Detecting Miscoded Diabetes Diagnosis Codes in Electronic Health Records for Quality Improvement: Temporal Deep Learning Approach

Abstract

Background Diabetes affects more than 30 million patients across the United States. With such a large disease burden, even a small error in classification can be significant. Currently billing codes, assigned at the time of a medical encounter, are the “gold standard” reflecting the actual diseases present in an individual, and thus in aggregate reflect disease prevalence in the population. These codes are generated by highly trained coders and by health care providers but are not always accurate. Objective This work provides a scalable deep learning methodology to more accurately classify individuals with diabetes across multiple health care systems. Methods We leveraged a long short-term memory-dense neural network (LSTM-DNN) model to identify patients with or without diabetes using data from 5 acute care facilities with 187,187 patients and 275,407 encounters, incorporating data elements including laboratory test results, diagnostic/procedure codes, medications, demographic data, and admission information. Furthermore, a blinded physician panel reviewed discordant cases, providing an estimate of the total impact on the population. Results When predicting the documented diagnosis of diabetes, our model achieved an 84% F1 score, 96% area under the curve–receiver operating characteristic curve, and 91% average precision on a heterogeneous data set from 5 distinct health facilities. However, in 81% of cases where the model disagreed with the documented phenotype, a blinded physician panel agreed with the model. Taken together, this suggests that 4.3% of our studied population have either missing or improper diabetes diagnosis. Conclusions This study demonstrates that deep learning methods can improve clinical phenotyping even when patient data are noisy, sparse, and heterogeneous.

Can telehealth expansion boost health care utilization specifically for patients with substance use disorders relative to patients with other types of chronic disease?

“…Relatedly, use of ICD10 codes from claims to identify individuals with diabetes and SUDs has limitations. As a result, we may have misclassified individuals by type of diabetes or presence of an SUD [ 54 , 55 ]. Third, in using claims data, we may have missed disease-specific services if SUD or diabetes diagnoses were not recorded.…”

NER Sequence Embedding of Unified Medical Corpora to Incorporate Semantic Intelligence in Big Data Healthcare Diagnostics

“…In our study, we omitted the use of POS tagging and applied NER embedding only that minimized the code and kept it simple. Our experimental study took inspiration from the experiment run in [38] that predicted miscoded diabetes ICD-10 labels in a large EHR dataset extracted from CERNER Health Facts, a HIPPA-compliant repository maintained by the University of South California.…”

SACDNet: Towards Early Type 2 Diabetes Prediction with Uncertainty for Electronic Health Records

“…Data collected from EHRs is sparse, noisy, heterogeneous, and unstructured [56,70,55]. Such data also presents other challenges including the problem of missing data and class imbalance [71,72].…”