HSS Research Highlights AI’s Growing Role in Predicting Pain Risk and Improving Anesthesia Education
At the American Society of Regional Anesthesia and Acute Pain Medicine (ASRA) annual meeting, researchers from Hospital for Special Surgery (HSS) unveiled new findings demonstrating how artificial intelligence (AI) is reshaping both clinical prediction of postoperative pain and patient education in anesthesia. These studies reflect a broader shift toward data-driven, personalized care in perioperative medicine.
Two major research efforts were presented: one using machine learning to predict long-term pain after knee replacement surgery, and another analyzing patient search behavior to better understand public concerns about anesthesia. Together, these studies illustrate how AI can enhance both clinical decision-making and communication between physicians and patients.
Machine Learning Identifies Risk Factors for Persistent Pain After Knee Replacement
One of the most impactful studies presented by HSS researchers focused on predicting persistent postoperative pain (PPP) following total knee arthroplasty (TKA), commonly known as knee replacement surgery. Despite being a widely performed and generally successful procedure, approximately 20% of patients continue to experience significant pain months after surgery—pain that can severely affect daily functioning and overall quality of life.
Understanding Persistent Postoperative Pain
Persistent postoperative pain is typically defined as pain at the surgical site lasting three to six months after the procedure, with a severity level exceeding 4 on a 0–10 pain scale and significantly interfering with everyday activities. Identifying which patients are at risk has historically been challenging due to the complex interplay of biological, psychological, and surgical factors.
Leveraging Machine Learning for Deeper Insights
To address this challenge, the HSS team employed machine learning (ML), a subset of AI that analyzes large datasets to detect patterns and make predictions. Researchers analyzed data from 160 patients who had undergone knee replacement surgery, incorporating a wide range of variables—including clinical data, patient-reported pain levels, and biological markers from blood samples.
In total, 318 distinct features were evaluated using four different ML models. Among these, the XGBoost algorithm emerged as the most effective in identifying meaningful predictors of long-term pain.
Key Findings
The study uncovered several important predictors of persistent pain:
- Elevated inflammatory markers, particularly a molecule called TARC (thymus and activation-regulated chemokine), measured shortly after surgery
- Higher levels of preoperative pain, especially pain at rest
- Longer tourniquet time during surgery, which may influence tissue stress and inflammation
- Other inflammatory cytokines present in the bloodstream postoperatively
While some of these factors—such as preexisting pain and psychological conditions like anxiety or depression—were already known, the identification of TARC as a consistent predictor across all models was particularly notable.
A New Biological Signal
According to the researchers, TARC had not previously been a major focus in pain studies. Its consistent association with persistent pain across multiple models suggests it may play a significant role in postoperative inflammation and pain pathways.
This discovery highlights one of AI’s key advantages: the ability to uncover previously overlooked patterns or variables that traditional analysis methods might miss.
Toward Personalized Pain Management
The ultimate goal of this research is to improve patient outcomes by enabling more personalized care. By identifying high-risk individuals before or shortly after surgery, clinicians may be able to:
- Tailor pain management strategies
- Adjust surgical techniques where possible
- Provide earlier interventions to prevent chronic pain
- Set more accurate expectations for recovery
Although further validation is needed before these findings can be applied in routine clinical practice, the study represents a significant step toward precision medicine in pain management.
AI Study Reveals What Patients Want to Know About Anesthesia
In a separate study, HSS researchers turned their attention to patient behavior—specifically, how people seek information about anesthesia online. With many patients relying on internet searches to prepare for surgery, understanding these information needs is critical for improving communication and education.
Exploring Patient Search Behavior
The research team used AI to analyze common questions appearing in Google’s “People Also Ask” feature. Seven anesthesia-related search terms were used, including:
- Regional anesthesia
- Peripheral nerve block
- Epidural anesthesia
- Spinal anesthesia
- Pain block
From these searches, researchers collected 1,400 question-and-website pairs, representing the most frequently encountered queries and their associated sources.
Key Themes in Patient Questions
The analysis revealed that patients are primarily concerned with:
- Risks and complications of anesthesia
- Medication details, including side effects
- Awareness during sedation
- Duration and effectiveness of nerve blocks
- Recovery timelines and expectations
Interestingly, a significant number of questions also focused on technical aspects of anesthesia—suggesting that patients are seeking deeper understanding, not just basic reassurance.
