Assessing Retrieval-Augmented Large Language Model Performance in Emergency Department ICD-10-CM Coding Compared to Human Coders.

Background: Accurate medical coding is essential for clinical and administrative purposes but complicated, time-consuming, and biased. This study compares Retrieval-Augmented Generation (RAG)-enhanced LLMs to provider-assigned codes in producing ICD-10-CM codes from emergency department (ED) clinical records. Methods: Retrospective cohort study using 500 ED visits randomly selected from the Mount Sinai Health System between January and April 2024. The RAG system integrated past 1,038,066 ED visits data (2021-2023) into the LLMs' predictions to improve coding accuracy. Nine commercial and open-source LLMs were evaluated. The primary outcome was a head-to-head comparison of the ICD-10-CM codes generated by the RAG-enhanced LLMs and those assigned by the original providers. A panel of four physicians and two LLMs blindly reviewed the codes, comparing the RAG-enhanced LLM and provider-assigned codes on accuracy and specificity. Findings: RAG-enhanced LLMs demonstrated superior performance to provider coders in both the accuracy and specificity of code assignments. In a targeted evaluation of 200 cases where discrepancies existed between GPT-4 and provider-assigned codes, human reviewers favored GPT-4 for accuracy in 447 instances, compared to 277 instances where providers' codes were preferred (p<0.001). Similarly, GPT-4 was selected for its superior specificity in 509 cases, whereas human coders were preferred in only 181 cases (p<0.001). Smaller open-access models, such as Llama-3.1-70B, also demonstrated substantial scalability when enhanced with RAG, with 218 instances of accuracy preference compared to 90 for providers' codes. Furthermore, across all models, the exact match rate between LLM-generated and provider-assigned codes significantly improved following RAG integration, with Qwen-2-7B increasing from 0.8% to 17.6% and Gemma-2-9b-it improving from 7.2% to 26.4%. Interpretation: RAG-enhanced LLMs improve medical coding accuracy in EDs, suggesting clinical workflow applications. These findings show that generative AI can improve clinical outcomes and reduce administrative burdens. Funding: This work was supported in part through the computational and data resources and staff expertise provided by Scientific Computing and Data at the Icahn School of Medicine at Mount Sinai and supported by the Clinical and Translational Science Awards (CTSA) grant UL1TR004419 from the National Center for Advancing Translational Sciences. Research reported in this publication was also supported by the Office of Research Infrastructure of the National Institutes of Health under award number S10OD026880 and S10OD030463. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The funders played no role in study design, data collection, analysis and interpretation of data, or the writing of this manuscript. Twitter Summary: A study showed AI models with retrieval-augmented generation outperformed human doctors in ED diagnostic coding accuracy and specificity. Even smaller AI models perform favorably when using RAG. This suggests potential for reducing administrative burden in healthcare, improving coding efficiency, and enhancing clinical documentation.

Extracting social support and social isolation information from clinical psychiatry notes: comparing a rule-based natural language processing system and a large language model.

OBJECTIVES: Social support (SS) and social isolation (SI) are social determinants of health (SDOH) associated with psychiatric outcomes. In electronic health records (EHRs), individual-level SS/SI is typically documented in narrative clinical notes rather than as structured coded data. Natural language processing (NLP) algorithms can automate the otherwise labor-intensive process of extraction of such information. MATERIALS AND METHODS: Psychiatric encounter notes from Mount Sinai Health System (MSHS, n = 300) and Weill Cornell Medicine (WCM, n = 225) were annotated to create a gold-standard corpus. A rule-based system (RBS) involving lexicons and a large language model (LLM) using FLAN-T5-XL were developed to identify mentions of SS and SI and their subcategories (eg, social network, instrumental support, and loneliness). RESULTS: For extracting SS/SI, the RBS obtained higher macroaveraged F1-scores than the LLM at both MSHS (0.89 versus 0.65) and WCM (0.85 versus 0.82). For extracting the subcategories, the RBS also outperformed the LLM at both MSHS (0.90 versus 0.62) and WCM (0.82 versus 0.81). DISCUSSION AND CONCLUSION: Unexpectedly, the RBS outperformed the LLMs across all metrics. An intensive review demonstrates that this finding is due to the divergent approach taken by the RBS and LLM. The RBS was designed and refined to follow the same specific rules as the gold-standard annotations. Conversely, the LLM was more inclusive with categorization and conformed to common English-language understanding. Both approaches offer advantages, although additional replication studies are warranted.

