Electronic medical record phenotyping using the anchor and learn framework

Abstract

Background Electronic medical records (EMRs) hold a tremendous amount of information about patients that is relevant to determining the optimal approach to patient care. As medicine becomes increasingly precise, a patient’s electronic medical record phenotype will play an important role in triggering clinical decision support systems that can deliver personalized recommendations in real time. Learning with anchors presents a method of efficiently learning statistically driven phenotypes with minimal manual intervention.Materials and Methods We developed a phenotype library that uses both structured and unstructured data from the EMR to represent patients for real-time clinical decision support. Eight of the phenotypes were evaluated using retrospective EMR data on emergency department patients using a set of prospectively gathered gold standard labels.Results We built a phenotype library with 42 publicly available phenotype definitions. Using information from triage time, the phenotype classifiers have an area under the ROC curve (AUC) of infection 0.89, cancer 0.88, immunosuppressed 0.85, septic shock 0.93, nursing home 0.87, anticoagulated 0.83, cardiac etiology 0.89, and pneumonia 0.90. Using information available at the time of disposition from the emergency department, the AUC values are infection 0.91, cancer 0.95, immunosuppressed 0.90, septic shock 0.97, nursing home 0.91, anticoagulated 0.94, cardiac etiology 0.92, and pneumonia 0.97.Discussion The resulting phenotypes are interpretable and fast to build, and perform comparably to statistically learned phenotypes developed with 5000 manually labeled patients.Conclusion Learning with anchors is an attractive option for building a large public repository of phenotype definitions that can be used for a range of health IT applications, including real-time decision support.

EHR-based phenotyping: Bulk learning and evaluation

“…Methodologically speaking, however, learning phenotypes from noisy labels has two ongoing research directions that are interrelated, i.e. modeling phenotypes from anchor variables [44,45] and silver-standard training data [46]. We also note that from the perspective of learning shared representations of diseases (such as the abstraction feature representation in this study), contemporary phenotyping effort has led to a growing body of work that learns phenotypes from population-scale clinical data using the methodology of representation learning [47,48] including i) spectral learning such as non-negative tensor factorization [5], ii) probabilistic mixture models [8], and additionally, when temporal phenotypic patterns are considered, iii) unsupervised feature learning using autoencoders [32] and latent medical concepts [49], etc., and iv) deep learning [6].…”

Using anchors from free text in electronic health records to diagnose postoperative delirium

“…However, in practice, the conditional independence property does not have to be completely satisfied [12]. On the other hand, if property 1 is relaxed, the false positive rate will automatically increase.…”

“…To overcome this drawback Halpern et al proposed a very promising framework, with a large number of possible applications. In this framework, which we refer to as the anchor method (AM), one can learn phenotypes and predict clinical state variables from EHR unlabeled data only by specifying a few key observations called anchors [11,12]. An underlying assumption is that the presence of an anchor variable implies the presence of the latent label of interest.…”

Clinical Research Informatics: Contributions from 2016

“…A content summary of these selected papers can be found in the appendix of this synopsis. [2], the authors demonstrated the feasibility of utilizing semi-automatically labeled training sets to create phenotype models via machine learning, using a comprehensive representation of the patient medical record [3]. They validated the phenotype models in the context of Type 2 diabetes mellitus (T2DM) and Myocardial Infarcts (MI) using respectively the phenotype definitions of the eMERGE [1] and OMOP [4] initiatives.…”

“…Using the Halpern et al method based on "anchor" terms [2], they defined a list of keywords specific to the phenotypes of interest to semi-automatically generate noisy labeled training data. Then, a sample of 1,500 patient records -750 patient records for each phenotype having a "noisy" label for the phenotype and 750 controls taken in the extract disjoint with possible cases (silver standard) -was used to train the XPRESS (eXtraction of Phenotypes from Records using Silver Standards) model.…”