Clinical decision support system in emergency telephone triage: A scoping review of technical design, implementation and evaluation

OBJECTIVES

Emergency department overcrowding could be improved by upstream telephone triage. Emergency telephone triage aims at managing and orientating adequately patients as early as possible and distributing limited supply of staff and materials. This complex task could be improved with the use of Clinical decision support systems (CDSS). The aim of this scoping review was to identify literature gaps for the future development and evaluation of CDSS for Emergency telephone triage.

MATERIALS AND METHODS

We present here a scoping review of CDSS designed for emergency telephone triage, and compared them in terms of functional characteristics, technical design, health care implementation and methodologies used for evaluation, following the PRISMA-ScR guidelines.

RESULTS

Regarding design, 19 CDSS were retrieved: 12 were knowledge based CDSS (decisional algorithms built according to guidelines or clinical expertise) and 7 were data driven (statistical, machine learning, or deep learning models). Most of them aimed at assisting nurses or non-medical staff by providing patient orientation and/or severity/priority assessment. Eleven were implemented in real life, and only three were connected to the Electronic Health Record. Regarding evaluation, CDSS were assessed through various aspects: intrinsic characteristics, impact on clinical practice or user apprehension. Only one pragmatic trial and one randomized controlled trial were conducted.

CONCLUSION

This review highlights the potential of a hybrid system, user tailored, flexible, connected to the electronic health record, which could work with oral, video and digital data; and the need to evaluate CDSS on intrinsic characteristics and impact on clinical practice, iteratively at each distinct stage of the IT lifecycle.

Performance and clinical utility of a new supervised machine-learning pipeline in detecting rare ciliopathy patients based on deep phenotyping from electronic health records and semantic similarity

BACKGROUND

Rare diseases affect approximately 400 million people worldwide. Many of them suffer from delayed diagnosis. Among them, NPHP1-related renal ciliopathies need to be diagnosed as early as possible as potential treatments have been recently investigated with promising results. Our objective was to develop a supervised machine learning pipeline for the detection of NPHP1 ciliopathy patients from a large number of nephrology patients using electronic health records (EHRs).

METHODS AND RESULTS

We designed a pipeline combining a phenotyping module re-using unstructured EHR data, a semantic similarity module to address the phenotype dependence, a feature selection step to deal with high dimensionality, an undersampling step to address the class imbalance, and a classification step with multiple train-test split for the small number of rare cases. The pipeline was applied to thirty NPHP1 patients and 7231 controls and achieved good performances (sensitivity 86% with specificity 90%). A qualitative review of the EHRs of 40 misclassified controls showed that 25% had phenotypes belonging to the ciliopathy spectrum, which demonstrates the ability of our system to detect patients with similar conditions.

CONCLUSIONS

Our pipeline reached very encouraging performance scores for pre-diagnosing ciliopathy patients. The identified patients could then undergo genetic testing. The same data-driven approach can be adapted to other rare diseases facing underdiagnosis challenges.

A Radiomics Approach to Identify Immunologically Active Tumor in Patients with Head and Neck Squamous Cell Carcinomas

BACKGROUND

We recently developed a gene-expression-based HOT score to identify the hot/cold phenotype of head and neck squamous cell carcinomas (HNSCCs), which is associated with the response to immunotherapy. Our goal was to determine whether radiomic profiling from computed tomography (CT) scans can distinguish hot and cold HNSCC.

METHOD

We included 113 patients from The Cancer Genome Atlas (TCGA) and 20 patients from the Groupe Hospitalier Pitié-Salpêtrière (GHPS) with HNSCC, all with available pre-treatment CT scans. The hot/cold phenotype was computed for all patients using the HOT score. The IBEX software (version 4.11.9, accessed on 30 march 2020) was used to extract radiomic features from the delineated tumor region in both datasets, and the intraclass correlation coefficient (ICC) was computed to select robust features. Machine learning classifier models were trained and tested in the TCGA dataset and validated using the area under the receiver operator characteristic curve (AUC) in the GHPS cohort.

RESULTS

A total of 144 radiomic features with an ICC >0.9 was selected. An XGBoost model including these selected features showed the best performance prediction of the hot/cold phenotype with AUC = 0.86 in the GHPS validation dataset.

CONCLUSIONS AND RELEVANCE

We identified a relevant radiomic model to capture the overall hot/cold phenotype of HNSCC. This non-invasive approach could help with the identification of patients with HNSCC who may benefit from immunotherapy.

Assessment of the Early Detection of Anosmia and Ageusia Symptoms in COVID-19 on Twitter: Retrospective Study

BACKGROUND

During the unprecedented COVID-19 pandemic, social media has been extensively used to amplify the spread of information and to express personal health-related experiences regarding symptoms, including anosmia and ageusia, 2 symptoms that have been reported later than other symptoms.

OBJECTIVE

Our objective is to investigate to what extent Twitter users reported anosmia and ageusia symptoms in their tweets and if they connected them to COVID-19, to evaluate whether these symptoms could have been identified as COVID-19 symptoms earlier using Twitter rather than the official notice.

METHODS

We collected French tweets posted between January 1, 2020, and March 31, 2020, containing anosmia- or ageusia-related keywords. Symptoms were detected using fuzzy matching. The analysis consisted of 3 parts. First, we compared the coverage of anosmia and ageusia symptoms in Twitter and in traditional media to determine if the association between COVID-19 and anosmia or ageusia could have been identified earlier through Twitter. Second, we conducted a manual analysis of anosmia- and ageusia-related tweets to obtain quantitative and qualitative insights regarding their nature and to assess when the first associations between COVID-19 and these symptoms were established. We randomly annotated tweets from 2 periods: the early stage and the rapid spread stage of the epidemic. For each tweet, each symptom was annotated regarding 3 modalities: symptom (yes or no), associated with COVID-19 (yes, no, or unknown), and whether it was experienced by someone (yes, no, or unknown). Third, to evaluate if there was a global increase of tweets mentioning anosmia or ageusia in early 2020, corresponding to the beginning of the COVID-19 epidemic, we compared the tweets reporting experienced anosmia or ageusia between the first periods of 2019 and 2020.

RESULTS

In total, 832 (respectively 12,544) tweets containing anosmia (respectively ageusia) related keywords were extracted over the analysis period in 2020. The comparison to traditional media showed a strong correlation without any lag, which suggests an important reactivity of Twitter but no earlier detection on Twitter. The annotation of tweets from 2020 showed that tweets correlating anosmia or ageusia with COVID-19 could be found a few days before the official announcement. However, no association could be found during the first stage of the pandemic. Information about the temporality of symptoms and the psychological impact of these symptoms could be found in the tweets. The comparison between early 2020 and early 2019 showed no difference regarding the volumes of tweets.

CONCLUSIONS

Based on our analysis of French tweets, associations between COVID-19 and anosmia or ageusia by web users could have been found on Twitter just a few days before the official announcement but not during the early stage of the pandemic. Patients share qualitative information on Twitter regarding anosmia or ageusia symptoms that could be of interest for future analyses.

The genetic landscape and clinical spectrum of nephronophthisis and related ciliopathies

Nephronophthisis (NPH) is an autosomal-recessive ciliopathy representing one of the most frequent causes of kidney failure in childhood characterized by a broad clinical and genetic heterogeneity. Applied to one of the worldwide largest cohorts of patients with NPH, genetic analysis encompassing targeted and whole exome sequencing identified disease-causing variants in 600 patients from 496 families with a detection rate of 71%. Of 788 pathogenic variants, 40 known ciliopathy genes were identified. However, the majority of patients (53%) bore biallelic pathogenic variants in NPHP1. NPH-causing gene alterations affected all ciliary modules defined by structural and/or functional subdomains. Seventy six percent of these patients had progressed to kidney failure, of which 18% had an infantile form (under five years) and harbored variants affecting the Inversin compartment or intraflagellar transport complex A. Forty eight percent of patients showed a juvenile (5-15 years) and 34% a late-onset disease (over 15 years), the latter mostly carrying variants belonging to the Transition Zone module. Furthermore, while more than 85% of patients with an infantile form presented with extra-kidney manifestations, it only concerned half of juvenile and late onset cases. Eye involvement represented a predominant feature, followed by cerebellar hypoplasia and other brain abnormalities, liver and skeletal defects. The phenotypic variability were in a large part associated with mutation types, genes and corresponding ciliary modules with hypomorphic variants in ciliary genes playing a role in early steps of ciliogenesis associated with juvenile-to-late onset NPH forms. Thus, our data confirm a considerable proportion of late-onset NPH suggesting an underdiagnosis in adult chronic kidney disease.

Improving Patient Similarity Using Different Modalities of Phenotypes Extracted from Clinical Narratives

In the context of medical concept extraction, it is critical to determine if clinical signs or symptoms mentioned in the text were present or absent, experienced by the patient or their relatives. Previous studies have focused on the NLP aspect but not on how to leverage this supplemental information for clinical applications. In this paper, we aim to use the patient similarity networks framework to aggregate different phenotyping modalities. NLP techniques were applied to extract phenotypes and predict their modalities from 5470 narrative reports of 148 patients with ciliopathies (a group of rare diseases). Patient similarities were computed using each modality separately for aggregation and clustering. We found that aggregating negated phenotypes improved patient similarity, but further aggregating relatives' phenotypes worsened the result. We suggest that different modalities of phenotypes can contribute to patient similarity, but they should be aggregated carefully and with appropriate similarity metrics and aggregation models.

The Smart Data Extractor, a Clinician Friendly Solution to Accelerate and Improve the Data Collection During Clinical Trials

In medical research, the traditional way to collect data, i.e. browsing patient files, has been proven to induce bias, errors, human labor and costs. We propose a semi-automated system able to extract every type of data, including notes. The Smart Data Extractor pre-populates clinic research forms by following rules. We performed a cross-testing experiment to compare semi-automated to manual data collection. 20 target items had to be collected for 79 patients. The average time to complete one form was 6'81" for manual data collection and 3'22" with the Smart Data Extractor. There were also more mistakes during manual data collection (163 for the whole cohort) than with the Smart Data Extractor (46 for the whole cohort). We present an easy to use, understandable and agile solution to fill out clinical research forms. It reduces human effort and provides higher quality data, avoiding data re-entry and fatigue induced errors.

Putting undergraduate medical students in AI-CDSS designers' shoes: An innovative teaching method to develop digital health critical thinking

INTRODUCTION

Digital health programs are urgently needed to accelerate the adoption of Artificial Intelligence and Clinical Decision Support Systems (AI-CDSS) in clinical settings. However, such programs are still lacking for undergraduate medical students, and new approaches are required to prepare them for the arrival of new and unknown technologies. At University Paris Cité medical school, we designed an innovative program to develop the digital health critical thinking of undergraduate medical students that consisted of putting medical students in AI-CDSS designers' shoes.

METHODS

We followed the six steps of Kern's approach for curriculum development: identification of needs, definition of objectives, design of an educational strategy, implementation, development of an assessment and design of program evaluation.

RESULTS

A stand-alone and elective AI-CDSS program was implemented for fourth-year medical students. Each session was designed from an AI-CDSS designer viewpoint, with theoretical and practical teaching and brainstorming time on a project that consisted of designing an AI-CDSS in small groups. From 2021 to 2022, 15 students were enrolled: they rated the program 4.4/5, and 80% recommended it. Seventy-four percent considered that they had acquired new skills useful for clinical practice, and 66% felt more confident with technologies. The AI-CDSS program aroused great enthusiasm and strong engagement of students: 8 designed an AI-CDSS and wrote two scientific 5-page articles presented at the Medical Informatics Europe conference; 4 students were involved in a CDSS research project; 2 students asked for a hospital internship in digital health; and 1 decided to pursue PhD training.

DISCUSSION

Putting students in AI-CDSS designers' shoes seemed to be a fruitful and innovative strategy to develop digital health skills and critical thinking toward AI technologies. We expect that such programs could help future doctors work in rapidly evolving digitalized environments and position themselves as key leaders in digital health.

Using an Ontology to Derive a Sharable and Interoperable Relational Data Model for Heterogeneous Healthcare Data and Various Applications

BACKGROUND

 A large volume of heavily fragmented data is generated daily in different healthcare contexts and is stored using various structures with different semantics. This fragmentation and heterogeneity make secondary use of data a challenge. Data integration approaches that derive a common data model from sources or requirements have some advantages. However, these approaches are often built for a specific application where the research questions are known. Thus, the semantic and structural reconciliation is often not reusable nor reproducible. A recent integration approach using knowledge models has been developed with ontologies that provide a strong semantic foundation. Nonetheless, deriving a data model that captures the richness of the ontology to store data with their full semantic remains a challenging task.

OBJECTIVES

 This article addresses the following question: How to design a sharable and interoperable data model for storing heterogeneous healthcare data and their semantic to support various applications?

METHOD

 This article describes a method using an ontological knowledge model to automatically generate a data model for a domain of interest. The model can then be implemented in a relational database which efficiently enables the collection, storage, and retrieval of data while keeping semantic ontological annotations so that the same data can be extracted for various applications for further processing.

RESULTS

 This article (1) presents a comparison of existing methods for generating a relational data model from an ontology using 23 criteria, (2) describes standard conversion rules, and (3) presents O n t o R e l a , a prototype developed to demonstrate the conversion rules.

CONCLUSION

 This work is a first step toward automating and refining the generation of sharable and interoperable relational data models using ontologies with a freely available tool. The remaining challenges to cover all the ontology richness in the relational model are pointed out.

Prevalence of Contraindications to Nirmatrelvir-Ritonavir Among Hospitalized Patients With COVID-19 at Risk for Progression to Severe Disease

This cohort study examines the prevalence of contraindications to nirmatrelvir-ritonavir in patients hospitalized with COVID-19.

Determining the Set of Items to Include in Breast Operative Reports, Using Clustering Algorithms on Retrospective Data Extracted from Clinical DataWarehouse

Medical reports are key elements to guarantee the quality, and continuity of care but their quality remains an issue. Standardization and structuration of reports can increase their quality, but are usually based on expert opinions. Here, we hypothesize that a structured model of medical reports could be learnt using machine learning on retrospective medical reports extracted from clinical data warehouses (CDW). To investigate our hypothesis, we extracted breast cancer operative reports from our CDW. Each document was preprocessed and split into sentences. Clustering was performed using TFIDF, Paraphrase or Universal Sentence Encoder along with K-Means, DBSCAN, or Hierarchical clustering. The best couple was TFIDF/K-Means, providing a sentence coverage of 89 % on our dataset; and allowing to identify 7 main categories of items to include in breast cancer operative reports. These results are encouraging for a document preset creation task and should then be validated and implemented in real life.

Using Deep Learning to Improve Phenotyping from Clinical Reports

With the development of clinical databases and the ubiquity of EHRs, physicians and researchers alike have access to an unprecedented amount of data. Complexity of the available data has also increased since clinical reports are also included and require frameworks with natural language processing capabilities in order to process them and extract information not found in other types of documents. In the following work we implement a data processing pipeline performing phenotyping, disambiguation, negation and subject prediction on such reports. We compare it to an existing solution routinely used in a children's hospital with special focus on genetic diseases. We show that by replacing components based on rules and pattern matching with components leveraging deep learning models and fine-tuned word embeddings we obtain performance improvements of 7%, 10% and 27% in terms of F1 measure for each task. The solution we devised will help build more reliable decision support systems.

Design of an Ontology-Based Triage System for Patients with Chronic Pain

OBJECTIVE

Waiting time for a consultation for chronic pain is a widespread health problem. This paper presents the design of an ontology use to assess patients referred to a consultation for chronic pain.

METHODS

We designed OntoDol, an ontology of pain domain for patient triage based on priority degrees. Terms were extracted from clinical practice guidelines and mapped to SNOMED-CT concepts through the Python module Owlready2. Selected SNOMED-CT concepts, relationships, and the TIME ontology, were implemented in the ontology using Protégé. Decision rules were implemented with SWRL. We evaluated OntoDol on 5 virtual cases.

RESULTS

OntoDol contains 762 classes, 92 object properties and 18 SWRL rules to assign patients to 4 categories of priority. OntoDol was able to assert every case and classify them in the right category of priority.

CONCLUSION

Further works will extend OntoDol to other diseases and assess OntoDol with real world data from the hospital.

An Interactive Interface for Displaying Recommendations on Emergency Phone Triage in Pediatrics

Emergency phone triage aims at identifying quickly patients with critical emergencies. Patient triage is not an easy task, especially in situations involving children, mostly due to the lack of training and the lack of clinical guidelines for children. To overcome these issues, we aim at designing and assessing an interactive interface for displaying recommendations on emergency phone triage in pediatrics. Four medical students formalized local guidelines written by the SAMU of Paris, into a decision tree and designed an interface according to usability principles. The navigation within the interface was designed to allow the identification of critical emergencies at the beginning of the decision process, and thus ensuring a quick response in case of critical emergencies. The interface was assessed by 10 medical doctors: they appreciated the ergonomics (e.g., intuitive colors), and found easy to navigate through the interface. Nine of them would like to use this interface during phone call triage. In the future, this interface will be improved and implemented in emergency call centers.

viroCapt: A Bioinformatics Pipeline for Identifying Viral Insertion in Human Host Genome

INTRODUCTION

The implication of viruses in human cancers, as well as the emergence of next generation sequencing has permitted to investigate further their role and pathophysiology in the development of this disease. One such mechanism is the integration of portions of viral genomes in the human genome, as well as the specific action of viral oncogenes.inding integration sites and preserved oncogenes is still relying on heavy manual intervention.

METHODS

We developed an analysis and interpretation pipeline to determine viral insertions. Using data from directed viral capture, the pipeline conducts a crude genotyping phase to select reference viral genomes, identifies chimeric reads, extracts the putative human sequences to locate in the human reference genome, scores and ranks candidate junctions, and exports tabular and visual results.

RESULTS

We leverage common bioinformatics tools (bowtie2, samtools, blat), and a dedicated filtering and ranking algorithm, implemented in R, to infer candidate junctions and insertions. Static results (tables, figures) are produced, as well as an interactive interpretation tool developed as a shiny web app.

DISCUSSION

We validated this pipeline against published results of HPV, HBV, and AAV2 insertions and show good information retrieval.

Towards a Clinical Decision Support System for Helping Medical Students in Emergency Call Centers

In critical situations such as pandemic, medical students are often called to help in emergency call centers. However, they may encounter difficulties in phone triage because of a lack of medical skills. Here, we aim at developing a Clinical Decision Support System for helping medical students in phone call triage of pediatric patients. The system is based on the PAT (Pediatric Assessment Triangle) and local guidelines. It is composed of two interfaces. The first allows a quick assessment of severity signs, and the second provides recommendations and additional elements such as "elements to keep in mind" or "medical advice to give to patient". The system was evaluated by 20 medical students, with two fictive clinical cases. 75% of them found the content useful and clear, and the navigation easy. 65% would feel more reassured to have this system in emergency call centers. Further works are planned to improve the system before implementation in real-life.

Enriching UMLS-Based Phenotyping of Rare Diseases Using Deep-Learning: Evaluation on Jeune Syndrome

The wide adoption of Electronic Health Records (EHR) in hospitals provides unique opportunities for high throughput phenotyping of patients. The phenotype extraction from narrative reports can be performed by using either dictionary-based or data-driven methods. We developed a hybrid pipeline using deep learning to enrich the UMLS Metathesaurus for automatic detection of phenotypes from EHRs. The pipeline was evaluated on a French database of patients with a rare disease characterized by skeletal abnormalities, Jeune syndrome. The results showed a 2.5-fold improvement regarding the number of detected skeletal abnormalities compared to the baseline extraction using the standard release of UMLS. Our method can help enrich the coverage of the UMLS and improve phenotyping, especially for languages other than English.

Deep neural networks for simultaneously capturing public topics and sentiments during a pandemic. Application to a COVID-19 tweet dataset

Background

Public engagement is a key element for mitigating pandemics, and a good understanding of public opinion could help to encourage the successful adoption of public health measures by the population. In past years, deep learning has been increasingly applied to the analysis of text from social networks. However, most of the developed approaches can only capture topics or sentiments alone but not both together.

Objective

Here, we aimed to develop a new approach, based on deep neural networks, for simultaneously capturing public topics and sentiments and applied it to tweets sent just after the announcement of the COVID-19 pandemic by the World Health Organization (WHO).

Methods

A total of 1,386,496 tweets were collected, preprocessed, and split with a ratio of 80:20 into training and validation sets, respectively. We combined lexicons and convolutional neural networks to improve sentiment prediction. The trained model achieved an overall accuracy of 81% and a precision of 82% and was able to capture simultaneously the weighted words associated with a predicted sentiment intensity score. These outputs were then visualized via an interactive and customizable web interface based on a word cloud representation. Using word cloud analysis, we captured the main topics for extreme positive and negative sentiment intensity scores.

Results

In reaction to the announcement of the pandemic by the WHO, 6 negative and 5 positive topics were discussed on Twitter. Twitter users seemed to be worried about the international situation, economic consequences, and medical situation. Conversely, they seemed to be satisfied with the commitment of medical and social workers and with the collaboration between people.

Conclusions

We propose a new method based on deep neural networks for simultaneously extracting public topics and sentiments from tweets. This method could be helpful for monitoring public opinion during crises such as pandemics.

Patient-Patient Similarity-Based Screening of a Clinical Data Warehouse to Support Ciliopathy Diagnosis

A timely diagnosis is a key challenge for many rare diseases. As an expanding group of rare and severe monogenic disorders with a broad spectrum of clinical manifestations, ciliopathies, notably renal ciliopathies, suffer from important underdiagnosis issues. Our objective is to develop an approach for screening large-scale clinical data warehouses and detecting patients with similar clinical manifestations to those from diagnosed ciliopathy patients. We expect that the top-ranked similar patients will benefit from genetic testing for an early diagnosis. The dependence and relatedness between phenotypes were taken into account in our similarity model through medical concept embedding. The relevance of each phenotype to each patient was also considered by adjusted aggregation of phenotype similarity into patient similarity. A ranking model based on the best-subtype-average similarity was proposed to address the phenotypic overlapping and heterogeneity of ciliopathies. Our results showed that using less than one-tenth of learning sources, our language and center specific embedding provided comparable or better performances than other existing medical concept embeddings. Combined with the best-subtype-average ranking model, our patient-patient similarity-based screening approach was demonstrated effective in two large scale unbalanced datasets containing approximately 10,000 and 60,000 controls with kidney manifestations in the clinical data warehouse (about 2 and 0.4% of prevalence, respectively). Our approach will offer the opportunity to identify candidate patients who could go through genetic testing for ciliopathy. Earlier diagnosis, before irreversible end-stage kidney disease, will enable these patients to benefit from appropriate follow-up and novel treatments that could alleviate kidney dysfunction.

Mining Electronic Health Records for Drugs Associated With 28-day Mortality in COVID-19: Pharmacopoeia-wide Association Study (PharmWAS)

BACKGROUND

Patients hospitalized for a given condition may be receiving other treatments for other contemporary conditions or comorbidities. The use of such observational clinical data for pharmacological hypothesis generation is appealing in the context of an emerging disease but particularly challenging due to the presence of drug indication bias.

OBJECTIVE

With this study, our main objective was the development and validation of a fully data-driven pipeline that would address this challenge. Our secondary objective was to generate pharmacological hypotheses in patients with COVID-19 and demonstrate the clinical relevance of the pipeline.

METHODS

We developed a pharmacopeia-wide association study (PharmWAS) pipeline inspired from the PheWAS methodology, which systematically screens for associations between the whole pharmacopeia and a clinical phenotype. First, a fully data-driven procedure based on adaptive least absolute shrinkage and selection operator (LASSO) determined drug-specific adjustment sets. Second, we computed several measures of association, including robust methods based on propensity scores (PSs) to control indication bias. Finally, we applied the Benjamini and Hochberg procedure of the false discovery rate (FDR). We applied this method in a multicenter retrospective cohort study using electronic medical records from 16 university hospitals of the Greater Paris area. We included all adult patients between 18 and 95 years old hospitalized in conventional wards for COVID-19 between February 1, 2020, and June 15, 2021. We investigated the association between drug prescription within 48 hours from admission and 28-day mortality. We validated our data-driven pipeline against a knowledge-based pipeline on 3 treatments of reference, for which experts agreed on the expected association with mortality. We then demonstrated its clinical relevance by screening all drugs prescribed in more than 100 patients to generate pharmacological hypotheses.

RESULTS

A total of 5783 patients were included in the analysis. The median age at admission was 69.2 (IQR 56.7-81.1) years, and 3390 (58.62%) of the patients were male. The performance of our automated pipeline was comparable or better for controlling bias than the knowledge-based adjustment set for 3 reference drugs: dexamethasone, phloroglucinol, and paracetamol. After correction for multiple testing, 4 drugs were associated with increased in-hospital mortality. Among these, diazepam and tramadol were the only ones not discarded by automated diagnostics, with adjusted odds ratios of 2.51 (95% CI 1.52-4.16, Q=.1) and 1.94 (95% CI 1.32-2.85, Q=.02), respectively.

CONCLUSIONS

Our innovative approach proved useful in generating pharmacological hypotheses in an outbreak setting, without requiring a priori knowledge of the disease. Our systematic analysis of early prescribed treatments from patients hospitalized for COVID-19 showed that diazepam and tramadol are associated with increased 28-day mortality. Whether these drugs could worsen COVID-19 needs to be further assessed.

Hydroxyzine Use and Mortality in Patients Hospitalized for COVID-19: A Multicenter Observational Study

(1) Background: Based on its antiviral activity, anti-inflammatory properties, and functional inhibition effects on the acid sphingomyelinase/ceramide system (FIASMA), we sought to examine the potential usefulness of the H1 antihistamine hydroxyzine in patients hospitalized for COVID-19. (2) Methods: In a multicenter observational study, we included 15,103 adults hospitalized for COVID-19, of which 164 (1.1%) received hydroxyzine within the first 48 h of hospitalization, administered orally at a median daily dose of 25.0 mg (SD = 29.5). We compared mortality rates between patients who received hydroxyzine at hospital admission and those who did not, using a multivariable logistic regression model adjusting for patients' characteristics, medical conditions, and use of other medications. (3) Results: This analysis showed a significant association between hydroxyzine use and reduced mortality (AOR, 0.51; 95%CI, 0.29-0.88, p = 0.016). This association was similar in multiple sensitivity analyses. (4) Conclusions: In this retrospective observational multicenter study, the use of the FIASMA hydroxyzine was associated with reduced mortality in patients hospitalized for COVID-19. Double-blind placebo-controlled randomized clinical trials of hydroxyzine for COVID-19 are needed to confirm these results, as are studies to examine the potential usefulness of this medication for outpatients and as post-exposure prophylaxis for individuals at high risk for severe COVID-19.

A framework for validating AI in precision medicine: considerations from the European ITFoC consortium

BACKGROUND

Artificial intelligence (AI) has the potential to transform our healthcare systems significantly. New AI technologies based on machine learning approaches should play a key role in clinical decision-making in the future. However, their implementation in health care settings remains limited, mostly due to a lack of robust validation procedures. There is a need to develop reliable assessment frameworks for the clinical validation of AI. We present here an approach for assessing AI for predicting treatment response in triple-negative breast cancer (TNBC), using real-world data and molecular -omics data from clinical data warehouses and biobanks.

METHODS

The European "ITFoC (Information Technology for the Future Of Cancer)" consortium designed a framework for the clinical validation of AI technologies for predicting treatment response in oncology.

RESULTS

This framework is based on seven key steps specifying: (1) the intended use of AI, (2) the target population, (3) the timing of AI evaluation, (4) the datasets used for evaluation, (5) the procedures used for ensuring data safety (including data quality, privacy and security), (6) the metrics used for measuring performance, and (7) the procedures used to ensure that the AI is explainable. This framework forms the basis of a validation platform that we are building for the "ITFoC Challenge". This community-wide competition will make it possible to assess and compare AI algorithms for predicting the response to TNBC treatments with external real-world datasets.

CONCLUSIONS

The predictive performance and safety of AI technologies must be assessed in a robust, unbiased and transparent manner before their implementation in healthcare settings. We believe that the consideration of the ITFoC consortium will contribute to the safe transfer and implementation of AI in clinical settings, in the context of precision oncology and personalized care.

Dexamethasone use and mortality in hospitalized patients with coronavirus disease 2019: A multicentre retrospective observational study

AIMS

To examine the association between dexamethasone use and mortality among patients hospitalized for COVID-19.

METHODS

We examined the association between dexamethasone use and mortality at AP-HP Greater Paris University hospitals. Study baseline was defined as the date of hospital admission. The primary endpoint was time to death. We compared this endpoint between patients who received dexamethasone and those who did not in time-to-event analyses adjusted for patient characteristics (such as age, sex and comorbidity) and clinical and biological markers of clinical severity of COVID-19, and stratified by the need for respiratory support, i.e. mechanical ventilation or oxygen. The primary analysis was a multivariable Cox regression model.

RESULTS

Of 12 217 adult patients hospitalized with a positive COVID-19 reverse transcriptase-polymerase chain reaction test, 171 (1.4%) received dexamethasone orally or by intravenous perfusion during the visit. Among patients who required respiratory support, the end-point occurred in 10/63 (15.9%) patients who received dexamethasone and 298/1129 (26.4%) patients who did not. In this group, there was a significant association between dexamethasone use and reduced mortality in the primary analysis (hazard ratio, 0.46; 95% confidence interval 0.22-0.96, P = .039). Among patients who did not require respiratory support, there was no significant association between dexamethasone use and the endpoint.

CONCLUSIONS

In this multicentre observational study, dexamethasone use administered either orally or by intravenous injection at a cumulative dose between 60 mg and 150 mg was associated with reduced mortality among patients with COVID-19 requiring respiratory support.

Association between antidepressant use and reduced risk of intubation or death in hospitalized patients with COVID-19: results from an observational study

A prior meta-analysis showed that antidepressant use in major depressive disorder was associated with reduced plasma levels of several pro-inflammatory mediators, which have been associated with severe COVID-19. Recent studies also suggest that several antidepressants may inhibit acid sphingomyelinase activity, which may prevent the infection of epithelial cells with SARS-CoV-2, and that the SSRI fluoxetine may exert in-vitro antiviral effects on SARS-CoV-2. We examined the potential usefulness of antidepressant use in patients hospitalized for COVID-19 in an observational multicenter retrospective cohort study conducted at AP-HP Greater Paris University hospitals. Of 7230 adults hospitalized for COVID-19, 345 patients (4.8%) received an antidepressant within 48 h of hospital admission. The primary endpoint was a composite of intubation or death. We compared this endpoint between patients who received antidepressants and those who did not in time-to-event analyses adjusted for patient characteristics, clinical and biological markers of disease severity, and other psychotropic medications. The primary analysis was a multivariable Cox model with inverse probability weighting. This analysis showed a significant association between antidepressant use and reduced risk of intubation or death (HR, 0.56; 95% CI, 0.43-0.73, p < 0.001). This association remained significant in multiple sensitivity analyses. Exploratory analyses suggest that this association was also significant for SSRI and non-SSRI antidepressants, and for fluoxetine, paroxetine, escitalopram, venlafaxine, and mirtazapine (all p < 0.05). These results suggest that antidepressant use could be associated with lower risk of death or intubation in patients hospitalized for COVID-19. Double-blind controlled randomized clinical trials of antidepressant medications for COVID-19 are needed.

Low-income neighbourhood was a key determinant of severe COVID-19 incidence during the first wave of the epidemic in Paris

BACKGROUND

Previous studies have demonstrated that socioeconomic factors are associated with COVID-19 incidence. In this study, we analysed a broad range of socioeconomic indicators in relation to hospitalised cases in the Paris area.

METHODS

We extracted 303 socioeconomic indicators from French census data for 855 residential units in Paris and assessed their association with COVID-19 hospitalisation risk.

FINDINGS

The indicators most associated with hospitalisation risk were the third decile of population income (OR=9.10, 95% CI 4.98 to 18.39), followed by the primary residence rate (OR=5.87, 95% CI 3.46 to 10.61), rate of active workers in unskilled occupations (OR=5.04, 95% CI 3.03 to 8.85) and rate of women over 15 years old with no diploma (OR=5.04, 95% CI 3.03 to 8.85). Of note, population demographics were considerably less associated with hospitalisation risk. Among these indicators, the rate of women aged between 45 and 59 years (OR=2.17, 95% CI 1.40 to 3.44) exhibited the greatest level of association, whereas population density was not associated. Overall, 86% of COVID-19 hospitalised cases occurred within the 45% most deprived areas.

INTERPRETATION

Studying a broad range of socioeconomic indicators using census data and hospitalisation data as a readily available and large resource can provide real-time indirect information on populations with a high incidence of COVID-19.

A COVID-19 Decision Support System for Phone Call Triage, Designed by and for Medical Students

During spring 2020, SARS-CoV-2 pandemic induced shortage of medical equipment, hospital capacity and staff. To tackle this issue, medical students have been strongly involved in early patient triage through medical phone call regulation. Here, we present an intelligent web-based decision support system for COVID-19 phone call regulation, developed by and for, medical students to help them during this difficult but crucial task. The system is divided into 5 tabs. The first tab displays administrative information, clinical data related to life-threatening emergency, and personalized recommendations for patient management. The second tab displays a PDF report summarizing the clinical situation; the third tab displays the guidelines used for establishing the recommendations, and the fourth tab displays the overall algorithm in the form of a decision tree. The fifth tab provided a short user guide. The system was assessed by 21 medical staff. More than 90% of them appreciated the navigation and the interface, and found the content relevant. 90,5% of them would like to use it during the medical regulation. In the future, we plan to use this system during simulation-based medical learning for the initial medical training of medical students.

Reorganisation of GP surgeries during the COVID-19 outbreak: analysis of guidelines from 15 countries

BACKGROUND

General practitioners (GPs) play a key role in managing the COVID-19 outbreak. However, they may encounter difficulties adapting their practices to the pandemic. We provide here an analysis of guidelines for the reorganisation of GP surgeries during the beginning of the pandemic from 15 countries.

METHODS

A network of GPs collaborated together in a three-step process: (i) identification of key recommendations of GP surgery reorganisation, according to WHO, CDC and health professional resources from health care facilities; (ii) collection of key recommendations included in the guidelines published in 15 countries; (iii) analysis, comparison and synthesis of the results.

RESULTS

Recommendations for the reorganisation of GP surgeries of four types were identified: (i) reorganisation of GP consultations (cancelation of non-urgent consultations, follow-up via e-consultations), (ii) reorganisation of GP surgeries (area partitioning, visual alerts and signs, strict hygiene measures), (iii) reorganisation of medical examinations by GPs (equipment, hygiene, partial clinical examinations, patient education), (iv) reorganisation of GP staff (equipment, management, meetings, collaboration with the local community).

CONCLUSIONS

We provide here an analysis of guidelines for the reorganisation of GP surgeries during the beginning of the COVID-19 outbreak from 15 countries. These guidelines focus principally on clinical care, with less attention paid to staff management, and the area of epidemiological surveillance and research is largely neglected. The differences of guidelines between countries and the difficulty to apply them in routine care, highlight the need of advanced research in primary care. Thereby, primary care would be able to provide recommendations adapted to the real-world settings and with stronger evidence, which is especially necessary during pandemics.

The Role of Radiomics in Lung Cancer: From Screening to Treatment and Follow-Up

Purpose

Lung cancer represents the first cause of cancer-related death in the world. Radiomics studies arise rapidly in this late decade. The aim of this review is to identify important recent publications to be synthesized into a comprehensive review of the current status of radiomics in lung cancer at each step of the patients’ care.

Methods

A literature review was conducted using PubMed/Medline for search of relevant peer-reviewed publications from January 2012 to June 2020

Results

We identified several studies at each point of patient’s care: detection and classification of lung nodules (n=16), determination of histology and genomic (n=10) and finally treatment outcomes predictions (=23). We reported the methodology of those studies and their results and discuss the limitations and the progress to be made for clinical routine applications.

Conclusion

Promising perspectives arise from machine learning applications and radiomics based models in lung cancers, yet further data are necessary for their implementation in daily care. Multicentric collaboration and attention to quality and reproductivity of radiomics studies should be further consider.

Concerns Discussed on Chinese and French Social Media During the COVID-19 Lockdown: Comparative Infodemiology Study Based on Topic Modeling

Background

During the COVID-19 pandemic, numerous countries, including China and France, have implemented lockdown measures that have been effective in controlling the epidemic. However, little is known about the impact of these measures on the population as expressed on social media from different cultural contexts.

Objective

This study aims to assess and compare the evolution of the topics discussed on Chinese and French social media during the COVID-19 lockdown.

Methods

We extracted posts containing COVID-19–related or lockdown-related keywords in the most commonly used microblogging social media platforms (ie, Weibo in China and Twitter in France) from 1 week before lockdown to the lifting of the lockdown. A topic model was applied independently for three periods (prelockdown, early lockdown, and mid to late lockdown) to assess the evolution of the topics discussed on Chinese and French social media.

Results

A total of 6395; 23,422; and 141,643 Chinese Weibo messages, and 34,327; 119,919; and 282,965 French tweets were extracted in the prelockdown, early lockdown, and mid to late lockdown periods, respectively, in China and France. Four categories of topics were discussed in a continuously evolving way in all three periods: epidemic news and everyday life, scientific information, public measures, and solidarity and encouragement. The most represented category over all periods in both countries was epidemic news and everyday life. Scientific information was far more discussed on Weibo than in French tweets. Misinformation circulated through social media in both countries; however, it was more concerned with the virus and epidemic in China, whereas it was more concerned with the lockdown measures in France. Regarding public measures, more criticisms were identified in French tweets than on Weibo. Advantages and data privacy concerns regarding tracing apps were also addressed in French tweets. All these differences were explained by the different uses of social media, the different timelines of the epidemic, and the different cultural contexts in these two countries.

Conclusions

This study is the first to compare the social media content in eastern and western countries during the unprecedented COVID-19 lockdown. Using general COVID-19–related social media data, our results describe common and different public reactions, behaviors, and concerns in China and France, even covering the topics identified in prior studies focusing on specific interests. We believe our study can help characterize country-specific public needs and appropriately address them during an outbreak.

Hybrid Deep Learning for Medication-Related Information Extraction From Clinical Texts in French: MedExt Algorithm Development Study

BACKGROUND

Information related to patient medication is crucial for health care; however, up to 80% of the information resides solely in unstructured text. Manual extraction is difficult and time-consuming, and there is not a lot of research on natural language processing extracting medical information from unstructured text from French corpora.

OBJECTIVE

We aimed to develop a system to extract medication-related information from clinical text written in French.

METHODS

We developed a hybrid system combining an expert rule-based system, contextual word embedding (embedding for language model) trained on clinical notes, and a deep recurrent neural network (bidirectional long short term memory-conditional random field). The task consisted of extracting drug mentions and their related information (eg, dosage, frequency, duration, route, condition). We manually annotated 320 clinical notes from a French clinical data warehouse to train and evaluate the model. We compared the performance of our approach to those of standard approaches: rule-based or machine learning only and classic word embeddings. We evaluated the models using token-level recall, precision, and F-measure.

RESULTS

The overall F-measure was 89.9% (precision 90.8; recall: 89.2) when combining expert rules and contextualized embeddings, compared to 88.1% (precision 89.5; recall 87.2) without expert rules or contextualized embeddings. The F-measures for each category were 95.3% for medication name, 64.4% for drug class mentions, 95.3% for dosage, 92.2% for frequency, 78.8% for duration, and 62.2% for condition of the intake.

CONCLUSIONS

Associating expert rules, deep contextualized embedding, and deep neural networks improved medication information extraction. Our results revealed a synergy when associating expert knowledge and latent knowledge.

Facing new challenges to informed consent processes in the context of translational research: the case in CARPEM consortium

BACKGROUND

In the context of translational research, researchers have increasingly been using biological samples and data in fundamental research phases. To explore informed consent practices, we conducted a retrospective study on informed consent documents that were used for CARPEM's translational research programs. This review focused on detailing their form, their informational content, and the adequacy of these documents with the international ethical principles and participants' rights.

METHODS

Informed consent forms (ICFs) were collected from CARPEM investigators. A content analysis focused on information related to biological samples and data treatment (context of sampling and collect, aims, reuse, consent renewal), including the type of consent. An automatic assessment of the readability of the ICFs were performed with the IT program "Flesch Score".

RESULTS

29 ICFs from 25 of 49 studies were analyzed after selection criteria were applied. Three types of consent were identified: 11 broad consents, six specific consents, and two opt-out consents. The Flesch Scores showed that most of the documents were too complex to be fully understood by most of the potential research participants. Most of the biological samples were collected during the healthcare routine, but the information content about secondary use of biological samples varied between ICFs. All documents mentioned personal data treatment but information about their reuse was not standardized in the ICFs.

CONCLUSIONS

Our review of current IC procedures of CARPEM showed that practices could be improved considering new translational research methods. "Old fashion written ICFs" should be adapted to the translational research approach, to better respect individual rights and international research ethics principles. In this context, theoretically, a digital tool allowing dynamic information and consent of participants, through an electronic interactive platform may be a good way to promote more active participation in research. Nevertheless, its feasibility in the complex environment of biological samples and data research remains to prove. The way of a combination of a broad consent followed by dynamic information may be alternatively tested.

Can reproducibility be improved in clinical natural language processing? A study of 7 clinical NLP suites

Background

The increasing complexity of data streams and computational processes in modern clinical health information systems makes reproducibility challenging. Clinical natural language processing (NLP) pipelines are routinely leveraged for the secondary use of data. Workflow management systems (WMS) have been widely used in bioinformatics to handle the reproducibility bottleneck.

Objective

To evaluate if WMS and other bioinformatics practices could impact the reproducibility of clinical NLP frameworks.

Materials and Methods

Based on the literature across multiple researcho fields (NLP, bioinformatics and clinical informatics) we selected articles which (1) review reproducibility practices and (2) highlight a set of rules or guidelines to ensure tool or pipeline reproducibility. We aggregate insight from the literature to define reproducibility recommendations. Finally, we assess the compliance of 7 NLP frameworks to the recommendations.

Results

We identified 40 reproducibility features from 8 selected articles. Frameworks based on WMS match more than 50% of features (26 features for LAPPS Grid, 22 features for OpenMinted) compared to 18 features for current clinical NLP framework (cTakes, CLAMP) and 17 features for GATE, ScispaCy, and Textflows.

Discussion

34 recommendations are endorsed by at least 2 articles from our selection. Overall, 15 features were adopted by every NLP Framework. Nevertheless, frameworks based on WMS had a better compliance with the features.

Conclusion

NLP frameworks could benefit from lessons learned from the bioinformatics field (eg, public repositories of curated tools and workflows or use of containers for shareability) to enhance the reproducibility in a clinical setting.

The Epidemiology of Patients' Email Addresses in a French University Hospital: Case-Control Study

Background

Health care professionals are caught between the wish of patients to speed up health-related communication via emails and the need for protecting health information.

Objective

We aimed to analyze the demographic characteristics of patients providing an email, and study the distribution of emails’ domain names.