Clinical associations with treatment resistance in depression: An electronic health record study.

Treatment resistance is common in major depressive disorder (MDD), yet clinical risk factors are not well understood. Using a discovery-replication design, we conducted phenome-wide association studies (PheWASs) of MDD treatment resistance in two electronic health record (EHR)-linked biobanks. The PheWAS included participants with an MDD diagnosis in the EHR and at least one antidepressant (AD) prescription. Participant lifetime diagnoses were mapped to phecodes. PheWASs were conducted for three treatment resistance outcomes based on AD prescription data: number of unique ADs prescribed, ≥1 and ≥2 CE switches. Of the 180 phecodes significantly associated with these outcomes in the discovery cohort (n = 12,558), 71 replicated (n = 8,206). In addition to identifying known clinical factors for treatment resistance in MDD, the total unique AD prescriptions was associated with additional clinical variables including irritable bowel syndrome, gastroesophageal reflux disease, symptomatic menopause, and spondylosis. We calculated polygenic risk of specific-associated conditions and tested their association with AD outcomes revealing that genetic risk for many of these conditions is also associated with the total unique AD prescriptions. The number of unique ADs prescribed, which is easily assessed in EHRs, provides a more nuanced measure of treatment resistance, and may facilitate future research and clinical application in this area.

Derivation, external and clinical validation of a deep learning approach for detecting intracranial hypertension.

Increased intracranial pressure (ICP) ≥15 mmHg is associated with adverse neurological outcomes, but needs invasive intracranial monitoring. Using the publicly available MIMIC-III Waveform Database (2000-2013) from Boston, we developed an artificial intelligence-derived biomarker for elevated ICP (aICP) for adult patients. aICP uses routinely collected extracranial waveform data as input, reducing the need for invasive monitoring. We externally validated aICP with an independent dataset from the Mount Sinai Hospital (2020-2022) in New York City. The AUROC, accuracy, sensitivity, and specificity on the external validation dataset were 0.80 (95% CI, 0.80-0.80), 73.8% (95% CI, 72.0-75.6%), 73.5% (95% CI 72.5-74.5%), and 73.0% (95% CI, 72.0-74.0%), respectively. We also present an exploratory analysis showing aICP predictions are associated with clinical phenotypes. A ten-percentile increment was associated with brain malignancy (OR = 1.68; 95% CI, 1.09-2.60), intracerebral hemorrhage (OR = 1.18; 95% CI, 1.07-1.32), and craniotomy (OR = 1.43; 95% CI, 1.12-1.84; P < 0.05 for all).

Development and Validation of a Policy Tree Approach for Optimizing Intravenous Fluids in Critically Ill Patients with Sepsis and Acute Kidney Injury.

Purpose: Intravenous fluids are mainstay of management of acute kidney injury (AKI) after sepsis but can cause fluid overload. Recent literature shows that restrictive fluid strategy may be beneficial in some patients with AKI, however, identifying these patients is challenging. We aimed to develop and validate a machine learning algorithm to identify patients who would benefit from a restrictive fluid strategy. Methods: We included patients with sepsis who developed AKI within 48 hours of ICU admission and defined restrictive fluid strategy as receiving <500mL fluids within 24 hours after AKI. Our primary outcome was early AKI reversal within 48 hours of AKI onset, and secondary outcomes included sustained AKI reversal and major adverse kidney events (MAKE) at discharge. We used a causal forest, a machine learning algorithm to estimate individual treatment effects and policy tree algorithm to identify patients who would benefit by restrictive fluid strategy. We developed the algorithm in MIMIC-IV and validated it in eICU database. Results: Among 2,091 patients in the external validation cohort, policy tree recommended restrictive fluids for 88.2%. Among these, patients who received restrictive fluids demonstrated significantly higher rate of early AKI reversal (48.2% vs 39.6%, p<0.001), sustained AKI reversal (36.7% vs 27.4%, p<0.001) and lower rates of MAKE by discharge (29.3% vs 35.1%, p=0.019). These results were consistent in adjusted analysis. Conclusion: Policy tree based on causal machine learning can identify septic patients with AKI who benefit from a restrictive fluid strategy. This approach needs to be validated in prospective trials.

Differentiation of Prior SARS-CoV-2 Infection and Postacute Sequelae by Standard Clinical Laboratory Measurements in the RECOVER Cohort.

