Dataset Information

Assessing electronic health record phenotypes against gold-standard diagnostic criteria for diabetes mellitus.

ABSTRACT: Objective:We assessed the sensitivity and specificity of 8 electronic health record (EHR)-based phenotypes for diabetes mellitus against gold-standard American Diabetes Association (ADA) diagnostic criteria via chart review by clinical experts. Materials and Methods:We identified EHR-based diabetes phenotype definitions that were developed for various purposes by a variety of users, including academic medical centers, Medicare, the New York City Health Department, and pharmacy benefit managers. We applied these definitions to a sample of 173?503 patients with records in the Duke Health System Enterprise Data Warehouse and at least 1 visit over a 5-year period (2007-2011). Of these patients, 22?679 (13%) met the criteria of 1 or more of the selected diabetes phenotype definitions. A statistically balanced sample of these patients was selected for chart review by clinical experts to determine the presence or absence of type 2 diabetes in the sample. Results:The sensitivity (62-94%) and specificity (95-99%) of EHR-based type 2 diabetes phenotypes (compared with the gold standard ADA criteria via chart review) varied depending on the component criteria and timing of observations and measurements. Discussion and Conclusions:Researchers using EHR-based phenotype definitions should clearly specify the characteristics that comprise the definition, variations of ADA criteria, and how different phenotype definitions and components impact the patient populations retrieved and the intended application. Careful attention to phenotype definitions is critical if the promise of leveraging EHR data to improve individual and population health is to be fulfilled.

SUBMITTER: Spratt SE

PROVIDER: S-EPMC6080723 | biostudies-literature | 2017 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Assessing electronic health record phenotypes against gold-standard diagnostic criteria for diabetes mellitus.

Spratt Susan E SE Pereira Katherine K Granger Bradi B BB Batch Bryan C BC Phelan Matthew M Pencina Michael M Miranda Marie Lynn ML Boulware Ebony E Lucas Joseph E JE Nelson Charlotte L CL Neely Benjamin B Goldstein Benjamin A BA Barth Pamela P Richesson Rachel L RL Riley Isaretta L IL Corsino Leonor L McPeek Hinz Eugenia R ER Rusincovitch Shelley S Green Jennifer J Barton Anna Beth AB Kelley Carly C Hyland Kristen K Tang Monica M Elliott Amanda A Ruel Ewa E Clark Alexander A Mabrey Melanie M Morrissey Kay Lyn KL Rao Jyothi J Hong Beatrice B Pierre-Louis Marjorie M Kelly Katherine K Jelesoff Nicole N

Journal of the American Medical Informatics Association : JAMIA 20170401 e1

<h4>Objective</h4>We assessed the sensitivity and specificity of 8 electronic health record (EHR)-based phenotypes for diabetes mellitus against gold-standard American Diabetes Association (ADA) diagnostic criteria via chart review by clinical experts.<h4>Materials and methods</h4>We identified EHR-based diabetes phenotype definitions that were developed for various purposes by a variety of users, including academic medical centers, Medicare, the New York City Health Department, and pharmacy ben ...[more]

PMID: 27616701

Similar Datasets

Project description:Although standardized terminologies such as the International Classification of Diseases have been in use in medicine for over a century, efforts in the dental profession to standardize dental diagnostic terms have not achieved widespread acceptance. To address this gap, a standardized dental diagnostic terminology, the EZCodes, was developed in 2009. Fifteen dental education institutions in the United States and Europe have implemented the EZCodes dental diagnostic terminology. This article reports on the utilization and valid entry of the EZCodes at three of the dental schools that have adopted this standardized dental diagnostic terminology. Electronic data on the use of procedure codes with diagnostic terms from the three schools over a period from July 2010 to June 2011 were aggregated. The diagnostic term and procedure code pairs were adjudicated by three calibrated dentists. Analyses were conducted to gain insight into the utilization and valid entry of the EZCodes diagnostic terminology in the one-year period. Error proportions in the entry of diagnostic term (and by diagnostic category) were also computed. In the twelve-month period, 29,965 diagnostic terms and 249,411 procedure codes were entered at the three institutions resulting in a utilization proportion of 12 percent. Caries and periodontics were the most frequently used categories. More than 1,000 of the available 1,321 diagnostic terms were never used. Overall, 60.5 percent of the EZCodes entries were found to be valid. The results demonstrate low utilization of EZCodes in an electronic health record and raise the need for specific training of dental providers on the importance of using dental diagnostic terminology and specifically how to use the terms in the electronic record. These findings will serve to increase the use/correct use of the EZCodes dental diagnostic terminology and ultimately create a reliable platform for undertaking clinical, outcomes, and quality improvement-related research.

