewer (7.6 percentage points; 95% CI, 6.1-9.2 percentage points), a lower number of tablets prescribed (-2.1 tablets; 95% CI, -3.3 to -0.9 tablets), and lower morphine milligram equivalents (MME) prescribed (-14.6 MME; 95% CI, -22.6 to -6.6 MME). Within 30 days after the new prescription, significant differences remained in the number of tablets prescribed (-2.7 tablets; 95% CI, -4.8 to -0.6 tablets), but not MME (-15.8 MME; 95% CI, -33.8 to 2.2 MME). Within this 30-day period, there were no significant differences between the arms in health service use.Conclusions and relevanceIn this study, implementation of a uniform reduced default dispense quantity of 10 tablets for opioid analgesic prescriptions led to a modest reduction in the quantity prescribed initially, without significantly increasing health service use. However, during 30 days after implementation, the influence on prescribing was mixed. Reducing EHR default dispense quantities for opioid analgesics is a feasible strategy that can be widely disseminated and may modestly reduce prescribing.Trial registrationClinicalTrials.gov Identifier: NCT03003832.

Project description:ObjectivesRecent sepsis studies have defined patients as "infected" using a combination of culture and antibiotic orders rather than billing data. However, the accuracy of these definitions is unclear. We aimed to compare the accuracy of different established criteria for identifying infected patients using detailed chart review.DesignRetrospective observational study.SettingSix hospitals from three health systems in Illinois.PatientsAdult admissions with blood culture or antibiotic orders, or Angus International Classification of Diseases infection codes and death were eligible for study inclusion as potentially infected patients. Nine-hundred to 1,000 of these admissions were randomly selected from each health system for chart review, and a proportional number of patients who did not meet chart review eligibility criteria were also included and deemed not infected.InterventionsNone.Measurements and main resultsThe accuracy of published billing code criteria by Angus et al and electronic health record criteria by Rhee et al and Seymour et al (Sepsis-3) was determined using the manual chart review results as the gold standard. A total of 5,215 patients were included, with 2,874 encounters analyzed via chart review and a proportional 2,341 added who did not meet chart review eligibility criteria. In the study cohort, 27.5% of admissions had at least one infection. This was most similar to the percentage of admissions with blood culture orders (26.8%), Angus infection criteria (28.7%), and the Sepsis-3 criteria (30.4%). Sepsis-3 criteria was the most sensitive (81%), followed by Angus (77%) and Rhee (52%), while Rhee (97%) and Angus (90%) were more specific than the Sepsis-3 criteria (89%). Results were similar for patients with organ dysfunction during their admission.ConclusionsPublished criteria have a wide range of accuracy for identifying infected patients, with the Sepsis-3 criteria being the most sensitive and Rhee criteria being the most specific. These findings have important implications for studies investigating the burden of sepsis on a local and national level.

Dataset Information

Recurrent Neural Networks for Early Detection of Heart Failure From Longitudinal Electronic Health Record Data: Implications for Temporal Modeling With Respect to Time Before Diagnosis, Data Density, Data Quantity, and Data Type.

Publications

Recurrent Neural Networks for Early Detection of Heart Failure From Longitudinal Electronic Health Record Data: Implications for Temporal Modeling With Respect to Time Before Diagnosis, Data Density, Data Quantity, and Data Type.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets