Top Themes: Real-World Evidence (RWE), Cost-Effectiveness Analysis (CEA), and Quality of Life (QoL) were the dominant topics.
Big Pharma: Over 300 studies were presented, with Pfizer authoring 52.
AI Trends: AI was a key emerging topic, with numerous studies exploring its application in healthcare decision-making, systematic reviews, and economic modeling.
HTA and ICER: Over 20 studies examined trends related to Health Technology Assessment (HTA) and ICER (Institute for Clinical and Economic Review).
IRA-Related Research: 13 studies focused on the Inflation Reduction Act and its implications for drug pricing and access.
Key Focus: These three themes dominated the research landscape, with studies addressing their application in assessing drug efficacy, patient outcomes, and healthcare costs. Many studies presented real-world evidence to supplement clinical trial data, showing the value of using real-world data to inform treatment guidelines.
Examples of Studies:
Studies explored how RWE is being used to demonstrate the effectiveness of treatments in real clinical settings, often showing greater flexibility and adaptability compared to traditional clinical trials.
The importance of QoL measures was emphasized in patient-centered care and in evaluating the effectiveness of long-term treatments.
AI Tools for Literature Reviews:
A study on using AI to assist with Systematic Literature Reviews (SLR) showed a significant reduction in the time required for screening articles, with AI tools like GPT-4 achieving 94.91% agreement with human reviewers. This demonstrates the potential of AI to streamline the traditionally labor-intensive process of literature reviews, speeding up research while maintaining accuracy.
AI tools were shown to save around 50% of the time in certain screening processes, making them useful for large-scale reviews while maintaining high sensitivity and specificity.
AI in Mammography Screening:
A study focused on the economic impact of AI in screening mammography suggested that augmentative AI could save healthcare facilities significant resources by reducing time spent on reading mammograms. For facilities screening 10,000 women per year, AI tools could save up to $71,000 annually by improving throughput and reducing the time needed to review normal exams.
Natural Language Processing (NLP) in Emergency Medicine:
NLP was shown to be highly effective in emergency medicine, particularly in radiologic interpretation, where it achieved a sensitivity of 93% and specificity of 96%. This suggests that NLP can help enhance the accuracy and efficiency of unstructured data processing in emergency settings, improving patient care and streamlining workflows.
ICER’s Health Technology Assessment (HTA) Methods:
Over 20 studies examined ICER HTA methods, focusing on the potential improvements in health equity and how ICER’s value frameworks compare with those of other HTA bodies like NICE.
A detailed assessment highlighted the challenges in balancing cost-effectiveness with broader social and equity concerns in health assessments.
Comparison of ICER and NICE:
Several studies compared ICER and NICE, highlighting differences in value assessments, particularly in rare diseases. For example, a study on treatments for non-alcoholic steatohepatitis (NASH) showed how different HTA bodies approach emerging treatments and the balance between innovation and cost-effectiveness.
Medicare Price Negotiation:
A series of studies examined the Inflation Reduction Act and its impact on drug pricing, particularly focusing on Medicare. One study explored how the IRA's negotiation timelines might delay access to drugs for orphan indications, while another assessed the impact on diabetes medications for Medicare beneficiaries.
A study also delved into the broader cost-savings generated by the IRA, with a focus on how the act’s provisions would influence commercial payers.
AI in Economic Modeling:
AI's role in updating economic modeling reports was explored, with an accuracy of 94.3% for AI-adapted reports compared to 98.5% for manually adapted reports. The findings suggest that AI can significantly reduce the time required for updating technical reports used in health technology assessments.
AI was also used to automate health economic evidence analysis, showing potential in handling large datasets and helping researchers generate insights more efficiently.
Drug Discovery Using AI and Machine Learning:
AI and ML were used to discover peptides and molecules for various conditions, including type 2 diabetes and cancer. Studies validated their potential for drug development, accelerating the discovery process and identifying promising molecules for future treatments.
Study Overview:
AI, specifically GPT-4, was used to review 2,085 rheumatology notes to identify 606 flare events in patients with SLE. The results showed a 97.5% agreement between AI and human reviewers, demonstrating AI's potential in improving the efficiency and accuracy of chart reviews in SLE and other diseases where flare events are critical to patient management.
These studies highlight the growing role of AI, RWE, and economic analyses in shaping the future of healthcare research, improving patient outcomes, and optimizing resource use.