BACKGROUND: There are currently no validated clinical biomarkers of postacute sequelae of SARS-CoV-2 infection (PASC). OBJECTIVE: To investigate clinical laboratory markers of SARS-CoV-2 and PASC. DESIGN: Propensity score-weighted linear regression models were fitted to evaluate differences in mean laboratory measures by prior infection and PASC index (≥12 vs. 0). (ClinicalTrials.gov: NCT05172024). SETTING: 83 enrolling sites. PARTICIPANTS: RECOVER-Adult cohort participants with or without SARS-CoV-2 infection with a study visit and laboratory measures 6 months after the index date (or at enrollment if >6 months after the index date). Participants were excluded if the 6-month visit occurred within 30 days of reinfection. MEASUREMENTS: Participants completed questionnaires and standard clinical laboratory tests. RESULTS: Among 10 094 participants, 8746 had prior SARS-CoV-2 infection, 1348 were uninfected, 1880 had a PASC index of 12 or higher, and 3351 had a PASC index of zero. After propensity score adjustment, participants with prior infection had a lower mean platelet count (265.9 × 109 cells/L [95% CI, 264.5 to 267.4 × 109 cells/L]) than participants without known prior infection (275.2 × 109 cells/L [CI, 268.5 to 282.0 × 109 cells/L]), as well as higher mean hemoglobin A1c (HbA1c) level (5.58% [CI, 5.56% to 5.60%] vs. 5.46% [CI, 5.40% to 5.51%]) and urinary albumin-creatinine ratio (81.9 mg/g [CI, 67.5 to 96.2 mg/g] vs. 43.0 mg/g [CI, 25.4 to 60.6 mg/g]), although differences were of modest clinical significance. The difference in HbA1c levels was attenuated after participants with preexisting diabetes were excluded. Among participants with prior infection, no meaningful differences in mean laboratory values were found between those with a PASC index of 12 or higher and those with a PASC index of zero. LIMITATION: Whether differences in laboratory markers represent consequences of or risk factors for SARS-CoV-2 infection could not be determined. CONCLUSION: Overall, no evidence was found that any of the 25 routine clinical laboratory values assessed in this study could serve as a clinically useful biomarker of PASC. PRIMARY FUNDING SOURCE: National Institutes of Health.

Applications of large language models in psychiatry: a systematic review.

Background: With their unmatched ability to interpret and engage with human language and context, large language models (LLMs) hint at the potential to bridge AI and human cognitive processes. This review explores the current application of LLMs, such as ChatGPT, in the field of psychiatry. Methods: We followed PRISMA guidelines and searched through PubMed, Embase, Web of Science, and Scopus, up until March 2024. Results: From 771 retrieved articles, we included 16 that directly examine LLMs' use in psychiatry. LLMs, particularly ChatGPT and GPT-4, showed diverse applications in clinical reasoning, social media, and education within psychiatry. They can assist in diagnosing mental health issues, managing depression, evaluating suicide risk, and supporting education in the field. However, our review also points out their limitations, such as difficulties with complex cases and potential underestimation of suicide risks. Conclusion: Early research in psychiatry reveals LLMs' versatile applications, from diagnostic support to educational roles. Given the rapid pace of advancement, future investigations are poised to explore the extent to which these models might redefine traditional roles in mental health care.

Divergent landscapes of A-to-I editing in postmortem and living human brain.

Adenosine-to-inosine (A-to-I) editing is a prevalent post-transcriptional RNA modification within the brain. Yet, most research has relied on postmortem samples, assuming it is an accurate representation of RNA biology in the living brain. We challenge this assumption by comparing A-to-I editing between postmortem and living prefrontal cortical tissues. Major differences were found, with over 70,000 A-to-I sites showing higher editing levels in postmortem tissues. Increased A-to-I editing in postmortem tissues is linked to higher ADAR and ADARB1 expression, is more pronounced in non-neuronal cells, and indicative of postmortem activation of inflammation and hypoxia. Higher A-to-I editing in living tissues marks sites that are evolutionarily preserved, synaptic, developmentally timed, and disrupted in neurological conditions. Common genetic variants were also found to differentially affect A-to-I editing levels in living versus postmortem tissues. Collectively, these discoveries offer more nuanced and accurate insights into the regulatory mechanisms of RNA editing in the human brain.