Methods

We used the information system of the European Hospital Georges Pompidou (HEGP) to identify patients who provided an email address. We used a 1:1 matching strategy to study the demographic characteristics of the patients associated with the presence of an email, and described the characteristics of the emails used (in terms of types of emails—free, business, and personal).

Results

Overall, 4.22% (41,004/971,822) of the total population of patients provided an email address. The year of last contact with the patient is the strongest driver of the presence of an email address (odds ratio [OR] 20.8, 95% CI 18.9-22.9). Patients more likely to provide an email address were treated for chronic conditions and were more likely born between 1950 and 1969 (taking patients born before 1950 as reference [OR 1.60, 95% CI 1.54-1.67], and compared to those born after 1990 [OR 0.56, 95% CI 0.53-0.59]). Of the 41,004 email addresses collected, 37,779 were associated with known email providers, 31,005 email addresses were associated with Google, Microsoft, Orange, and Yahoo!, 2878 with business emails addresses, and 347 email addresses with personalized domain names.

Conclusions

Emails have been collected only recently in our institution. The importance of the year of last contact probably reflects this recent change in contact information collection policy. The demographic characteristics and especially the age distribution are likely the result of a population bias in the hospital: patients providing email are more likely to be treated for chronic diseases. A risk analysis of the use of email revealed several situations that could constitute a breach of privacy that is both likely and with major consequences. Patients treated for chronic diseases are more likely to provide an email address, and are also more at risk in case of privacy breach. Several common situations could expose their private information. We recommend a very restrictive use of the emails for health communication.

Association of Antihypertensive Agents with the Risk of In-Hospital Death in Patients with Covid-19

PURPOSE

The role of angiotensin receptor blockers (ARB), angiotensin-converting enzyme inhibitors (ACEi), or other antihypertensive agents in the case of Covid-19 remains controversial. We aimed to investigate the association between antihypertensive agent exposure and in-hospital mortality in patients with Covid-19.

METHODS

We performed a retrospective multicenter cohort study on patients hospitalized between February 1 and May 15, 2020. All patients had been followed up for at least 30 days.

RESULTS

Of the 8078 hospitalized patients for Covid-19, 3686 (45.6%) had hypertension and were included in the study. In this population, the median age was 75.4 (IQR, 21.5) years and 57.1% were male. Overall in-hospital 30-day mortality was 23.1%. The main antihypertensive pharmacological classes used were calcium channel blockers (CCB) (n=1624, 44.1%), beta-blockers (n=1389, 37.7%), ARB (n=1154, 31.3%), and ACEi (n=998, 27.1%). The risk of mortality was lower in CCB (aOR, 0.83 [0.70-0.99]) and beta-blockers (aOR, 0.80 [0.67-0.95]) users and non-significant in ARB (aOR, 0.88 [0.72-1.06]) and ACEi (aOR, 0.83 [0.68-1.02]) users, compared to non-users. These results remain consistent for patients receiving CCB, beta-blocker, or ARB as monotherapies.

CONCLUSION

This large multicenter retrospective of Covid-19 patients with hypertension found a reduced mortality among CCB and beta-blockers users, suggesting a putative protective effect. Our findings did not show any association between the use of renin-angiotensin-aldosterone system inhibitors and the risk of in-hospital death. Although they need to be confirmed in further studies, these results support the continuation of antihypertensive agents in patients with Covid-19, in line with the current guidelines.

International electronic health record-derived COVID-19 clinical course profiles: the 4CE consortium

We leveraged the largely untapped resource of electronic health record data to address critical clinical and epidemiological questions about Coronavirus Disease 2019 (COVID-19). To do this, we formed an international consortium (4CE) of 96 hospitals across five countries (www.covidclinical.net). Contributors utilized the Informatics for Integrating Biology and the Bedside (i2b2) or Observational Medical Outcomes Partnership (OMOP) platforms to map to a common data model. The group focused on temporal changes in key laboratory test values. Harmonized data were analyzed locally and converted to a shared aggregate form for rapid analysis and visualization of regional differences and global commonalities. Data covered 27,584 COVID-19 cases with 187,802 laboratory tests. Case counts and laboratory trajectories were concordant with existing literature. Laboratory tests at the time of diagnosis showed hospital-level differences equivalent to country-level variation across the consortium partners. Despite the limitations of decentralized data generation, we established a framework to capture the trajectory of COVID-19 disease in patients and their response to interventions.

International electronic health record-derived COVID-19 clinical course profiles: the 4CE consortium

We leveraged the largely untapped resource of electronic health record data to address critical clinical and epidemiological questions about Coronavirus Disease 2019 (COVID-19). To do this, we formed an international consortium (4CE) of 96 hospitals across five countries (www.covidclinical.net). Contributors utilized the Informatics for Integrating Biology and the Bedside (i2b2) or Observational Medical Outcomes Partnership (OMOP) platforms to map to a common data model. The group focused on temporal changes in key laboratory test values. Harmonized data were analyzed locally and converted to a shared aggregate form for rapid analysis and visualization of regional differences and global commonalities. Data covered 27,584 COVID-19 cases with 187,802 laboratory tests. Case counts and laboratory trajectories were concordant with existing literature. Laboratory tests at the time of diagnosis showed hospital-level differences equivalent to country-level variation across the consortium partners. Despite the limitations of decentralized data generation, we established a framework to capture the trajectory of COVID-19 disease in patients and their response to interventions.

Natural Language Processing for Rapid Response to Emergent Diseases: Case Study of Calcium Channel Blockers and Hypertension in the COVID-19 Pandemic

BACKGROUND

A novel disease poses special challenges for informatics solutions. Biomedical informatics relies for the most part on structured data, which require a preexisting data or knowledge model; however, novel diseases do not have preexisting knowledge models. In an emergent epidemic, language processing can enable rapid conversion of unstructured text to a novel knowledge model. However, although this idea has often been suggested, no opportunity has arisen to actually test it in real time. The current coronavirus disease (COVID-19) pandemic presents such an opportunity.

OBJECTIVE

The aim of this study was to evaluate the added value of information from clinical text in response to emergent diseases using natural language processing (NLP).

METHODS

We explored the effects of long-term treatment by calcium channel blockers on the outcomes of COVID-19 infection in patients with high blood pressure during in-patient hospital stays using two sources of information: data available strictly from structured electronic health records (EHRs) and data available through structured EHRs and text mining.

RESULTS

In this multicenter study involving 39 hospitals, text mining increased the statistical power sufficiently to change a negative result for an adjusted hazard ratio to a positive one. Compared to the baseline structured data, the number of patients available for inclusion in the study increased by 2.95 times, the amount of available information on medications increased by 7.2 times, and the amount of additional phenotypic information increased by 11.9 times.

CONCLUSIONS

In our study, use of calcium channel blockers was associated with decreased in-hospital mortality in patients with COVID-19 infection. This finding was obtained by quickly adapting an NLP pipeline to the domain of the novel disease; the adapted pipeline still performed sufficiently to extract useful information. When that information was used to supplement existing structured data, the sample size could be increased sufficiently to see treatment effects that were not previously statistically detectable.

The Impact of Specialized Corpora for Word Embeddings in Natural Langage Understanding

Recent studies in the biomedical domain suggest that learning statistical word representations (static or contextualized word embeddings) on large corpora of specialized data improve the results on downstream natural language processing (NLP) tasks. In this paper, we explore the impact of the data source of word representations on a natural language understanding task. We compared embeddings learned with Fasttext (static embedding) and ELMo (contextualized embedding) representations, learned either on the general domain (Wikipedia) or on specialized data (electronic health records, EHR). The best results were obtained with ELMo representations learned on EHR data for the two sub-tasks (+7% and +4% of gain in F1-score). Moreover, ELMo representations were trained with only a fraction of the data used for Fasttext.

