High-throughput multimodal automated phenotyping (MAP) with application to PheWAS

Liao, Katherine P.; Sun, Jiehuan; Cai, Tianrun; Link, Nicholas; Hong, Chuan; Huang, Jie; Huffman, Jennifer E.; Gronsbell, Jessica; Zhang, Yichi; Ho, Yuk Lam; Castro, Víctor M.; Gainer, Vivian S.; Murphy, Shawn N.; O’Donnell, Christopher J.; Gaziano, J. Michael; Cho, Kelly; Szolovits, Peter; Kohane, Isaac S.; Yu, Sheng

doi:10.1093/jamia/ocz066

Cited by 71 publications

(37 citation statements)

References 35 publications

(41 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another limitation was the use of ICD codes to assign diagnoses in the studied VA cystic kidney and liver cohort. Although ICD codes are commonly used for disease phenotyping based on electronic medical records, their accuracy is disease-dependent and often inferior to machine learning-based complex methods 15 . While it would be ideal to have diagnoses among these patients confirmed by genotyping, imaging, and family history data, such a depth of information will unlikely be fully retrievable from general medical records.…”

Section: Discussionmentioning

confidence: 99%

Autosomal Dominant Polycystic Kidney Disease does not significantly alter major COVID-19 outcomes among veterans

Cui

Gallini

Jasien

et al. 2020

Preprint

View full text Add to dashboard Cite

Chronic kidney disease (CKD), as well as its common causes (e.g., diabetes and obesity), are recognized risk factors for severe COVID-19 illness. To explore whether the most common inherited cause of CKD, autosomal dominant polycystic kidney disease (ADPKD), is also an independent risk factor, we studied data from the VA health system and the VA COVID-19-shared resources (e.g., ICD codes, demographics, pre-existing conditions, pre-testing symptoms, and post-testing outcomes). Among 61 COVID-19-positive ADPKD patients, 21 (34.4%) were hospitalized, 10 (16.4%) were admitted to ICU, 4 (6.6%) required ventilator, and 4 (6.6%) died by August 18, 2020. These rates were comparable to patients with other cystic kidney diseases and cystic liver-only diseases. ADPKD was not a significant risk factor for any of the four outcomes in multivariable logistic regression analyses when compared with other cystic kidney diseases and cystic liver-only diseases. In contrast, diabetes was a significant risk factor for hospitalization [OR 2.30 (1.61, 3.30), p<0.001], ICU admission [OR 2.23 (1.47, 3.42), p<0.001], and ventilator requirement [OR 2.20 (1.27, 3.88), p=0.005]. Black race significantly increased the risk for ventilator requirement [OR 2.00 (1.18, 3.44), p=0.011] and mortality [OR 1.60 (1.02, 2.51), p=0.040]. We also examined the outcome of starting dialysis after COVID-19 confirmation. The main risk factor for starting dialysis was CKD [OR 6.37 (2.43, 16.7)] and Black race [OR 3.47 (1.48, 8.1)]. After controlling for CKD, ADPKD did not significantly increase the risk for newly starting dialysis comparing with other cystic kidney diseases and cystic liver-only diseases. In summary, ADPKD did not significantly alter major COVID-19 outcomes among veterans when compared to other cystic kidney and liver patients.

show abstract

Section: Discussionmentioning

confidence: 99%

Autosomal Dominant Polycystic Kidney Disease does not significantly alter major COVID-19 outcomes among veterans

Cui

Gallini

Jasien

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Phenotyping with PheMap has several advantages compared with previously reported approaches to high-throughput phenotyping. 8 , 9 , 43 Compared to ontological approaches to high-throughput phenotyping, 43 PheMap provides a way to quantify the importance of relationships between phenotypes and medical concepts through NLP and the TF-IDF statistic. In addition to diagnosis codes, PheMap incorporates other medical information into the phenotype score, including symptoms, medications, laboratory tests, and procedures, which has been shown to improve phenotyping.…”

Section: Discussionmentioning

confidence: 99%

PheMap: a multi-resource knowledge base for high-throughput phenotyping within electronic health records

Zheng

Feng

Kerchberger

et al. 2020

Journal of the American Medical Informatics Association

View full text Add to dashboard Cite

Objective Developing algorithms to extract phenotypes from electronic health records (EHRs) can be challenging and time-consuming. We developed PheMap, a high-throughput phenotyping approach that leverages multiple independent, online resources to streamline the phenotyping process within EHRs. Materials and Methods PheMap is a knowledge base of medical concepts with quantified relationships to phenotypes that have been extracted by natural language processing from publicly available resources. PheMap searches EHRs for each phenotype’s quantified concepts and uses them to calculate an individual’s probability of having this phenotype. We compared PheMap to clinician-validated phenotyping algorithms from the Electronic Medical Records and Genomics (eMERGE) network for type 2 diabetes mellitus (T2DM), dementia, and hypothyroidism using 84 821 individuals from Vanderbilt Univeresity Medical Center's BioVU DNA Biobank. We implemented PheMap-based phenotypes for genome-wide association studies (GWAS) for T2DM, dementia, and hypothyroidism, and phenome-wide association studies (PheWAS) for variants in FTO, HLA-DRB1, and TCF7L2. Results In this initial iteration, the PheMap knowledge base contains quantified concepts for 841 disease phenotypes. For T2DM, dementia, and hypothyroidism, the accuracy of the PheMap phenotypes were >97% using a 50% threshold and eMERGE case-control status as a reference standard. In the GWAS analyses, PheMap-derived phenotype probabilities replicated 43 of 51 previously reported disease-associated variants for the 3 phenotypes. For 9 of the 11 top associations, PheMap provided an equivalent or more significant P value than eMERGE-based phenotypes. The PheMap-based PheWAS showed comparable or better performance to a traditional phecode-based PheWAS. PheMap is publicly available online. Conclusions PheMap significantly streamlines the process of extracting research-quality phenotype information from EHRs, with comparable or better performance to current phenotyping approaches.

show abstract

“…Let , and . These two key features can be mapped automatically as in Liao et al [21] using existing knowledge sources including the PheWAS catalogue [12] and the Unified Medical Language System (UMLS). Additional candidate features – including counts of other ICD codes, NLP features, drug prescriptions, lab tests, and procedure codes – can be identified automatically without GLabels via existing methods such as the SAFE method.…”

Section: Methodsmentioning

confidence: 99%

“…not requring GLabels) computational phenotyping methods. [17, 18, 19, 20, 21, 22, 23, 24] The class of “weakly supervised” methods, which train supervised classifiers using noisy labels generated from key surrogate features in the data rather than expensive GLabels, has proven particularly powerful to this end. For instance, the “anchor and learn” approach trains a regularized logistic regression model on imperfect labels derived from ‘anchor’ features with high PPVs.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

sureLDA: A Multi-Disease Automated Phenotyping Method for the Electronic Health Record

Ahuja

Zhou

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Objective: A major bottleneck hindering utilization of electronic health record (EHR) data for translational research is the lack of precise phenotype labels. Chart review as well as rule-based and supervised phenotyping approaches require laborious expert input, hampering applicability to studies that require many phenotypes to be defined and labeled de novo. Though ICD codes are often used as surrogates for true labels in this setting, these sometimes suffer from poor specificity. We propose a fully automated topic modeling algorithm to simultaneously annotate multiple phenotypes.Methods: sureLDA is a label-free multidimensional phenotyping method. It first uses the PheNorm algorithm to initialize probabilities based on two surrogate features for each target phenotype, and then leverages these probabilities to constrain the Latent Dirichlet Allocation (LDA) topic model to generate phenotype-specific topics. Finally, it combines phenotype-feature counts with surrogates via clustering ensemble to yield final phenotype probabilities.Results: sureLDA achieves reliably high accuracy and precision across a range of simulated and real-world phenotypes. Its performance is robust to phenotype prevalence and relative informativeness of surogate versus non-surrogate features. It also exhibits powerful feature selection properties. Discussion: sureLDA combines attractive properties of PheNorm and LDA to achieve high accuracy and precision robust to diverse phenotype characteristics. It offers particular improvement for phenotypes insufficiently captured by a few surrogate features. Moreover, sureLDA's feature selection ability enables it to handle high feature dimensions and produce interpretable computational phenotypes. Conclusion: sureLDA is well suited toward large-scale EHR phenotyping for highly multi-phenotype applications such as PheWAS. , using a slightly modified PheNorm algorithm. In brief, standard PheNorm estimates ߨ by (i) normalizing ‫ܫ‬ ‫ܥ‬ ‫ܦ‬ ,

show abstract

High-throughput multimodal automated phenotyping (MAP) with application to PheWAS

Cited by 71 publications

References 35 publications

Autosomal Dominant Polycystic Kidney Disease does not significantly alter major COVID-19 outcomes among veterans

Autosomal Dominant Polycystic Kidney Disease does not significantly alter major COVID-19 outcomes among veterans

PheMap: a multi-resource knowledge base for high-throughput phenotyping within electronic health records

sureLDA: A Multi-Disease Automated Phenotyping Method for the Electronic Health Record

Contact Info

Product

Resources

About