Using an artificial intelligence tool incorporating natural language processing to identify patients with a diagnosis of ANCA-associated vasculitis in electronic health records.

BACKGROUND: Because anti-neutrophil cytoplasmatic antibody (ANCA)-associated vasculitis (AAV) is a rare, life-threatening, auto-immune disease, conducting research is difficult but essential. A long-lasting challenge is to identify rare AAV patients within the electronic-health-record (EHR)-system to facilitate real-world research. Artificial intelligence (AI)-search tools using natural language processing (NLP) for text-mining are increasingly postulated as a solution. METHODS: We employed an AI-tool that combined text-mining with NLP-based exclusion, to accurately identify rare AAV patients within large EHR-systems (>2.000.000 records). We developed an identification method in an academic center with an established AAV-training set (n = 203) and validated the method in a non-academic center with an AAV-validation set (n = 84). To assess accuracy anonymized patient records were manually reviewed. RESULTS: Based on an iterative process, a text-mining search was developed on disease description, laboratory measurements, medication and specialisms. In the training center, 608 patients were identified with a sensitivity of 97.0 % (95%CI [93.7, 98.9]) and positive predictive value (PPV) of 56.9 % (95%CI [52.9, 60.1]). NLP-based exclusion resulted in 444 patients increasing PPV to 77.9 % (95%CI [73.7, 81.7]) while sensitivity remained 96.3 % (95%CI [93.8, 98.0]). In the validation center, text-mining identified 333 patients (sensitivity 97.6 % (95%CI [91.6, 99.7]), PPV 58.2 % (95%CI [52.8, 63.6])) and NLP-based exclusion resulted in 223 patients, increasing PPV to 86.1 % (95%CI [80.9, 90.4]) with 98.0 % (95%CI [94.9, 99.4]) sensitivity. Our identification method outperformed ICD-10-coding predominantly in identifying MPO+ and organ-limited AAV patients. CONCLUSIONS: Our study highlights the advantages of implementing AI, notably NLP, to accurately identify rare AAV patients within large EHR-systems and demonstrates the applicability and transportability. Therefore, this method can reduce efforts to identify AAV patients and accelerate real-world research, while avoiding bias by ICD-10-coding.

Recommendation to implementation of remote patient monitoring in rheumatology: lessons learned and barriers to take.

Remote patient monitoring (RPM) leverages advanced technology to monitor and manage patients' health remotely and continuously. In 2022 European Alliance of Associations for Rheumatology (EULAR) points-to-consider for remote care were published to foster adoption of RPM, providing guidelines on where to position RPM in our practices. Sample papers and studies describe the value of RPM. But for many rheumatologists, the unanswered question remains the 'how to?' implement RPM.Using the successful, though not frictionless example of the Southmead rheumatology department, we address three types of barriers for the implementation of RPM: service, clinician and patients, with subsequent learning points that could be helpful for new teams planning to implement RPM. These address, but are not limited to, data governance, selecting high quality cost-effective solutions and ensuring compliance with data protection regulations. In addition, we describe five lacunas that could further improve RPM when addressed: establishing quality standards, creating a comprehensive database of available RPM tools, integrating data with electronic patient records, addressing reimbursement uncertainties and improving digital literacy among patients and healthcare professionals.

The Application of Genetic Risk Scores in Rheumatic Diseases: A Perspective.

Modest effect sizes have limited the clinical applicability of genetic associations with rheumatic diseases. Genetic risk scores (GRSs) have emerged as a promising solution to translate genetics into useful tools. In this review, we provide an overview of the recent literature on GRSs in rheumatic diseases. We describe six categories for which GRSs are used: (a) disease (outcome) prediction, (b) genetic commonalities between diseases, (c) disease differentiation, (d) interplay between genetics and environmental factors, (e) heritability and transferability, and (f) detecting causal relationships between traits. In our review of the literature, we identified current lacunas and opportunities for future work. First, the shortage of non-European genetic data restricts the application of many GRSs to European populations. Next, many GRSs are tested in settings enriched for cases that limit the transferability to real life. If intended for clinical application, GRSs are ideally tested in the relevant setting. Finally, there is much to elucidate regarding the co-occurrence of clinical traits to identify shared causal paths and elucidate relationships between the diseases. GRSs are useful instruments for this. Overall, the ever-continuing research on GRSs gives a hopeful outlook into the future of GRSs and indicates significant progress in their potential applications.

Rapid Response to Remdesivir in Hospitalised COVID-19 Patients: A Propensity Score Weighted Multicentre Cohort Study.

INTRODUCTION: Remdesivir is a registered treatment for hospitalised patients with COVID-19 that has moderate clinical effectiveness. Anecdotally, some patients' respiratory insufficiency seemed to recover particularly rapidly after initiation of remdesivir. In this study, we investigated if this rapid improvement was caused by remdesivir, and which patient characteristics might predict a rapid clinical improvement in response to remdesivir. METHODS: This was a multicentre observational cohort study of hospitalised patients with COVID-19 who required supplemental oxygen and were treated with dexamethasone. Rapid clinical improvement in response to treatment was defined by a reduction of at least 1 L of supplemental oxygen per minute or discharge from the hospital within 72 h after admission. Inverse probability of treatment-weighted logistic regression modelling was used to assess the association between remdesivir and rapid clinical improvement. Secondary endpoints included in-hospital mortality, ICU admission rate and hospitalisation duration. RESULTS: Of 871 patients included, 445 were treated with remdesivir. There was no influence of remdesivir on the occurrence of rapid clinical improvement (62% vs 61% OR 1.05, 95% CI 0.79-1.40; p = 0.76). The in-hospital mortality was lower (14.7% vs 19.8% OR 0.70, 95% CI 0.48-1.02; p = 0.06) for the remdesivir-treated patients. Rapid clinical improvement occurred more often in patients with low C-reactive protein (≤ 75 mg/L) and short duration of symptoms prior to hospitalisation (< 7 days) (OR 2.84, 95% CI 1.07-7.56). CONCLUSION: Remdesivir generally does not increase the incidence of rapid clinical improvement in hospitalised patients with COVID-19, but it might have an effect in patients with short duration of symptoms and limited signs of systemic inflammation.

Population-based user-perceived experience of Rheumatic?: a novel digital symptom-checker in rheumatology.

OBJECTIVE: Digital symptom-checkers (SCs) have potential to improve rheumatology triage and reduce diagnostic delays. In addition to being accurate, SCs should be user friendly and meet patient's needs. Here, we examined usability and acceptance of Rheumatic?-a new and freely available online SC (currently with >44 000 users)-in a real-world setting. METHODS: Study participants were recruited from an ongoing prospective study, and included people ≥18 years with musculoskeletal complaints completing Rheumatic? online. The user experience survey comprised five usability and acceptability questions (11-point rating scale), and an open-ended question regarding improvement of Rheumatic? Data were analysed in R using t-test or Wilcoxon rank test (group comparisons), or linear regression (continuous variables). RESULTS: A total of 12 712 people completed the user experience survey. The study population had a normal age distribution, with a peak at 50-59 years, and 78% women. A majority found Rheumatic? useful (78%), thought the questionnaire gave them an opportunity to describe their complaints well (76%), and would recommend Rheumatic? to friends and other patients (74%). Main shortcoming was that 36% thought there were too many questions. Still, 39% suggested more detailed questions, and only 2% suggested a reduction of questions. CONCLUSION: Based on real-world data from the largest user evaluation study of a digital SC in rheumatology, we conclude that Rheumatic? is well accepted by women and men with rheumatic complaints, in all investigated age groups. Wide-scale adoption of Rheumatic?, therefore, seems feasible, with promising scientific and clinical implications on the horizon.

The Role of Genetics in Clinically Suspect Arthralgia and Rheumatoid Arthritis Development: A Large Cross-Sectional Study.

OBJECTIVE: To investigate whether established genetic predictors for rheumatoid arthritis (RA) differentiate healthy controls, patients with clinically suspect arthralgia (CSA), and RA patients. METHODS: Using analyses of variance, chi-square tests, and mean risk difference analyses, we investigated the association of an RA polygenic risk score (PRS) and HLA shared epitope (HLA-SE) with all participant groups, both unstratified and stratified for anti-citrullinated protein antibody (ACPA) status. We used 3 separate data sets sampled from the same Dutch population (1,015 healthy controls, 479 CSA patients, and 1,146 early classified RA patients). CSA patients were assessed for conversion to inflammatory arthritis over a period of 2 years, after which they were classified as either CSA converters (n = 84) or CSA nonconverters (n = 395). RESULTS: The PRS was increased in RA patients (mean ± SD PRS 1.31 ± 0.96) compared to the complete CSA group (1.07 ± 0.94) and compared to CSA converters (1.12 ± 0.94). In ACPA- strata, PRS distributions differed strongly when comparing the complete CSA group (mean ± SD PRS 1.05 ± 0.94) and CSA converters (0.97 ± 0.87) to RA patients (1.20 ± 0.94), while in the ACPA+ strata, the complete CSA group (1.25 ± 0.99) differed clearly from healthy controls (1.05 ± 0.94) and RA patients (1.41 ± 0.96). HLA-SE was more prevalent in the RA group (prevalence 0.64) than the complete CSA group (0.45), with small differences between RA patients and CSA converters (0.64 versus 0.60) and larger differences between CSA converters and CSA nonconverters (0.60 versus 0.42). HLA-SE prevalence differed more strongly within the ACPA+ strata as follows: healthy controls (prevalence 0.43), CSA nonconverters (0.48), complete CSA group (0.59), CSA converters (0.66), and RA patients (0.79). CONCLUSION: We observed that genetic predisposition increased across pre-RA participant groups. The RA PRS differed in early classified RA and inflammatory pre-disease stages, regardless of ACPA stratification. HLA-SE prevalence differed between arthritis patients, particularly ACPA+ patients, and healthy controls. Genetics seem to fulfill different etiologic roles.

From real-world electronic health record data to real-world results using artificial intelligence.

With the worldwide digitalisation of medical records, electronic health records (EHRs) have become an increasingly important source of real-world data (RWD). RWD can complement traditional study designs because it captures almost the complete variety of patients, leading to more generalisable results. For rheumatology, these data are particularly interesting as our diseases are uncommon and often take years to develop. In this review, we discuss the following concepts related to the use of EHR for research and considerations for translation into clinical care: EHR data contain a broad collection of healthcare data covering the multitude of real-life patients and the healthcare processes related to their care. Machine learning (ML) is a powerful method that allows us to leverage a large amount of heterogeneous clinical data for clinical algorithms, but requires extensive training, testing, and validation. Patterns discovered in EHR data using ML are applicable to real life settings, however, are also prone to capturing the local EHR structure and limiting generalisability outside the EHR(s) from which they were developed. Population studies on EHR necessitates knowledge on the factors influencing the data available in the EHR to circumvent biases, for example, access to medical care, insurance status. In summary, EHR data represent a rapidly growing and key resource for real-world studies. However, transforming RWD EHR data for research and for real-world evidence using ML requires knowledge of the EHR system and their differences from existing observational data to ensure that studies incorporate rigorous methods that acknowledge or address factors such as access to care, noise in the data, missingness and indication bias.

Genetic regulation of serum IgA levels and susceptibility to common immune, infectious, kidney, and cardio-metabolic traits.

Immunoglobulin A (IgA) mediates mucosal responses to food antigens and the intestinal microbiome and is involved in susceptibility to mucosal pathogens, celiac disease, inflammatory bowel disease, and IgA nephropathy. We performed a genome-wide association study of serum IgA levels in 41,263 individuals of diverse ancestries and identified 20 genome-wide significant loci, including 9 known and 11 novel loci. Co-localization analyses with expression QTLs prioritized candidate genes for 14 of 20 significant loci. Most loci encoded genes that produced immune defects and IgA abnormalities when genetically manipulated in mice. We also observed positive genetic correlations of serum IgA levels with IgA nephropathy, type 2 diabetes, and body mass index, and negative correlations with celiac disease, inflammatory bowel disease, and several infections. Mendelian randomization supported elevated serum IgA as a causal factor in IgA nephropathy. African ancestry was consistently associated with higher serum IgA levels and greater frequency of IgA-increasing alleles compared to other ancestries. Our findings provide novel insights into the genetic regulation of IgA levels and its potential role in human disease.

E-health as a sine qua non for modern healthcare.

In each era we need to balance between being able to provide care with our "technical skill, scientific knowledge, and human understanding" (Harrison's Principles of Internal Medicine, 1950) to the individual patient and simultaneously ensure that our healthcare serves all. With the increasing demand of healthcare by an aging population and the lack of specialists, accessible healthcare within a reasonable time frame is not always guaranteed. E-health provides solutions for current situations where we do not meet our own aims of good healthcare, such as restrictions in access to care and a reduction in care availability by a reducing workforce. In addition, telemedicine offers opportunities to improve our healthcare beyond what is possible by in person visits. However, e-health is often viewed as an deficient version of healthcare of low quality. We disagree with this view. In this article we will discuss how to position e-health in the current situation of healthcare, given the continuing rapid development of digital technologies and the changing needs of healthcare professionals and patients. We will address the evolution of e-health towards connected and intelligent systems and the stakeholders perspective, aiming to open up the discussion on e-Health.

Development and validation of an early warning model for hospitalized COVID-19 patients: a multi-center retrospective cohort study.

BACKGROUND: Timely identification of deteriorating COVID-19 patients is needed to guide changes in clinical management and admission to intensive care units (ICUs). There is significant concern that widely used Early warning scores (EWSs) underestimate illness severity in COVID-19 patients and therefore, we developed an early warning model specifically for COVID-19 patients. METHODS: We retrospectively collected electronic medical record data to extract predictors and used these to fit a random forest model. To simulate the situation in which the model would have been developed after the first and implemented during the second COVID-19 'wave' in the Netherlands, we performed a temporal validation by splitting all included patients into groups admitted before and after August 1, 2020. Furthermore, we propose a method for dynamic model updating to retain model performance over time. We evaluated model discrimination and calibration, performed a decision curve analysis, and quantified the importance of predictors using SHapley Additive exPlanations values. RESULTS: We included 3514 COVID-19 patient admissions from six Dutch hospitals between February 2020 and May 2021, and included a total of 18 predictors for model fitting. The model showed a higher discriminative performance in terms of partial area under the receiver operating characteristic curve (0.82 [0.80-0.84]) compared to the National early warning score (0.72 [0.69-0.74]) and the Modified early warning score (0.67 [0.65-0.69]), a greater net benefit over a range of clinically relevant model thresholds, and relatively good calibration (intercept = 0.03 [- 0.09 to 0.14], slope = 0.79 [0.73-0.86]). CONCLUSIONS: This study shows the potential benefit of moving from early warning models for the general inpatient population to models for specific patient groups. Further (independent) validation of the model is needed.

External validation of risk scores to predict in-hospital mortality in patients hospitalized due to coronavirus disease 2019.

BACKGROUND: The coronavirus disease 2019 (COVID-19) presents an urgent threat to global health. Prediction models that accurately estimate mortality risk in hospitalized patients could assist medical staff in treatment and allocating limited resources. AIMS: To externally validate two promising previously published risk scores that predict in-hospital mortality among hospitalized COVID-19 patients. METHODS: Two prospective cohorts were available; a cohort of 1028 patients admitted to one of nine hospitals in Lombardy, Italy (the Lombardy cohort) and a cohort of 432 patients admitted to a hospital in Leiden, the Netherlands (the Leiden cohort). The endpoint was in-hospital mortality. All patients were adult and tested COVID-19 PCR-positive. Model discrimination and calibration were assessed. RESULTS: The C-statistic of the 4C mortality score was good in the Lombardy cohort (0.85, 95CI: 0.82-0.89) and in the Leiden cohort (0.87, 95CI: 0.80-0.94). Model calibration was acceptable in the Lombardy cohort but poor in the Leiden cohort due to the model systematically overpredicting the mortality risk for all patients. The C-statistic of the CURB-65 score was good in the Lombardy cohort (0.80, 95CI: 0.75-0.85) and in the Leiden cohort (0.82, 95CI: 0.76-0.88). The mortality rate in the CURB-65 development cohort was much lower than the mortality rate in the Lombardy cohort. A similar but less pronounced trend was found for patients in the Leiden cohort. CONCLUSION: Although performances did not differ greatly, the 4C mortality score showed the best performance. However, because of quickly changing circumstances, model recalibration may be necessary before using the 4C mortality score.

Rheumatic?-A Digital Diagnostic Decision Support Tool for Individuals Suspecting Rheumatic Diseases: A Multicenter Pilot Validation Study.

Introduction: Digital diagnostic decision support tools promise to accelerate diagnosis and increase health care efficiency in rheumatology. Rheumatic? is an online tool developed by specialists in rheumatology and general medicine together with patients and patient organizations. It calculates a risk score for several rheumatic diseases. We ran a pilot study retrospectively testing Rheumatic? for its ability to differentiate symptoms from existing or emerging immune-mediated rheumatic diseases from other rheumatic and musculoskeletal complaints and disorders in patients visiting rheumatology clinics. Materials and Methods: The performance of Rheumatic? was tested using in three university rheumatology centers: (A) patients at Risk for RA (Karolinska Institutet, n = 50 individuals with musculoskeletal complaints and anti-citrullinated protein antibody positivity) (B) patients with early joint swelling [dataset B (Erlangen) n = 52]. (C) Patients with early arthritis where the clinician considered it likely to be of auto-immune origin [dataset C (Leiden) n = 73]. In dataset A we tested whether Rheumatic? could predict the development of arthritis. In dataset B and C we tested whether Rheumatic? could predict the development of an immune-mediated rheumatic diseases. We examined the discriminative power of the total score with the Wilcoxon rank test and the area-under-the-receiver-operating-characteristic curve (AUC-ROC). Next, we calculated the test characteristics for these patients passing the first or second expert-based Rheumatic? scoring threshold. Results: The total test scores differentiated between: (A) Individuals developing arthritis or not, median 245 vs. 163, P < 0.0001, AUC-ROC = 75.3; (B) patients with an immune-mediated arthritic disease or not median 191 vs. 107, P < 0.0001, AUC-ROC = 79.0; but less patients with an immune-mediated arthritic disease or not amongst those where the clinician already considered an immune mediated disease most likely (median 262 vs. 212, P < 0.0001, AUC-ROC = 53.6). Threshold-1 (advising to visit primary care doctor) was highly specific in dataset A and B (0.72, 0.87, and 0.23, respectively) and sensitive (0.67, 0.61, and 0.67). Threshold-2 (advising to visit rheumatologic care) was very specific in all three centers but not very sensitive: specificity of 1.0, 0.96, and 0.91, sensitivity 0.05, 0.07, 0.14 in dataset A, B, and C, respectively. Conclusion: Rheumatic? is a web-based patient-centered multilingual diagnostic tool capable of differentiating immune-mediated rheumatic conditions from other musculoskeletal problems. The current scoring system needs to be further optimized.

A framework for employing longitudinally collected multicenter electronic health records to stratify heterogeneous patient populations on disease history.

OBJECTIVE: To facilitate patient disease subset and risk factor identification by constructing a pipeline which is generalizable, provides easily interpretable results, and allows replication by overcoming electronic health records (EHRs) batch effects. MATERIAL AND METHODS: We used 1872 billing codes in EHRs of 102 880 patients from 12 healthcare systems. Using tools borrowed from single-cell omics, we mitigated center-specific batch effects and performed clustering to identify patients with highly similar medical history patterns across the various centers. Our visualization method (PheSpec) depicts the phenotypic profile of clusters, applies a novel filtering of noninformative codes (Ranked Scope Pervasion), and indicates the most distinguishing features. RESULTS: We observed 114 clinically meaningful profiles, for example, linking prostate hyperplasia with cancer and diabetes with cardiovascular problems and grouping pediatric developmental disorders. Our framework identified disease subsets, exemplified by 6 "other headache" clusters, where phenotypic profiles suggested different underlying mechanisms: migraine, convulsion, injury, eye problems, joint pain, and pituitary gland disorders. Phenotypic patterns replicated well, with high correlations of ≥0.75 to an average of 6 (2-8) of the 12 different cohorts, demonstrating the consistency with which our method discovers disease history profiles. DISCUSSION: Costly clinical research ventures should be based on solid hypotheses. We repurpose methods from single-cell omics to build these hypotheses from observational EHR data, distilling useful information from complex data. CONCLUSION: We establish a generalizable pipeline for the identification and replication of clinically meaningful (sub)phenotypes from widely available high-dimensional billing codes. This approach overcomes datatype problems and produces comprehensive visualizations of validation-ready phenotypes.

Determining in which pre-arthritis stage HLA-shared epitope alleles and smoking exert their effect on the development of rheumatoid arthritis.

OBJECTIVES: The human leukocyte antigen-shared epitope (HLA-SE) alleles and smoking are the most prominent genetic and environmental risk factors for rheumatoid arthritis (RA). However, at which pre-arthritis stage (asymptomatic/symptomatic) they exert their effect is unknown. We aimed to determine whether HLA-SE and smoking are involved in the onset of autoantibody positivity, symptoms (clinically suspect arthralgia (CSA)) and/or progression to clinical arthritis. METHODS: We performed meta-analyses on results from the literature on associations of HLA-SE and smoking with anti-citrullinated protein antibodies (ACPAs) in the asymptomatic population. Next, we studied associations of HLA-SE and smoking with autoantibody positivity at CSA onset and with progression to clinical inflammatory arthritis (IA) during follow-up. Associations in ACPA-positive patients with CSA were validated in meta-analyses with other arthralgia cohorts. Analyses were repeated for rheumatoid factor (RF), anti-carbamylated protein antibodies (anti-CarP) and anti-acetylated protein antibodies (AAPA). RESULTS: Meta-analyses showed that HLA-SE is not associated with ACPA positivity in the asymptomatic population (OR 1.06 (95% CI:0.69 to 1.64)), whereas smoking was associated (OR 1.37 (95% CI: 1.15 to 1.63)). At CSA onset, both HLA-SE and smoking associated with ACPA positivity (OR 2.08 (95% CI: 1.24 to 3.49), OR 2.41 (95% CI: 1.31 to 4.43)). During follow-up, HLA-SE associated with IA development (HR 1.86 (95% CI: 1.23 to 2.82)), in contrast to smoking. This was confirmed in meta-analyses in ACPA-positive arthralgia (HR 1.52 (95% CI: 1.08 to 2.15)). HLA-SE and smoking were not associated with RF, anti-CarP or AAPA-positivity at CSA onset. Longitudinally, AAPA associated with IA development independent from ACPA and RF (HR 1.79 (95% CI: 1.02 to 3.16)), anti-CarP did not. CONCLUSIONS: HLA-SE and smoking act at different stages: smoking confers risk for ACPA and symptom development, whereas HLA-SE mediates symptom and IA development. These data enhance the understanding of the timing of the key risk factors in the development of RA.

The Disconnect Between Development and Intended Use of Clinical Prediction Models for Covid-19: A Systematic Review and Real-World Data Illustration.

Background: The SARS-CoV-2 pandemic has boosted the appearance of clinical predictions models in medical literature. Many of these models aim to provide guidance for decision making on treatment initiation. Special consideration on how to account for post-baseline treatments is needed when developing such models. We examined how post-baseline treatment was handled in published Covid-19 clinical prediction models and we illustrated how much estimated risks may differ according to how treatment is handled. Methods: Firstly, we reviewed 33 Covid-19 prognostic models published in literature in the period up to 5 May 2020. We extracted: (1) the reported intended use of the model; (2) how treatment was incorporated during model development and (3) whether the chosen analysis strategy was in agreement with the intended use. Secondly, we used nationwide Dutch data on hospitalized patients who tested positive for SARS-CoV-2 in 2020 to illustrate how estimated mortality risks will differ when using four different analysis strategies to model ICU treatment. Results: Of the 33 papers, 21 (64%) had misalignment between intended use and analysis strategy, 7 (21%) were unclear about the estimated risk and only 5 (15%) had clear alignment between intended use and analysis strategy. We showed with real data how different approaches to post-baseline treatment yield different estimated mortality risks, ranging between 33 and 46% for a 75 year-old patient with two medical conditions. Conclusions: Misalignment between intended use and analysis strategy is common in reported Covid-19 clinical prediction models. This can lead to considerable under or overestimation of intended risks.

New risk model is able to identify patients with a low risk of progression in systemic sclerosis.

OBJECTIVES: To develop a prediction model to guide annual assessment of systemic sclerosis (SSc) patients tailored in accordance to disease activity. METHODS: A machine learning approach was used to develop a model that can identify patients without disease progression. SSc patients included in the prospective Leiden SSc cohort and fulfilling the ACR/EULAR 2013 criteria were included. Disease progression was defined as progression in ≥1 organ system, and/or start of immunosuppression or death. Using elastic-net-regularisation, and including 90 independent clinical variables (100% complete), we trained the model on 75% and validated it on 25% of the patients, optimising on negative predictive value (NPV) to minimise the likelihood of missing progression. Probability cutoffs were identified for low and high risk for disease progression by expert assessment. RESULTS: Of the 492 SSc patients (follow-up range: 2-10 years), disease progression during follow-up was observed in 52% (median time 4.9 years). Performance of the model in the test set showed an AUC-ROC of 0.66. Probability score cutoffs were defined: low risk for disease progression (<0.197, NPV:1.0; 29% of patients), intermediate risk (0.197-0.223, NPV:0.82; 27%) and high risk (>0.223, NPV:0.78; 44%). The relevant variables for the model were: previous use of cyclophosphamide or corticosteroids, start with immunosuppressive drugs, previous gastrointestinal progression, previous cardiovascular event, pulmonary arterial hypertension, modified Rodnan Skin Score, creatine kinase and diffusing capacity for carbon monoxide. CONCLUSION: Our machine-learning-assisted model for progression enabled us to classify 29% of SSc patients as 'low risk'. In this group, annual assessment programmes could be less extensive than indicated by international guidelines.

Medical Records-Based Genetic Studies of the Complement System.

BACKGROUND: Genetic variants in complement genes have been associated with a wide range of human disease states, but well-powered genetic association studies of complement activation have not been performed in large multiethnic cohorts. METHODS: We performed medical records-based genome-wide and phenome-wide association studies for plasma C3 and C4 levels among participants of the Electronic Medical Records and Genomics (eMERGE) network. RESULTS: In a GWAS for C3 levels in 3949 individuals, we detected two genome-wide significant loci: chr.1q31.3 (CFH locus; rs3753396-A; ß=0.20; 95% CI, 0.14 to 0.25; P=1.52x10-11) and chr.19p13.3 (C3 locus; rs11569470-G; ß=0.19; 95% CI, 0.13 to 0.24; P=1.29x10-8). These two loci explained approximately 2% of variance in C3 levels. GWAS for C4 levels involved 3998 individuals and revealed a genome-wide significant locus at chr.6p21.32 (C4 locus; rs3135353-C; ß=0.40; 95% CI, 0.34 to 0.45; P=4.58x10-35). This locus explained approximately 13% of variance in C4 levels. The multiallelic copy number variant analysis defined two structural genomic C4 variants with large effect on blood C4 levels: C4-BS (ß=-0.36; 95% CI, -0.42 to -0.30; P=2.98x10-22) and C4-AL-BS (ß=0.25; 95% CI, 0.21 to 0.29; P=8.11x10-23). Overall, C4 levels were strongly correlated with copy numbers of C4A and C4B genes. In comprehensive phenome-wide association studies involving 102,138 eMERGE participants, we cataloged a full spectrum of autoimmune, cardiometabolic, and kidney diseases genetically related to systemic complement activation. CONCLUSIONS: We discovered genetic determinants of plasma C3 and C4 levels using eMERGE genomic data linked to electronic medical records. Genetic variants regulating C3 and C4 levels have large effects and multiple clinical correlations across the spectrum of complement-related diseases in humans.