Robust replication of genotype-phenotype associations across multiple diseases in an electronic medical record

Leveraging undecided cases in chart-reviewed phenotypes to enhance EHR-based association studies

OBJECTIVES

In electronic health record (EHR)-based association studies, phenotyping algorithms efficiently classify patient clinical outcomes into binary categories but are susceptible to misclassification errors. The gold standard, manual chart review, involves clinicians determining the true disease status based on their assessment of health records. These clinicians-labeled phenotypes are labor-intensive and typically limited to a small subset of patients, potentially introducing a third "undecided" category when phenotypes are indeterminate. We aim to effectively integrate the algorithm-derived and chart-reviewed outcomes when both are available in EHR-based association studies.

MATERIAL AND METHODS

We propose an augmented estimation method that combines the binary algorithm-derived phenotypes for the entire cohort with the trinary chart-reviewed phenotypes for a small, selected subset. Additionally, a cost-effective outcome-dependent sampling strategy is used to address the rare disease scenarios. The proposed trinary chart-reviewed phenotype integrated cost-effective augmented estimation (TriCA) was evaluated across a wide range of simulation settings and real-world applications, including using EHR data on Alzheimer's disease and related dementias (ADRD) from the OneFlorida + Clinical Research Network, and using cohort data on second breast cancer events (SBCE) from the Kaiser Permanente Washington.

RESULTS

Compared to estimation based on random sampling, our augmented method improved mean square error by up to 28.3% in simulation studies; compared to estimation using only trinary chart-reviewed phenotypes, our method improved efficiency by up to 33.3% in ADRD data and 50.8% in SBCE data.

DISCUSSION

Our simulation studies and real-world applications demonstrate that, compared to existing methods, the proposed method provides unbiased estimates with higher statistical efficiency.

CONCLUSION

The proposed method effectively combined binary algorithm-derived phenotypes for the whole cohort with trinary chart-reviewed outcomes for a limited validation set, making it applicable to a broader range of applications and enhancing risk factor identification in EHR-based association studies.

Artificial intelligence in autoimmune diseases: a bibliometric exploration of the past two decades

Objective

Autoimmune diseases have long been recognized for their intricate nature and elusive mechanisms, presenting significant challenges in both diagnosis and treatment. The advent of artificial intelligence technology has opened up new possibilities for understanding, diagnosing, predicting, and managing autoimmune disorders. This study aims to explore the current state and emerging trends in the field through bibliometric analysis, providing guidance for future research directions.

Methods

The study employed the Web of Science Core Collection database for data acquisition and performed bibliometric analysis using CiteSpace, HistCite Pro, and VOSviewer.

Results

Over the past two decades, 1,695 publications emerged in this research field, including 1,409 research articles and 286 reviews. This investigation unveils the global development landscape predominantly led by the United States and China. The research identifies key institutions, such as Brigham & Women's Hospital, influential journals like the Annals of the Rheumatic Diseases, distinguished authors including Katherine P. Liao, and pivotal articles. It visually maps out the research clusters' evolutionary path over time and explores their applications in patient identification, risk factors, prognosis assessment, diagnosis, classification of disease subtypes, monitoring and decision support, and drug discovery.

Conclusion

AI is increasingly recognized for its potential in the field of autoimmune diseases, yet it continues to face numerous challenges, including insufficient model validation and difficulties in data integration and computational power. Significant advancements have been demanded to enhance diagnostic precision, improve treatment methodologies, and establish robust frameworks for data protection, thereby facilitating more effective management of these complex conditions.

Evaluating the impact of modeling choices on the performance of integrated genetic and clinical models

PURPOSE

The value of genetic information for improving the performance of clinical risk prediction models has yielded variable conclusions. Many methodological decisions have the potential to contribute to differential results. We performed multiple modeling experiments integrating clinical and demographic data from electronic health records with genetic data to understand which decisions may affect performance.

METHODS

Clinical data in the form of structured diagnostic codes, medications, procedural codes, and demographics were extracted from 2 large independent health systems, and polygenic risk scores (PRS) were generated across all patients of European ancestry with genetic data in the corresponding biobanks. Crohn's disease was studied based on its substantial genetic component, established electronic health records-based definition, and sufficient prevalence for training and testing. We investigated the impact of choices regarding the PRS integration method, training sample, model complexity, and performance metrics.

RESULTS

Overall, our results showed that including PRS resulted in higher performance, but this gain was only robust in situations with limited clinical information. We found consistent performance increases from more compute-intensive models, such as random forest, but the impact of other decisions varied by site.

CONCLUSION

This work highlights the importance of considering methodological decision points in interpreting the impact of PRS on prediction performance in clinical models.

Bidirectional Risk Modulator and Modifier Variant of Dilated and Hypertrophic Cardiomyopathy in BAG3

Importance

The genetic factors that modulate the reduced penetrance and variable expressivity of heritable dilated cardiomyopathy (DCM) are largely unknown. BAG3 genetic variants have been implicated in both DCM and hypertrophic cardiomyopathy (HCM), nominating BAG3 as a gene that harbors potential modifier variants in DCM.

Objective

To interrogate the clinical traits and diseases associated with BAG3 coding variation.

Design, Setting, and Participants

This was a cross-sectional study in the Penn Medicine BioBank (PMBB) enrolling patients of the University of Pennsylvania Health System's clinical practice sites from 2014 to 2023. Whole-exome sequencing (WES) was linked to electronic health record (EHR) data to associate BAG3 coding variants with EHR phenotypes. This was a health care population-based study including individuals of European and African genetic ancestry in the PMBB with WES linked to EHR phenotypes, with replication studies in BioVU, UK Biobank, MyCode, and DCM Precision Medicine Study.

Exposures

Carrier status for BAG3 coding variants.

Main Outcomes and Measures

Association of BAG3 coding variation with clinical diagnoses, echocardiographic traits, and longitudinal outcomes.

Results

In PMBB (n = 43 731; median [IQR] age, 65 [50-76] years; 21 907 female [50.1%]), among 30 324 European and 11 198 African individuals, the common C151R variant was associated with decreased risk for DCM (odds ratio [OR], 0.85; 95% CI, 0.78-0.92) and simultaneous increased risk for HCM (OR, 1.59; 95% CI, 1.25-2.02), which was confirmed in the replication cohorts. C151R carriers exhibited improved longitudinal outcomes compared with noncarriers as assessed by age at death (hazard ratio [HR], 0.85; 95% CI, 0.74-0.96; median [IQR] age, 71.8 [63.1-80.7] in carriers and 70.3 [61.6-79.2] in noncarriers) and heart transplant (HR, 0.81; 95% CI, 0.66-0.99; median [IQR] age, 56.7 [46.1-63.1] in carriers and 55.6 [45.2-62.9] in noncarriers). C151R was associated with reduced risk of DCM (OR, 0.42; 95% CI, 0.24-0.74) and heart failure (OR, 0.27; 95% CI, 0.14-0.50) among individuals harboring truncating TTN variants in exons with high cardiac expression (n = 358).

Conclusions and Relevance

BAG3 C151R was identified as a bidirectional modulator of risk along the DCM-HCM spectrum, as well as an important genetic modifier variant in TTN-mediated DCM. This work expands on the understanding of the etiology and penetrance of DCM, suggesting that BAG3 C151R is an important genetic modifier variant contributing to the variable expressivity of DCM, warranting further exploration of its mechanisms and of genetic modifiers in DCM more broadly.

Electronic health record biobank cohort recapitulates an association between the MUC5B promoter polymorphism and ARDS in critically ill adults

Background

Large population-based DNA biobanks linked to electronic health records (EHRs) may provide novel opportunities to identify genetic drivers of ARDS.

Research Question

Can we develop an EHR-based algorithm to identify ARDS in a biobank database, and can this validate a previously reported ARDS genetic risk factor?

Study Design and Methods

We analyzed two parallel genotyped cohorts: a prospective biomarker cohort of critically ill adults (VALID), and a retrospective cohort of hospitalized participants enrolled in a de-identified EHR biobank (BioVU). ARDS was identified by clinician-investigator review in VALID and an EHR algorithm in BioVU (EHR-ARDS). We tested the association between the MUC5B promoter polymorphism rs35705950 with development of ARDS, and assessed if age modified this genetic association in each cohort.

Results

In VALID, 2,795 patients were included, age was 55 [43, 66] (median [IQR]) years, and 718 (25.7%) developed ARDS. In BioVU, 9,025 hospitalized participants were included, age was 60 [48, 70] years, and 1,056 (11.7%) developed EHR-ARDS. We observed a significant age-related interaction effect on ARDS in VALID: among older patients, rs35705950 was associated with increased ARDS risk (OR: 1.44; 95%CI 1.08-1.92; p=0.012) whereas among younger patients this effect was absent (OR: 0.84; 95%CI: 0.62-1.14; p=0.26). In BioVU, rs35705950 was associated with increased risk for EHR-ARDS among all participants (OR: 1.20; 95%CI: 1.00-1.43, p=0.043) and this did not vary by age. The polymorphism was also associated worse oxygenation in mechanically ventilated BioVU participants, but had no association with oxygenation in VALID.

Interpretation

The MUC5B promoter polymorphism was associated with ARDS in two cohorts of at-risk adults. Although age-related effect modification was observed only in VALID, BioVU identified a consistent association between MUC5B and ARDS risk regardless of age, and a novel association with oxygenation impairment. Our study highlights the potential for EHR biobanks to enable precision-medicine ARDS studies.

Transcriptome-wide association study and Mendelian randomization in pancreatic cancer identifies susceptibility genes and causal relationships with type 2 diabetes and venous thromboembolism

BACKGROUND

Two important questions regarding the genetics of pancreatic adenocarcinoma (PDAC) are 1. Which germline genetic variants influence the incidence of this cancer; and 2. Whether PDAC causally predisposes to associated non-malignant phenotypes, such as type 2 diabetes (T2D) and venous thromboembolism (VTE).

METHODS

In this study of 8803 patients with PDAC and 67,523 controls, we first performed a large-scale transcriptome-wide association study to investigate the association between genetically determined gene expression in normal pancreas tissue and PDAC risk. Secondly, we used Mendelian Randomization (MR) to analyse the causal relationships among PDAC, T2D (74,124 cases and 824,006 controls) and VTE (30,234 cases and 172,122 controls).

FINDINGS

Sixteen genes showed an association with PDAC risk (FDR <0.10), including six genes not yet reported for PDAC risk (PPIP5K2, TFR2, HNF4G, LRRC10B, PRC1 and FBXL20) and ten previously reported genes (INHBA, SMC2, ABO, PDX1, MTMR6, ACOT2, PGAP3, STARD3, GSDMB, ADAM33). MR provided support for a causal effect of PDAC on T2D using genetic instruments in the HNF4G and PDX1 loci, and unidirectional causality of VTE on PDAC involving the ABO locus (OR 2.12, P < 1e-7). No evidence of a causal effect of PDAC on VTE was found.

INTERPRETATION

These analyses identified candidate susceptibility genes and disease relationships for PDAC that warrant further investigation. HNF4G and PDX1 may induce PDAC-associated diabetes, whereas ABO may induce the causative effect of VTE on PDAC.

FUNDING

National Institutes of Health (USA).

Association Between Remnant Cholesterol and Risk of Incident Atrial Fibrillation: Population-Based Evidence From a Large-Scale Prospective Cohort Study

BACKGROUND

Evidence for the relationship between remnant cholesterol (RC) and incident atrial fibrillation (AF) risk remains sparse and limited.

METHODS AND RESULTS

Participants were enrolled between 2006 and 2010 and followed up to 2021. The multivariable Cox proportional hazards model was used to examine the relationship between RC quartiles and risk of incident AF. Subgroup analyses and sensitivity analyses were performed to explore the potential modification of the association and the robustness of the main findings. A total of 422 316 participants (mean age, 56 years; 54% women) were included for analyses. During a median follow-up of 11.9 years (first quartile-third quartile, 11.6-13.2 years), there were 24 774 AF events documented with an incidence of 4.92 events per 1000 person-years (95% CI, 4.86-4.98). Participants in higher RC quartiles had a lower risk of incident AF than those in the lowest quartile (first quartile): hazard ratio (HR)=0.96 (95% CI, 0.91-1.00) for second quartile; HR=0.92 (95% CI, 0.88-0.96) for third quartile; and HR=0.85 (95% CI, 0.81-0.89) for fourth quartile (P for trend <0.001). The association between RC quartiles and risk of incident AF was stronger in participants aged ≥65 years, in men, and in participants without history of diabetes when compared with control groups (P<0.001 for interaction).

CONCLUSIONS

On the basis of data from this large-scale prospective cohort study, elevated RC was associated with a lower risk of incident AF.

Phenomewide Association Study of Health Outcomes Associated With the Genetic Correlates of 25 Hydroxyvitamin D Concentration and Vitamin D Binding Protein Concentration

While it is known that vitamin D deficiency is associated with adverse bone outcomes, it remains unclear whether low vitamin D status may increase the risk of a wider range of health outcomes. We had the opportunity to explore the association between common genetic variants associated with both 25 hydroxyvitamin D (25OHD) and the vitamin D binding protein (DBP, encoded by the GC gene) with a comprehensive range of health disorders and laboratory tests in a large academic medical center. We used summary statistics for 25OHD and DBP to generate polygenic scores (PGS) for 66,482 participants with primarily European ancestry and 13,285 participants with primarily African ancestry from the Vanderbilt University Medical Center Biobank (BioVU). We examined the predictive properties of PGS25OHD, and two scores related to DBP concentration with respect to 1322 health-related phenotypes and 315 laboratory-measured phenotypes from electronic health records. In those with European ancestry: (a) the PGS25OHD and PGSDBP scores, and individual SNPs rs4588 and rs7041 were associated with both 25OHD concentration and 1,25 dihydroxyvitamin D concentrations; (b) higher PGS25OHD was associated with decreased concentrations of triglycerides and cholesterol, and reduced risks of vitamin D deficiency, disorders of lipid metabolism, and diabetes. In general, the findings for the African ancestry group were consistent with findings from the European ancestry analyses. Our study confirms the utility of PGS and two key variants within the GC gene (rs4588 and rs7041) to predict the risk of vitamin D deficiency in clinical settings and highlights the shared biology between vitamin D-related genetic pathways a range of health outcomes.

Evaluating the impact of modeling choices on the performance of integrated genetic and clinical models

The value of genetic information for improving the performance of clinical risk prediction models has yielded variable conclusions. Many methodological decisions have the potential to contribute to differential results across studies. Here, we performed multiple modeling experiments integrating clinical and demographic data from electronic health records (EHR) and genetic data to understand which decision points may affect performance. Clinical data in the form of structured diagnostic codes, medications, procedural codes, and demographics were extracted from two large independent health systems and polygenic risk scores (PRS) were generated across all patients with genetic data in the corresponding biobanks. Crohn's disease was used as the model phenotype based on its substantial genetic component, established EHR-based definition, and sufficient prevalence for model training and testing. We investigated the impact of PRS integration method, as well as choices regarding training sample, model complexity, and performance metrics. Overall, our results show that including PRS resulted in higher performance by some metrics but the gain in performance was only robust when combined with demographic data alone. Improvements were inconsistent or negligible after including additional clinical information. The impact of genetic information on performance also varied by PRS integration method, with a small improvement in some cases from combining PRS with the output of a clinical model (late-fusion) compared to its inclusion an additional feature (early-fusion). The effects of other modeling decisions varied between institutions though performance increased with more compute-intensive models such as random forest. This work highlights the importance of considering methodological decision points in interpreting the impact on prediction performance when including PRS information in clinical models.

Genome-Wide Association Study of CKD Progression

SIGNIFICANCE STATEMENT

Rapid progression of CKD is associated with poor clinical outcomes. Most previous studies looking for genetic factors associated with low eGFR have used cross-sectional data. The authors conducted a meta-analysis of genome-wide association studies of eGFR decline among 116,870 participants with CKD, focusing on longitudinal data. They identified three loci (two of them novel) associated with longitudinal eGFR decline. In addition to the known UMOD/PDILT locus, variants within BICC1 were associated with significant differences in longitudinal eGFR slope. Variants within HEATR4 also were associated with differences in eGFR decline, but only among Black/African American individuals without diabetes. These findings help characterize molecular mechanisms of eGFR decline in CKD and may inform new therapeutic approaches for progressive kidney disease.

BACKGROUND

Rapid progression of CKD is associated with poor clinical outcomes. Despite extensive study of the genetics of cross-sectional eGFR, only a few loci associated with eGFR decline over time have been identified.

METHODS

We performed a meta-analysis of genome-wide association studies of eGFR decline among 116,870 participants with CKD-defined by two outpatient eGFR measurements of <60 ml/min per 1.73 m 2 , obtained 90-365 days apart-from the Million Veteran Program and Vanderbilt University Medical Center's DNA biobank. The primary outcome was the annualized relative slope in outpatient eGFR. Analyses were stratified by ethnicity and diabetes status and meta-analyzed thereafter.

RESULTS

In cross-ancestry meta-analysis, the strongest association was rs77924615, near UMOD / PDILT ; each copy of the G allele was associated with a 0.30%/yr faster eGFR decline ( P = 4.9×10 -27 ). We also observed an association within BICC1 (rs11592748), where every additional minor allele was associated with a 0.13%/yr slower eGFR decline ( P = 5.6×10 -9 ). Among participants without diabetes, the strongest association was the UMOD/PDILT variant rs36060036, associated with a 0.27%/yr faster eGFR decline per copy of the C allele ( P = 1.9×10 -17 ). Among Black participants, a significantly faster eGFR decline was associated with variant rs16996674 near APOL1 (R 2 =0.29 with the G1 high-risk genotype); among Black participants with diabetes, lead variant rs11624911 near HEATR4 also was associated with a significantly faster eGFR decline. We also nominally replicated loci with known associations with eGFR decline, near PRKAG2, FGF5, and C15ORF54.

CONCLUSIONS

Three loci were significantly associated with longitudinal eGFR change at genome-wide significance. These findings help characterize molecular mechanisms of eGFR decline and may contribute to the development of new therapeutic approaches for progressive CKD.

A flexible symbolic regression method for constructing interpretable clinical prediction models

Machine learning (ML) models trained for triggering clinical decision support (CDS) are typically either accurate or interpretable but not both. Scaling CDS to the panoply of clinical use cases while mitigating risks to patients will require many ML models be intuitively interpretable for clinicians. To this end, we adapted a symbolic regression method, coined the feature engineering automation tool (FEAT), to train concise and accurate models from high-dimensional electronic health record (EHR) data. We first present an in-depth application of FEAT to classify hypertension, hypertension with unexplained hypokalemia, and apparent treatment-resistant hypertension (aTRH) using EHR data for 1200 subjects receiving longitudinal care in a large healthcare system. FEAT models trained to predict phenotypes adjudicated by chart review had equivalent or higher discriminative performance (p < 0.001) and were at least three times smaller (p < 1 × 10-6) than other potentially interpretable models. For aTRH, FEAT generated a six-feature, highly discriminative (positive predictive value = 0.70, sensitivity = 0.62), and clinically intuitive model. To assess the generalizability of the approach, we tested FEAT on 25 benchmark clinical phenotyping tasks using the MIMIC-III critical care database. Under comparable dimensionality constraints, FEAT's models exhibited higher area under the receiver-operating curve scores than penalized linear models across tasks (p < 6 × 10-6). In summary, FEAT can train EHR prediction models that are both intuitively interpretable and accurate, which should facilitate safe and effective scaling of ML-triggered CDS to the panoply of potential clinical use cases and healthcare practices.

Polygenic Contributions to Chronic Overlapping Pain Conditions in a Large Electronic Health Record Sample

Chronic overlapping pain conditions (COPCs) are believed to share common etiological mechanisms involving central sensitization. Genetic and environmental factors putatively combine to influence susceptibility to central sensitization and COPCs. This study employed a genome-wide polygenic risk score approach to evaluate genetic influences on 8 common COPCs. COPCs were identified by International Classification of Disease codes in Vanderbilt's deidentified clinical biorepository (BioVU), with each COPC condition empirically weighted for the level of central sensitization based on prior work. A centralized pain score (CPS) was calculated for 55,340 individuals by summing the weighted number of COPCs. Overall, 12,502 individuals (22.6%) were diagnosed with at least 1 COPC, with females exhibiting nearly twice the mean CPS as males. To assess the genetic influence on centralized pain in COPCs, 6 pain polygenic risk scores (PRSs) were developed using UK Biobank data to predict 6 pain criteria (no pain, neck/shoulder, abdomen, hip, knee, low back pain). These PRSs were then deployed in the BioVU cohort to test for association with CPS. In regression models adjusted for age, sex, and BMI, all pain PRSs except hip pain were significantly associated with CPS. Our findings support a shared polygenic influence across COPCs potentially involving central sensitization mechanisms. PERSPECTIVE: This study used a polygenic risk score approach to investigate genetic influences on chronic overlapping pain conditions. Significant findings in this study provide evidence supporting previous hypotheses that a shared polygenic influence involving central sensitization may underly chronic overlapping pain conditions and can guide future biomarker and risk assessment research.

Effects of 25-Hydroxyvitamin D Levels on Renal Function: A Bidirectional Mendelian Randomization Study

CONTEXT

Observational studies investigating the role of vitamin D in renal function have yielded inconsistent results.

OBJECTIVE

We tested whether 25-hydroxyvitamin D (25[OH]D) serum levels are associated with renal function, and inversely, whether altered renal function causes changes in 25(OH)D, using Mendelian randomization (MR).

METHODS

In this two-sample MR study, we used single nucleotide polymorphisms (SNP) associated with 25(OH)D in 443 734 Europeans and evaluated their effects on estimated glomerular filtration rate (eGFR), blood urea nitrogen (BUN), chronic kidney disease (CKD) risk and progression in genome-wide association studies totaling over 1 million Europeans. To control for pleiotropy, we also used SNPs solely in DHCR7, CYP2R1, and GC, all genes with known roles in vitamin D metabolism. We performed a reverse MR, using SNPs for the above indices of renal function to study causal effects on 25(OH)D levels.

RESULTS

We did not find robust evidence supporting effects of 25(OH)D on eGFR, BUN, and CKD or its progression. Our inverse variance weighted MR demonstrated a 0.56 decrease in standardized log-transformed 25(OH)D (95% CI -0.73, -0.41; P = 2.89 × 10-12) per unit increase in log-transformed eGFR. Increased BUN was associated with increased 25(OH)D (β = 0.25, 95% CI 0.15, 0.36; P = 4.12 × 10-6 per unit increase in log-transformed BUN). Finally, genetically predicted CKD conferred a 0.05 increase in standardized log-transformed 25(OH)D level (95% CI 0.04, 0.06; P = 1.06 × 10-13). Other MR methods confirmed the findings of the main analyses.

CONCLUSION

Genetically predicted CKD, increased BUN, and decreased eGFR are associated with increased 25(OH)D levels, but we found no causal effect of 25(OH)D on renal function in Europeans.

Patient-specific comorbidities as prognostic variables for survival in myelofibrosis

Treatment decisions in primary myelofibrosis (PMF) are guided by numerous prognostic systems. Patient-specific comorbidities have influence on treatment-related survival and are considered in clinical contexts but have not been routinely incorporated into current prognostic models. We hypothesized that patient-specific comorbidities would inform prognosis and could be incorporated into a quantitative score. All patients with PMF or secondary myelofibrosis with available DNA and comprehensive electronic health record (EHR) data treated at Vanderbilt University Medical Center between 1995 and 2016 were identified within Vanderbilt's Synthetic Derivative and BioVU Biobank. We recapitulated established PMF risk scores (eg, Dynamic International Prognostic Scoring System [DIPSS], DIPSS plus, Genetics-Based Prognostic Scoring System, Mutation-Enhanced International Prognostic Scoring System 70+) and comorbidities through EHR chart extraction and next-generation sequencing on biobanked peripheral blood DNA. The impact of comorbidities was assessed via DIPSS-adjusted overall survival using Bonferroni correction. Comorbidities associated with inferior survival include renal failure/dysfunction (hazard ratio [HR], 4.3; 95% confidence interval [95% CI], 2.1-8.9; P = .0001), intracranial hemorrhage (HR, 28.7; 95% CI, 7.0-116.8; P = 2.83e-06), invasive fungal infection (HR, 41.2; 95% CI, 7.2-235.2; P = 2.90e-05), and chronic encephalopathy (HR, 15.1; 95% CI, 3.8-59.4; P = .0001). The extended DIPSS model including all 4 significant comorbidities showed a significantly higher discriminating power (C-index 0.81; 95% CI, 0.78-0.84) than the original DIPSS model (C-index 0.73; 95% CI, 0.70-0.77). In summary, we repurposed an institutional biobank to identify and risk-classify an uncommon hematologic malignancy by established (eg, DIPSS) and other clinical and pathologic factors (eg, comorbidities) in an unbiased fashion. The inclusion of comorbidities into risk evaluation may augment prognostic capability of future genetics-based scoring systems.

How T118M peripheral myelin protein 22 predisposes humans to Charcot-Marie-Tooth disease

Data from gnomAD indicate that a missense mutation encoding the T118M variation in human peripheral myelin protein 22 (PMP22) is found in roughly one of every 75 genomes of western European lineage (1:120 in the overall human population). It is unusual among PMP22 variants that cause Charcot-Marie-Tooth (CMT) disease in that it is not 100% penetrant. Here, we conducted cellular and biophysical studies to determine why T118M PMP22 predisposes humans to CMT, but with only incomplete penetrance. We found that T118M PMP22 is prone to mistraffic but differs even from the WT protein in that increased expression levels do not result in a reduction in trafficking efficiency. Moreover, the T118M mutant exhibits a reduced tendency to form large intracellular aggregates relative to other disease mutants and even WT PMP22. NMR spectroscopy revealed that the structure and dynamics of T118M PMP22 resembled those of WT. These results show that the main consequence of T118M PMP22 in WT/T118M heterozygous individuals is a reduction in surface-trafficked PMP22, unaccompanied by formation of toxic intracellular aggregates. This explains the incomplete disease penetrance and the mild neuropathy observed for WT/T118M CMT cases. We also analyzed BioVU, a biobank linked to deidentified electronic medical records, and found a statistically robust association of the T118M mutation with the occurrence of long and/or repeated episodes of carpal tunnel syndrome. Collectively, our results illuminate the cellular effects of the T118M PMP22 variation leading to CMT disease and indicate a second disorder for which it is a risk factor.

PhenoExam: gene set analyses through integration of different phenotype databases

BACKGROUND

Gene set enrichment analysis (detecting phenotypic terms that emerge as significant in a set of genes) plays an important role in bioinformatics focused on diseases of genetic basis. To facilitate phenotype-oriented gene set analysis, we developed PhenoExam, a freely available R package for tool developers and a web interface for users, which performs: (1) phenotype and disease enrichment analysis on a gene set; (2) measures statistically significant phenotype similarities between gene sets and (3) detects significant differential phenotypes or disease terms across different databases.

RESULTS

PhenoExam generates sensitive and accurate phenotype enrichment analyses. It is also effective in segregating gene sets or Mendelian diseases with very similar phenotypes. We tested the tool with two similar diseases (Parkinson and dystonia), to show phenotype-level similarities but also potentially interesting differences. Moreover, we used PhenoExam to validate computationally predicted new genes potentially associated with epilepsy.

CONCLUSIONS

We developed PhenoExam, a freely available R package and Web application, which performs phenotype enrichment and disease enrichment analysis on gene set G, measures statistically significant phenotype similarities between pairs of gene sets G and G' and detects statistically significant exclusive phenotypes or disease terms, across different databases. We proved with simulations and real cases that it is useful to distinguish between gene sets or diseases with very similar phenotypes. Github R package URL is https://github.com/alexcis95/PhenoExam . Shiny App URL is https://alejandrocisterna.shinyapps.io/phenoexamweb/ .

Quantifying the phenome-wide disease burden of obesity using electronic health records and genomics

OBJECTIVE

High BMI is associated with many comorbidities and mortality. This study aimed to elucidate the overall clinical risk of obesity using a genome- and phenome-wide approach.

METHODS

This study performed a phenome-wide association study of BMI using a clinical cohort of 736,726 adults. This was followed by genetic association studies using two separate cohorts: one consisting of 65,174 adults in the Electronic Medical Records and Genomics (eMERGE) Network and another with 405,432 participants in the UK Biobank.

RESULTS

Class 3 obesity was associated with 433 phenotypes, representing 59.3% of all billing codes in individuals with severe obesity. A genome-wide polygenic risk score for BMI, accounting for 7.5% of variance in BMI, was associated with 296 clinical diseases, including strong associations with type 2 diabetes, sleep apnea, hypertension, and chronic liver disease. In all three cohorts, 199 phenotypes were associated with class 3 obesity and polygenic risk for obesity, including novel associations such as increased risk of renal failure, venous insufficiency, and gastroesophageal reflux.

CONCLUSIONS

This combined genomic and phenomic systematic approach demonstrated that obesity has a strong genetic predisposition and is associated with a considerable burden of disease across all disease classes.

HCLC-FC: A novel statistical method for phenome-wide association studies

The emergence of genetic data coupled to longitudinal electronic medical records (EMRs) offers the possibility of phenome-wide association studies (PheWAS). In PheWAS, the whole phenome can be divided into numerous phenotypic categories according to the genetic architecture across phenotypes. Currently, statistical analyses for PheWAS are mainly univariate analyses, which test the association between one genetic variant and one phenotype at a time. In this article, we derived a novel and powerful multivariate method for PheWAS. The proposed method involves three steps. In the first step, we apply the bottom-up hierarchical clustering method to partition a large number of phenotypes into disjoint clusters within each phenotypic category. In the second step, the clustering linear combination method is used to combine test statistics within each category based on the phenotypic clusters and obtain p-values from each phenotypic category. In the third step, we propose a new false discovery rate (FDR) control approach. We perform extensive simulation studies to compare the performance of our method with that of other existing methods. The results show that our proposed method controls FDR very well and outperforms other methods we compared with. We also apply the proposed approach to a set of EMR-based phenotypes across more than 300,000 samples from the UK Biobank. We find that the proposed approach not only can well-control FDR at a nominal level but also successfully identify 1,244 significant SNPs that are reported to be associated with some phenotypes in the GWAS catalog. Our open-access tools and instructions on how to implement HCLC-FC are available at https://github.com/XiaoyuLiang/HCLCFC.

Interaction of obesity polygenic score with lifestyle risk factors in an electronic health record biobank

BACKGROUND

Genetic and lifestyle factors have considerable effects on obesity and related diseases, yet their effects in a clinical cohort are unknown. This study in a patient biobank examined associations of a BMI polygenic risk score (PRS), and its interactions with lifestyle risk factors, with clinically measured BMI and clinical phenotypes.

METHODS

The Mass General Brigham (MGB) Biobank is a hospital-based cohort with electronic health record, genetic, and lifestyle data. A PRS for obesity was generated using 97 genetic variants for BMI. An obesity lifestyle risk index using survey responses to obesogenic lifestyle risk factors (alcohol, education, exercise, sleep, smoking, and shift work) was used to dichotomize the cohort into high and low obesogenic index based on the population median. Height and weight were measured at a clinical visit. Multivariable linear cross-sectional associations of the PRS with BMI and interactions with the obesity lifestyle risk index were conducted. In phenome-wide association analyses (PheWAS), similar logistic models were conducted for 675 disease outcomes derived from billing codes.

RESULTS

Thirty-three thousand five hundred eleven patients were analyzed (53.1% female; age 60.0 years; BMI 28.3 kg/m²), of which 17,040 completed the lifestyle survey (57.5% female; age: 60.2; BMI: 28.1 (6.2) kg/m²). Each standard deviation increment in the PRS was associated with 0.83 kg/m² unit increase in BMI (95% confidence interval (CI) =0.76, 0.90). There was an interaction between the obesity PRS and obesity lifestyle risk index on BMI. The difference in BMI between those with a high and low obesogenic index was 3.18 kg/m² in patients in the highest decile of PRS, whereas that difference was only 1.55 kg/m² in patients in the lowest decile of PRS. In PheWAS, the obesity PRS was associated with 40 diseases spanning endocrine/metabolic, circulatory, and 8 other disease groups. No interactions were evident between the PRS and the index on disease outcomes.

CONCLUSIONS

In this hospital-based clinical biobank, obesity risk conferred by common genetic variants was associated with elevated BMI and this risk was attenuated by a healthier patient lifestyle. Continued consideration of the role of lifestyle in the context of genetic predisposition in healthcare settings is necessary to quantify the extent to which modifiable lifestyle risk factors may moderate genetic predisposition and inform clinical action to achieve personalized medicine.

Assessment of Natural Language Processing Methods for Ascertaining the Expanded Disability Status Scale Score From the Electronic Health Records of Patients With Multiple Sclerosis: Algorithm Development and Validation Study

Background

The Expanded Disability Status Scale (EDSS) score is a widely used measure to monitor disability progression in people with multiple sclerosis (MS). However, extracting and deriving the EDSS score from unstructured electronic health records can be time-consuming.

Objective

We aimed to compare rule-based and deep learning natural language processing algorithms for detecting and predicting the total EDSS score and EDSS functional system subscores from the electronic health records of patients with MS.

Methods

We studied 17,452 electronic health records of 4906 MS patients followed at one of Canada’s largest MS clinics between June 2015 and July 2019. We randomly divided the records into training (80%) and test (20%) data sets, and compared the performance characteristics of 3 natural language processing models. First, we applied a rule-based approach, extracting the EDSS score from sentences containing the keyword “EDSS.” Next, we trained a convolutional neural network (CNN) model to predict the 19 half-step increments of the EDSS score. Finally, we used a combined rule-based–CNN model. For each approach, we determined the accuracy, precision, recall, and F-score compared with the reference standard, which was manually labeled EDSS scores in the clinic database.

Results

Overall, the combined keyword-CNN model demonstrated the best performance, with accuracy, precision, recall, and an F-score of 0.90, 0.83, 0.83, and 0.83 respectively. Respective figures for the rule-based and CNN models individually were 0.57, 0.91, 0.65, and 0.70, and 0.86, 0.70, 0.70, and 0.70. Because of missing data, the model performance for EDSS subscores was lower than that for the total EDSS score. Performance improved when considering notes with known values of the EDSS subscores.

Conclusions

A combined keyword-CNN natural language processing model can extract and accurately predict EDSS scores from patient records. This approach can be automated for efficient information extraction in clinical and research settings.

SAT: a Surrogate-Assisted Two-wave case boosting sampling method, with application to EHR-based association studies

OBJECTIVES

Electronic health records (EHRs) enable investigation of the association between phenotypes and risk factors. However, studies solely relying on potentially error-prone EHR-derived phenotypes (ie, surrogates) are subject to bias. Analyses of low prevalence phenotypes may also suffer from poor efficiency. Existing methods typically focus on one of these issues but seldom address both. This study aims to simultaneously address both issues by developing new sampling methods to select an optimal subsample to collect gold standard phenotypes for improving the accuracy of association estimation.

MATERIALS AND METHODS

We develop a surrogate-assisted two-wave (SAT) sampling method, where a surrogate-guided sampling (SGS) procedure and a modified optimal subsampling procedure motivated from A-optimality criterion (OSMAC) are employed sequentially, to select a subsample for outcome validation through manual chart review subject to budget constraints. A model is then fitted based on the subsample with the true phenotypes. Simulation studies and an application to an EHR dataset of breast cancer survivors are conducted to demonstrate the effectiveness of SAT.

RESULTS

We found that the subsample selected with the proposed method contains informative observations that effectively reduce the mean squared error of the resultant estimator of the association.

CONCLUSIONS

The proposed approach can handle the problem brought by the rarity of cases and misclassification of the surrogate in phenotype-absent EHR-based association studies. With a well-behaved surrogate, SAT successfully boosts the case prevalence in the subsample and improves the efficiency of estimation.

Embedding electronic health records onto a knowledge network recognizes prodromal features of multiple sclerosis and predicts diagnosis

OBJECTIVE

Early identification of chronic diseases is a pillar of precision medicine as it can lead to improved outcomes, reduction of disease burden, and lower healthcare costs. Predictions of a patient's health trajectory have been improved through the application of machine learning approaches to electronic health records (EHRs). However, these methods have traditionally relied on "black box" algorithms that can process large amounts of data but are unable to incorporate domain knowledge, thus limiting their predictive and explanatory power. Here, we present a method for incorporating domain knowledge into clinical classifications by embedding individual patient data into a biomedical knowledge graph.

MATERIALS AND METHODS

A modified version of the Page rank algorithm was implemented to embed millions of deidentified EHRs into a biomedical knowledge graph (SPOKE). This resulted in high-dimensional, knowledge-guided patient health signatures (ie, SPOKEsigs) that were subsequently used as features in a random forest environment to classify patients at risk of developing a chronic disease.

RESULTS

Our model predicted disease status of 5752 subjects 3 years before being diagnosed with multiple sclerosis (MS) (AUC = 0.83). SPOKEsigs outperformed predictions using EHRs alone, and the biological drivers of the classifiers provided insight into the underpinnings of prodromal MS.

CONCLUSION

Using data from EHR as input, SPOKEsigs describe patients at both the clinical and biological levels. We provide a clinical use case for detecting MS up to 5 years prior to their documented diagnosis in the clinic and illustrate the biological features that distinguish the prodromal MS state.

Evidence of a Causal Relationship between Serum Thyroid-Stimulating Hormone and Osteoporotic Bone Fractures

OBJECTIVE

We aimed to validate the association of genome-wide association study (GWAS)-identified loci and polygenic risk score with serum thyroid-stimulating hormone (TSH) concentrations and the diagnosis of hypothyroidism. Then, the causal relationship between serum TSH and osteoporotic bone fracture risk was tested.

METHODS

A cross-sectional study was done among patients of European Caucasian ethnicity recruited in Tayside (Scotland, UK). Electronic medical records (EMRs) were used to identify patients and average serum TSH concentration and linked to genetic biobank data. Genetic associations were performed by linear and logistic regression models. One-sample Mendelian randomization (MR) was used to test causality of serum TSH on bone fracture risk.

RESULTS

Replication in 9,452 euthyroid individuals confirmed known loci previously reported. The 58 polymorphisms accounted for 11.08% of the TSH variation (p < 1e-04). TSH-GRS was directly associated with the risk of hypothyroidism with an odds ratio (OR) of 1.98 for the highest quartile compared to the first quartile (p = 2.2e-12). MR analysis of 5,599 individuals showed that compared with those in the lowest tertile of the TSH-GRS, men in the highest tertile had a decreased risk of osteoporotic bone fracture (OR = 0.59, p = 2.4e-03), while no difference in a similar comparison was observed in women (OR = 0.93, p = 0.61). Sensitivity analysis yielded similar results.

CONCLUSIONS

EMRs linked to genomic data in large populations allow replication of GWAS discoveries without additional genotyping costs. This study suggests that genetically raised serum TSH concentrations are causally associated with decreased bone fracture risk in men.

A cost-effective chart review sampling design to account for phenotyping error in electronic health records (EHR) data

OBJECTIVES

Electronic health records (EHR) are commonly used for the identification of novel risk factors for disease, often referred to as an association study. A major challenge to EHR-based association studies is phenotyping error in EHR-derived outcomes. A manual chart review of phenotypes is necessary for unbiased evaluation of risk factor associations. However, this process is time-consuming and expensive. The objective of this paper is to develop an outcome-dependent sampling approach for designing manual chart review, where EHR-derived phenotypes can be used to guide the selection of charts to be reviewed in order to maximize statistical efficiency in the subsequent estimation of risk factor associations.

MATERIALS AND METHODS

After applying outcome-dependent sampling, an augmented estimator can be constructed by optimally combining the chart-reviewed phenotypes from the selected patients with the error-prone EHR-derived phenotype. We conducted simulation studies to evaluate the proposed method and applied our method to data on colon cancer recurrence in a cohort of patients treated for a primary colon cancer in the Kaiser Permanente Washington (KPW) healthcare system.

RESULTS

Simulations verify the coverage probability of the proposed method and show that, when disease prevalence is less than 30%, the proposed method has smaller variance than an existing method where the validation set for chart review is uniformly sampled. In addition, from design perspective, the proposed method is able to achieve the same statistical power with 50% fewer charts to be validated than the uniform sampling method, thus, leading to a substantial efficiency gain in chart review. These findings were also confirmed by the application of the competing methods to the KPW colon cancer data.

DISCUSSION

Our simulation studies and analysis of data from KPW demonstrate that, compared to an existing uniform sampling method, the proposed outcome-dependent method can lead to a more efficient chart review sampling design and unbiased association estimates with higher statistical efficiency.

CONCLUSION

The proposed method not only optimally combines phenotypes from chart review with EHR-derived phenotypes but also suggests an efficient design for conducting chart review, with the goal of improving the efficiency of estimated risk factor associations using EHR data.

Two data-driven approaches to identifying the spectrum of problematic opioid use: A pilot study within a chronic pain cohort

BACKGROUND

Although electronic health records (EHR) have significant potential for the study of opioid use disorders (OUD), detecting OUD in clinical data is challenging. Models using EHR data to predict OUD often rely on case/control classifications focused on extreme opioid use. There is a need to expand this work to characterize the spectrum of problematic opioid use.

METHODS

Using a large academic medical center database, we developed 2 data-driven methods of OUD detection: (1) a Comorbidity Score developed from a Phenome-Wide Association Study of phenotypes associated with OUD and (2) a Text-based Score using natural language processing to identify OUD-related concepts in clinical notes. We evaluated the performance of both scores against a manual review with correlation coefficients, Wilcoxon rank sum tests, and area-under the receiver operating characteristic curves. Records with the highest Comorbidity and Text-based scores were re-evaluated by manual review to explore discrepancies.

RESULTS

Both the Comorbidity and Text-based OUD risk scores were significantly elevated in the patients judged as High Evidence for OUD in the manual review compared to those with No Evidence (p = 1.3E-5 and 1.3E-6, respectively). The risk scores were positively correlated with each other (rho = 0.52, p < 0.001). AUCs for the Comorbidity and Text-based scores were high (0.79 and 0.76, respectively). Follow-up manual review of discrepant findings revealed strengths of data-driven methods over manual review, and opportunities for improvement in risk assessment.

CONCLUSION

Risk scores comprising comorbidities and text offer differing but synergistic insights into characterizing problematic opioid use. This pilot project establishes a foundation for more robust work in the future.

A cross-population atlas of genetic associations for 220 human phenotypes

Current genome-wide association studies do not yet capture sufficient diversity in populations and scope of phenotypes. To expand an atlas of genetic associations in non-European populations, we conducted 220 deep-phenotype genome-wide association studies (diseases, biomarkers and medication usage) in BioBank Japan (n = 179,000), by incorporating past medical history and text-mining of electronic medical records. Meta-analyses with the UK Biobank and FinnGen (n_total = 628,000) identified ~5,000 new loci, which improved the resolution of the genomic map of human traits. This atlas elucidated the landscape of pleiotropy as represented by the major histocompatibility complex locus, where we conducted HLA fine-mapping. Finally, we performed statistical decomposition of matrices of phenome-wide summary statistics, and identified latent genetic components, which pinpointed responsible variants and biological mechanisms underlying current disease classifications across populations. The decomposed components enabled genetically informed subtyping of similar diseases (for example, allergic diseases). Our study suggests a potential avenue for hypothesis-free re-investigation of human diseases through genetics.

The Role of Electronic Health Records in Advancing Genomic Medicine

Recent advances in genomic technology and widespread adoption of electronic health records (EHRs) have accelerated the development of genomic medicine, bringing promising research findings from genome science into clinical practice. Genomic and phenomic data, accrued across large populations through biobanks linked to EHRs, have enabled the study of genetic variation at a phenome-wide scale. Through new quantitative techniques, pleiotropy can be explored with phenome-wide association studies, the occurrence of common complex diseases can be predicted using the cumulative influence of many genetic variants (polygenic risk scores), and undiagnosed Mendelian syndromes can be identified using EHR-based phenotypic signatures (phenotype risk scores). In this review, we trace the role of EHRs from the development of genome-wide analytic techniques to translational efforts to test these new interventions to the clinic. Throughout, we describe the challenges that remain when combining EHRs with genetics to improve clinical care.

Using Phecodes for Research with the Electronic Health Record: From PheWAS to PheRS

Electronic health records (EHRs) are a rich source of data for researchers, but extracting meaningful information out of this highly complex data source is challenging. Phecodes represent one strategy for defining phenotypes for research using EHR data. They are a high-throughput phenotyping tool based on ICD (International Classification of Diseases) codes that can be used to rapidly define the case/control status of thousands of clinically meaningful diseases and conditions. Phecodes were originally developed to conduct phenome-wide association studies to scan for phenotypic associations with common genetic variants. Since then, phecodes have been used to support a wide range of EHR-based phenotyping methods, including the phenotype risk score. This review aims to comprehensively describe the development, validation, and applications of phecodes and suggest some future directions for phecodes and high-throughput phenotyping.

Ascertaining Framingham heart failure phenotype from inpatient electronic health record data using natural language processing: a multicentre Atherosclerosis Risk in Communities (ARIC) validation study

OBJECTIVES

Using free-text clinical notes and reports from hospitalised patients, determine the performance of natural language processing (NLP) ascertainment of Framingham heart failure (HF) criteria and phenotype.

STUDY DESIGN

A retrospective observational study design of patients hospitalised in 2015 from four hospitals participating in the Atherosclerosis Risk in Communities (ARIC) study was used to determine NLP performance in the ascertainment of Framingham HF criteria and phenotype.

SETTING

Four ARIC study hospitals, each representing an ARIC study region in the USA.

PARTICIPANTS

A stratified random sample of hospitalisations identified using a broad range of International Classification of Disease, ninth revision, diagnostic codes indicative of an HF event and occurring during 2015 was drawn for this study. A randomly selected set of 394 hospitalisations was used as the derivation dataset and 406 hospitalisations was used as the validation dataset.

INTERVENTION

Use of NLP on free-text clinical notes and reports to ascertain Framingham HF criteria and phenotype.

PRIMARY AND SECONDARY OUTCOME MEASURES

NLP performance as measured by sensitivity, specificity, positive-predictive value (PPV) and agreement in ascertainment of Framingham HF criteria and phenotype. Manual medical record review by trained ARIC abstractors was used as the reference standard.

RESULTS

Overall, performance of NLP ascertainment of Framingham HF phenotype in the validation dataset was good, with 78.8%, 81.7%, 84.4% and 80.0% for sensitivity, specificity, PPV and agreement, respectively.

CONCLUSIONS

By decreasing the need for manual chart review, our results on the use of NLP to ascertain Framingham HF phenotype from free-text electronic health record data suggest that validated NLP technology holds the potential for significantly improving the feasibility and efficiency of conducting large-scale epidemiologic surveillance of HF prevalence and incidence.

A Mendelian Randomization Approach Using 3-HMG-Coenzyme-A Reductase Gene Variation to Evaluate the Association of Statin-Induced Low-Density Lipoprotein Cholesterol Lowering With Noncardiovascular Disease Phenotypes

IMPORTANCE

Observational studies suggest that statins, which inhibit 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase, may be associated with beneficial effects in many noncardiovascular diseases.

OBJECTIVE

To construct a weighted HMG-CoA reductase (HMGCR) gene genetic risk score (GRS) using variants in the HMGCR gene affecting low-density lipoprotein cholesterol as an instrumental variable for mendelian randomization analyses to test associations with candidate noncardiovascular phenotypes previously associated with statin use in observational studies.

DESIGN, SETTING, AND PARTICIPANTS

This cohort study included 53 385 unrelated adults of European ancestry with genome-wide genotypes available from BioVU (a practice-based biobank, used for discovery) and 30 444 unrelated adults with European ancestry available in the Electronic Medical Records and Genomics (eMERGE; a research consortium that conducts genetic research using electronic medical records, used for replication). The study was conducted from February 6, 2015, through April 31, 2019; data analysis was performed from August 26, 2019, through December 22, 2020.

INTERVENTIONS

An HMGCR GRS was calculated.

MAIN OUTCOMES AND MEASURES

The association between the HMGCR GRS and the presence or absence of 22 noncardiovascular phenotypes previously associated with statin use in clinical studies.

RESULTS

Of the 53 385 individuals in BioVU, 29 958 (56.1%) were women; mean (SD) age was 59.9 (15.6) years. The finding between the HMGCR GRS and the noncardiovascular phenotypes of interest in this cohort was significant only for type 2 diabetes. An HMGCR GRS equivalent to a 10-mg/dL decrease in the low-density lipoprotein cholesterol level was associated with an increased risk of type 2 diabetes (odds ratio [OR], 1.09; 95% CI, 1.04-1.15; P = 5.58 × 10-4). The HMGCR GRS was not associated with other phenotypes; the closest were increased risk of Parkinson disease (OR, 1.30; 95% CI, 1.07-1.58; P = .007) and kidney failure (OR, 1.18; 95% CI, 1.05-1.34; P = .008). Of the 30 444 individuals in eMERGE, 16 736 (55.0%) were women; mean (SD) age was 68.7 (15.4) years. The association between the HMGCR GRS and type 2 diabetes was replicated in this cohort (OR, 1.09; 95% CI, 1.01-1.17; P = .02); however, the HMGCR GRS was not associated with Parkinson disease (OR, 0.93; 95% CI, 0.75-1.16; P = .53) and kidney failure (OR, 1.18; 95% CI, 0.98-1.41; P = .08) in the eMERGE cohort.

CONCLUSIONS AND RELEVANCE

A mendelian randomization approach using variants in the HMGCR gene replicated the association between statin use and increased type 2 diabetes risk but provided no strong evidence for pleiotropic effects of statin-induced decrease of the low-density lipoprotein cholesterol level on other diseases.

Development of an Algorithm to Identify Cases of Nonalcoholic Steatohepatitis Cirrhosis in the Electronic Health Record

BACKGROUND AND AIMS

Current genetic research of nonalcoholic steatohepatitis (NASH) cirrhosis is limited by our ability to accurately identify cases on a large scale. Our objective was to develop and validate an electronic health record (EHR) algorithm to accurately identify cases of NASH cirrhosis in the EHR.

METHODS

We used Clinical Query 2, a search tool at Beth Israel Deaconess Medical Center, to create a pool of potential NASH cirrhosis cases (n = 5415). We created a training set of 300 randomly selected patients for chart review to confirm cases of NASH cirrhosis. Test characteristics of different algorithms, consisting of diagnosis codes, laboratory values, anthropomorphic measurements, and medication records, were calculated. The algorithms with the highest positive predictive value (PPV) and the highest F score with a PPV ≥ 80% were selected for internal validation using a separate random set of 100 patients from the potential NASH cirrhosis pool. These were then externally validated in another random set of 100 individuals using the research patient data registry tool at Massachusetts General Hospital.

RESULTS

The algorithm with the highest PPV of 100% on internal validation and 92% on external validation consisted of ≥ 3 counts of "cirrhosis, no mention of alcohol" (571.5, K74.6) and ≥ 3 counts of "nonalcoholic fatty liver" (571.8-571.9, K75.81, K76.0) codes in the absence of any diagnosis codes for other common causes of chronic liver disease.

CONCLUSIONS

We developed and validated an EHR algorithm using diagnosis codes that accurately identifies patients with NASH cirrhosis.

Sleep health, diseases, and pain syndromes: findings from an electronic health record biobank

STUDY OBJECTIVES

Implementation of electronic health record biobanks has facilitated linkage between clinical and questionnaire data and enabled assessments of relationships between sleep health and diseases in phenome-wide association studies (PheWAS). In the Mass General Brigham Biobank, a large health system-based study, we aimed to systematically catalog associations between time in bed, sleep timing, and weekly variability with clinical phenotypes derived from ICD-9/10 codes.

METHODS

Self-reported habitual bed and wake times were used to derive variables: short (<7 hours) and long (≥9 hours) time in bed, sleep midpoint, social jetlag, and sleep debt. Logistic regression and Cox proportional hazards models were used to test cross-sectional and prospective associations, respectively, adjusted for age, gender, race/ethnicity, and employment status and further adjusted for body mass index.

RESULTS

In cross-sectional analysis (n = 34,651), sleep variable associations were most notable for circulatory system, mental disorders, and endocrine/metabolic phenotypes. We observed the strongest associations for short time in bed with obesity, for long time in bed and sleep midpoint with major depressive disorder, for social jetlag with hypercholesterolemia, and for sleep debt with acne. In prospective analysis (n = 24,065), we observed short time in bed associations with higher incidence of acute pain and later sleep midpoint and higher sleep debt and social jetlag associations with higher incidence of major depressive disorder.

CONCLUSIONS

Our analysis reinforced that sleep health is a multidimensional construct, corroborated robust known findings from traditional cohort studies, and supported the application of PheWAS as a promising tool for advancing sleep research. Considering the exploratory nature of PheWAS, careful interrogation of novel findings is imperative.

Electronic Medical Record-Based Case Phenotyping for the Charlson Conditions: Scoping Review

BACKGROUND

Electronic medical records (EMRs) contain large amounts of rich clinical information. Developing EMR-based case definitions, also known as EMR phenotyping, is an active area of research that has implications for epidemiology, clinical care, and health services research.

OBJECTIVE

This review aims to describe and assess the present landscape of EMR-based case phenotyping for the Charlson conditions.

METHODS

A scoping review of EMR-based algorithms for defining the Charlson comorbidity index conditions was completed. This study covered articles published between January 2000 and April 2020, both inclusive. Embase (Excerpta Medica database) and MEDLINE (Medical Literature Analysis and Retrieval System Online) were searched using keywords developed in the following 3 domains: terms related to EMR, terms related to case finding, and disease-specific terms. The manuscript follows the Preferred Reporting Items for Systematic reviews and Meta-analyses extension for Scoping Reviews (PRISMA) guidelines.

RESULTS

A total of 274 articles representing 299 algorithms were assessed and summarized. Most studies were undertaken in the United States (181/299, 60.5%), followed by the United Kingdom (42/299, 14.0%) and Canada (15/299, 5.0%). These algorithms were mostly developed either in primary care (103/299, 34.4%) or inpatient (168/299, 56.2%) settings. Diabetes, congestive heart failure, myocardial infarction, and rheumatology had the highest number of developed algorithms. Data-driven and clinical rule-based approaches have been identified. EMR-based phenotype and algorithm development reflect the data access allowed by respective health systems, and algorithms vary in their performance.

CONCLUSIONS

Recognizing similarities and differences in health systems, data collection strategies, extraction, data release protocols, and existing clinical pathways is critical to algorithm development strategies. Several strategies to assist with phenotype-based case definitions have been proposed.

Depression in multiple sclerosis patients associated with risk variant near NEGR1

BACKGROUND

A substantial number of patients diagnosed with multiple sclerosis (MS) suffer from depression in addition to physical symptoms and disability. Recent evidence suggests a stronger relationship may exist between MS and depression than previously thought, in which a diagnosis of depression may be prodromic to the development of MS.

METHODS

A genome-wide association study (GWAS) was performed to identify genetic variants associated with the development of depression in a cohort of MS patients. The control group (n = 1180) was composed of MS patients with no diagnoses of depression as determined by ICD-9 and ICD-10 billing codes present in the electronic health record (EHR). Separate analyses were performed for three different case groups: 1) MS patients having a depression diagnosis at any time (n = 182), 2) MS patients having a depression diagnosis one year pre-MS diagnosis (n = 27), and 3) MS patients having a depression diagnosis one year post-MS diagnosis (n = 130). Logistic regression analyses were also performed to test for associations between the development of depression and an APOE tagging variant, as APOE was previously linked to depressive affect in MS. An additional logistic regression analysis tested for associations between depression in MS patients and SNPs associated with depression in the general population. Pathway enrichment analyses were also conducted to identify pathways that link the two diseases.

RESULTS

GWAS identified no novel associations between variants and a diagnosis of depression relative to a diagnosis of MS. One variant, rs1432639, associated with depression in the general population, was significantly associated with the development of depression post-MS diagnosis. The APOE-related SNPs were not associated with depression in this study population. An IGF1 pathway approached statistical significance in patients diagnosed with depression prior to a diagnosis of MS.

CONCLUSION

rs1432639 and the IGF1 pathway provide evidence for a genetic link between MS and depression that warrants further research.

Development and evaluation of an EHR-based computable phenotype for identification of pediatric Crohn's disease patients in a National Pediatric Learning Health System

Objectives

To develop and evaluate the classification accuracy of a computable phenotype for pediatric Crohn's disease using electronic health record data from PEDSnet, a large, multi‐institutional research network and Learning Health System.

Study Design

Using clinician and informatician input, algorithms were developed using combinations of diagnostic and medication data drawn from the PEDSnet clinical dataset which is comprised of 5.6 million children from eight U.S. academic children's health systems. Six test algorithms (four cases, two non‐cases) that combined use of specific medications for Crohn's disease plus the presence of Crohn's diagnosis were initially tested against the entire PEDSnet dataset. From these, three were selected for performance assessment using manual chart review (primary case algorithm, n = 360, primary non‐case algorithm, n = 360, and alternative case algorithm, n = 80). Non‐cases were patients having gastrointestinal diagnoses other than inflammatory bowel disease. Sensitivity, specificity, and positive predictive value (PPV) were assessed for the primary case and primary non‐case algorithms.

Results

Of the six algorithms tested, the least restrictive algorithm requiring just ≥1 Crohn's diagnosis code yielded 11 950 cases across PEDSnet (prevalence 21/10 000). The most restrictive algorithm requiring ≥3 Crohn's disease diagnoses plus at least one medication yielded 7868 patients (prevalence 14/10 000). The most restrictive algorithm had the highest PPV (95%) and high sensitivity (91%) and specificity (94%). False positives were due primarily to a diagnosis reversal (from Crohn's disease to ulcerative colitis) or having a diagnosis of “indeterminate colitis.” False negatives were rare.

Conclusions

Using diagnosis codes and medications available from PEDSnet, we developed a computable phenotype for pediatric Crohn's disease that had high specificity, sensitivity and predictive value. This process will be of use for developing computable phenotypes for other pediatric diseases, to facilitate cohort identification for retrospective and prospective studies, and to optimize clinical care through the PEDSnet Learning Health System.

Use of ICD-10 diagnosis codes to identify seropositive and seronegative rheumatoid arthritis when lab results are not available

BACKGROUND

Rheumatoid factor (RF) and anti-cyclic citrullinated peptide (CCP) antibody tests are often measured at the time of rheumatoid arthritis (RA) diagnosis but may not be repeated and therefore not available in electronic health record (EHR) data; lab test results are unavailable in most administrative claims databases. ICD10 coding allows discrimination between rheumatoid factor positive (M05) ("seropositive") and seronegative (M06) patients, but the validity of these codes has not been examined.

METHODS

Using the ACR's Rheumatology Informatics System for Effectiveness (RISE) EHR-based registry and U.S. MarketScan data where some patients have lab test results, we assembled two cohorts. Seropositive RA was defined having a M05 diagnosis code on the second rheumatologist encounter, M06 similarly identified seronegative RA, and RF and anti-CCP lab test results were the gold standard. We calculated sensitivity (Se) and positive predicted value (PPV) of the M05/M06 diagnosis codes.

RESULTS

We identified 43,581 eligible RA patients (RISE) and 1185 (MarketScan) with RF or anti-CCP lab test results available. Using M05 as the proxy for seropositive RA, sensitivity = 0.76, PPV = 0.82 in RISE, and Se = 0.73, PPV = 0.84 in MarketScan. Results for M06 as a proxy for seronegative RA were comparable in RISE, albeit somewhat lower in MarketScan. Over 3 consecutive visits, approximately 90% of RA patients were coded consistently using either M05 or M06 at each visit.

CONCLUSION

Under ICD10, M05 and M06 diagnosis codes are reasonable proxies to identify seropositive and seronegative RA with high sensitivity and positive predictive values if lab test results are not available.

Implicit bias of encoded variables: frameworks for addressing structured bias in EHR-GWAS data

The 'discovery' stage of genome-wide association studies required amassing large, homogeneous cohorts. In order to attain clinically useful insights, we must now consider the presentation of disease within our clinics and, by extension, within our medical records. Large-scale use of electronic health record (EHR) data can help to understand phenotypes in a scalable manner, incorporating lifelong and whole-phenome context. However, extending analyses to incorporate EHR and biobank-based analyses will require careful consideration of phenotype definition. Judgements and clinical decisions that occur 'outside' the system inevitably contain some degree of bias and become encoded in EHR data. Any algorithmic approach to phenotypic characterization that assumes non-biased variables will generate compounded biased conclusions. Here, we discuss and illustrate potential biases inherent within EHR analyses, how these may be compounded across time and suggest frameworks for large-scale phenotypic analysis to minimize and uncover encoded bias.

Initial changes in peak aortic jet velocity and mean gradient predict progression to severe aortic stenosis

Background

There is significant interindividual variability in the rate of aortic stenosis (AS) progression that is not accounted for in the current surveillance algorithms. We sought to examine the association between changes in peak aortic jet velocity (Vmax) and mean gradient (MG) among patients with mild or moderate AS and risk of progression to severe disease.

Methods

Adult subjects referred for echocardiography at a single academic referral center with a diagnosis of mild or moderate AS and ≥2 additional surveillance echocardiograms were included in the study. Changes in Vmax and MG between the first two echocardiograms were indexed to time and tested for association with future progression to severe AS.

Results

Among three hundred and sixty-four subjects, the median time between first and second echocardiograms was 1.3 years and initial changes in Vmax and MG indexed to time were +0.16 m/s per year and +1.44 mmHg per year, respectively. Fifty-three (15%) and fifty-six (15%) subjects progressed to severe AS defined by Vmax and MG, respectively. In multivariable logistic regression, initial increase in Vmax (OR = 4.19, 95% CI 1.93–9.10, p < 0.001) and initial increase in MG (OR = 1.12, 95% CI 1.06–1.18, p < 0.001) were associated with progression to severe AS.

Conclusions

Initial changes in Vmax and MG among patients with mild or moderate AS are strongly associated with risk of progression to severe AS and may help guide individualized surveillance strategies.

The Role of Bone Morphogenetic Protein Signaling in Non-Alcoholic Fatty Liver Disease

Non-alcoholic fatty liver disease (NAFLD) affects over 30% of adults in the United States. Bone morphogenetic protein (BMP) signaling is known to contribute to hepatic fibrosis, but the role of BMP signaling in the development of NAFLD is unclear. In this study, treatment with either of two BMP inhibitors reduced hepatic triglyceride content in diabetic (db/db) mice. BMP inhibitor-induced decrease in hepatic triglyceride levels was associated with decreased mRNA encoding Dgat2, an enzyme integral to triglyceride synthesis. Treatment of hepatoma cells with BMP2 induced DGAT2 expression and activity via intracellular SMAD signaling. In humans we identified a rare missense single nucleotide polymorphism in the BMP type 1 receptor ALK6 (rs34970181;R371Q) associated with a 2.1-fold increase in the prevalence of NAFLD. In vitro analyses revealed R371Q:ALK6 is a previously unknown constitutively active receptor. These data show that BMP signaling is an important determinant of NAFLD in a murine model and is associated with NAFLD in humans.

Impact of Diverse Data Sources on Computational Phenotyping

Electronic health records (EHRs) are widely adopted with a great potential to serve as a rich, integrated source of phenotype information. Computational phenotyping, which extracts phenotypes from EHR data automatically, can accelerate the adoption and utilization of phenotype-driven efforts to advance scientific discovery and improve healthcare delivery. A list of computational phenotyping algorithms has been published but data fragmentation, i.e., incomplete data within one single data source, has been raised as an inherent limitation of computational phenotyping. In this study, we investigated the impact of diverse data sources on two published computational phenotyping algorithms, rheumatoid arthritis (RA) and type 2 diabetes mellitus (T2DM), using Mayo EHRs and Rochester Epidemiology Project (REP) which links medical records from multiple health care systems. Results showed that both RA (less prevalent) and T2DM (more prevalent) case selections were markedly impacted by data fragmentation, with positive predictive value (PPV) of 91.4 and 92.4%, false-negative rate (FNR) of 26.6 and 14% in Mayo data, respectively, PPV of 97.2 and 98.3%, FNR of 5.2 and 3.3% in REP. T2DM controls also contain biases, with PPV of 91.2% and FNR of 1.2% for Mayo. We further elaborated underlying reasons impacting the performance.

A bibliometric analysis and visualization of medical data mining research

BACKGROUND

Data mining technology used in the field of medicine has been widely studied by scholars all over the world. But there is little research on medical data mining (MDM) from the perspectives of bibliometrics and visualization, and the research topics and development trends in this field are still unclear.

METHODS

This paper has applied bibliometric visualization software tools, VOSviewer 1.6.10 and CiteSpace V, to study the citation characteristics, international cooperation, author cooperation, and geographical distribution of the MDM.

RESULTS

A total of 1575 documents are obtained, and the most frequent document type is article (1376). SHAN NH is the most productive author, with the highest number of publications of 12, and the Gillies's article (750 times citation) is the most cited paper. The most productive country and institution in MDM is the USA (559) and US FDA (35), respectively. The Journal of Biomedical Informatics, Expert Systems with Applications and Journal of Medical Systems are the most productive journals, which reflected the nature of the research, and keywords "classification (790)" and "system (576)" have the strongest strength. The hot topics in MDM are drug discovery, medical imaging, vaccine safety, and so on. The 3 frontier topics are reporting system, precision medicine, and inflammation, and would be the foci of future research.

CONCLUSION

The present study provides a panoramic view of data mining methods applied in medicine by visualization and bibliometrics. Analysis of authors, journals, institutions, and countries could provide reference for researchers who are fresh to the field in different ways. Researchers may also consider the emerging trends when deciding the direction of their study.

Heart Failure Risk Associated With Rheumatoid Arthritis-Related Chronic Inflammation

Background

Inflammation may contribute to incident heart failure (HF). Rheumatoid arthritis (RA), a prototypic inflammatory condition, may serve as a model for understanding inflammation‐related HF risk.

Methods and Results

Using the Vanderbilt University Medical Center electronic health record, we retrospectively identified 9889 patients with RA and 9889 control patients without autoimmune disease matched for age, sex, and race. Prevalent HF at entry into the electronic health record or preceding RA diagnosis was excluded. Incident HF was ascertained using International Classification of Diseases, Ninth Revision (ICD‐9), codes and medications. Over 177 566 person‐years of follow‐up, patients with RA were at 21% greater risk of HF (95% CI, 3–42%) independent of traditional cardiovascular risk factors. Among patients with RA, higher CRP (C‐reactive protein) was associated with greater HF risk (P<0.001), while the anti‐inflammatory drug methotrexate was associated with ≈25% lower HF risk (P=0.021). In a second cohort (n=115) of prospectively enrolled patients with and without RA, we performed proteomics and cardiac magnetic resonance imaging to discover circulating markers of inflammation associated with cardiac structure and function. Artemin levels were higher in patients with RA compared with controls (P=0.009), and higher artemin levels were associated with worse ventricular end‐systolic elastance and ventricular‐vascular coupling ratio (P=0.044 and P=0.031, respectively).

Conclusions

RA, a prototypic chronic inflammatory condition, is associated with increased risk of HF. Among patients with RA, higher levels of CRP were associated with greater HF risk, while methotrexate was associated with lower risk.

Pleiotropy-Based Decomposition of Genetic Risk Scores: Association and Interaction Analysis for Type 2 Diabetes and CAD

Genetic risk for a disease in the population may be represented as a genetic risk score (GRS) constructed as the sum of inherited risk alleles, weighted by allelic effects established in an independent population. While this formulation captures overall genetic risk, it typically does not address risk due to specific biological mechanisms or pathways that may nevertheless be important for interpretation or treatment response. Here, a GRS for disease is resolved into independent or nearly independent components pertaining to biological mechanisms inferred from pleiotropic relationships. The component GRSs' weights are derived from the singular value decomposition (SVD) of the matrix of appropriately scaled genetic effects, i.e., beta coefficients, of the disease variants across a panel of the disease-related phenotypes. The SVD-based formalism also associates combinations of disease-related phenotypes with inferred disease pathways. Applied to incident type 2 diabetes (T2D) in the Women's Genome Health Study (N = 23,294), component GRSs discriminate glycemic control and lipid-based genetic risk, while revealing significant interactions between specific components and BMI or physical activity, the latter not observed with a GRS for overall T2D genetic liability. Applied to coronary artery disease (CAD) in both the WGHS and in JUPITER (N = 8,749), a randomized trial of rosuvastatin for primary prevention of CVD, component GRSs discriminate genetic risk associated with LDL-C from risk associated with reciprocal genetic effects on triglycerides and HDL-C. They also inform the pharmacogenetics of statin treatment by demonstrating that benefit from rosuvastatin is as strongly related to genetic risk from triglycerides and HDL-C as from LDL-C.

Family Study Designs Informed by Tumor Heterogeneity and Multi-Cancer Pleiotropies: The Power of the Utah Population Database

BACKGROUND

Previously, family-based designs and high-risk pedigrees have illustrated value for the discovery of high- and intermediate-risk germline breast cancer susceptibility genes. However, genetic heterogeneity is a major obstacle hindering progress. New strategies and analytic approaches will be necessary to make further advances. One opportunity with the potential to address heterogeneity via improved characterization of disease is the growing availability of multisource databases. Specific to advances involving family-based designs are resources that include family structure, such as the Utah Population Database (UPDB). To illustrate the broad utility and potential power of multisource databases, we describe two different novel family-based approaches to reduce heterogeneity in the UPDB.

METHODS

Our first approach focuses on using pedigree-informed breast tumor phenotypes in gene mapping. Our second approach focuses on the identification of families with similar pleiotropies. We use a novel network-inspired clustering technique to explore multi-cancer signatures for high-risk breast cancer families.

RESULTS

Our first approach identifies a genome-wide significant breast cancer locus at 2q13 [P = 1.6 × 10-8, logarithm of the odds (LOD) equivalent 6.64]. In the region, IL1A and IL1B are of particular interest, key cytokine genes involved in inflammation. Our second approach identifies five multi-cancer risk patterns. These clusters include expected coaggregations (such as breast cancer with prostate cancer, ovarian cancer, and melanoma), and also identify novel patterns, including coaggregation with uterine, thyroid, and bladder cancers.

CONCLUSIONS

Our results suggest pedigree-informed tumor phenotypes can map genes for breast cancer, and that various different cancer pleiotropies exist for high-risk breast cancer pedigrees.

IMPACT

Both methods illustrate the potential for decreasing etiologic heterogeneity that large, population-based multisource databases can provide.See all articles in this CEBP Focus section, "Modernizing Population Science."

The emerging landscape of health research based on biobanks linked to electronic health records: Existing resources, statistical challenges, and potential opportunities

Biobanks linked to electronic health records provide rich resources for health-related research. With improvements in administrative and informatics infrastructure, the availability and utility of data from biobanks have dramatically increased. In this paper, we first aim to characterize the current landscape of available biobanks and to describe specific biobanks, including their place of origin, size, and data types. The development and accessibility of large-scale biorepositories provide the opportunity to accelerate agnostic searches, expedite discoveries, and conduct hypothesis-generating studies of disease-treatment, disease-exposure, and disease-gene associations. Rather than designing and implementing a single study focused on a few targeted hypotheses, researchers can potentially use biobanks' existing resources to answer an expanded selection of exploratory questions as quickly as they can analyze them. However, there are many obvious and subtle challenges with the design and analysis of biobank-based studies. Our second aim is to discuss statistical issues related to biobank research such as study design, sampling strategy, phenotype identification, and missing data. We focus our discussion on biobanks that are linked to electronic health records. Some of the analytic issues are illustrated using data from the Michigan Genomics Initiative and UK Biobank, two biobanks with two different recruitment mechanisms. We summarize the current body of literature for addressing these challenges and discuss some standing open problems. This work complements and extends recent reviews about biobank-based research and serves as a resource catalog with analytical and practical guidance for statisticians, epidemiologists, and other medical researchers pursuing research using biobanks.

An augmented estimation procedure for EHR-based association studies accounting for differential misclassification

OBJECTIVES

The ability to identify novel risk factors for health outcomes is a key strength of electronic health record (EHR)-based research. However, the validity of such studies is limited by error in EHR-derived phenotypes. The objective of this study was to develop a novel procedure for reducing bias in estimated associations between risk factors and phenotypes in EHR data.

MATERIALS AND METHODS

The proposed method combines the strengths of a gold-standard phenotype obtained through manual chart review for a small validation set of patients and an automatically-derived phenotype that is available for all patients but is potentially error-prone (hereafter referred to as the algorithm-derived phenotype). An augmented estimator of associations is obtained by optimally combining these 2 phenotypes. We conducted simulation studies to evaluate the performance of the augmented estimator and conducted an analysis of risk factors for second breast cancer events using data on a cohort from Kaiser Permanente Washington.

RESULTS

The proposed method was shown to reduce bias relative to an estimator using only the algorithm-derived phenotype and reduce variance compared to an estimator using only the validation data.

DISCUSSION

Our simulation studies and real data application demonstrate that, compared to the estimator using validation data only, the augmented estimator has lower variance (ie, higher statistical efficiency). Compared to the estimator using error-prone EHR-derived phenotypes, the augmented estimator has smaller bias.

CONCLUSIONS

The proposed estimator can effectively combine an error-prone phenotype with gold-standard data from a limited chart review in order to improve analyses of risk factors using EHR data.

Phenome-based approach identifies RIC1-linked Mendelian syndrome through zebrafish models, biobank associations and clinical studies

Discovery of genotype-phenotype relationships remains a major challenge in clinical medicine. Here, we combined three sources of phenotypic data to uncover a novel mechanism for rare and common diseases resulting from collagen secretion deficits. Using zebrafish genetic screen, we identified the ric1 gene to be essential for skeletal biology. Using a gene-based phenome-wide association study (PheWAS) in the EHR-linked BioVU biobank, we show that reduced genetically determined expression of RIC1 is associated with musculoskeletal and dental conditions. Whole exome sequencing (WES) identified individuals homozygous-by-descent for a rare variant in RIC1, and, through a guided clinical re-evaluation, they were discovered to share signs with the BioVU-associated phenome. We named this novel Mendelian syndrome CATIFA (Cleft lip, cAtaract, Tooth abnormality, Intellectual disability, Facial dysmorphism, ADHD), and revealed further disease mechanisms. This gene-based PheWAS-guided approach can accelerate the discovery of clinically relevant disease phenome and associated biological mechanisms.

Association of Genetic Risk of Obesity with Postoperative Complications Using Mendelian Randomization

BACKGROUND

The extent to which obesity and genetics determine postoperative complications is incompletely understood.

METHODS

We performed a retrospective study using two population cohorts with electronic health record (EHR) data. The first included 736,726 adults with body mass index (BMI) recorded between 1990 and 2017 at Vanderbilt University Medical Center. The second cohort consisted of 65,174 individuals from 12 institutions contributing EHR and genome-wide genotyping data to the Electronic Medical Records and Genomics (eMERGE) Network. Pairwise logistic regression analyses were used to measure the association of BMI categories with postoperative complications derived from International Classification of Disease-9 codes, including postoperative infection, incisional hernia, and intestinal obstruction. A genetic risk score was constructed from 97 obesity-risk single-nucleotide polymorphisms for a Mendelian randomization study to determine the association of genetic risk of obesity on postoperative complications. Logistic regression analyses were adjusted for sex, age, site, and race/principal components.

RESULTS

Individuals with overweight or obese BMI (≥25 kg/m2) had increased risk of incisional hernia (odds ratio [OR] 1.7-5.5, p < 3.1 × 10-20), and people with obesity (BMI ≥ 30 kg/m2) had increased risk of postoperative infection (OR 1.2-2.3, p < 2.5 × 10-5). In the eMERGE cohort, genetically predicted BMI was associated with incisional hernia (OR 2.1 [95% CI 1.8-2.5], p = 1.4 × 10-6) and postoperative infection (OR 1.6 [95% CI 1.4-1.9], p = 3.1 × 10-6). Association findings were similar after limitation of the cohorts to those who underwent abdominal procedures.

CONCLUSIONS

Clinical and Mendelian randomization studies suggest that obesity, as measured by BMI, is associated with the development of postoperative incisional hernia and infection.

PheProb: probabilistic phenotyping using diagnosis codes to improve power for genetic association studies

Objective

Standard approaches for large scale phenotypic screens using electronic health record (EHR) data apply thresholds, such as ≥2 diagnosis codes, to define subjects as having a phenotype. However, the variation in the accuracy of diagnosis codes can impair the power of such screens. Our objective was to develop and evaluate an approach which converts diagnosis codes into a probability of a phenotype (PheProb). We hypothesized that this alternate approach for defining phenotypes would improve power for genetic association studies.

Methods

The PheProb approach employs unsupervised clustering to separate patients into 2 groups based on diagnosis codes. Subjects are assigned a probability of having the phenotype based on the number of diagnosis codes. This approach was developed using simulated EHR data and tested in a real world EHR cohort. In the latter, we tested the association between low density lipoprotein cholesterol (LDL-C) genetic risk alleles known for association with hyperlipidemia and hyperlipidemia codes (ICD-9 272.x). PheProb and thresholding approaches were compared.

Results

Among n = 1462 subjects in the real world EHR cohort, the threshold-based p-values for association between the genetic risk score (GRS) and hyperlipidemia were 0.126 (≥1 code), 0.123 (≥2 codes), and 0.142 (≥3 codes). The PheProb approach produced the expected significant association between the GRS and hyperlipidemia: p = .001.

Conclusions

PheProb improves statistical power for association studies relative to standard thresholding approaches by leveraging information about the phenotype in the billing code counts. The PheProb approach has direct applications where efficient approaches are required, such as in Phenome-Wide Association Studies.