Divergent landscapes of A-to-I editing in postmortem and living human brain.

Adenosine-to-inosine (A-to-I) editing is a prevalent post-transcriptional RNA modification within the brain. Yet, most research has relied on postmortem samples, assuming it is an accurate representation of RNA biology in the living brain. We challenge this assumption by comparing A-to-I editing between postmortem and living prefrontal cortical tissues. Major differences were found, with over 70,000 A-to-I sites showing higher editing levels in postmortem tissues. Increased A-to-I editing in postmortem tissues is linked to higher ADAR1 and ADARB1 expression, is more pronounced in non-neuronal cells, and indicative of postmortem activation of inflammation and hypoxia. Higher A-to-I editing in living tissues marks sites that are evolutionarily preserved, synaptic, developmentally timed, and disrupted in neurological conditions. Common genetic variants were also found to differentially affect A-to-I editing levels in living versus postmortem tissues. Collectively, these discoveries illuminate the nuanced functions and intricate regulatory mechanisms of RNA editing within the human brain.

Evaluating the accuracy of a state-of-the-art large language model for prediction of admissions from the emergency room.

BACKGROUND: Artificial intelligence (AI) and large language models (LLMs) can play a critical role in emergency room operations by augmenting decision-making about patient admission. However, there are no studies for LLMs using real-world data and scenarios, in comparison to and being informed by traditional supervised machine learning (ML) models. We evaluated the performance of GPT-4 for predicting patient admissions from emergency department (ED) visits. We compared performance to traditional ML models both naively and when informed by few-shot examples and/or numerical probabilities. METHODS: We conducted a retrospective study using electronic health records across 7 NYC hospitals. We trained Bio-Clinical-BERT and XGBoost (XGB) models on unstructured and structured data, respectively, and created an ensemble model reflecting ML performance. We then assessed GPT-4 capabilities in many scenarios: through Zero-shot, Few-shot with and without retrieval-augmented generation (RAG), and with and without ML numerical probabilities. RESULTS: The Ensemble ML model achieved an area under the receiver operating characteristic curve (AUC) of 0.88, an area under the precision-recall curve (AUPRC) of 0.72 and an accuracy of 82.9%. The naïve GPT-4's performance (0.79 AUC, 0.48 AUPRC, and 77.5% accuracy) showed substantial improvement when given limited, relevant data to learn from (ie, RAG) and underlying ML probabilities (0.87 AUC, 0.71 AUPRC, and 83.1% accuracy). Interestingly, RAG alone boosted performance to near peak levels (0.82 AUC, 0.56 AUPRC, and 81.3% accuracy). CONCLUSIONS: The naïve LLM had limited performance but showed significant improvement in predicting ED admissions when supplemented with real-world examples to learn from, particularly through RAG, and/or numerical probabilities from traditional ML models. Its peak performance, although slightly lower than the pure ML model, is noteworthy given its potential for providing reasoning behind predictions. Further refinement of LLMs with real-world data is necessary for successful integration as decision-support tools in care settings.

Characterizing cell type specific transcriptional differences between the living and postmortem human brain.

Single-nucleus RNA sequencing (snRNA-seq) is often used to define gene expression patterns characteristic of brain cell types as well as to identify cell type specific gene expression signatures of neurological and mental illnesses in postmortem human brains. As methods to obtain brain tissue from living individuals emerge, it is essential to characterize gene expression differences associated with tissue originating from either living or postmortem subjects using snRNA-seq, and to assess whether and how such differences may impact snRNA-seq studies of brain tissue. To address this, human prefrontal cortex single nuclei gene expression was generated and compared between 31 samples from living individuals and 21 postmortem samples. The same cell types were consistently identified in living and postmortem nuclei, though for each cell type, a large proportion of genes were differentially expressed between samples from postmortem and living individuals. Notably, estimation of cell type proportions by cell type deconvolution of pseudo-bulk data was found to be more accurate in samples from living individuals. To allow for future integration of living and postmortem brain gene expression, a model was developed that quantifies from gene expression data the probability a human brain tissue sample was obtained postmortem. These probabilities are established as a means to statistically account for the gene expression differences between samples from living and postmortem individuals. Together, the results presented here provide a deep characterization of both differences between snRNA-seq derived from samples from living and postmortem individuals, as well as qualify and account for their effect on common analyses performed on this type of data.

