AI Case Study Module Updated Mobile Copy

Explore the transformative capabilities of IQVIA Healthcare-grade AI

Discovery

Clinical

Safety

Market Access Med Affairs

Brand

Sales

Research & Development

Real World

Commercial

Create biomarker database

Increase clinical trial participation

Decrease risk liability

Predict treatment response

Identify qualified patient leads

Measure HCP response to marketing

Tap to read our customer success stories

Biomarker database created with IQVIA Natural Language Processing (NLP)

This led 1M facts to be extracted and integrated—including 12,000 genes 3,500 diseases, and 340k associations

Challenge: Diagnostic researchers from a top 10 pharma company needed to overcome barriers preventing ready access to disease-marker associations. Solution: They developed an automated workflow for IQVIA Natural Language Processing (NLP) to extract facts from MEDLINE, NCBI GeneRIFs (reference-into-function), and in-house sources. This data was loaded into their knowledge base “DiMA” for search portal and dashboards. Results: The dashboard graphic user interface (GUI) allows end users to filter information by marker or disease. It also offers visualisation of associations (expression, genetic variation, negative associations) to improve effectiveness.

seamlessly extracted and integrated—including:

12,000 genes

3,500 diseases

and

340k associations

1M facts

Research & Development Discovery

Increasing early identification and recruitment of patients with Alzheimer’s Disease

This led 1M facts to be extracted and integrated—including 12,000 genes 3,500 diseases, and 340k associations

Challenge: More than 100+ Alzheimer’s disease (AD) clinical trials have been unsuccessfully completed over the past two decades due to difficulties identifying, recruiting and retaining qualified participants. As a result, AD treatments are limited. Solution: Using predictive analysis and machine learning to quickly sort through data assets—including claims, electronic medical records, and prescription data—we identified patient populations with AD. Results: These findings, which achieved 80% predictive algorithm precision, are now making a valuable contribution to trial referral networks to further development of new treatments.

(76% through primary care settings) for Alzheimer’s Disease

223K PATIENTS

Research & Development Clinical

predicted as high risk

Improving the capture of adverse events in MedDRA

This led 1M facts to be extracted and integrated—including 12,000 genes 3,500 diseases, and 340k associations

Challenge: Adverse events (AE) are often captured in natural language that doesn’t readily fit into MedDRA classifications. This requires someone to manually code the AE to determine where it fits. Solution: We built IQVIA Natural Language Processing (NLP) query strategies that work with Oracle Argus safety case management system to extract AEs described in verbatim language in individual case safety reports (ICSRs) and automatically translate them into the MedDRA classifications. Results: This NLP solution doubled the amount of auto-coding possible (from only 30% auto-coded AEs to over 70%), while reducing the manual time required and improving coding consistently, thus ultimately decreasing risk liability.

Eliminated the need to manually code

(saving 1-30 minutes each) for one pharma company

1k Adverse Events

Real World Safety

monthly

Improving early diagnosis and prediction of treatment response in people with Alzheimer’s Disease (AD)

This led 1M facts to be extracted and integrated—including 12,000 genes 3,500 diseases, and 340k associations

Challenge: We needed to develop better ways to identify people with AD at an earlier stage and predict how they will respond to treatment. Solution: Working with a large-scale electronic medical record (EMR) vendor and treating physicians, we developed a predictive tool that integrates into the EMR systems to support physician-patient engagement. Results: This tool helps clinicians better determine optimal treatment type and dosage for people with AD. It also has potential to engage patients in helping develop effective treatment plans.

Real-world data from

helps identify key clinical and demographic predictors of Alzheimer's Disease

40 specialist centers

Real World Market Access & Medical Affairs

and

25k non-identified

patients

Using Artificial Intelligence & Machine Learning to expand patient pathways

This led 1M facts to be extracted and integrated—including 12,000 genes 3,500 diseases, and 340k associations

Challenge: A pharma company with a specialty biologic for severe asthma wanted to expand the reach of a rules-based trigger program to better identify eligible patients in a more time-effective way. Solution: We helped the client implement a more flexible, dynamic, and predictive targeting model and worked closely with the client brand, medical, sales ops, analytics, and legal teams to ensure a seamless transition. Results: This helped the company get better leads more efficiently to meet its targets over a 3-year deployment period.

Commercial Brand

increased precision

increase in new patient initiation

20%

return on investment

~$5.8M

Tailoring multi-channel marketing to better engage healthcare professionals (HCPs)

This led 1M facts to be extracted and integrated—including 12,000 genes 3,500 diseases, and 340k associations

Challenge: Every clinician has different preferences for communication methods and channels, which can make it complicated to identify the best approach for the target audience. Solution: We applied customized machine learning techniques to dynamically measure HCP responsiveness to promotions using a wealth of data, plus sensitivity scores by channel and time, to prioritize which physicians to reach, and the best channel to reach them. Results: We identified who, when, and how to effectively promote across channels. This allowed us to optimize channel synergy and investment.

Commercial Sales

within 7 months of implementation

revenue increase at no additional cost

$20M+

Patient Engagement and Support

Improve patient experience

Informing personalized patient engagement strategies with generative AI

This led 1M facts to be extracted and integrated—including 12,000 genes 3,500 diseases, and 340k associations

Challenge: Effective patient engagement requires a personalized experience. This can be achieved by creating evidence-based personas, but this often requires analyzing large volumes of data which is time-consuming and prone to errors. Solution: Using generative AI, we quickly and accurately analyze vast amounts of unstructured patient data to map patient journeys and generate insights. These insights help create physician and patient personas that can further be used to generate interactive AI avatars that let us test engagement strategies to improve patient experience. Results: This approach to using generative AI enables us to create more accurate personas and streamlines omnichannel content development, leading to improved patient communication, experience, and outcomes.

Commercial Patient Engagement and Support

Generative AI transforms insights into detailed personas and custom content, leading to more relevant and cohesive omnichannel experiences.

Enhanced engagement

Increased satisfaction

Improved outcomes

Patient Engagement and Support

Informing personalized patient engagement strategies with generative AI

Commercial Patient Engagement and Support