Project description:AIMS/HYPOTHESIS:Detection and management of gestational diabetes mellitus (GDM) are crucial to reduce the risk of pregnancy-related complications for both mother and child. In 2013, the WHO adopted new diagnostic criteria for GDM to improve pregnancy outcomes. However, the evidence supporting these criteria is limited. Consequently, these new criteria have not yet been endorsed in the Netherlands. The aim of this study was to determine the impact of these criteria on the number of GDM diagnoses and pregnancy outcomes. METHODS:Data were available from 10,642 women who underwent a 75 g OGTT because of risk factors or signs suggestive of GDM. Women were treated if diagnosed with GDM according to the WHO 1999 criteria. Data on pregnancy outcomes were obtained from extensive chart reviews from 4,431 women and were compared between women with normal glucose tolerance (NGT) and women classified into the following groups: (1) GDM according to WHO 1999 criteria; (2) GDM according to WHO 2013 criteria; (3) GDM according to WHO 2013 fasting glucose threshold, but not WHO 1999 criteria; and (4) GDM according to WHO 1999 2 h plasma glucose threshold (2HG), but not WHO 2013 criteria. RESULTS:Applying the new WHO 2013 criteria would have increased the number of diagnoses by 45% (32% vs 22%) in this population of women at higher risk for GDM. In comparison with women with NGT, women classified as having GDM based only on the WHO 2013 threshold for fasting glucose, who were not treated for GDM, were more likely to have been obese (46.1% vs 28.1%, p?<?0.001) and hypertensive (3.3% vs 1.2%, p?<?0.001) before pregnancy, and to have had higher rates of gestational hypertension (7.8% vs 4.9%, p?=?0.003), planned Caesarean section (10.3% vs 6.5%, p?=?0.001) and induction of labour (34.8% vs 28.0%, p?=?0.001). In addition, their neonates were more likely to have had an Apgar score <7 at 5 min (4.4% vs 2.6%, p?=?0.015) and to have been admitted to the Neonatology Department (15.0% vs 11.1%, p?=?0.004). The number of large for gestational age (LGA) neonates was not significantly different between the two groups. Women potentially missed owing to the higher 2HG threshold set by WHO 2013 had similar pregnancy outcomes to women with NGT. These women were all treated for GDM with diet and 20.5% received additional insulin. CONCLUSIONS/INTERPRETATION:Applying the WHO 2013 criteria will have a major impact on the number of GDM diagnoses. Using the fasting glucose threshold set by WHO 2013 identifies a group of women with an increased risk of adverse outcomes compared with women with NGT. We therefore support the use of a lower fasting glucose threshold in the Dutch national guideline for GDM diagnosis. However, adopting the WHO 2013 criteria with a higher 2HG threshold would exclude women in whom treatment for GDM seems to be effective.

Project description:BackgroundIdentifying cases of disease in primary care electronic medical records (EMRs) is important for surveillance, research, quality improvement and clinical care. We aimed to develop and validate a case definition for type 1 diabetes mellitus using EMRs.MethodsFor this exploratory study, we used EMR data from the Southern Alberta Primary Care Network within the Canadian Primary Care Sentinel Surveillance Network (CPCSSN), for the period 2008 to 2016. For patients identified as having diabetes mellitus according to the existing CPCSSN case definition, we asked family physicians to confirm the diabetes subtype, to create the reference standard. We used 3 decision-tree classification algorithms and least absolute shrinkage and selection operator logistic regression to identify variables that correctly distinguished between type 1 and type 2 diabetes cases.ResultsWe identified a total of 1309 people with type 1 or type 2 diabetes, 110 of whom were confirmed by their physicians as having type 1 diabetes. Two machine learning algorithms were useful in identifying these cases in the EMRs. The first algorithm used "type 1" text words or age less than 22 years at time of initial diabetes diagnosis; this algorithm had sensitivity 42.7% (95% confidence interval [CI] 33.5%-52.5%), specificity 99.3% (95% CI 98.6%-99.7%), positive predictive value 85.5% (95% CI 72.8%-93.1%) and negative predictive value 94.9% (95% CI 93.5%-96.1%). The second algorithm used a combination of free-text terms, insulin prescriptions and age; it had sensitivity 87.3% (95% CI 79.2%-92.6%), specificity 85.4% (95% CI 83.2%-87.3%), positive predictive value 35.6% (95% CI 29.9%-41.6%) and negative predictive value 98.6% (95% CI 97.7%-99.2%).InterpretationWe used machine learning to develop and validate 2 case definitions that achieve different goals in distinguishing between type 1 and type 2 diabetes in CPCSSN data. Further validation and testing with a larger and more diverse sample are recommended.

Project description:BACKGROUND:Stroke severity is an important predictor of patient outcomes and is commonly measured with the National Institutes of Health Stroke Scale (NIHSS) scores. Because these scores are often recorded as free text in physician reports, structured real-world evidence databases seldom include the severity. The aim of this study was to use machine learning models to impute NIHSS scores for all patients with newly diagnosed stroke from multi-institution electronic health record (EHR) data. METHODS:NIHSS scores available in the Optum© de-identified Integrated Claims-Clinical dataset were extracted from physician notes by applying natural language processing (NLP) methods. The cohort analyzed in the study consists of the 7149 patients with an inpatient or emergency room diagnosis of ischemic stroke, hemorrhagic stroke, or transient ischemic attack and a corresponding NLP-extracted NIHSS score. A subset of these patients (n?=?1033, 14%) were held out for independent validation of model performance and the remaining patients (n?=?6116, 86%) were used for training the model. Several machine learning models were evaluated, and parameters optimized using cross-validation on the training set. The model with optimal performance, a random forest model, was ultimately evaluated on the holdout set. RESULTS:Leveraging machine learning we identified the main factors in electronic health record data for assessing stroke severity, including death within the same month as stroke occurrence, length of hospital stay following stroke occurrence, aphagia/dysphagia diagnosis, hemiplegia diagnosis, and whether a patient was discharged to home or self-care. Comparing the imputed NIHSS scores to the NLP-extracted NIHSS scores on the holdout data set yielded an R2 (coefficient of determination) of 0.57, an R (Pearson correlation coefficient) of 0.76, and a root-mean-squared error of 4.5. CONCLUSIONS:Machine learning models built on EHR data can be used to determine proxies for stroke severity. This enables severity to be incorporated in studies of stroke patient outcomes using administrative and EHR databases.

Dataset Information

Assessing electronic health record phenotypes against gold-standard diagnostic criteria for diabetes mellitus.

Publications

Assessing electronic health record phenotypes against gold-standard diagnostic criteria for diabetes mellitus.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets