Clinical Information Retrieval: A Literature Review

Clinical information retrieval (IR) plays a vital role in modern healthcare by facilitating efficient access and analysis of medical literature for clinicians and researchers. This scoping review aims to offer a comprehensive overview of the current state of clinical IR research and identify gaps and potential opportunities for future studies in this field. The main objective was to assess and analyze the existing literature on clinical IR, focusing on the methods, techniques, and tools employed for effective retrieval and analysis of medical information. Adhering to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, we conducted an extensive search across databases such as Ovid Embase, Ovid Medline, Scopus, ACM Digital Library, IEEE Xplore, and Web of Science, covering publications from January 1, 2010, to January 4, 2023. The rigorous screening process led to the inclusion of 184 papers in our review. Our findings provide a detailed analysis of the clinical IR research landscape, covering aspects like publication trends, data sources, methodologies, evaluation metrics, and applications. The review identifies key research gaps in clinical IR methods such as indexing, ranking, and query expansion, offering insights and opportunities for future studies in clinical IR, thus serving as a guiding framework for upcoming research efforts in this rapidly evolving field. The study also underscores an imperative for innovative research on advanced clinical IR systems capable of fast semantic vector search and adoption of neural IR techniques for effective retrieval of information from unstructured electronic health records (EHRs).

Supplementary Information

The online version contains supplementary material available at 10.1007/s41666-024-00159-4.

Can the Dose of Belimumab be Reduced in Patients with Systemic Lupus Erythematosus?

OBJECTIVES

The aims of this study were to investigate the prevalence of dose reduction in patients with SLE treated with belimumab (BEL) in Spain, analyze treatment modalities, and determine impact on control of disease activity.

METHODS

Retrospective longitudinal and multicentre study of SLE patients treated with BEL. Data on disease activity, treatments and outcomes were recorded before and after reduction (6-12 months), and they were compared.

RESULTS

A total of 324 patients were included. The dose was reduced in 29 patients (8.9%). The dosing interval was increased in 9 patients receiving subcutaneous BEL and in 6 patients receiving intravenous BEL. The dose per administration was reduced in 16 patients.Pre-reduction status was remission (2021 DORIS) in 15/26 patients (57.7%) and LLDAS in 23/26 patients (88.5%). After reduction, 2/24 patients (8.3%) and 3/22 patients (13.6%) lost remission at 6 months and 12 months, respectively (not statistically significant [NS]). As for LLDAS, 2/23 patients (8.7%) and 2/21 patients (9.5%) lost their status at 6 and 12 months, respectively (NS). Significantly fewer patients were taking glucocorticoids (GCs) at their 12-month visit, although the median dose of GCs was higher at the 12-month visit (5 [0.62-8.75] vs 2.5 [0-5] at baseline).

CONCLUSION

Doses of BEL can be reduced with no relevant changes in disease activity-at least in the short term-in a significant percentage of patients, and most maintain the reduced dose. However, increased clinical or serologic activity may be observed in some patients. Consequently, tighter post-reduction follow-up is advisable.

Natural language processing of clinical notes enables early inborn error of immunity risk ascertainment

Background

There are now approximately 450 discrete inborn errors of immunity (IEI) described; however, diagnostic rates remain suboptimal. Use of structured health record data has proven useful for patient detection but may be augmented by natural language processing (NLP). Here we present a machine learning model that can distinguish patients from controls significantly in advance of ultimate diagnosis date.

Objective

We sought to create an NLP machine learning algorithm that could identify IEI patients early during the disease course and shorten the diagnostic odyssey.

Methods

Our approach involved extracting a large corpus of IEI patient clinical-note text from a major referral center's electronic health record (EHR) system and a matched control corpus for comparison. We built text classifiers with simple machine learning methods and trained them on progressively longer time epochs before date of diagnosis.

Results

The top performing NLP algorithm effectively distinguished cases from controls robustly 36 months before ultimate clinical diagnosis (area under precision recall curve > 0.95). Corpus analysis demonstrated that statistically enriched, IEI-relevant terms were evident 24+ months before diagnosis, validating that clinical notes can provide a signal for early prediction of IEI.

Conclusion

Mining EHR notes with NLP holds promise for improving early IEI patient detection.

Improving large language models for clinical named entity recognition via prompt engineering

IMPORTANCE

The study highlights the potential of large language models, specifically GPT-3.5 and GPT-4, in processing complex clinical data and extracting meaningful information with minimal training data. By developing and refining prompt-based strategies, we can significantly enhance the models' performance, making them viable tools for clinical NER tasks and possibly reducing the reliance on extensive annotated datasets.

OBJECTIVES

This study quantifies the capabilities of GPT-3.5 and GPT-4 for clinical named entity recognition (NER) tasks and proposes task-specific prompts to improve their performance.

MATERIALS AND METHODS

We evaluated these models on 2 clinical NER tasks: (1) to extract medical problems, treatments, and tests from clinical notes in the MTSamples corpus, following the 2010 i2b2 concept extraction shared task, and (2) to identify nervous system disorder-related adverse events from safety reports in the vaccine adverse event reporting system (VAERS). To improve the GPT models' performance, we developed a clinical task-specific prompt framework that includes (1) baseline prompts with task description and format specification, (2) annotation guideline-based prompts, (3) error analysis-based instructions, and (4) annotated samples for few-shot learning. We assessed each prompt's effectiveness and compared the models to BioClinicalBERT.

RESULTS

Using baseline prompts, GPT-3.5 and GPT-4 achieved relaxed F1 scores of 0.634, 0.804 for MTSamples and 0.301, 0.593 for VAERS. Additional prompt components consistently improved model performance. When all 4 components were used, GPT-3.5 and GPT-4 achieved relaxed F1 socres of 0.794, 0.861 for MTSamples and 0.676, 0.736 for VAERS, demonstrating the effectiveness of our prompt framework. Although these results trail BioClinicalBERT (F1 of 0.901 for the MTSamples dataset and 0.802 for the VAERS), it is very promising considering few training samples are needed.

DISCUSSION

The study's findings suggest a promising direction in leveraging LLMs for clinical NER tasks. However, while the performance of GPT models improved with task-specific prompts, there's a need for further development and refinement. LLMs like GPT-4 show potential in achieving close performance to state-of-the-art models like BioClinicalBERT, but they still require careful prompt engineering and understanding of task-specific knowledge. The study also underscores the importance of evaluation schemas that accurately reflect the capabilities and performance of LLMs in clinical settings.

CONCLUSION

While direct application of GPT models to clinical NER tasks falls short of optimal performance, our task-specific prompt framework, incorporating medical knowledge and training samples, significantly enhances GPT models' feasibility for potential clinical applications.

Text Classification of Cancer Clinical Trial Eligibility Criteria

Automatic identification of clinical trials for which a patient is eligible is complicated by the fact that trial eligibility are stated in natural language. A potential solution to this problem is to employ text classification methods for common types of eligibility criteria. In this study, we focus on seven common exclusion criteria in cancer trials: prior malignancy, human immunodeficiency virus, hepatitis B, hepatitis C, psychiatric illness, drug/substance abuse, and autoimmune illness. Our dataset consists of 764 phase III cancer trials with these exclusions annotated at the trial level. We experiment with common transformer models as well as a new pre-trained clinical trial BERT model. Our results demonstrate the feasibility of automatically classifying common exclusion criteria. Additionally, we demonstrate the value of a pre-trained language model specifically for clinical trials, which yield the highest average performance across all criteria.

Current use of drains and management of seroma following mastectomy and axillary surgery: results of a United Kingdom national practice survey

PURPOSE

Up to 40% of the 56,000 women diagnosed with breast cancer each year in the UK undergo mastectomy. Seroma formation following surgery is common, may delay wound healing, and be uncomfortable or delay the start of adjuvant treatment. Multiple strategies to reduce seroma formation include surgical drains, flap fixation and external compression exist but evidence to support best practice is lacking. We aimed to survey UK breast surgeons to determine current practice to inform the feasibility of undertaking a future trial.

METHODS

An online survey was developed and circulated to UK breast surgeons via professional and trainee associations and social media to explore current attitudes to drain use and management of post-operative seroma. Simple descriptive statistics were used to summarise the results.

RESULTS

The majority of surgeons (82/97, 85%) reported using drains either routinely (38, 39%) or in certain circumstances (44, 45%). Other methods for reducing seroma such as flap fixation were less commonly used. Wide variation was reported in the assessment and management of post-operative seromas. Over half (47/91, 52%) of respondents felt there was some uncertainty about drain use after mastectomy and axillary surgery and two-thirds (59/91, 65%) felt that a trial evaluating the use of drains vs no drains after simple breast cancer surgery was needed.

CONCLUSIONS

There is a need for a large-scale UK-based RCT to determine if, when and in whom drains are necessary following mastectomy and axillary surgery. This work will inform the design and conduct of a future trial.

Sequencing conversational turns in peer interactions: An integrated approach for evidence-based conversational agent for just-in-time nicotine cravings intervention

Background

Risky health behaviors place an enormous toll on public health systems. While relapse prevention support is integrated with most behavior modification programs, the results are suboptimal. Recent advances in artificial intelligence (AI) applications provide us with unique opportunities to develop just-in-time adaptive behavior change solutions.

Methods

In this study, we present an innovative framework, grounded in behavioral theory, and enhanced with social media sequencing and communications scenario builder to architect a conversational agent (CA) specialized in the prevention of relapses in the context of tobacco cessation. We modeled peer interaction data (n = 1000) using the taxonomy of behavior change techniques (BCTs) and speech act (SA) theory to uncover the socio-behavioral and linguistic context embedded within the online social discourse. Further, we uncovered the sequential patterns of BCTs and SAs from social conversations (n = 339,067). We utilized grounded theory-based techniques for extracting the scenarios that best describe individuals' needs and mapped them into the architecture of the virtual CA.

Results

The frequently occurring sequential patterns for BCTs were comparison of behavior and feedback and monitoring; for SAs were directive and assertion. Five cravings-related scenarios describing users' needs as they deal with nicotine cravings were identified along with the kinds of behavior change constructs that are being elicited within those scenarios.

Conclusions

AI-led virtual CAs focusing on behavior change need to employ data-driven and theory-linked approaches to address issues related to engagement, sustainability, and acceptance. The sequential patterns of theory and intent manifestations need to be considered when developing effective behavior change CAs.

Multiorgan MRI findings after hospitalisation with COVID-19 in the UK (C-MORE): a prospective, multicentre, observational cohort study

INTRODUCTION

The multiorgan impact of moderate to severe coronavirus infections in the post-acute phase is still poorly understood. We aimed to evaluate the excess burden of multiorgan abnormalities after hospitalisation with COVID-19, evaluate their determinants, and explore associations with patient-related outcome measures.

METHODS

In a prospective, UK-wide, multicentre MRI follow-up study (C-MORE), adults (aged ≥18 years) discharged from hospital following COVID-19 who were included in Tier 2 of the Post-hospitalisation COVID-19 study (PHOSP-COVID) and contemporary controls with no evidence of previous COVID-19 (SARS-CoV-2 nucleocapsid antibody negative) underwent multiorgan MRI (lungs, heart, brain, liver, and kidneys) with quantitative and qualitative assessment of images and clinical adjudication when relevant. Individuals with end-stage renal failure or contraindications to MRI were excluded. Participants also underwent detailed recording of symptoms, and physiological and biochemical tests. The primary outcome was the excess burden of multiorgan abnormalities (two or more organs) relative to controls, with further adjustments for potential confounders. The C-MORE study is ongoing and is registered with ClinicalTrials.gov, NCT04510025.

FINDINGS

Of 2710 participants in Tier 2 of PHOSP-COVID, 531 were recruited across 13 UK-wide C-MORE sites. After exclusions, 259 C-MORE patients (mean age 57 years [SD 12]; 158 [61%] male and 101 [39%] female) who were discharged from hospital with PCR-confirmed or clinically diagnosed COVID-19 between March 1, 2020, and Nov 1, 2021, and 52 non-COVID-19 controls from the community (mean age 49 years [SD 14]; 30 [58%] male and 22 [42%] female) were included in the analysis. Patients were assessed at a median of 5·0 months (IQR 4·2-6·3) after hospital discharge. Compared with non-COVID-19 controls, patients were older, living with more obesity, and had more comorbidities. Multiorgan abnormalities on MRI were more frequent in patients than in controls (157 [61%] of 259 vs 14 [27%] of 52; p<0·0001) and independently associated with COVID-19 status (odds ratio [OR] 2·9 [95% CI 1·5-5·8]; padjusted=0·0023) after adjusting for relevant confounders. Compared with controls, patients were more likely to have MRI evidence of lung abnormalities (p=0·0001; parenchymal abnormalities), brain abnormalities (p<0·0001; more white matter hyperintensities and regional brain volume reduction), and kidney abnormalities (p=0·014; lower medullary T1 and loss of corticomedullary differentiation), whereas cardiac and liver MRI abnormalities were similar between patients and controls. Patients with multiorgan abnormalities were older (difference in mean age 7 years [95% CI 4-10]; mean age of 59·8 years [SD 11·7] with multiorgan abnormalities vs mean age of 52·8 years [11·9] without multiorgan abnormalities; p<0·0001), more likely to have three or more comorbidities (OR 2·47 [1·32-4·82]; padjusted=0·0059), and more likely to have a more severe acute infection (acute CRP >5mg/L, OR 3·55 [1·23-11·88]; padjusted=0·025) than those without multiorgan abnormalities. Presence of lung MRI abnormalities was associated with a two-fold higher risk of chest tightness, and multiorgan MRI abnormalities were associated with severe and very severe persistent physical and mental health impairment (PHOSP-COVID symptom clusters) after hospitalisation.

INTERPRETATION

After hospitalisation for COVID-19, people are at risk of multiorgan abnormalities in the medium term. Our findings emphasise the need for proactive multidisciplinary care pathways, with the potential for imaging to guide surveillance frequency and therapeutic stratification.

FUNDING

UK Research and Innovation and National Institute for Health Research.

Weakly supervised spatial relation extraction from radiology reports

Objective

Weak supervision holds significant promise to improve clinical natural language processing by leveraging domain resources and expertise instead of large manually annotated datasets alone. Here, our objective is to evaluate a weak supervision approach to extract spatial information from radiology reports.

Materials and Methods

Our weak supervision approach is based on data programming that uses rules (or labeling functions) relying on domain-specific dictionaries and radiology language characteristics to generate weak labels. The labels correspond to different spatial relations that are critical to understanding radiology reports. These weak labels are then used to fine-tune a pretrained Bidirectional Encoder Representations from Transformers (BERT) model.

Results

Our weakly supervised BERT model provided satisfactory results in extracting spatial relations without manual annotations for training (spatial trigger F1: 72.89, relation F1: 52.47). When this model is further fine-tuned on manual annotations (relation F1: 68.76), performance surpasses the fully supervised state-of-the-art.

Discussion

To our knowledge, this is the first work to automatically create detailed weak labels corresponding to radiological information of clinical significance. Our data programming approach is (1) adaptable as the labeling functions can be updated with relatively little manual effort to incorporate more variations in radiology language reporting formats and (2) generalizable as these functions can be applied across multiple radiology subdomains in most cases.

Conclusions

We demonstrate a weakly supervision model performs sufficiently well in identifying a variety of relations from radiology text without manual annotations, while exceeding state-of-the-art results when annotated data are available.

quEHRy: a question answering system to query electronic health records

OBJECTIVE

We propose a system, quEHRy, to retrieve precise, interpretable answers to natural language questions from structured data in electronic health records (EHRs).

MATERIALS AND METHODS

We develop/synthesize the main components of quEHRy: concept normalization (MetaMap), time frame classification (new), semantic parsing (existing), visualization with question understanding (new), and query module for FHIR mapping/processing (new). We evaluate quEHRy on 2 clinical question answering (QA) datasets. We evaluate each component separately as well as holistically to gain deeper insights. We also conduct a thorough error analysis for a crucial subcomponent, medical concept normalization.

RESULTS

Using gold concepts, the precision of quEHRy is 98.33% and 90.91% for the 2 datasets, while the overall accuracy was 97.41% and 87.75%. Precision was 94.03% and 87.79% even after employing an automated medical concept extraction system (MetaMap). Most incorrectly predicted medical concepts were broader in nature than gold-annotated concepts (representative of the ones present in EHRs), eg, Diabetes versus Diabetes Mellitus, Non-Insulin-Dependent.

DISCUSSION

The primary performance barrier to deployment of the system is due to errors in medical concept extraction (a component not studied in this article), which affects the downstream generation of correct logical structures. This indicates the need to build QA-specific clinical concept normalizers that understand EHR context to extract the "relevant" medical concepts from questions.

CONCLUSION

We present an end-to-end QA system that allows information access from EHRs using natural language and returns an exact, verifiable answer. Our proposed system is high-precision and interpretable, checking off the requirements for clinical use.

Logical definition-based identification of potential missing concepts in SNOMED CT

BACKGROUND

Biomedical ontologies are representations of biomedical knowledge that provide terms with precisely defined meanings. They play a vital role in facilitating biomedical research in a cross-disciplinary manner. Quality issues of biomedical ontologies will hinder their effective usage. One such quality issue is missing concepts. In this study, we introduce a logical definition-based approach to identify potential missing concepts in SNOMED CT. A unique contribution of our approach is that it is capable of obtaining both logical definitions and fully specified names for potential missing concepts.

METHOD

The logical definitions of unrelated pairs of fully defined concepts in non-lattice subgraphs that indicate quality issues are intersected to generate the logical definitions of potential missing concepts. A text summarization model (called PEGASUS) is fine-tuned to predict the fully specified names of the potential missing concepts from their generated logical definitions. Furthermore, the identified potential missing concepts are validated using external resources including the Unified Medical Language System (UMLS), biomedical literature in PubMed, and a newer version of SNOMED CT.

RESULTS

From the March 2021 US Edition of SNOMED CT, we obtained a total of 30,313 unique logical definitions for potential missing concepts through the intersecting process. We fine-tuned a PEGASUS summarization model with 289,169 training instances and tested it on 36,146 instances. The model achieved 72.83 of ROUGE-1, 51.06 of ROUGE-2, and 71.76 of ROUGE-L on the test dataset. The model correctly predicted 11,549 out of 36,146 fully specified names in the test dataset. Applying the fine-tuned model on the 30,313 unique logical definitions, 23,031 total potential missing concepts were identified. Out of these, a total of 2,312 (10.04%) were automatically validated by either of the three resources.

CONCLUSIONS

The results showed that our logical definition-based approach for identification of potential missing concepts in SNOMED CT is encouraging. Nevertheless, there is still room for improving the performance of naming concepts based on logical definitions.

Application of an ontology for model cards to generate computable artifacts for linking machine learning information from biomedical research

Model card reports provide a transparent description of machine learning models which includes information about their evaluation, limitations, intended use, etc. Federal health agencies have expressed an interest in model cards report for research studies using machine-learning based AI. Previously, we have developed an ontology model for model card reports to structure and formalize these reports. In this paper, we demonstrate a Java-based library (OWL API, FaCT++) that leverages our ontology to publish computable model card reports. We discuss future directions and other use cases that highlight applicability and feasibility of ontology-driven systems to support FAIR challenges.

SELF-SUPERVISED LEARNING WITH RADIOLOGY REPORTS, A COMPARATIVE ANALYSIS OF STRATEGIES FOR LARGE VESSEL OCCLUSION AND BRAIN CTA IMAGES

Scarcity of labels for medical images is a significant barrier for training representation learning approaches based on deep neural networks. This limitation is also present when using imaging data collected during routine clinical care stored in picture archiving communication systems (PACS), as these data rarely have attached the high-quality labels required for medical image computing tasks. However, medical images extracted from PACS are commonly coupled with descriptive radiology reports that contain significant information and could be leveraged to pre-train imaging models, which could serve as starting points for further task-specific fine-tuning. In this work, we perform a head-to-head comparison of three different self-supervised strategies to pre-train the same imaging model on 3D brain computed tomography angiogram (CTA) images, with large vessel occlusion (LVO) detection as the downstream task. These strategies evaluate two natural language processing (NLP) approaches, one to extract 100 explicit radiology concepts (Rad-SpatialNet) and the other to create general-purpose radiology reports embeddings (DistilBERT). In addition, we experiment with learning radiology concepts directly or by using a recent self-supervised learning approach (CLIP) that learns by ranking the distance between language and image vector embeddings. The LVO detection task was selected because it requires 3D imaging data, is clinically important, and requires the algorithm to learn outputs not explicitly stated in the radiology report. Pre-training was performed on an unlabeled dataset containing 1,542 3D CTA - reports pairs. The downstream task was tested on a labeled dataset of 402 subjects for LVO. We find that the pre-training performed with CLIP-based strategies improve the performance of the imaging model to detect LVO compared to a model trained only on the labeled data. The best performance was achieved by pre-training using the explicit radiology concepts and CLIP strategy.

Discerning conversational context in online health communities for personalized digital behavior change solutions using Pragmatics to Reveal Intent in Social Media (PRISM) framework

BACKGROUND

Online health communities (OHCs) have emerged as prominent platforms for behavior modification, and the digitization of online peer interactions has afforded researchers with unique opportunities to model multilevel mechanisms that drive behavior change. Existing studies, however, have been limited by a lack of methods that allow the capture of conversational context and socio-behavioral dynamics at scale, as manifested in these digital platforms.

OBJECTIVE

We develop, evaluate, and apply a novel methodological framework, Pragmatics to Reveal Intent in Social Media (PRISM), to facilitate granular characterization of peer interactions by combining multidimensional facets of human communication.

METHODS

We developed and applied PRISM to analyze peer interactions (N = 2.23 million) in QuitNet, an OHC for tobacco cessation. First, we generated a labeled set of peer interactions (n = 2,005) through manual annotation along three dimensions: communication themes (CTs), behavior change techniques (BCTs), and speech acts (SAs). Second, we used deep learning models to apply our qualitative codes at scale. Third, we applied our validated model to perform a retrospective analysis. Finally, using social network analysis (SNA), we portrayed large-scale patterns and relationships among the aforementioned communication dimensions embedded in peer interactions in QuitNet.

RESULTS

Qualitative analysis showed that the themes of social support and behavioral progress were common. The most used BCTs were feedback and monitoring and comparison of behavior, and users most commonly expressed their intentions using SAs-expressive and emotion. With additional in-domain pre-training, bidirectional encoder representations from Transformers (BERT) outperformed other deep learning models on the classification tasks. Content-specific SNA revealed that users' engagement or abstinence status is associated with the prevalence of various categories of BCTs and SAs, which also was evident from the visualization of network structures.

CONCLUSIONS

Our study describes the interplay of multilevel characteristics of online communication and their association with individual health behaviors.

Are physical performance and frailty assessments useful in targeting and improving access to adjuvant therapy in patients undergoing resection for pancreatic cancer?

BACKGROUND

Many patients fail to receive adjuvant chemotherapy following pancreatic cancer surgery. This study implemented a multimodal, multidisciplinary approach to improving recovery after pancreatoduodenectomy (the 'Fast Recovery' programme) and measured its impact on adjuvant chemotherapy uptake and nutritional decline. The predictive accuracies of a bundle of frailty and physical performance assessments, with respect to the recipient of adjuvant chemotherapy, were also evaluated.

RESULTS

The N = 44 patients treated after the introduction of the 'Fast Recovery' programme were not found to have a significantly higher adjuvant chemotherapy uptake than the N = 409 treated before the pathway change (80.5 vs. 74.3%, p = 0.452), but did have a significantly lower average weight loss at six weeks post-operatively (mean: 4.3 vs. 6.9 kg, p = 0.013). Of the pre-operative frailty and physical performance assessments tested, the 6-min walk test was found to be the strongest predictor of the receipt of adjuvant chemotherapy (area under the ROC curve: 0.91, p = 0.001); all patients achieving distances ≥ 360 m went on to receive adjuvant chemotherapy, compared to 33% of those walking < 360 m.

CONCLUSIONS

The multimodal 'Fast Recovery' programme was not found to significantly improve access to adjuvant chemotherapy, but did appear to have benefits in reducing nutritional decline. Pre-operative assessments were found to be useful in identifying patients at risk of non-receipt of adjuvant therapies, with markers of physical performance appearing to be the best predictors. As such, these markers could be useful in targeting pre- and post-habilitation measures, such as physiotherapy and improved dietetic support.

Cohort profile: the ESC EURObservational Research Programme Non-ST-segment elevation myocardial infraction (NSTEMI) Registry

AIMS

The European Society of Cardiology (ESC) EURObservational Research Programme (EORP) Non-ST-segment elevation myocardial infarction (NSTEMI) Registry aims to identify international patterns in NSTEMI management in clinical practice and outcomes against the 2015 ESC Guidelines for the management of acute coronary syndromes in patients presenting without ST-segment-elevation.

METHODS AND RESULTS

Consecutively hospitalised adult NSTEMI patients (n = 3620) were enrolled between 11 March 2019 and 6 March 2021, and individual patient data prospectively collected at 287 centres in 59 participating countries during a two-week enrolment period per centre. The registry collected data relating to baseline characteristics, major outcomes (in-hospital death, acute heart failure, cardiogenic shock, bleeding, stroke/transient ischaemic attack, and 30-day mortality) and guideline-recommended NSTEMI care interventions: electrocardiogram pre- or in-hospital, pre-hospitalization receipt of aspirin, echocardiography, coronary angiography, referral to cardiac rehabilitation, smoking cessation advice, dietary advice, and prescription on discharge of aspirin, P2Y12 inhibition, angiotensin converting enzyme inhibitor (ACEi)/angiotensin receptor blocker (ARB), beta-blocker, and statin.

CONCLUSION

The EORP NSTEMI Registry is an international, prospective registry of care and outcomes of patients treated for NSTEMI, which will provide unique insights into the contemporary management of hospitalised NSTEMI patients, compliance with ESC 2015 NSTEMI Guidelines, and identify potential barriers to optimal management of this common clinical presentation associated with significant morbidity and mortality.

Parathyroid hormone changes in infants investigated for inflicted injury; an observational retrospective single centre cohort study

BACKGROUND

Biochemical and haematological testing is recommended in the United Kingdom when inflicted injury is suspected. We examined the associations of test results with radiologically-confirmed fracture(s), and between test results, in a large retrospective observational cohort.

METHODS

Infants up to age two years presenting with suspected inflicted injury, without clinically or radiologically apparent bone disease, and where a skeletal survey was undertaken during the period 1st August 2013 to 31st December 2020, were included. Biochemical parameters: corrected calcium (cCa); phosphate (P); alkaline phosphatase (ALP); parathyroid hormone (PTH); 25-hydroxyvitamin D (25D); and haematological parameters: haemoglobin (Hb); mean corpuscular haemoglobin (MCH); mean corpuscular haemoglobin content (MCHC); mean corpuscular volume (MCV); platelet count were collated together with the results of the radiological assessments.

FINDINGS

Of 332 eligible infants (190 male), 142 (84 male) had fracture(s) and/or intracranial injury. Mean PTH in the non-fracture group (n measured 50/190) was 27.3 ng/l; in those with intracranial injury alone (n measured 9/23) was 39.4 ng/l; in those with fracture alone (n measured 62/84) was 45.0 ng/l; and in those with fracture and intracranial injury (n measured 20/35) 51.8 ng/l. F-test of multiple means = 0.0369. There was no difference in 25D between the groups.

INTERPRETATION

PTH was raised in infants who had fracture(s), intracranial injury or both. A single raised PTH may not necessarily be an indicator of prior disturbed skeletal health in these circumstances. The relevance of vitamin D status and interpretation of data from biochemical testing should be informed by the overall presentation in suspected inflicted injury cases. A single raised PTH may be a consequence of the child's injuries rather than prior disturbed bone health.

Epidemiology and impact of frailty in patients with atrial fibrillation in Europe

BACKGROUND

Frailty is a medical syndrome characterised by reduced physiological reserve and increased vulnerability to stressors. Data regarding the relationship between frailty and atrial fibrillation (AF) are still inconsistent.

OBJECTIVES

We aim to perform a comprehensive evaluation of frailty in a large European cohort of AF patients.

METHODS

A 40-item frailty index (FI) was built according to the accumulation of deficits model in the AF patients enrolled in the ESC-EHRA EORP-AF General Long-Term Registry. Association of baseline characteristics, clinical management, quality of life, healthcare resources use and risk of outcomes with frailty was examined.

RESULTS

Among 10,177 patients [mean age (standard deviation) 69.0 (11.4) years, 4,103 (40.3%) females], 6,066 (59.6%) were pre-frail and 2,172 (21.3%) were frail, whereas only 1,939 (19.1%) were considered robust. Baseline thromboembolic and bleeding risks were independently associated with increasing FI. Frail patients with AF were less likely to be treated with oral anticoagulants (OACs) (odds ratio 0.70, 95% confidence interval 0.55-0.89), especially with non-vitamin K antagonist OACs and managed with a rhythm control strategy, compared with robust patients. Increasing frailty was associated with a higher risk for all outcomes examined, with a non-linear exponential relationship. The use of OAC was associated with a lower risk of outcomes, except in patients with very/extremely high frailty.

CONCLUSIONS

In this large cohort of AF patients, there was a high burden of frailty, influencing clinical management and risk of adverse outcomes. The clinical benefit of OAC is maintained in patients with high frailty, but not in very high/extremely frail ones.

Clinical characteristics with inflammation profiling of long COVID and association with 1-year recovery following hospitalisation in the UK: a prospective observational study

BACKGROUND

No effective pharmacological or non-pharmacological interventions exist for patients with long COVID. We aimed to describe recovery 1 year after hospital discharge for COVID-19, identify factors associated with patient-perceived recovery, and identify potential therapeutic targets by describing the underlying inflammatory profiles of the previously described recovery clusters at 5 months after hospital discharge.

METHODS

The Post-hospitalisation COVID-19 study (PHOSP-COVID) is a prospective, longitudinal cohort study recruiting adults (aged ≥18 years) discharged from hospital with COVID-19 across the UK. Recovery was assessed using patient-reported outcome measures, physical performance, and organ function at 5 months and 1 year after hospital discharge, and stratified by both patient-perceived recovery and recovery cluster. Hierarchical logistic regression modelling was performed for patient-perceived recovery at 1 year. Cluster analysis was done using the clustering large applications k-medoids approach using clinical outcomes at 5 months. Inflammatory protein profiling was analysed from plasma at the 5-month visit. This study is registered on the ISRCTN Registry, ISRCTN10980107, and recruitment is ongoing.

FINDINGS

2320 participants discharged from hospital between March 7, 2020, and April 18, 2021, were assessed at 5 months after discharge and 807 (32·7%) participants completed both the 5-month and 1-year visits. 279 (35·6%) of these 807 patients were women and 505 (64·4%) were men, with a mean age of 58·7 (SD 12·5) years, and 224 (27·8%) had received invasive mechanical ventilation (WHO class 7-9). The proportion of patients reporting full recovery was unchanged between 5 months (501 [25·5%] of 1965) and 1 year (232 [28·9%] of 804). Factors associated with being less likely to report full recovery at 1 year were female sex (odds ratio 0·68 [95% CI 0·46-0·99]), obesity (0·50 [0·34-0·74]) and invasive mechanical ventilation (0·42 [0·23-0·76]). Cluster analysis (n=1636) corroborated the previously reported four clusters: very severe, severe, moderate with cognitive impairment, and mild, relating to the severity of physical health, mental health, and cognitive impairment at 5 months. We found increased inflammatory mediators of tissue damage and repair in both the very severe and the moderate with cognitive impairment clusters compared with the mild cluster, including IL-6 concentration, which was increased in both comparisons (n=626 participants). We found a substantial deficit in median EQ-5D-5L utility index from before COVID-19 (retrospective assessment; 0·88 [IQR 0·74-1·00]), at 5 months (0·74 [0·64-0·88]) to 1 year (0·75 [0·62-0·88]), with minimal improvements across all outcome measures at 1 year after discharge in the whole cohort and within each of the four clusters.

INTERPRETATION

The sequelae of a hospital admission with COVID-19 were substantial 1 year after discharge across a range of health domains, with the minority in our cohort feeling fully recovered. Patient-perceived health-related quality of life was reduced at 1 year compared with before hospital admission. Systematic inflammation and obesity are potential treatable traits that warrant further investigation in clinical trials.

FUNDING

UK Research and Innovation and National Institute for Health Research.

Toward a standard formal semantic representation of the model card report

BACKGROUND

Model card reports aim to provide informative and transparent description of machine learning models to stakeholders. This report document is of interest to the National Institutes of Health's Bridge2AI initiative to address the FAIR challenges with artificial intelligence-based machine learning models for biomedical research. We present our early undertaking in developing an ontology for capturing the conceptual-level information embedded in model card reports.

RESULTS

Sourcing from existing ontologies and developing the core framework, we generated the Model Card Report Ontology. Our development efforts yielded an OWL2-based artifact that represents and formalizes model card report information. The current release of this ontology utilizes standard concepts and properties from OBO Foundry ontologies. Also, the software reasoner indicated no logical inconsistencies with the ontology. With sample model cards of machine learning models for bioinformatics research (HIV social networks and adverse outcome prediction for stent implantation), we showed the coverage and usefulness of our model in transforming static model card reports to a computable format for machine-based processing.

CONCLUSIONS

The benefit of our work is that it utilizes expansive and standard terminologies and scientific rigor promoted by biomedical ontologists, as well as, generating an avenue to make model cards machine-readable using semantic web technology. Our future goal is to assess the veracity of our model and later expand the model to include additional concepts to address terminological gaps. We discuss tools and software that will utilize our ontology for potential application services.

The impulsive horse: comparing genetic, physiological and behavioral indicators to that of human addiction

Stress and genotype elicit changes in impulse control in a range of species that are attributable to adaptations in both the central and peripheral nervous system. We examined aspects of this mechanism in the horse by assessing the effect of a dopamine receptor genotype (DRD4) and central dopaminergic tone (measured via spontaneous blink rate [SBR] and behavioral initiation rate [BIR]), on measures of impulsivity, compulsivity (3-choice serial reaction time task) and sympathetic/ parasympathetic system balance (heart rate variability [HRV]). Genotype did not have a significant effect on any of the parameters measured. SBR but not BIR correlated significantly with levels of impulsivity. There was no clear association of HRV parameters with either measures of central dopaminergic activity or impulsivity/compulsivity. Overall, some elements of the data suggest that the horse may be a useful animal model for assessing the genetic and environmental factors that lead to the physiological and behavioral phenotype of human addiction, particularly when considering the relationship between central dopaminergic tone and impulsivity.

AB1094 SAFETY AND EFFICACY OF VACCINES FOR SARS-CoV-2 IN PATIENTS WITH RHEUMATIC AND IMMUNE-MEDIATED INFLAMMATORY DISEASES: DATA FROM THE ARGENTINEAN REGISTRY SAR-CoVAC

Background

Currently there is little information on the efficacy and safety of SARS-CoV-2 vaccination in patients with immune-mediated diseases and/or under immunosuppressive treatment in our country, where different types of vaccines and mix regimens are used. For this reason, the Argentine Society of Rheumatology (SAR) with the Argentine Society of Psoriasis (SOARPSO) set out to develop a national register of patients with rheumatic and immune-mediated inflammatory diseases (IMIDs) who have received a SARS-CoV-2 vaccine in order to assess their efficacy and safety in this population.

Objectives

To assess SARS-CoV-2 vaccine efficacy and safety in patients with rheumatic and IMIDs.

Methods

SAR-CoVAC is a national, multicenter and observational registry. Adult patients with a diagnosis of rheumatic or IMIDs who have been vaccinated for SARS-CoV-2 were consecutively included between June 1st and September 17th, 2021. Sociodemographic data, comorbidities, underlying rheumatic or IMIDs, treatments received and their modification prior to vaccination and history of SARS-CoV-2 infection were recorded. In addition, the date and place of vaccination, type of vaccine applied, scheme and indication will be registered. Finally, adverse events (AE), as well as SARS-CoV-2 infection after the application of the vaccine were documented

Results

A total of 1234 patients were included, 79% were female, with a mean age of 57.8 (SD 14.1) years. The most frequent diseases were rheumatoid arthritis (41.2%), osteoarthritis (14.5%), psoriasis (12.7%) and spondyloarthritis (12.3%). Most of them were in remission (28.5%) and low disease activity (41.4%). At the time of vaccination, 21% were receiving glucocorticoid treatment, 35.7% methotrexate, 29.7% biological (b) Disease Modifying Anti-Rheumatic Drugs (DMARDs) and 5.4% JAK inhibitors. Before vaccine application 16.9% had had a SARS-CoV-2 infection.

Regarding the first dose of the vaccine, the most of the patients (51.1%) received Gam-COVID-Vac, followed by ChAdOx1 nCoV-19 (32.8%) and BBIBP-CorV (14.5%). In a lesser proportion, BNT162b2 (0.6%), Ad26.COV2.S (0.2%) and CoronaVac (0.2%) vaccines were used. Almost half of them (48.8%) completed the scheme, 12.5% were mix regimenes, the most frequent being Gam-COVID-Vac / mRNA-1273. The median time between doses was 51days (IQR 53).

More than a quarter (25.9%) of the patients reported at least one AE after the first dose and 15.9% after the second. The flu-like syndrome and local hypersensitivity were the most frequent manifestations. There was one case of mild anaphylaxis. No patient was hospitalized. Altogether, the incidence of AE was 246.5 events/1000 doses. BBIBP-CorV presented significantly lower incidence of AE in comparison with the other types of vaccines. (118.5 events/1000 doses, p<0.002 in all cases)

Regarding efficacy, 63 events of SARS-CoV-2 infection were reported after vaccination, 19% occurred before 14 days post-vaccination, 57.1% after the first dose (>14 days) and 23.8% after the second. In most cases (85.9%) the infection was asymptomatic or had an outpatient course and 2 died due to COVID-19.

Conclusion

In this national cohort of patients with rheumatic and IMIDs vaccinated for SARS-CoV-2, the most widely used vaccines were Gam-COVID-Vac and ChAdOx1 nCoV-19, approximately half completed the schedule and in most cases homologously. A quarter of the patients presented some AE, while 5.1% presented SARS-CoV-2 infection after vaccination, in most cases mild.

Disclosure of Interests

None declared

POS1199 IS PSORIATIC ARTHRITIS A RISK FACTOR FOR SEVERE COVID -19 INFECTION? DATA FROM THE ARGENTINIAN REGISTRY SAR-COVID

Background

Comorbidities, particularly cardio-metabolic disorders, are highly prevalent in patients with psoriatic arthritis (PsA) and they were associated with an increased risk of atherosclerotic cardiovascular disease, which have been associated with higher morbidity and mortality. Whether PsA enhances the risk of SARS-CoV-2 infection or affects the disease outcome remains to be ascertained.

Objectives

To describe the sociodemographic, clinical and treatment characteristics of patients with PsA with confirmed SARS-CoV-2 infection from the SAR-COVID registry and to identify the variables associated with poor COVID-19 outcomes, comparing them with those with rheumatoid arthritis (RA).

Methods

Cross-sectional observational study including patients ≥18 years old, with diagnosis of PsA (CASPAR criteria) and RA (ACR / EULAR 2010 criteria), who had confirmed SARS-CoV-2 infection (RT-PCR or serology) from the SAR-COVID registry. Recruitment period was between August 13, 2020 and July 31, 2021. Sociodemographic variables, comorbidities, and treatments were analyzed. To assess the severity of the infection, the ordinal scale of the National Institute of Allergy and Infectious Diseases (NIAID)1 was used, and it was considered that a patient met the primary outcome, if they presented criteria of categories 5 or higher on the severity scale. For this analysis, Chi2 test, Fisher’s test, Student’s test or Wilcoxon test, and binomial logistic regression using NIAID>=5 as dependent variable were performed.

Results

A total of 129 PsA patients and 808 with RA were included. Clinical characteristics are shown in Table 1. Regarding PsA treatment, 12.4% of PsA were receiving IL-17 inhibitors, 5.4% IL12-23 inhibitors, one patient apremilast and one abatacept. The frequency of NIAID≥5 was comparable between groups (PsA 19.5% vs RA 20.1%; p=0.976). (Figure 1).

Table 1.

Characteristics of patients with PsA and RA who presented COVID-19 in the SAR-COVID registry.

Psoriatic arthritis (n=129)Rheumatoid arthritis (n=808)P valueTotal (n=937)Age (years), mean (SD)51.7 (12.7)53.1 (12.9)0.23952.9 (12.9)Female72 (55.8)684 (84.7)<0.001756 (80.7)Comorbidities65 (50.4)355 (43.9)0.203420 (44.8) Obesity (BMI ≥30)19 (15.2)102 (13.4)0.692121 (13.7) Morbid obesity (BMI ≥40)1 (0.8)10 (1.3)111 (1.25) Hypertension35 (28.5)205 (26.8)0.783240 (27.0) Diabetes16 (13.0)67 (8.8)0.18883 (9.39) Dyslipidemia24 (19.5)102 (13.5)0.106126 (14.4) Cardiovascular or cerebrovascular disease5 (11.4)32 (3.9)0.03337 (4.2)Two or more comorbidities55 (42.6)219 (27.1)<0.001274 (29.2)Current smoking4 (3.6)60 (8.4)0.7964 (7.7)High disease activity0 (0)29 (3.8)0.02729 (3.23)Glucocorticoids treatment5 (20.0)95 (60.1)<0.001100 (54.6)Conventional DMARDs47 (36.4)443 (54.8)<0.001490 (52.3)Biologic DMARDs60 (46.5)193 (23.9)<0.001253 (27.0)JAK inhibitors4 (3.10)72 (8.9)0.03876 (8.1)Full recovery of COVID-19105 (84.0)644 (81.7)0.127749 (82.0)COVID-19 complications16 (12.5)68 (8.7)0.22784 (9.2)Death due to COVID-191 (0.8)34 (4.3)0.07435 (3.8)

Notes=values n (%) unless otherwise indicated; BMI: Body Mass Index; DMARDs: disease-modifying antirheumatic drugs; JAK inhibitors: Janus kinase inhibitors.

PsA patients with NIAID≥5 in comparison with NIAID<5 were older (58.6±11.4 vs 50±12.5; p=0.002), had more frequently hypertension (52.2% vs 23%; p=0.011) and dyslipidemia (39.1% vs 15%; p=0.017). In the multivariate analysis, age (OR 1.06; 95% CI 1.02–1.11) was associated with a worse outcome of the COVID-19 (NIAID≥5) in patients with PsA, while those who received methotrexate (OR 0.34; 95% CI 0.11–0.92) and biological DMARDs (OR 0.28; 95% CI 0.09–0.78) had a better outcome.

Conclusion

Although PsA patients have a higher frequency of cardiovascular and metabolic comorbidities than those with RA, the COVID-19 severity was similar. Most of the patients had mild SARS-CoV-2 infection and a low death rate.

References

[1]Beigel JH, et al. Remdesivir for the Treatment of Covid-19 - Final Report. N Engl J Med. 2020 Nov 5;383(19):1813-1826.

Disclosure of Interests

None declared

POS1204 SARS-CoV-2 INFECTION AFTER VACCINATION IN PATIENTS WITH RHEUMATIC DISEASES FROM ARGENTINA

Background

Patients with rheumatic diseases (RD) have been excluded from SARS-CoV-2 vaccine trials. Though data appear to show safety and efficacy, mostly evidence remains in mRNA vaccines. However in our country, adenovirus and inactivated vaccines, as well as heterologous schemes are frequently used.

Objectives

To describe clinical characteristics and outcomes of SARS-CoV-2 infection after vaccination in patients with RD from de the SAR-CoVAC registry and to compare them with patients who got infected before vaccination. Additionally, factors associated with COVID-19 unfavorable outcome were assessed.

Methods

Adult patients with RD who have been vaccinated for SARS-CoV-2 were consecutively included between June 1st and December 21st, 2021. Confirmed SARS-CoV-2 infection (RT-PCR o serology) was reported by the treated physician. Infection after an incomplete scheme was defined when the event was diagnosed at least 14 days after first dose; and after a complete scheme when it occurred > 14 days after second dose. Homologous scheme is defined by two same doses of vaccine and heterologous by two different doses. Patients with previous SARS-CoV-2 infection were excluded. To compare SARS-CoV-2 infection characteristics in not vaccinated patients, subjects from the SAR-COVID registry, which includes patients with RD and SARS-CoV-2 infection, were matched 2:1 by gender, age and RD.

WHO-Ordinal Scale ≥5 was used to define unfavorable infection outcome. Descriptive statics, Chi2 test, Fischer test, T test and ANOVA were used.

Results

A total of 1350 patients from the SAR COVAC registry were included, 67 (5%) presented SARS-CoV-2 infection after vaccination. The later were mostly (72%) females with a mean age of 57 (SD 15) years old. The most frequent RD were rheumatoid arthritis (41%), psoriatic arthritis (12%) and systemic lupus erythematosus (10%). At vaccination, most of them (75%) had low disease activity or remission, 19% were taking steroids, 39% methotrexate, 27% bDMARDs and 6% JAK inhibitors.

A total of 11 (16%) patients had SARS-CoV-2 infection <14 days after the first vaccine dose, 39 (58%) after an incomplete scheme and 17 (25 %) following a complete one. In the incomplete scheme group, 59% received Gam-COVID-Vac, 31% ChAdOx1 nCov-19 and 10% BBIBP-CorV; and in patients with complete scheme 47%, 24% and 29%, respectively. No event was reported after a complete heterologous scheme. No significant differences regarding sociodemoghraphic characteristics, RD, disease treatment, type of vaccine and regimen was found between in those with infection and those without it.

After vaccination only 8 (12%) of the patients who got infected had an unfavorable course, 88% of them following an incomplete scheme (5 received Gam-COVID-Vac, 1 ChAdOx1 nCov-19 and 1 BBIBP-CorV) and one subject after a complete homologous Gam-COVID-Vac scheme. Having an unfavorable outcome of SARS-CoV-2 infection was associated to: male gender [63% vs 24%, p=0.036], older age [mean 70 years (SD 7) vs 55 years (SD 15), p=0.005], being Caucasian [100% vs 54%, p=0.018], higher education [mean 17 years (SD 4) vs 12 years (SD 4), p=0.010], the presence of comorbidities [100% vs 39%, p=0.001, having pulmonary disease [37% vs 5%, p=0.019], dyslipidemia [63% vs 17%, p=0.011] and arterial hypertension [63% vs 24%, p=0.036], RD, treatments, disease activity and types of vaccines received were comparable between groups.

When comparing patients with and without vaccination prior SARS-CoV-2 infection, those who received at least one dose of vaccine had less frequently severe COVID-19 (12% vs 24%, p=0.067) and presented lower mortality due to COVID-19 (3% vs 6%, p=0.498). However these differences did not reach statistical significance.

Conclusion

In the SAR-CoVAC registry 5% of the patients had SARS-CoV-2 infection after vaccination, most of them mild and 25% after a complete scheme. Any vaccine was associated with severe COVID-19. When comparing with non-vaccinated patients, those with at least one dose, had less frequently severe disease and died due COVID-19.

Disclosure of Interests

None declared

POS1200 DIFFERENCES BETWEEN THE FIRST AND THE SECOND WAVE OF SARS-COV-2 INFECTION IN PATIENTS WITH IMMUNE-MEDIATED INFLAMMATORY DISEASES IN ARGENTINA: DATA FROM THE SAR-COVID REGISTRY

Background

In Argentina we have witnessed two COVID 19 waves between 2020 and 2021. The first wave occurred during the spring of 2020 and it was related to the wild type of the virus, the second occurred during the fall/winter of 2021 when the gamma variant showed a clear predominance. During the first wave, patient with rheumatic diseases showed a higher frequency of hospitalization and mortality (4% vs 0.26%) when compared to the general population1; at that time, however, vaccination was not yet available.

Objectives

To compare sociodemographic and disease characteristics, course and outcomes of SARS-CoV-2 infection in patients with immune-mediated/autoinflammatory diseases (IMADs) during the first and second waves in Argentina.

Methods

SAR-COVID is a national, multicenter, longitudinal and observational registry, in which patients ≥18 years of age, with a diagnosis of a rheumatic disease who had confirmed SARS-CoV-2 infection (RT-PCR or positive serology) were consecutively included since August 2020. For the purpose of this report, only patients with IMADs who had SARS-CoV-2 infection during the first wave (defined as cases occurred between March 2020 and March 2021) and the second wave (cases occurred between April and August 2021) were examined. Sociodemographic characteristics, disease diagnosis and activity, comorbidities, immunosuppressive treatment and COVID 19 clinical characteristics, complications and outcomes: hospitalization, intensive care unit (ICU) admission, use of mechanical ventilation and death were compared among groups. Descriptive statistical analysis was performed. Variables were compared with Chi squared test and Student T test or Mann Whitney test. Multivariable logistic regression models with forward and backward selection method, using hospitalization, ICU admission and death as dependent variables were carried out.

Results

A total of 1777 patients were included, 1342 from the first wave and 435 of the second one. Patients had a mean (SD) age of 50.7 (14.2) years and 81% were female. Both groups of patients were similar in terms of socio-demographic features, disease diagnosis, disease activity, the use of glucocorticoids ≥ 10 mg/day and the immunosuppressive drugs (Table 1 below). Patients infected during the first wave have higher frequency of comorbidities (49% vs 41%; p= 0.004). Hospitalizations due to COVID 19 (31% vs 20%; p <0.001) and ICU admissions (9% vs 5%; p= 0.009) were higher during the first wave. No differences in the use of mechanical ventilation (16% vs 16%; p= 0.97) nor in the mortality rate (5% vs 4%; p= 0.41) were observed. In the multivariable analysis, after adjusting for demographics, clinical features and immunosuppressive treatment, patients infected during the second wave were 40% less likely to be hospitalized (OR= 0.6, IC95% 0.4-0.8) and to be admitted to the ICU (OR= 0.6, IC95% 0.3-0.9).

Table 1.Variable (% or Mean – SD)First wave(n=1342)Second wave(n=435)p ValueFemale gender81800.7Age (years)51.0 (14.5)50.0 (13.3)0.2Disease diagnosis Rheumatoid arthritis46461 Ankylosing spondylitis10110.8 Systemic lupus erythematosus171850.9 Systemic Scleroderma551 Sjögren´s syndrome650.7 Inflammatory myopathies330.5 Vasculitis430.4Disease activity High430.5Use of immune modulatorsDMARDcs53560.2DMARDts460.1DMARDb82821Use of glucocorticoids ≥10 mg12120.9Comorbidities49410.004

Conclusion

The impact of COVID 19 in Argentina, in terms of mortality in patients with IMADs was still higher compared to the general population during the second wave. However, the frequency of hospitalizations and ICU admissions was lower. These findings could be explained by the introduction of the SARS COV 2 vaccination and, probably, by the cumulative knowledge and management improvement of this infection among physicians.

References

[1]Isnardi CA et al. Epidemiology and outcomes of patients with rheumatic diseases and SARS-COV-2 infection: data from the argentinean SAR-COVID Registry. Ann Rheum Dis, 2021, suppl 1, 887.

Disclosure of Interests

None declared

AB1101 PREVALENCE OF LONG COVID IN RHEUMATIC DISEASE PATIENTS: ANALYSIS OF SAR COVID REGISTRY

Background

Persistent symptoms after acute COVID have been described previously. Main symptoms reported are fatigue, arthralgias, myalgias and mental sickness. Definition and methods vary widely.1

Objectives

To asses prevalence and related factors to long COVID in a retrospective cohort of patients with rheumatic diseases from Argentina.

Methods

A total of 1915 patients were registered from August 18th, 2020 to July 29th, 2021. Patients > 18 years old, with rheumatic disease and confirmed infection by SARS-CoV-2 (antigen or RT-PCR) were included. Those dead, with unknown outcome, wrong date or missing data were excluded. Demographic data, comorbidities, rheumatic disease, and characteristics of SARS-CoV-2 infection were recorded.

Long COVID was defined according to NICE guidelines (persistent symptoms for more than 4 weeks, without alternative diagnosis). Long COVID symptoms were defined by rheumatologist. Severity of infection was classified according to WHO ordinal scale.

We used descriptive statistics, univariate model (Student’s test, chi square test, ANOVA) and multivariate logistic regression analysis.

Results

230 (12%) had long COVID. Median age was 51 (IQR 40-61]) years, 82% were females, 51% were not caucasian. Median of education was 13.3 years (IQR 12 – 16), 79 % had private health insurance and 55 % were employed. Nearly half (n=762, 46%) had comorbidities, the most prevalent was hypertension (n=396, 24%).

The most frequent rheumatic diseases were rheumatoid arthritis (n=719, 42%) and systemic lupus erythematosus (n=280, 16 %). Most were in low activity/remission (79%), used Conventional DMARD (n=773 patients, 45%) and steroids (n=588, 34%) at low dose (n=415, 71%).

Main laboratory findings were abnormal D-dimer (n=94, 28%) and leukopenia (n=93, 26%). Most patients had a WHO ordinal scale < 5 (n=1472, 86%). Median of hospitalization at intensive care unit (ICU) was 8 days [IQR 5, 13]. Treatment for SARS-CoV-2 infection (steroids, anticoagulation, azithromycin, convalescent plasma) was used in 461 (27%) patients.

Most of long COVID (n= 152, 69%) reported 1 symptom, the most frequent was fatigue (n= 55, 22%). Figure 1.

Univariate analysis is presented in Table 1. In multivariate logistic regression analysis non-caucasian ethnicity OR 1.44 (1.07-1.95), years of education OR 1.05 (1-1.09), treatment with cyclophosphamide OR 11.35 (1.56-112.97), symptoms of COVID – 19 OR 13.26 (2.75-242.08), severity scale WHO ≥ 5 OR 2.46 (1.68-3.57), and ICU hospitalization days OR 1.09 (1.05-1.14) were factors associated to long COVID.

Table 1.

Univariate analysis of long COVID syndrome in SAR – COVID registry

VariableAcute COVID n=1486Long COVID n=221P valueAge, years, median [IQR]51 [40, 60]54 [42, 62]0.032Caucasian, n (%)744 (48)132 (53)0.227Female sex, n (%)1242 (80)215 (86)0.066Education, years, median [IQR]12 [10, 17]13 [12, 16]-Private health insurance, n (%)1161 (79)181 (82)0.325Smoking, n (%)381 (25)71 (29)0.224Comorbidities, n (%)650 (45)108 (52)0.066Dyslipidemia, n (%)173 (12)39 (19)0.008Hypertension, n (%)332 (23)60 (29)0.053Low activity/remission disease, n (%)1140 (80)179 (77)1Rheumatoid arthritis, n (%)623 (42)96 (42)1Systemic lupus erythematosus, n (%)243 (16)37 (16)0.996DMARD, n (%)664 (45)109 (47)0.486Cyclophosphamide, n (%)3 (0.2)3 (1)0.035Rituximab, n (%)19 (1)9 (34)0.008Lymphocyte66 (23)19 (30)0.011<1.500 / mm3, n (%)Ferritin > 2000 ng/ml, n (%)32 (11)16 (25)0.011ICU hospitalization, days,7 [4, 10]10 [8, 24]<0.001median [IQR]Treatment for COVID-19, n (%)394 (27)91 (41)<0.001

Conclusion

Prevalence of long COVID was 12%. Non-caucasian ethnicity, higher education, treatment with cyclophosphamide, symptoms of COVID – 19, severe disease and ICU hospitalization days were related to long COVID.

References

[1]Cabrera Martimbianco AL, Pacheco RL, Bagattini ÂM, Riera R. Frequency, signs and symptoms, and criteria adopted for long COVID-19: A systematic review. Int J Clin Pract.

Disclosure of Interests

None declared

POS1238 GLUCOCORTICOIDS, RITUXIMAB AND THE PRESENCE OF INTERSTITIAL LUNG DISEASE ARE ASSOCIATED WITH POOR OUTCOMES OF THE SARS-COV-2 INFECTION IN PATIENTS WITH RHEUMATOID ARTHRITIS: DATA FROM THE NATIONAL REGISTRY SAR-COVID

Background

High disease activity, treatment with glucocorticoids (GC) and rituximab (RTX), have been related to worse outcomes of COVID-19.

Objectives

To assess the clinical characteristics and severity of the SARS-CoV-2 infection in patients with rheumatoid arthritis (RA) included in the SAR-COVID registry and to identify factors associated with poor outcomes.

Methods

SAR-COVID is a national, longitudinal and observational registry. Patients of ≥18 years old, with diagnosis of RA (ACR-EULAR criteria 2010) who had confirmed SARS-CoV-2 infection (RT-PCR or positive serology) were included between 13-8-20 and 31-7-21. Sociodemographic and clinical data, comorbidities, disease activity and treatment at the moment of the SARS-CoV-2 infection were collected. Additionally, infection symptoms, complications, medical interventions and treatments for COVID-19 were registered. Infection severity was assessed using the WHO-ordinal scale (WHO-OS)1. A cut-off value of ≥5 identified patients with severe COVID-19 and those who died.

Statistical analysis: Descriptive statistics. Chi2 or Fischer test, Student T test or Mann-Whitney and Kruskal Wallis or ANOVA, as appropriate. Multiple logistic regression model.

Results

A total of 801 patients were included, with a mean age of 53.1 ± 12.9 years, most of them were female (84.5%) and the median (m) disease duration was 8 years (IQR 4-14). One third were in remission and 46.4% had comorbidities, being the most frequent, hypertension (26.9 %), dyslipidemia (13.5 %), obesity (13.4 %) and diabetes (8.9%). Moreover, 3.2% had interstitial lung disease (ILD) associated with RA. At SARS-CoV-2 diagnosis, 42.5% were receiving glucocorticoids (GC), 73.9% conventional (c) disease modifying antirheumatic drugs (DMARD), 24% biologic (b) DMARD and 9.1% targeted synthetic (ts) DMARD. Among bDMARD, the most frequently used were TNF inhibitors (17%), followed by abatacept (2.8%), IL-6 inhibitors (2.4%) and rituximab (RTX) (2.1%). During the SARS-CoV-2 infection, 95.8% had symptoms, 27% required hospitalization, 7.9% presented complications and 4.4% died due to COVID-19. Severe disease and death (WHO-OS≥5) was present in 7.5% of the patients. They were older (62.9±12.5 vs 52.2±12.7, p<0.001), and they had more frequently ILD (18.5% vs 2%, p<0.001), comorbidities (82.5% vs 43.7%, p<0.001), ≥2 comorbidities (60.3% vs 25.8%, p<0.001), treatment with GC (61% vs 40.7%, p=0.04) and RTX (8.3% vs 1.6%, p=0.007). Conversely, the use of cDMARD and TNF inhibitors was more frequent in patients with WHO-OS<5, nevertheless this difference was not significant. Disease activity was comparable between groups. In multivariable analysis, older age, the presence of diabetes, ILD, the use of GC and RTX were significantly associated with WHO-OS≥5 (Figure 1). Furthermore, older age (65.7±10.8 vs 52.4±12.8, p<0.001), the presence of comorbidities (87.9% vs 44.7%, p<0.001), chronic obstructive pulmonary disease (21.9% vs 5.2%, p=0.002), diabetes (30.3% vs 7.9%, p<0.001), hypertension (57.6% vs 25.6%, p<0.001), cardiovascular disease (15.6% vs 3.2%, p=0.005), cancer (9.1% vs 1.3%, p=0.001), ILD (23.3% vs 2.4%, p<0.001) and the use of GC (61.8% vs 41.4%, p=0.02) were associated with mortality. Older age [OR 1.1 IC95% 1.06-1.13] and the use of GC 5-10 mg/day [OR 4.6 IC95% 1.8-11.6] remained significantly associated with death due to COVID-19.

Figure 1.

Factors associated with severe disease and death due to COVID-19 (WHO-OS≥5) in patients with rheumatoid arthritis. Multivariable analysis. (ref.: reference; PDN: prednisone; OR: odds ratio; CI: confidence interval)

Conclusion

Treatment with RTX and GC, as well as older age, the presence of diabetes and ILD were associated with poor COVID-19 outcomes in this national cohort of patients with RA. Older patients and those taking GC had a higher mortality rate.

References

[1]World Health Organization coronavirus disease (COVID-19) Therapeutic Trial Synopsis Draft 2020.

Disclosure of Interests

None declared

POS1201 SAFETY OF SARS-COV-2 VACCINES IN PATIENTS WITH RHEUMATIC DISEASES: DATA FROM THE NATIONAL REGISTRY SAR-CoVAC FROM ARGENTINA

Background

Patients with rheumatic diseases (RD) have been excluded from SARS-CoV-2 vaccine trials, though data appear to show safety and efficacy, mostly evidence remains in mRNA vaccines. In our country, adenovirus-vector, inactivated and heterologous scheme vaccines are frequently used.

Objectives

To describe the safety of SARS-CoV-2 vaccines in patients with RD from the national registry SAR-CoVAC and to assess sociodemographic and clinical factors associated to AE and disease flares after vaccination.

Methods

Adult patients with RD who have been vaccinated for SARS-CoV-2 from de Argentine Society of Rheumatology Vaccine Registry (SAR-CoVAC) were consecutively included between June 1st and December 21st, 2021, This is a national multicentric observational registry that includes patients that have received at least one dose of any SARS-CoV-2 available vaccines in Argentina. Data is voluntarily collected by the treating physician. Naranjo scale was use to assess the association between the AE and vaccination.

Homologous and heterologous schedules were defined according to whether both vaccines received were the same or different, respectively. Descriptive statics, Chi2 test, Fischer test, T test, ANOVA and multivariate regression logistic model were used.

Results

A total of 1679 patients, with 2795 SARS-CoV-2 vaccine doses were included. Vaccines more frequently used were: Gam-COVID-Vac (1227 doses, 44%), ChAdOx1 nCov-19 (872 doses, 31%), BBIBP-CorV (482 doses, 17%) and mRAN-1273 (172 doses, 6%). Altogether, 510 EA were experienced by 449 (27%) patients. Pseudo-flu syndrome was the most frequent (11%), followed by injection site reaction (7%). They were significantly more frequent after the first dose in comparison to the second one (13% vs 7% and 9% vs 5%, respectively, p<0.001 in both cases). All were mild or moderate and no patient was hospitalized due to an AE. One case of moderate anaphylaxis was reported by a patient who received Gam-COVID-Vac. No cases of vaccine-induced thrombotic thrombocytopenia were observed. There were 25 disease flares reported, 17 (68%) cases of arthritis. Among patients with two doses, those with heterologous schedule presented AE more frequent after the second dose (39% vs 17%).

Total incidence of EA was 182.5 events/10 00 doses, it was significantly lower for BBIBP-CorV (105.9 events/1000 dosis, p<0.002 for all cases). The higher incidence of AE was observed for mRAN-1273 (261.6 events/1000 doses) and ChAdOx1 nCov-19 (232.8 events/1000 doses).

Patients with AE were younger [mean 55 years (SD 14) vs 59 years (SD 14), p <0.010], not Caucasian ethnicity [48% vs 35%, p<0.001], had higher education level [mean 13.8 years (SD 4) vs 11.9 years (SD 5), p<0.001], were more frequently employed [54% vs 44%, p<0.001], lived mostly in urban area [99% vs 95% p <0.001, had more frequently dyslipidemia [38% vs 28% p 0.012], and less frequently arterial hypertension [49% vs 65%, p<0.001]. Systemic lupus erythematosus [11% vs 7%, p=0.039] and Sjögren syndrome [6% vs 1.8%, p<0.001] were more frequent among them, while non inflammatory diseases were less prevalent [19% vs 31%, p<0.001]. They were taking steroids [24 vs 18%, p=0.007], antimalarials [17% vs 10%, p<0.001] and methotrexate [41% vs 31%, p <0.001] more frequently.

In the multivariable analysis, mRAN-1273 and ChAdOx1 nCov-19 were associated with AE, while BBIBP-CorV with lower probability of having one. (Figure 1)

Figure 1.

Variables associated with the development of AE. Multivariate logistic regression model

Conclusion

The incidence of AE was 1825 events/1000 doses, were significantly higher for mRAN-1273 and ChAdOx1 nCov-19 and lower for BBIBP-CorV. Most common AE was pseudo-flu syndrome. Female sex, being younger, higher education level, ChAdOx1 nCov-19 and mRAN-1273 vaccines, the use of methotrexate and antimalarials were related of EA in patients with RD.

References

[1]Sattui SE et al. Early experience of COVID-19 vaccination in adults with systemic rheumatic diseases: results from the COVID-19 Global Rheumatology Alliance Vaccine Survey. RMD Open 2021;7.

Disclosure of Interests

None declared

POS0655 SURVIVAL AND SAFETY OF BIOLOGICAL AND TARGETED SYNTHETIC THERAPIES AS REGARDS TO AGE GROUPS. BIOBADASAR 3.0 REGISTRY

Background

Advances in rheumatology and new therapeutic options have certainly impacted patient survival, changing the age range, from youth to seniors. The differences between the age groups could influence the evolution of the disease and the adverse events (AEs) related to the treatments. There are few real-world data on the safety and efficacy of treatments in different age groups.

Objectives

To evaluate the frequency of AEs and the survival of treatments according to the age in patients with rheumatoid arthritis (RA), psoriatic arthritis (PsA) or ankylosing spondylitis (AS).

Methods

Retrospective, observational, multicenter study of real-life data of patients included in the BIOBADASAR 3.0 registry; exposed and not exposed to original biological treatments (b-DMARDs), biosimilars, targeted synthetic drugs (ts-DMARDs). The unexposed group received treatment with conventional disease-modifying drugs (cDMARDs). A Kaplan-Meier and Log Rank Test analysis was performed to study AEs-free survival and treatment in different age groups (young people <25; young adults 25-34; mature adults 34-65; old adults >65). Factors related to treatment survival were evaluated using Cox regression models.

Results

5,297 patients were included, 80.3% female, mean age 43.7 years (SD 15.6) and median disease progression 14.3 [IQR 11.5]. RA 4658 (87.9%); APs 490 (9.25%) and EA 149 (2.8%). The main reason for treatment discontinuation was ineffectiveness, in 624 patients in the exposed group and in 53 (2.5%) patients in control group, followed by the presence of AEs in 352 (11.2%) and 83 (3.9%), respectively (p=0.001).

A mean Charlson Score of 0.268 (SD 0.6) in the exposed group and 0.306 (SD 0.7) in the control group (p=0.095). Median EAs-free survival in the exposed group was 12.5 years [IQR 16.6] while in controls was 28 years [IQR 11], p<0.0001. Median AEs-free survival was 12 years (IQR 11) in young people, 11.5 years [IQR: 4.9] in young adults, 10 years [IQR: 3.25] in mature adults and 7.6 years [IQR: 6] in old adults with a difference statistically significant (p>0.017). The exposed group presented a median treatment survival in years of 11.25 years [IQR: 10] in young people; 12.5 years [IQR: 4.7] in young adults, 7.5 years [IQR: 12.1] in mature adults and 4.5 years [IQR: 1.14] in old adults (p>0.0001). Considering only the first line of treatment, a median survival of 11.5 years [IQR: 10] was evidenced in the age group <25; 12 years [IQR: 2.6] between 25-34 years old, 10 years [IQR: 12] in the group between 34-65 years old and 5.5 years [IQR: 1.14] in the group > 65 years old (p>0.004). (Figure 1). Considering the second line of treatment, the differences between the groups were not statistically significant (p=0.57). In the multivariate regression model for patients with RA, the factors with the greatest impact on treatment survival were female sex (HR 1.3, 95% CI 1.2-1.4), old age (HR 1.01, 95% CI 1.008-1.01), treatment with steroids (HR 1.19, 95% CI1.1-1.2) and longer disease duration (HR 1.01, 95% CI1.01 – 1.02).

Conclusion

In the present study we were able to demonstrate a greater occurrence of AEs in old adults and mature adults compared to young people and young adults. Conversely, survival for b-DMARDs and ts-DMARDs were greater in youth and young adults. In patients with RA, female sex, corticosteroid therapy, old aged and longer disease duration were associated with treatment discontinuation.

References

[1]Souto A, et al. Rate of discontinuation and drug survival of biologic therapies in rheumatoid arthritis: a systematic review and meta-analysis of drug registries and health care databases. Rheumatology (Oxford). 2016;55(3):523–34.

[2]Ray D, et al. Immune senescence, epigenetics and autoimmunity. Clin Immunol. 2018 Nov;196:59-63. doi: 10.1016/j.clim.2018.04.002. Epub 2018 Apr 11.

[3]Vela P, et al. Influence of age on the occurrence of adverse events in rheumatic patients at the onset of biological treatment: data from the BIOBADASER III register. Arthritis Res Ther. 2020 Jun 15;22(1):143. doi: 10.1186/s13075-020-02231-x.

Disclosure of Interests

None declared

Benefits of, Barriers to, and Needs for an Artificial Intelligence-Powered Medication Information Voice Chatbot for Older Adults: Interview Study With Geriatrics Experts

BACKGROUND

One of the most complicated medical needs of older adults is managing their complex medication regimens. However, the use of technology to aid older adults in this endeavor is impeded by the fact that their technological capabilities are lower than those of much of the rest of the population. What is needed to help manage medications is a technology that seamlessly integrates within their comfort levels, such as artificial intelligence agents.

OBJECTIVE

This study aimed to assess the benefits, barriers, and information needs that can be provided by an artificial intelligence-powered medication information voice chatbot for older adults.

METHODS

A total of 8 semistructured interviews were conducted with geriatrics experts. All interviews were audio-recorded and transcribed. Each interview was coded by 2 investigators (2 among ML, PR, METR, and KR) using a semiopen coding method for qualitative analysis, and reconciliation was performed by a third investigator. All codes were organized into the benefit/nonbenefit, barrier/nonbarrier, and need categories. Iterative recoding and member checking were performed until convergence was reached for all interviews.

RESULTS

The greatest benefits of a medication information voice-based chatbot would be helping to overcome the vision and dexterity hurdles experienced by most older adults, as it uses voice-based technology. It also helps to increase older adults' medication knowledge and adherence and supports their overall health. The main barriers were technology familiarity and cost, especially in lower socioeconomic older adults, as well as security and privacy concerns. It was noted however that technology familiarity was not an insurmountable barrier for older adults aged 65 to 75 years, who mostly owned smartphones, whereas older adults aged >75 years may have never been major users of technology in the first place. The most important needs were to be usable, to help patients with reminders, and to provide information on medication side effects and use instructions.

CONCLUSIONS

Our needs analysis results derived from expert interviews clarify that a voice-based chatbot could be beneficial in improving adherence and overall health if it is built to serve the many medication information needs of older adults, such as reminders and instructions. However, the chatbot must be usable and affordable for its widespread use.

Deep learning-based NLP data pipeline for EHR-scanned document information extraction

Objective

Scanned documents in electronic health records (EHR) have been a challenge for decades, and are expected to stay in the foreseeable future. Current approaches for processing include image preprocessing, optical character recognition (OCR), and natural language processing (NLP). However, there is limited work evaluating the interaction of image preprocessing methods, NLP models, and document layout.

Materials and Methods

We evaluated 2 key indicators for sleep apnea, Apnea hypopnea index (AHI) and oxygen saturation (SaO2), from 955 scanned sleep study reports. Image preprocessing methods include gray-scaling, dilating, eroding, and contrast. OCR was implemented with Tesseract. Seven traditional machine learning models and 3 deep learning models were evaluated. We also evaluated combinations of image preprocessing methods, and 2 deep learning architectures (with and without structured input providing document layout information), with the goal of optimizing end-to-end performance.

Results

Our proposed method using ClinicalBERT reached an AUROC of 0.9743 and document accuracy of 94.76% for AHI, and an AUROC of 0.9523 and document accuracy of 91.61% for SaO2.

Discussion

There are multiple, inter-related steps to extract meaningful information from scanned reports. While it would be infeasible to experiment with all possible option combinations, we experimented with several of the most critical steps for information extraction, including image processing and NLP. Given that scanned documents will likely be part of healthcare for years to come, it is critical to develop NLP systems to extract key information from this data.

Conclusion

We demonstrated the proper use of image preprocessing and document layout could be beneficial to scanned document processing.

Closing the loop: automatically identifying abnormal imaging results in scanned documents

OBJECTIVES

Scanned documents (SDs), while common in electronic health records and potentially rich in clinically relevant information, rarely fit well with clinician workflow. Here, we identify scanned imaging reports requiring follow-up with high recall and practically useful precision.

MATERIALS AND METHODS

We focused on identifying imaging findings for 3 common causes of malpractice claims: (1) potentially malignant breast (mammography) and (2) lung (chest computed tomography [CT]) lesions and (3) long-bone fracture (X-ray) reports. We train our ClinicalBERT-based pipeline on existing typed/dictated reports classified manually or using ICD-10 codes, evaluate using a test set of manually classified SDs, and compare against string-matching (baseline approach).

RESULTS

A total of 393 mammograms, 305 chest CT, and 683 bone X-ray reports were manually reviewed. The string-matching approach had an F1 of 0.667. For mammograms, chest CTs, and bone X-rays, respectively: models trained on manually classified training data and optimized for F1 reached an F1 of 0.900, 0.905, and 0.817, while separate models optimized for recall achieved a recall of 1.000 with precisions of 0.727, 0.518, and 0.275. Models trained on ICD-10-labelled data and optimized for F1 achieved F1 scores of 0.647, 0.830, and 0.643, while those optimized for recall achieved a recall of 1.0 with precisions of 0.407, 0.683, and 0.358.

DISCUSSION

Our pipeline can identify abnormal reports with potentially useful performance and so decrease the manual effort required to screen for abnormal findings that require follow-up.

CONCLUSION

It is possible to automatically identify clinically significant abnormalities in SDs with high recall and practically useful precision in a generalizable and minimally laborious way.

Toward a Neural Semantic Parsing System for EHR Question Answering

Clinical semantic parsing (SP) is an important step toward identifying the exact information need (as a machine-understandable logical form) from a natural language query aimed at retrieving information from electronic health records (EHRs). Current approaches to clinical SP are largely based on traditional machine learning and require hand-building a lexicon. The recent advancements in neural SP show a promise for building a robust and flexible semantic parser without much human effort. Thus, in this paper, we aim to systematically assess the performance of two such neural SP models for EHR question answering (QA). We found that the performance of these advanced neural models on two clinical SP datasets is promising given their ease of application and generalizability. Our error analysis surfaces the common types of errors made by these models and has the potential to inform future research into improving the performance of neural SP models for EHR QA.

Patient and Provider Perspectives on Medication Non-adherence Among Patients with Depression and/or Diabetes in Diverse Community Settings - A Qualitative Analysis

BACKGROUND

Diabetes and depression affect a significant percentage of the world's total population, and the management of these conditions is critical for reducing the global burden of disease. Medication adherence is crucial for improving diabetes and depression outcomes, and research is needed to elucidate barriers to medication adherence, including the intentionality of non-adherence, to intervene effectively. The purpose of this study was to explore the perspectives of patients and health care providers on intentional and unintentional medication adherence among patients with depression and diabetes through a series of focus groups conducted across clinical settings in a large urban area.

METHODS

This qualitative study utilized a grounded theory approach to thematically analyze qualitative data using the framework method. Four focus groups in total were conducted, two with patients and two with providers, over a one-year period using a semi-structured facilitation instrument containing open-ended questions about experiences, perceptions and beliefs about medication adherence.

RESULTS

Across the focus groups, communication difficulties between patients and providers resulting in medication non-adherence was a primary theme that emerged. Concerns about medication side effects and beliefs about medication effectiveness were identified as perceptual barriers related to intentional medication non-adherence. Practical barriers to medication adherence, including medication costs, forgetting to take medications and polypharmacy, emerged as themes related to unintentional medication non-adherence.

CONCLUSION

The study findings contribute to a growing body of research suggesting health system changes are needed to improve provider education and implement multicomponent interventions to improve medication adherence among patients with depression and/or diabetes, both chronic illnesses accounting for significant disease burden globally.

O-P01 Potential Utility of Intraoperative Fluid Amylase Measurement During Pancreaticoduodenectomy

Background

Postoperative pancreatic fistula (POPF) after pancreaticoduodenectomy (PD) is a source of major morbidity and mortality. Early diagnosis and treatment of POPF is mandatory to improve patient outcomes, and clinical risk scores may be combined with postoperative drain fluid amylase (DFA) values to stratify patients.

The aim of this study was to determine if intraoperative fluid amylase values (IFA) correlate with DFA1 and POPF.

Methods

In consecutive patients undergoing PD between February and November 2020, intraoperative samples of intra-abdominal fluid adjacent to the pancreatic anastomosis were taken and sent for fluid amylase measurement prior to abdominal closure. Data regarding patient demographics, postoperative DFA values, complications and mortality were prospectively collected.

Results

Patient Demographics: Data was obtained for 52 patients with a median alternative Fistula Risk Score (aFRS) of 9.4. Postoperative complications occurred in 20 patients (38%), including five Clavien grade 3+. There were eight POPFs and two patients died (pneumonia/sepsis). There was significant correlation between IFA and DFA1 (Pearson’s correlation: R2=0.713; p &lt; 0.001) and DFA3 (p &lt; 0.001), and median IFA was higher in patients with POPF than patients without (1232.5 vs. 122; p = 0.0003). IFA&gt;260 U/l predicted POPF with sensitivity, specificity, PPV and NPV of 88%, 75%, 39% and 97%, respectively. The incidence of POPF was 43% in high risk (high aFRS/IFA) and 0% in low risk patients (low aFRS/IFA).

Complications:

Conclusions

Intraoperative fluid amylase closely correlated with postoperative pancreatic fistula, and may be a useful adjunct to clinical risk scores to stratify patients during pancreatico-duodenectomy, allowing targeted intervention to reduce the clinical impact of pancreatic fistula.It is possible to detect fluid amylase adjacent to the completed pancreatic anastomosis in patients undergoing PD. Presence of IFA correlates with postoperative DFA and POPF.

IFA increases the accuracy of the alternate Fistula Risk Score in predicting POPF. Low risk patients with a low IFA may be suitable for a ‘no drain’ strategy, whilst patients with a high IFA may benefit from intraoperative mitigation strategies to reduce the incidence and/or severity of a postoperative pancreatic fistula.

Dental EHR-infused Persona Ontologies to Enrich Dental Dialogue Interaction of Agents

The quality of patient-provider communication can predict the healthcare outcomes in patients, and therefore, training dental providers to handle the communication effort with patients is crucial. In our previous work, we developed an ontology model that can standardize and represent patient-provider communication, which can later be integrated in conversational agents as tools for dental communication training. In this study, we embark on enriching our previous model with an ontology of patient personas to portray and express types of dental patient archetypes. The Ontology of Patient Personas that we developed was rooted in terminologies from an OBO Foundry ontology and dental electronic health record data elements. We discuss how this ontology aims to enhance the aforementioned dialogue ontology and future direction in executing our model in software agents to train dental students.

Fine-grained spatial information extraction in radiology as two-turn question answering

OBJECTIVES

Radiology reports contain important clinical information that can be used to automatically construct fine-grained labels for applications requiring deep phenotyping. We propose a two-turn question answering (QA) method based on a transformer language model, BERT, for extracting detailed spatial information from radiology reports. We aim to demonstrate the advantage that a multi-turn QA framework provides over sequence-based methods for extracting fine-grained information.

METHODS

Our proposed method identifies spatial and descriptor information by answering queries given a radiology report text. We frame the extraction problem such that all the main radiology entities (e.g., finding, device, anatomy) and the spatial trigger terms (denoting the presence of a spatial relation between finding/device and anatomical location) are identified in the first turn. In the subsequent turn, various other contextual information that acts as important spatial roles with respect to a spatial trigger term are extracted along with identifying the spatial and other descriptor terms qualifying a radiological entity. The queries are constructed using separate templates for the two turns and we employ two query variations in the second turn.

RESULTS

When compared to the best-reported work on this task using a traditional sequence tagging method, the two-turn QA model exceeds its performance on every component. This includes promising improvements of 12, 13, and 12 points in the average F1 scores for identifying the spatial triggers, Figure, and Ground frame elements, respectively.

DISCUSSION

Our experiments suggest that incorporating domain knowledge in the query (a general description about a frame element) helps in obtaining better results for some of the spatial and descriptive frame elements, especially in the case of the clinical pre-trained BERT model. We further highlight that the two-turn QA approach fits well for extracting information for complex schema where the objective is to identify all the frame elements linked to each spatial trigger and finding/device/anatomy entity, thereby enabling the extraction of more comprehensive information in the radiology domain.

CONCLUSION

Extracting fine-grained spatial information from text in the form of answering natural language queries holds potential in achieving better results when compared to more standard sequence labeling-based approaches.

Contemporary Neuroscience Core Curriculum for Medical Schools

Medical students need to understand core neuroscience principles as a foundation for their required clinical experiences in neurology. In fact, they need a solid neuroscience foundation for their clinical experiences in all other medical disciplines also because the nervous system plays such a critical role in the function of every organ system. Because of the rapid pace of neuroscience discoveries, it is unrealistic to expect students to master the entire field. It is also unnecessary, as students can expect to have ready access to electronic reference sources no matter where they practice. In the preclerkship phase of medical school, the focus should be on providing students with the foundational knowledge to use those resources effectively and interpret them correctly. This article describes an organizational framework for teaching the essential neuroscience background needed by all physicians. This is particularly germane at a time when many medical schools are reassessing traditional practices and instituting curricular changes such as competency-based approaches, earlier clinical immersion, and increased emphasis on active learning. This article reviews factors that should be considered when developing the preclerkship neuroscience curriculum, including goals and objectives for the curriculum, the general topics to include, teaching and assessment methodology, who should direct the course, and the areas of expertise of faculty who might be enlisted as teachers or content experts. These guidelines were developed by a work group of experienced educators appointed by the Undergraduate Education Subcommittee (UES) of the American Academy of Neurology (AAN). They were then successively reviewed, edited, and approved by the entire UES, the AAN Education Committee, and the AAN Board of Directors.

Searching for scientific evidence in a pandemic: An overview of TREC-COVID

We present an overview of the TREC-COVID Challenge, an information retrieval (IR) shared task to evaluate search on scientific literature related to COVID-19. The goals of TREC-COVID include the construction of a pandemic search test collection and the evaluation of IR methods for COVID-19. The challenge was conducted over five rounds from April to July 2020, with participation from 92 unique teams and 556 individual submissions. A total of 50 topics (sets of related queries) were used in the evaluation, starting at 30 topics for Round 1 and adding 5 new topics per round to target emerging topics at that state of the still-emerging pandemic. This paper provides a comprehensive overview of the structure and results of TREC-COVID. Specifically, the paper provides details on the background, task structure, topic structure, corpus, participation, pooling, assessment, judgments, results, top-performing systems, lessons learned, and benchmark datasets.

POS1188 COVID-19 IN PATIENTS WITH RHEUMATIC DISEASES: COMPARISON OF DATA FROM THE ARGENTINE REGISTRY (SAR-COVID), WITH THE LATIN AMERICAN AND GLOBAL REGISTRY (GLOBAL RHEUMATOLOGY ALLIANCE)

Background:

SARS CoV-2 infection has recently burst onto the global scene, and the knowledge of the course of this infection in patients with rheumatic diseases receiving immunomodulatory treatment is still insufficient. The Argentine Society of Rheumatology (SAR) designed a national registry called SAR-COVID in order to get to assess our reality.

Objectives:

To identify the particular characteristics of patients with rheumatic diseases and COVID-19 in Argentina (SAR-COVID Registry), and to compare them with the data reported at the Latin American and Global level (Global International Alliance Rheum-COVID Registry).

Methods:

A national, multicenter, prospective and observational registry was carried out. Patients older than 18 years, with a diagnosis of rheumatic disease and SARS-CoV-2 infection by PCR or serology, were included between August 13, 2020 and January 17, 2021. Demographic data, underlying rheumatic disease (activity of the disease, current treatment), comorbidities, clinical-laboratory characteristics of the SARS-CoV-2 infection, as well as received treatments (pharmacological, oxygen therapy / ventilatory support) and outcomes (hospitalization, mortality) were recorded. The characteristics of the included patients were compared with the data reported at the Latin American and global level. Descriptive statistics were performed. Comparisons between groups were made using ANOVA, chi2 or Fisher’s test, according to the type of variable.

Results:

Four hundred sixty-five patients from Argentina, 74 patients from Latin America and 583 from the rest of the world were included, mostly women (79.6%, 73% and 71% respectively), with a mean age of 50.2 (SD 15.3), 53.5 (DE 15.6) and 55.8 (15.5), years respectively. The most frequent rheumatic diseases in the three groups were rheumatoid arthritis (43.9%, 35%, and 39%) and systemic lupus erythematosus (16.1%, 22%, and 14%) (Table 1).

In Argentina, fewer patients received specific pharmacological treatment for COVID-19 (40.9%, 68% and 43% respectively, p <0.0001), and there was a lower requirement of NIMV / IMV (Non-Invasive Mechanical Ventilation/Invasive Mechanical Ventilation) than in the rest of Latin America and the world (10.5% vs 31% vs 13%, p <0.0001).

Hospitalization was lower in Argentina than in the rest of Latin America (37.4% vs 61% p 0.0002) and of the world (37.4% vs 45% p 0.0123), and mortality was numerically lower in Argentina, but without statistically significant differences between the three groups (6.9%, 12% and 11%; p 0.6311). Most of the patients, (86.9%) did not present any complications in Argentina, with a statistically significant difference with the rest of the groups (62% and 77%, p <0.0001) (Graph 1).

Conclusion:

The patients with rheumatic diseases and SARS-CoV-2 infection reported in this argentinian registry received less specific pharmacological treatment for COVID-19, presented fewer complications and required less ventilatory support, than those reported in the Latinoamerican and Global registry. However, no statistically significant differences were observed in terms of mortality.

Graph 1.

Main outcomes and evolution of patients with rheumatic disease and COVID-19.

References:

[1]Stokes, Erin K, Zambrano, Laura D, Anderson, Kayla N, et al. Coronavirus Disease 2019 Case Surveillance - United States, January 22-May 30, 2020. MMWR Morb Mortal Wkly Rep; 69(24): 759-765, 2020 Jun 19.

[2]Mehta P, McAuley DF, Brown M, et al. COVID-19: consider cytokine storm síndromes and immunosuppression. Lancet 2020;395:1033–4.

[3]Gianfrancesco M, et al. Characteristics associated with hospitalisation for COVID-19 in people with rheumatic disease: data from the COVID-19 Global Rheumatology Alliance physician-reported registry. Ann Rheum Dis 2020;79:859–866.

[4]Manuel F. Ugarte-Gil, et al. Characteristics associated with Covid-19 in patients with Rheumatic Disease in Latin America. Global Rheumatology. Septiembre 2020.

Disclosure of Interests:

Alvaro Andres Reyes Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Gelsomina Alle Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Romina Tanten Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Marina Scolnik Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Enrique Soriano Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Guillermo Berbotto Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Maria Haye Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, María Julieta Gamba Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Romina Nieto Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Mercedes García Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Veronica Savio Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Luciana Gonzalez Lucero Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Paula Alba Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Lorena Takashima Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, FABIAN RISUEÑO Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Luciana CASALLA Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Natalia Cucchiaro Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Ana Bertoli Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Sabrina POrta Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Carla Maldini Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Rosana Gallo Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Cecilia Goizueta Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Eugenia Picco Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Rosana Quintana Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Karen Roberts Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Carolina Ayelen Isnardi Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”, Guillermo Pons-Estel Grant/research support from: “Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.”

POS1186 EFFECT OF SOCIO-ECONOMIC STATUS AND EDUCATIONAL LEVEL ON COVID-19 OUTCOMES IN PATIENTS WITH RHEUMATIC DISEASES FROM ARGENTINA: DATA FROM THE SAR-COVID REGISTRY

Background:

SARS-CoV-2 infection can present with a broad clinical spectrum, from asymptomatic to lethal. Different risk factors have been recognized. Socio-economic status and educational level may affect access to the healthcare system and therefore COVID-19 infection outcome.

Objectives:

The aim of this study was to assess the association between socio-demographic status and educational level and SARS-CoV-2 outcomes, such as hospitalization, ICU admission, need for mechanical ventilation and death, in Argentinean patients with rheumatic diseases from the SAR-COVID Registry.

Methods:

We performed a cross-sectional study of consecutive adult patients with rheumatic diseases and SARS-CoV-2 infection included in the multicentric Argentinean SAR-COVID Registry. The following variables were included: gender, ethnicity, age, health insurance, educational level (under or over 12 years of education), socio-economic level according to Graffar Scale in high, medium-high, medium, medium-low, low; underlying rheumatic disease, its duration and treatment at the time of infection.

SARS-CoV-2 infection outcomes were: hospitalization, admission to ICU, mechanical ventilation requirement and death.

Statistical analysis was performed using Chi2, Fisher, T-test, ANOVA.

Results:

Five hundred and twenty-five patients were included, 422 (80.4%) were female, with a mean age of 51.3 years (SD 15.2). Most of them were caucasians (48%) or mestizos (43%) and 96.8% lived in an urban environment. Almost half of the patients (47%) were categorized as middle-class, 24% middle-high or high class, 21% middle-low or low. 48.4% of the patients were employed. Regarding educational level, 54% had more than 12 years of education.

The most prevalent rheumatic disease was Rheumatoid Arthritis (40.4%), followed by Systemic Lupus Erythematosus (14.9%), Sjögren (5.5%) and Psoriatic Arthritis (5.5%). Treatments used at the time of SARS-CoV-2 infection were corticosteroids (19%), cs-DMARDs (49%), and b- and ts-DMARDs (16%).

Overall hospitalization frequency was 35%, median hospital stay was 10 days (IQR 10 days), 11.6% were admitted to the ICU, 10% required mechanical ventilation and the global mortality was 8%.

Notably, patients with less than 12 years of education required mechanical ventilation more frequently than the more educated ones (11.9% vs. 5.6%, p=0.026) and showed a higher mortality due to COVID-19 (9% vs. 2.8%, p=0.0004).

Patients categorized as upper social classes (middle-high and high) were admitted to the hospital on a more frequent basis (74.4% of cases), when compared with middle class (64.4%) and middle-low and low class (58%) (p=0.77). Median duration of hospitalization for the aforementioned groups was 12.5 (IQR 17.3), 10 (IQR 9) and 10.5 (IQR 9.3) days respectively (p=0.60).

Patients with health insurance were found to be hospitalized more frequently in comparison to those without insurance (42.4% vs. 33.7%, p=0.14), but showed similar admission rates to the ICU (11.8% vs. 12.8%; p=0.78), need for mechanical ventilation (10.7% vs. 8.7%; p=0.70) and mortality (7.1% vs. 6.5%; p=0.99).

Caucasian patients had fewer hospital admissions when compared against other ethnicities (mestizos mostly) (26.1% vs. 43.4%; p<0.0001), but showed no statistically significant difference in need for mechanical ventilation 10.3% vs. 9.9% (p=0.99) or mortality 8.7% vs. 5.1% (p=0.15).

Conclusion:

Patients with lower educational level needed twice the frequency of mechanical ventilation, and showed thrice the mortality than those with more than 12 years of education.

Albeit patients in upper social stratus and those with health insurance were admitted to the hospital in a more frequent manner, no statistically significant differences were found regarding the need for ICU, mechanical ventilation or mortality.

Caucasians were hospitalized less frequently than mestizos, but had no significant differences in the other measured outcomes.

Disclosure of Interests:

Laura Sorrentino Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Jonathan Rebak Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Federico Maldonado Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Vanessa Viviana Castro Coello Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Alejandro Brigante Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Adriana Hamaui Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Diana Dubinsky Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Roberto Baez Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Cecilia Pisoni Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Carla Gobbi Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Leandro Carlevaris Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Romina Tanten Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Adriana Karina Cogo Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Maria DeLaVega Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Rodolfo Perez Alamino Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Maria Alicia Lazaro Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Mariana Pera Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Susana Isabel Pineda Vidal Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Maria Elena Calvo Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Debora Guaglianone Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Carla G Alonso Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Mara Guinsburg Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Cinthya Retamozo Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Carolina Aeschlimann Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Rosana Quintana Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Karen Roberts Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Carolina Ayelen Isnardi Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Guillermo Pons Estel Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.

AB0342 OBSTETRIC AND THROMBOTIC ANTIPHOSPHOLIPID SYNDROME: ARE THEY DIFFERENT ENTITIES?

Background:

Several studies showed two main clinical phenotypes of antiphospholipid syndrome (APS): thrombotic (TAPS) and obstetric APS (OAPS). Although they have the same autoantibody profile, one of them developed thrombosis and other one obstetric morbidity.

Objectives:

To study clinical, demographic and antibody profile in patients with TAPS and OAPS.

Methods:

we retrospectively evaluated TAPS and OAPS patients who were included in Argentine Antiphospholipid antibodies registry. We studied clinical, demographic and antibody profile in both groups.

Results:

238 patients were included in the registry. 201 (84.81%) of them were female. 122 (60.69 %) of them fullfilled APS Sydney classification criteria, 47 (38.52%) TAPS and 52 (42.62%) OAPS. 23 (18.85%) patients had both thrombotic and obstetric events so they were excluded in this analysis.

Arterial Hypertension (HBP) and Hyperlipidemia were more frequent in TAPS versus OAPS. Older age was found in TAPS as well as in association with Systemic lupus erythematosus (SLE). There was no difference in antibody profile between the 2 groups, and the Global Antiphospholipid Syndrome Score (aGAPSS) was higher in TAPS than OAPS.

18 (38.3%) of TAPS patients had at least 1 pregnancy. Mean number of pregnancies of TAPS was 2.5 (1.10) and 3.84 (1.86) in OAPS. Thrombotic events were not found in TAPS during pregnancy and puerperium. HBP and gestational diabetes (GD) and other pregnancy related comorbidities were found in TAPS.

OAPS (n=52)TAPS (n=47)POAPS(n=52)TAPS(n=18)SLE, n (%)11 (21.2)28 (59.6)0.0002N% (DE)N % (DE)aGAPSS, mean (RIQ)4 (5)8 (5)<0.0001Abortions (<10 weeks)3338.4 (36.2)826.5 (35.3)Age, mean (DE)39.3 (6.24)43.1 (13.5)<0.0001Live Birth3633.5 (28.1)1672.2 (34.7)HBP, n (%)5 (9.6)15 (31.9)0.0121 >37 weeks2116.4 (2.38)1661.1 (31.7)Hyperlipidemia, n (%)4 (7.7)12(25.5)0.0267Prematurity <37 >34 weeks118.76 (1.97)0-GD, n (%)3(5.8)2(4.3)0.9999Prematurity <34 weeks96.37 (1.49)11.39 (5.89)Obesity, n (%)8 (15.4)4 (8.5)0.2912Pre eclampsia >34-<37 weeks10.490 (0.0350)24.63 (0.138)Smoking, n (%)11 (21.1)13 (27.6)0.4019Placental Hematoma31.86 (8.18)12.78 (11.8)Sedentary lifestyle, n (%)16 (30.8)17 (36.2)0.8486Abruptio Placentae21.96 (9.80)24.17 (12.9)Triple Positivity59.648.50.8323 Normal delivery2017.8 (27.9)1447.7 (33.9)Double Positivity11.936.4Cesarean section2217.2 (21.6)211.1 (32.3)Simple Positivity2938.51940.4Urgent Cesarean section139.80 (20.3)412.5 (24.6)GD21.37 (7.49)12.78 (11.8)HBP63.46 (10.3)310.2 (26.3)

Conclusion:

Antibody profile was similar in TAPS and OAPS. However, clinical manifestations and cardiovascular risk were different. These results should be evaluated in prospective studies.

Disclosure of Interests:

None declared

POS1208 EPIDEMIOLOGY AND OUTCOMES OF PATIENTS WITH RHEUMATIC DISEASES AND SARS-CoV-2 INFECTION: DATA FROM THE ARGENTINEAN SAR-COVID REGISTRY

Background:

In the last time, many papers about SARS-CoV-2 have been published in the world. However, data from latinamerican patients is still scarce. In order to assess the impact of SARS-CoV-2 infection in patients with rheumatic diseases in our country and contribute to the global knowledge about the effect of immunosuppressive therapies in this group, the Argentine Society of Rheumatology has developed the National Registry of Patients with Rheumatic Diseases and COVID-19 (SAR-COVID).

Objectives:

The aim of this study was to evaluate clinical characteristics and outcomes of SARS-CoV-2 infection in patients with rheumatic diseases, treated or not with immunomodulators and/or immunosuppressants.

Methods:

SAR-COVID is a national, multicenter, prospective and observational registry, in which patients, ≥18 years of age, with a diagnosis of a rheumatic disease who had SARS-CoV-2 infection (PCR or positive serology) are consecutively included between August 13, 2020 and January 17, 2021. Sociodemographic data, comorbidities, underlying rheumatic disease and treatment, clinical characteristics, complications, laboratory and treatment of the SARS-CoV-2 infection were recorded. Hospitalization, mechanical ventilation requirements and death were assessed to evaluate COVID-19 outcome. Statistical analysis: Descriptive analysis. Chi2 or Fischer test and T test or Mann-Whitney U test or ANOVA, as appropriate. Multiple logistic regression.

Results:

A total of 525 patients were included, 80.4% were female, with a median age of 52 years (IQR 40-62). Comorbidities were reported in half of them (53.3%). The most frequent rheumatological diseases were rheumatoid arthritis (40.4%) and systemic lupus erythematosus (14.9%). At the time of the infection, most of them were in remission or in minimal/low disease activity (68.2%) and 72.9% were receiving immunosuppressive or immunomodulatory treatment.

Symptoms were present in 96% of the patients, the most frequent being fever (56.2%), cough (46.7%) and headache (39.2%). During infection, 35.1% received some pharmacological treatment, dexamethasone (20%) the most frequently used. One third (35.1%) of the patients were hospitalized, 11.6% were admitted to the ICU, 10.1% needed mechanical ventilation and 6.9% died due to COVID-19. Complications were reported in 12.4%, being acute respiratory distress syndrome the most prevalent (8.8%).

Patients over 65 years of age were more frequently hospitalized, admitted to the ICU, needed mechanical ventilation and died due to COVID-19 (50% vs 31.4%, 22% vs 9%, 16.3% vs 5.2%, 14% vs 5%, respectively; p<0.001 in all cases). Similar results were seen in patients with vasculitis (57.7% vs 33.9%, 46.2 vs 9.8%, 34.6% vs 6 %; 30.8% vs 5.6%, respectively; p< 0.001 in all cases) and those with moderate/high disease activity (55.7% vs 26.5%, 21.3 vs 7.8%, 17.2% vs 4.2 %; 17.2% vs 4.2 %, respectively; p< 0.001 in all cases). Patients with APS were more frequently admitted to the ICU (29.4% vs 11%, p= 0.037). The presence of comorbidities was associated with higher hospitalization (46% vs 22.6%, p<0.001), admission to the ICU (17.2% vs 5.9%, p<0.001) and mechanical ventilation (10.2% vs 4.6%, p= 0.028). Immunosuppressive treatment was not associated with worse outcomes.

Conclusion:

In this cohort of patients with a wide distribution of rheumatic diseases, we have found clinical characteristics similar to those reported by other international cohorts. Compared with national data, the mortality reported in these patients is higher. However, it should be noted that these are early data collected during isolation and that there may be an underreporting of asymptomatic patients or with mild symptoms who do not attend the rheumatologist.

Older patients, those with comorbidities, with vasculitis and with higher disease activity showed poor COVID-19 outcomes.

Disclosure of Interests:

Carolina Ayelen Isnardi Speakers bureau: Janssen, BMS, Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Rosana Quintana Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Karen Roberts Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Vanessa Viviana Castro Coello Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Alvaro Andres Reyes Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Yohana Tissera Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Micaela Cosatti Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Romina Rojas Tessel Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Julia Scafati Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Tatiana Barbich Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., María Soledad Gálvez Elkin Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Gustavo Fabian Rodriguez Gil Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Sebastian Moyano Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Marina Laura Werner Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Jonathan Rebak Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Julieta Morbiducci Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Victoria Martire Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., María Sol Castaño Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Carolina Dieguez Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Gisela Constanza Subils Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data., Guillermo Pons-Estel Grant/research support from: Unrestricted grants: Pfizer, Abbvie, Elea Phoenix. None of them have access to patient data.