By Brandon Wade
Numerous AI are being developed across healthcare to address clinical, financial, and operational unmet needs. The excitement around AI’s potential impact was a prominent theme at the Digital Orthopedics Conference in San Francisco (DOCSF) this past week. Nearly half of the presentations covered emerging AI technologies and use cases across the orthopedics space.
With all the buzz around AI for healthcare, why are we so excited about its applications in orthopedics?
- Clinical Unmet Needs Persist: Musculoskeletal care can be the most traumatic and impactful event(s) in a patient’s life with real risk of further care needed. Despite decades of advancements in implant design, procedural technique, and advent of novel tech (e.g., navigation, robotics), there is still work to be done to improve outcomes and patient experience.
- Top 3 Financial Burden: Musculoskeletal continues to rank in the top 3 in overall healthcare expenditures in the US, only outpaced by cardiovascular and ill-defined conditions. Recent measures to control costs (e.g., bundled payments through CJR) have largely failed to move the needle as much as hoped. Two-year outcomes from CMS’ CJR bundled payment model (run across nearly all hospitals in 67 MSAs in the US) resulted in a total cost savings of $17MM…a relative drop in the bucket.
- Operational Nightmare: Orthopedics is a business of equipment. The sheer volume of implants, instruments, surgical materials, etc. that are needed to successfully execute thousands of procedures per day is staggering. Ensuring the right materials are in the right places all while trying to fill ORs to maximum throughout is a logistical nightmare for OEMs, hospitals, OR admins, and HCPs.
The obvious areas for AI in MSK: The obvious applications for AI, and the setting of most development activity, is encompassing the acute episode of care as patients prepare for, undergo, and recover from surgery. These AI are likely to have positive impact, but will they be revolutionary? Not necessarily.
AI will absolutely help drive improvements in pre-surgical planning, make surgical navigation and robotics more accurate, guide intra-operative decision making, and help drive better adherence to rehab. However, as we saw in the CJR model, there may not be “revolutionary” juice left to squeeze around the acute episode of care. Surgeons, armed with the latest computer-assisted technologies (e.g., nav, robotics), can consistently perform at extremely high levels and ensure very solid outcomes. Will AI provide more benefits to less-skilled surgeons who only perform a few cases a year? Of course. But we are still skeptical of a true step-change in outcomes specifically driven by AI. We are excited about and applaud the significant AI development around the acute episode but must recognize that not all AI is created equal. A growing list of competitors will compete fiercely for what marginal room there is to drive improvements. Some notable developments happening in the space include:
- Practice Care Coordination (e.g., Ospitek, Memora Health, DocSpera)
- Intra-Operative Decision-Making (e.g., OrthoGrid)
- Intra-Operative Instrument Tracking (nSight)
- Post-Operative Care (e.g., Kemtai, Sword Health, Kaia, ViFive)
So where will AI truly be disruptive? Amara’s Law tells us that in the near-term…not much will be revolutionary. But, over time, technology is likely to outpace our expectations and deliver truly disruptive value. The two areas we are eager to keep close watch on are 1) AI working well upstream in the patient journey and 2) AI utilized directly by OEMs to address operational needs.
- AI Upstream in the Patient Journey: Where AI could truly be disruptive is upstream in the patient journey, well before a patient progresses in their MSK disease and needs an orthopedic surgeon. In this use case, AI will play a huge role as an augmentation to or replacement of primary care. AI tools will ingest large amounts of data to proactively identify at-risk patients and implement a series of care paradigms to slow the progression of MSK disease and delay, or even prevent entirely, the need for procedures. AI is already showing it can do this in other industries, like oncology, where AI tools have been trained to spot cancer risk years before eventual diagnosis from EMR data alone. Should AI do something similar in MSK, it will represent an existential threat to many of the stakeholders delivering orthopedic care as we know it. And while we believe every stakeholder would agree that this is the best course of action in the long-term, great change will be needed (e.g., MSK practices getting into primary and secondary prevention, consolidation of surgeons as procedure volumes decline). It’s a long way off, but with continued investment in product development, willingness to conduct pilots by at-risk stakeholders (e.g., Medicare, integrated delivery networks), and robust evidence-generation by AI companies the industry can get there.
- AI for OEMS: Orthopedics, like many other surgical specialties, requires a significant volume of equipment (instrumentation, implants, general surgical materials, etc.). Today, OEMs in the space spend significant time, energy, and money to ensure the right tools are in the right places for surgery, not to mention significant capital needed to produce enough equipment for surgeons across the US and globe. Hospitals and ASCs as well must contend every day with the logistical nightmare of coordinating scheduling of HCPs, patients, and sterilized equipment. While AI will absolutely improve facility coordination and logistics, the changes for OEMs could be revolutionary. Healthcare logistics has always lagged other equipment/material-intensive industries (e.g., automotive, big-box stores, grocery). AI has the potential to streamline orthopedics such that the right equipment is always ready in the right places at the right time with very little waste. Further, with decreasing complexity of instrument and implant trays needed for surgery and enhanced, AI-enabled support for surgeons in the OR, the OEMs in the space may finally be able to do away with the rep-intensive model. We are talking about eliminating significant drivers of cost in the orthopedics industry. The only question is whether these cost savings will trickle down to hospitals, payers, and ultimately, patients.
Very interested to hear what forward thinking leaders in the AI space have to say about where AI will deliver value in the world of orthopedics!
 “Spend on Disease Treatment.” Kaiser Family Foundation, 2019, https://www.healthsystemtracker.org/indicator/spending/spending-disease-treatment/
 “CJR Model Performance Year 2 (PY2) Evaluation Results.” CMS, 2019, https://www.hhs.gov/guidance/sites/default/files/hhs-guidance-documents/webinar%20cjr%20model%20performance%20year%202%20%28py2%29%20evaluation%20results%2009%2025%2019.pdf
 Placido, Davide, et al. “A Deep Learning Algorithm to Predict Risk of Pancreatic Cancer from Disease Trajectories.” Nature Medicine, 8 May 2023, http://www.nature.com/articles/s41591-023-02332-5.
Brandon Wade is a Vice President at Health Advances and works in the intersections between the Health IT and Digital Health, Biopharma, and MedTech practices with a specific focus on cross-sector healthtech strategy. Brandon co-leads the Health IT and Digital Health practice along with the Musculoskeletal practice.