IQVIA Virtual NLP Summit 2022

Needle in a haystack: How text mining helps us at Sanofi find the right Drug Label information - Priya Puthankar & Gabriela Marroquin (Sanofi) 

Sanofi have embarked on an exciting project to help teams across Regulatory Affairs and beyond, to work with drug label data from across different sources.

The aim is to reduce the time they spend on tedious tasks, help them find the right data points they need to answer a variety of questions, and create more value for their stakeholders.

Connected NLP - David Milward (IQVIA)

Recent developments of IQVIA NLP technology have aimed to simplify standard workflows to reliably deliver great results, whilst also providing an open platform to allow experimentation and variety in processing methods. This talk will describe some of the latest developments in both areas.

IQVIA NLP is not only used in fully automated workflows; there are also many semi-automated workflows where human curation is used to review results, or sample results to check the quality. HART (Human-Assisted Review Tool) enables subject matter experts to review documents pre-annotated with IQVIA NLP. Once reviewed and curated, results can be distributed to end-users. Curated results can also be used as training data for machine learning models that can be incorporated back into the platform via the NLP connector framework or the newly incorporated ONNX framework for neural network models.

Automating biomarker and phenotype extraction with natural language processing in a real-world precision oncology platform - Irfan Shah (Guardant Health)

In the area of precision oncology, the more availability of high-quality clinical data that enables additional stratified insights to be gleaned, the better. This is the driving force behind an initiative by Guardant Health in partnership with IQVIA, Inc. to use state of the art text mining and Natural Language Processing (NLP) to structure and normalize complex clinical variables from clinical documents in the real-world clinical-genomic platform—Guardant INFORM.

Built on Guardant360 liquid biopsy results, Guardant INFORM combines large volumes of genomic and clinical data to help accelerate research and development of next generation cancer therapeutics. Using NLP, key information such as TNM stage, biomarker profile, tumor histology, smoking history and performance status is now extracted from the patient narrative and transformed to normalized, structured data with precision of up to 100% for certain variables. By coupling deep phenotypic information with liquid biopsy results, more targeted drug development, better clinical trial optimization and more powerful post market research are all enabled.

Spanning the gap: connecting data scientists to unstructured information - Paul Milligan (IQVIA)

Around 80% of a data scientists time is spent cleaning and preparing their data before it is ready to be processed. For unstructured information, this figure could be even higher when you factor in the lack of schemas, term ambiguity, inconsistent formatting and terminologies, etc.

In the session, we will show how IQVIA NLP products can be used to obtain data from free text that has been cleaned, structured and normalized, allow users to spend their time on what matters: solving high-value business problems using the available tools and data.

The Learning Health (Record) System: Deep Phenotyping, AI, and the Future of Precision Health - Philip Payne (Wash U)

Healthcare systems are producing data at a rate that exceeds growth in any other industry. A huge amount (80%) of this data is unstructured.

Those who adopt technologies to untap the potential value of these data are best placed to understand their populations more. And are able to provide appropriate and tailored treatment and initiatives where they are needed.

The latest in AIMed’s webinar series will explore how leaders from Washington University School of Medicine, St Louis and Kaiser Permanente Northern California are using NLP to advance clinical care.

Data and AI now have leading roles to play in advancing precision medicine research by identifying early onset and treatments for disease. Dr Philip Payne will share how Washington University is building a set of NLP pipelines to extract high quality phenotypic data from the clinical narrative, to develop registries for patients with Alzheimer’s disease, breast cancer, diabetes and obesity. The use of natural language processing is key because close to 80% of the high value phenotypic data is encoded in the clinical narrative, not in structured or discrete fields in the electronic health record.
 

Precision population health and risk stratification – transform unstructured data across the healthcare ecosystem - Calum Yacoubian (IQVIA)

It has long been known that the unstructured data holds the answers to most questions in healthcare – but finding repeatable and scalable solutions to unlock this data across functions and applications has remained a challenge. In this session, we will learn how new components in the NLP toolkit can enable payers, providers and organizations processing medical records to transform processes, increase efficiencies and derive meaningful value to their organization.  

Deploying AI in Healthcare – from algorithm development to workflow integration - Nadea Leavitt & Matthew Hackenberg (IQVIA)

Artificial intelligence (AI) holds a promise of advancing our ability to improve patient care. Trained on routinely collected healthcare data, AI algorithms can identify patients at risk for a variety of medical events such as lingering misdiagnosis, disease progression, an upcoming severe adverse event, or non-adherence to treatment. Healthcare data can originate from structured fields (e.g., diagnoses captured via ICD-10) but 80% of data in the healthcare setting is unstructured, locked in sources such as radiology reports, clinical notes, discharge summaries and pathology notes.

Applying AI methods to these large data sets can provide a window of clarity for traditionally opaque patient predictors and support prioritization of patient engagement based on risk level and available clinical resources. This webinar will provide an overview of proven AI applications for medical event prediction, covering the lifecycle of solution development: data source selection, study design, training the AI algorithm, efficient integration into the clinical workflow, role of Social Determinants of Health as care gaps, etc. We will showcase real-world applications of such AI algorithms with measured impact to patient care across a variety of therapeutic areas.

Improving discreet cancer diagnoses from pathology reports with NLP: A case study in Cervical Intraepithelial Neoplasia - Soora Wi (Permanente Medical)

The terminology used by pathologists to describe and grade dysplasia and premalignant changes of the cervical epithelium has evolved over time. Unfortunately, coexistence of different classification systems combined with non-standardized interpretive text has created multiple layers of interpretive ambiguity. We used natural language processing (NLP) to automate and expedite translation of interpretive text to a single most severe, and thus actionable, cervical intraepithelial neoplasia (CIN) diagnosis. The algorithms that were developed and then applied to 35,847 unstructured cervical pathology reports assessed NLP performance in identifying the severest diagnosis, compared to expert manual review. NLP performance was determined by calculating precision (0.957), recall (0.925), and F score (0.94). Using NLP also significantly reduced the time to evaluate each monthly biopsy file from 30 hours to 0.5 hours.

The use of NLP rapidly and efficiently assigned a discrete, actionable diagnosis using CIN classification that can assist with clinical management of cervical pathology and disease. Moreover, discrete diagnostic data encoded as CIN terminology can enhance the efficiency of clinical research.

Getting started with NLP for Social Determinants of Health - Nirav Shah & Urmila Ravichandran (NorthShore Health System)

NorthShore has established itself as a preeminent health system in building and implementing analytical tools at scale to improve the quality and efficiency of care. A current goal is to build a learning health system. A learning health system is a system where knowledge generation processes are embedded in daily practices to produce continual improvements in care. At NorthShore, we are at the nascent stages of building a learning health system, with Natural Language Processing (NLP) as a starting point. 

We partnered with Linguamatics to understand how to leverage vast unstructured language fields that are currently of limited practical use but pose tremendous potential value to our data assets. With the COVID pandemic and the drive to push into value-based care, it is more important than ever that healthcare providers need to understand Social Determinants of Health (SDOH) to effectively and equitably care for their populations. It is estimated that SDOH can account for 60-70% of health outcomes and at an urban health system on the east coast where NLP was used to identify SDOH, 31% of their entire population had at least one SDOH that could be affecting their health. These socioeconomic, behavioral, and environmental factors are not routinely and discretely documented and available in structured fields but rather are documented in unstructured text notes. Identifying and extracting these data elements at the patient level will provide a valuable path to intervene on the social determinants that impact care and outcomes.  

Innovating drug development with natural language processing: focus on safety - Jane Reed (IQVIA)

Life science organizations face the challenge of handling ever-increasing volumes of text information within drug safety processes. Natural Language Processing (NLP) can be applied to optimize safety platforms and lower clinical development costs. NLP transforms unstructured text into structured data that can be rapidly analyzed or visualized.

This session will explore how these capabilities can be applied for safety case processing, medical coding (e.g. to MedDRA), publication search for potential Adverse Events situations and medical review of Adverse Events, and indeed, at every stage through the safety lifecycle of a drug.

Identification of potential targets for drug repurposing in neonatology - Paolo Grossi (Chiesi)

SKIR, together with members from Pre-Clinical departments, would like to investigate repurposing hypothesis to associate genes/proteins involved in a given disease towards drugs already known. Historically, such discoveries were the results of serendipity. The rapid growth in electronic clinical data and text mining tools makes it feasible to systematically obtain this kind of information. The overall process consists of two different workflows: first, we retrieved any genes related to the indication, sorting those by an internal defined Scoring System, based on descriptive and qualitative metadata. Comparing molecules with the internal knowledge, we were able to produce a Top Tier list of genes of interest. Later, we developed a specific query by correlating genes selection and all potential drugs as pharmaceutical substances possessing definite properties. Once again, we applied an analytical Scoring System to determine which could be most interesting drugs to use as repurposing target for future developments.

NLP for MedDRA Coding: Piloting in ICSR processing at CSL Behring - Martin Menke (CSL Behring)

For processing of adverse event reports, the information provided by a reporter in natural language is transferred (coded) into a standardized format to allow database processing. For the adverse event, indication, medical history, etc. the Medical Dictionary for Regulatory Activities – MedDRA must be used. Most of the coding is manual and time consuming. Only when the verbatim exactly matches a MedDRA term, coding is automatic (currently about 30%).

Natural language processing (NLP) is programming computers to process and analyze natural language data, to recognize relevant content and act upon it in a specific way, e.g. recognize a verbatim as an adverse event and assign a respective MedDRA code.

Panning for gold: Surfacing novel content to drive richer conversations - Hywel Evans (IQVIA)

As organizations engage with key stakeholders such as physicians, key opinion leaders and digital though leaders it is vital to be able to find relevant, specific content to answer questions and drive rich discussions. Over time these rich interactions can themselves form the basis for rich datasets that inform teams about emerging topics and effectiveness of communications. Find out more about text mining can unlock valuable data in pharma for stakeholder engagement.