Gaps in Patient Awareness
One surprising finding was that many patients are unaware of key aspects of anesthesia procedures. For example, some did not realize that they can remain awake during certain types of regional anesthesia, such as peripheral nerve blocks.
This highlights an important communication gap between clinicians and patients—one that AI-driven insights can help address.
Evaluating the Quality of Online Information
In addition to identifying common questions, the study also assessed the quality and sources of the information patients encounter online.
Distribution of Information Sources
The AI analysis categorized websites as follows:
- 55% academic or hospital-based sources
- 19% government websites
- 11% public or social media platforms
- Remaining sources included private medical practices and other sites
Accuracy and Reliability
Government and academic websites were found to provide the most accurate and reliable information. In contrast, private medical practice websites tended to score lower in terms of quality.
However, even when information was accurate, the source was not always clearly identifiable, which could affect patient trust and interpretation.
The Risk of Misinterpretation
Researchers emphasized that while online information can be helpful, it often lacks the nuance required for individual decision-making. For example, general comparisons between regional and general anesthesia may not apply to a specific patient’s condition, medical history, or surgical context.
This reinforces the importance of direct consultation with healthcare providers.
Improving Patient Education Through AI
The insights from this study have practical implications for how anesthesiologists communicate with patients.
Enhancing Clinical Conversations
Preoperative consultations are often brief, yet they require physicians to convey a large amount of information. Patients may forget to ask important questions or may not even know what to ask.
By understanding common concerns identified through AI analysis, clinicians can:
- Proactively address frequent questions
- Structure consultations more effectively
- Provide clearer explanations tailored to patient interests
Updating Educational Materials
The research team plans to use these findings to improve patient education resources by:
- Incorporating commonly asked questions
- Providing clearer explanations of procedures
- Offering materials in multiple languages
- Adjusting content for different literacy levels
Future Directions
The team also intends to continue using AI to evaluate how patients perceive and understand educational materials. Future research may explore:
- Differences in search behavior across languages and cultures
- How misinformation spreads online
- The effectiveness of targeted educational interventions
The Broader Impact of AI in Healthcare
Together, these studies demonstrate the expanding role of artificial intelligence in healthcare—not just as a clinical tool, but also as a means of understanding patient behavior and improving communication.
- AI can identify previously unknown biological markers associated with chronic pain
- Machine learning enables more accurate risk prediction for postoperative outcomes
- Analysis of search data reveals real-world patient concerns and knowledge gaps
- AI can help clinicians improve education and communication strategies
The research presented by HSS at the ASRA annual meeting underscores the transformative potential of AI in modern medicine. From predicting which patients are most likely to experience long-term pain after surgery to uncovering what patients truly want to know about anesthesia, these studies highlight how data-driven approaches can lead to more personalized, effective care.
As AI continues to evolve, its integration into clinical practice and patient engagement strategies is likely to deepen—offering new opportunities to improve outcomes, enhance understanding, and ultimately deliver more patient-centered healthcare.
About HSS
HSS is the world’s leading academic medical center focused on musculoskeletal health. At its core is Hospital for Special Surgery, nationally ranked No. 1 in orthopedics (for the 16th consecutive year), No. 3 in rheumatology by U.S. News & World Report (2024-2025), and the best pediatric orthopedic hospital in NY, NJ and CT by U.S. News & World Report “Best Children’s Hospitals” list (2025-2026). Founded in 1863, the Hospital has the lowest readmission rates in the nation for orthopedics, and among the lowest infection and complication rates. HSS was the first in New York State to receive Magnet Recognition for Excellence in Nursing Service from the American Nurses Credentialing Center five consecutive times. An affiliate of Weill Cornell Medical College, HSS has a main campus in New York City and facilities in New Jersey, Connecticut and in the Long Island and Westchester County regions of New York State, as well as in Florida. In addition to patient care, HSS leads the field in research, innovation and education. The HSS Research Institute comprises 20 laboratories and 300 staff members focused on leading the advancement of musculoskeletal health through prevention of degeneration, tissue repair and tissue regeneration. In addition, more than 200 HSS clinical investigators are working to improve patient outcomes through better ways to prevent, diagnose, and treat orthopedic, rheumatic and musculoskeletal diseases. The HSS Innovation Institute works to realize the potential of new drugs, therapeutics and devices.