Diagnosis support systems for rare diseases: a scoping review

INTRODUCTION

Rare diseases affect approximately 350 million people worldwide. Delayed diagnosis is frequent due to lack of knowledge of most clinicians and a small number of expert centers. Consequently, computerized diagnosis support systems have been developed to address these issues, with many relying on rare disease expertise and taking advantage of the increasing volume of generated and accessible health-related data. Our objective is to perform a review of all initiatives aiming to support the diagnosis of rare diseases.

METHODS

A scoping review was conducted based on methods proposed by Arksey and O'Malley. A charting form for relevant study analysis was developed and used to categorize data.

RESULTS

Sixty-eight studies were retained at the end of the charting process. Diagnosis targets varied from 1 rare disease to all rare diseases. Material used for diagnosis support consisted mostly of phenotype concepts, images or fluids. Fifty-seven percent of the studies used expert knowledge. Two-thirds of the studies relied on machine learning algorithms, and one-third used simple similarities. Manual algorithms were encountered as well. Most of the studies presented satisfying performance of evaluation by comparison with references or with external validation. Fourteen studies provided online tools, most of which aimed to support the diagnosis of all rare diseases by considering queries based on phenotype concepts.

CONCLUSION

Numerous solutions relying on different materials and use of various methodologies are emerging with satisfying preliminary results. However, the variability of approaches and evaluation processes complicates the comparison of results. Efforts should be made to adequately validate these tools and guarantee reproducibility and explicability.

Radiomics: A primer for the radiation oncologist

PURPOSE

Radiomics are a set of methods used to leverage medical imaging and extract quantitative features that can characterize a patient's phenotype. All modalities can be used with several different software packages. Specific informatics methods can then be used to create meaningful predictive models. In this review, we will explain the major steps of a radiomics analysis pipeline and then present the studies published in the context of radiation therapy.

METHODS

A literature review was performed on Medline using the search engine PubMed. The search strategy included the search terms "radiotherapy", "radiation oncology" and "radiomics". The search was conducted in July 2019 and reference lists of selected articles were hand searched for relevance to this review.

RESULTS

A typical radiomics workflow always includes five steps: imaging and segmenting, data curation and preparation, feature extraction, exploration and selection and finally modeling. In radiation oncology, radiomics studies have been published to explore different clinical outcome in lung (n=5), head and neck (n=5), esophageal (n=3), rectal (n=3), pancreatic (n=2) cancer and brain metastases (n=2). The quality of these retrospective studies is heterogeneous and their results have not been translated to the clinic.

CONCLUSION

Radiomics has a great potential to predict clinical outcome and better personalize treatment. But the field is still young and constantly evolving. Improvement in bias reduction techniques and multicenter studies will hopefully allow more robust and generalizable models.

Healthcare trajectory of children with rare bone disease attending pediatric emergency departments

Background

Children with rare bone diseases (RBDs), whether medically complex or not, raise multiple issues in emergency situations. The healthcare burden of children with RBD in emergency structures remains unknown. The objective of this study was to describe the place of the pediatric emergency department (PED) in the healthcare of children with RBD.

Methods

We performed a retrospective single-center cohort study at a French university hospital. We included all children under the age of 18 years with RBD who visited the PED in 2017. By cross-checking data from the hospital clinical data warehouse, we were able to trace the healthcare trajectories of the patients. The main outcome of interest was the incidence (IR) of a second healthcare visit (HCV) within 30 days of the index visit to the PED. The secondary outcomes were the IR of planned and unplanned second HCVs and the proportion of patients classified as having chronic medically complex (CMC) disease at the PED visit.

Results

The 141 visits to the PED were followed by 84 s HCVs, giving an IR of 0.60 [95% CI: 0.48–0.74]. These second HCVs were planned in 60 cases (IR = 0.43 [95% CI: 0.33–0.55]) and unplanned in 24 (IR = 0.17 [95% CI: 0.11–0.25]). Patients with CMC diseases accounted for 59 index visits (42%) and 43 s HCVs (51%). Multivariate analysis including CMC status as an independent variable, with adjustment for age, yielded an incidence rate ratio (IRR) of second HCVs of 1.51 [95% CI: 0.98–2.32]. The IRR of planned second HCVs was 1.20 [95% CI: 0.76–1.90] and that of unplanned second HCVs was 2.81 [95% CI: 1.20–6.58].

Conclusion

An index PED visit is often associated with further HCVs in patients with RBD. The IRR of unplanned second HCVs was high, highlighting the major burden of HCVs for patients with chronic and severe disease.

What can millions of laboratory test results tell us about the temporal aspect of data quality? Study of data spanning 17 years in a clinical data warehouse

OBJECTIVE

To identify common temporal evolution profiles in biological data and propose a semi-automated method to these patterns in a clinical data warehouse (CDW).

MATERIALS AND METHODS

We leveraged the CDW of the European Hospital Georges Pompidou and tracked the evolution of 192 biological parameters over a period of 17 years (for 445,000 + patients, and 131 million laboratory test results).

RESULTS

We identified three common profiles of evolution: discretization, breakpoints, and trends. We developed computational and statistical methods to identify these profiles in the CDW. Overall, of the 192 observed biological parameters (87,814,136 values), 135 presented at least one evolution. We identified breakpoints in 30 distinct parameters, discretizations in 32, and trends in 79.

DISCUSSION AND CONCLUSION

our method allowed the identification of several temporal events in the data. Considering the distribution over time of these events, we identified probable causes for the observed profiles: instruments or software upgrades and changes in computation formulas. We evaluated the potential impact for data reuse. Finally, we formulated recommendations to enable safe use and sharing of biological data collection to limit the impact of data evolution in retrospective and federated studies (e.g. the annotation of laboratory parameters presenting breakpoints or trends).

Do You Need Embeddings Trained on a Massive Specialized Corpus for Your Clinical Natural Language Processing Task?

We explore the impact of data source on word representations for different NLP tasks in the clinical domain in French (natural language understanding and text classification). We compared word embeddings (Fasttext) and language models (ELMo), learned either on the general domain (Wikipedia) or on specialized data (electronic health records, EHR). The best results were obtained with ELMo representations learned on EHR data for one of the two tasks(+7% and +8% of gain in F1-score).

Evaluating the Impact of Text Duplications on a Corpus of More than 600,000 Clinical Narratives in a French Hospital

A significant part of medical knowledge is stored as unstructured free text. However, clinical narratives are known to contain duplicated sections due to clinicians' copy/paste parts of a former report into a new one. In this study, we aim at evaluating the duplications found within patient records in more than 650,000 French clinical narratives. We adapted a method to identify efficiently duplicated zones in a reasonable time. We evaluated the potential impact of duplications in two use cases: the presence of (i) treatments and/or (ii) relative dates. We identified an average rate of duplication of 33%. We found that 20% of the document contained drugs mentioned only in duplicated zones and that 1.45% of the document contained mentions of relative dates in duplicated zone, that could potentially lead to erroneous interpretation. We suggest the systematic identification and annotation of duplicated zones in clinical narratives for information extraction and temporal-oriented tasks.