Neuropsychiatric complications of coronavirus disease 2019: Mount Sinai Health System cohort study.

OBJECTIVE: To describe the frequency of neuropsychiatric complications among hospitalized patients with coronavirus disease 2019 (COVID-19) and their association with pre-existing comorbidities and clinical outcomes. METHODS: We retrospectively identified all patients hospitalized with COVID-19 within a large multicenter New York City health system between March 15, 2020 and May 17, 2021 and randomly selected a representative cohort for detailed chart review. Clinical data, including the occurrence of neuropsychiatric complications (categorized as either altered mental status [AMS] or other neuropsychiatric complications) and in-hospital mortality, were extracted using an electronic medical record database and individual chart review. Associations between neuropsychiatric complications, comorbidities, laboratory findings, and in-hospital mortality were assessed using multivariate logistic regression. RESULTS: Our study cohort consisted of 974 patients, the majority were admitted during the first wave of the pandemic. Patients were treated with anticoagulation (88.4%), glucocorticoids (24.8%), and remdesivir (10.5%); 18.6% experienced severe COVID-19 pneumonia (evidenced by ventilator requirement). Neuropsychiatric complications occurred in 58.8% of patients; 39.8% experienced AMS; and 19.0% experienced at least one other complication (seizures in 1.4%, ischemic stroke in 1.6%, hemorrhagic stroke in 1.0%) or symptom (headache in 11.4%, anxiety in 6.8%, ataxia in 6.3%). Higher odds of mortality, which occurred in 22.0%, were associated with AMS, ventilator support, increasing age, and higher serum inflammatory marker levels. Anticoagulant therapy was associated with lower odds of mortality and AMS. CONCLUSION: Neuropsychiatric complications of COVID-19, especially AMS, were common, varied, and associated with in-hospital mortality in a diverse multicenter cohort at an epicenter of the COVID-19 pandemic.

Integrated multiomics implicates dysregulation of ECM and cell adhesion pathways as drivers of severe COVID-associated kidney injury.

COVID-19 has been a significant public health concern for the last four years; however, little is known about the mechanisms that lead to severe COVID-associated kidney injury. In this multicenter study, we combined quantitative deep urinary proteomics and machine learning to predict severe acute outcomes in hospitalized COVID-19 patients. Using a 10-fold cross-validated random forest algorithm, we identified a set of urinary proteins that demonstrated predictive power for both discovery and validation set with 87% and 79% accuracy, respectively. These predictive urinary biomarkers were recapitulated in non-COVID acute kidney injury revealing overlapping injury mechanisms. We further combined orthogonal multiomics datasets to understand the mechanisms that drive severe COVID-associated kidney injury. Functional overlap and network analysis of urinary proteomics, plasma proteomics and urine sediment single-cell RNA sequencing showed that extracellular matrix and autophagy-associated pathways were uniquely impacted in severe COVID-19. Differentially abundant proteins associated with these pathways exhibited high expression in cells in the juxtamedullary nephron, endothelial cells, and podocytes, indicating that these kidney cell types could be potential targets. Further, single-cell transcriptomic analysis of kidney organoids infected with SARS-CoV-2 revealed dysregulation of extracellular matrix organization in multiple nephron segments, recapitulating the clinically observed fibrotic response across multiomics datasets. Ligand-receptor interaction analysis of the podocyte and tubule organoid clusters showed significant reduction and loss of interaction between integrins and basement membrane receptors in the infected kidney organoids. Collectively, these data suggest that extracellular matrix degradation and adhesion-associated mechanisms could be a main driver of COVID-associated kidney injury and severe outcomes.

Dopamine and serotonin in human substantia nigra track social context and value signals during economic exchange.

Dopamine and serotonin are hypothesized to guide social behaviours. In humans, however, we have not yet been able to study neuromodulator dynamics as social interaction unfolds. Here, we obtained subsecond estimates of dopamine and serotonin from human substantia nigra pars reticulata during the ultimatum game. Participants, who were patients with Parkinson's disease undergoing awake brain surgery, had to accept or reject monetary offers of varying fairness from human and computer players. They rejected more offers in the human than the computer condition, an effect of social context associated with higher overall levels of dopamine but not serotonin. Regardless of the social context, relative changes in dopamine tracked trial-by-trial changes in offer value-akin to reward prediction errors-whereas serotonin tracked the current offer value. These results show that dopamine and serotonin fluctuations in one of the basal ganglia's main output structures reflect distinct social context and value signals.

A novel method leveraging time series data to improve subphenotyping and application in critically ill patients with COVID-19.

Computational subphenotyping, a data-driven approach to understanding disease subtypes, is a prominent topic in medical research. Numerous ongoing studies are dedicated to developing advanced computational subphenotyping methods for cross-sectional data. However, the potential of time-series data has been underexplored until now. Here, we propose a Multivariate Levenshtein Distance (MLD) that can account for address correlation in multiple discrete features over time-series data. Our algorithm has two distinct components: it integrates an optimal threshold score to enhance the sensitivity in discriminating between pairs of instances, and the MLD itself. We have applied the proposed distance metrics on the k-means clustering algorithm to derive temporal subphenotypes from time-series data of biomarkers and treatment administrations from 1039 critically ill patients with COVID-19 and compare its effectiveness to standard methods. In conclusion, the Multivariate Levenshtein Distance metric is a novel method to quantify the distance from multiple discrete features over time-series data and demonstrates superior clustering performance among competing time-series distance metrics.

Derivation, External Validation and Clinical Implications of a deep learning approach for intracranial pressure estimation using non-cranial waveform measurements.

Importance: Increased intracranial pressure (ICP) is associated with adverse neurological outcomes, but needs invasive monitoring. Objective: Development and validation of an AI approach for detecting increased ICP (aICP) using only non-invasive extracranial physiological waveform data. Design: Retrospective diagnostic study of AI-assisted detection of increased ICP. We developed an AI model using exclusively extracranial waveforms, externally validated it and assessed associations with clinical outcomes. Setting: MIMIC-III Waveform Database (2000-2013), a database derived from patients admitted to an ICU in an academic Boston hospital, was used for development of the aICP model, and to report association with neurologic outcomes. Data from Mount Sinai Hospital (2020-2022) in New York City was used for external validation. Participants: Patients were included if they were older than 18 years, and were monitored with electrocardiograms, arterial blood pressure, respiratory impedance plethysmography and pulse oximetry. Patients who additionally had intracranial pressure monitoring were used for development (N=157) and external validation (N=56). Patients without intracranial monitors were used for association with outcomes (N=1694). Exposures: Extracranial waveforms including electrocardiogram, arterial blood pressure, plethysmography and SpO2. Main Outcomes and Measures: Intracranial pressure > 15 mmHg. Measures were Area under receiver operating characteristic curves (AUROCs), sensitivity, specificity, and accuracy at threshold of 0.5. We calculated odds ratios and p-values for phenotype association. Results: The AUROC was 0.91 (95% CI, 0.90-0.91) on testing and 0.80 (95% CI, 0.80-0.80) on external validation. aICP had accuracy, sensitivity, and specificity of 73.8% (95% CI, 72.0%-75.6%), 99.5% (95% CI 99.3%-99.6%), and 76.9% (95% CI, 74.0-79.8%) on external validation. A ten-percentile increment was associated with stroke (OR=2.12; 95% CI, 1.27-3.13), brain malignancy (OR=1.68; 95% CI, 1.09-2.60), subdural hemorrhage (OR=1.66; 95% CI, 1.07-2.57), intracerebral hemorrhage (OR=1.18; 95% CI, 1.07-1.32), and procedures like percutaneous brain biopsy (OR=1.58; 95% CI, 1.15-2.18) and craniotomy (OR = 1.43; 95% CI, 1.12-1.84; P < 0.05 for all). Conclusions and Relevance: aICP provides accurate, non-invasive estimation of increased ICP, and is associated with neurological outcomes and neurosurgical procedures in patients without intracranial monitoring.

Persistent complement dysregulation with signs of thromboinflammation in active Long Covid.

Long Covid is a debilitating condition of unknown etiology. We performed multimodal proteomics analyses of blood serum from COVID-19 patients followed up to 12 months after confirmed severe acute respiratory syndrome coronavirus 2 infection. Analysis of >6500 proteins in 268 longitudinal samples revealed dysregulated activation of the complement system, an innate immune protection and homeostasis mechanism, in individuals experiencing Long Covid. Thus, active Long Covid was characterized by terminal complement system dysregulation and ongoing activation of the alternative and classical complement pathways, the latter associated with increased antibody titers against several herpesviruses possibly stimulating this pathway. Moreover, markers of hemolysis, tissue injury, platelet activation, and monocyte-platelet aggregates were increased in Long Covid. Machine learning confirmed complement and thromboinflammatory proteins as top biomarkers, warranting diagnostic and therapeutic interrogation of these systems.

Quantitative Prediction of Right Ventricular Size and Function From the ECG.

BACKGROUND: Right ventricular ejection fraction (RVEF) and end-diastolic volume (RVEDV) are not readily assessed through traditional modalities. Deep learning-enabled ECG analysis for estimation of right ventricular (RV) size or function is unexplored. METHODS AND RESULTS: We trained a deep learning-ECG model to predict RV dilation (RVEDV >120 mL/m2), RV dysfunction (RVEF ≤40%), and numerical RVEDV and RVEF from a 12-lead ECG paired with reference-standard cardiac magnetic resonance imaging volumetric measurements in UK Biobank (UKBB; n=42 938). We fine-tuned in a multicenter health system (MSHoriginal [Mount Sinai Hospital]; n=3019) with prospective validation over 4 months (MSHvalidation; n=115). We evaluated performance with area under the receiver operating characteristic curve for categorical and mean absolute error for continuous measures overall and in key subgroups. We assessed the association of RVEF prediction with transplant-free survival with Cox proportional hazards models. The prevalence of RV dysfunction for UKBB/MSHoriginal/MSHvalidation cohorts was 1.0%/18.0%/15.7%, respectively. RV dysfunction model area under the receiver operating characteristic curve for UKBB/MSHoriginal/MSHvalidation cohorts was 0.86/0.81/0.77, respectively. The prevalence of RV dilation for UKBB/MSHoriginal/MSHvalidation cohorts was 1.6%/10.6%/4.3%. RV dilation model area under the receiver operating characteristic curve for UKBB/MSHoriginal/MSHvalidation cohorts was 0.91/0.81/0.92, respectively. MSHoriginal mean absolute error was RVEF=7.8% and RVEDV=17.6 mL/m2. The performance of the RVEF model was similar in key subgroups including with and without left ventricular dysfunction. Over a median follow-up of 2.3 years, predicted RVEF was associated with adjusted transplant-free survival (hazard ratio, 1.40 for each 10% decrease; P=0.031). CONCLUSIONS: Deep learning-ECG analysis can identify significant cardiac magnetic resonance imaging RV dysfunction and dilation with good performance. Predicted RVEF is associated with clinical outcome.

The genetic contribution to the comorbidity of depression and anxiety: a multi-site electronic health records study of almost 178 000 people.

BACKGROUND: Depression and anxiety are common and highly comorbid, and their comorbidity is associated with poorer outcomes posing clinical and public health concerns. We evaluated the polygenic contribution to comorbid depression and anxiety, and to each in isolation. METHODS: Diagnostic codes were extracted from electronic health records for four biobanks [N = 177 865 including 138 632 European (77.9%), 25 612 African (14.4%), and 13 621 Hispanic (7.7%) ancestry participants]. The outcome was a four-level variable representing the depression/anxiety diagnosis group: neither, depression-only, anxiety-only, and comorbid. Multinomial regression was used to test for association of depression and anxiety polygenic risk scores (PRSs) with the outcome while adjusting for principal components of ancestry. RESULTS: In total, 132 960 patients had neither diagnosis (74.8%), 16 092 depression-only (9.0%), 13 098 anxiety-only (7.4%), and 16 584 comorbid (9.3%). In the European meta-analysis across biobanks, both PRSs were higher in each diagnosis group compared to controls. Notably, depression-PRS (OR 1.20 per s.d. increase in PRS; 95% CI 1.18-1.23) and anxiety-PRS (OR 1.07; 95% CI 1.05-1.09) had the largest effect when the comorbid group was compared with controls. Furthermore, the depression-PRS was significantly higher in the comorbid group than the depression-only group (OR 1.09; 95% CI 1.06-1.12) and the anxiety-only group (OR 1.15; 95% CI 1.11-1.19) and was significantly higher in the depression-only group than the anxiety-only group (OR 1.06; 95% CI 1.02-1.09), showing a genetic risk gradient across the conditions and the comorbidity. CONCLUSIONS: This study suggests that depression and anxiety have partially independent genetic liabilities and the genetic vulnerabilities to depression and anxiety make distinct contributions to comorbid depression and anxiety.