Page 40 - Read Online

Page 276 Yoseph et al. Art Int Surg 2024;4:267-77 https://dx.doi.org/10.20517/ais.2024.38

While our study provides valuable insights, it also raises several important research questions that warrant
further exploration. Future studies could investigate how LLMs perform across various medical specialties
and how they manage more complex or sensitive patient inquiries. Understanding the impact of LLM-
generated responses on patient decision making is another exciting area for future research. There is also
great potential in determining how a combined model that integrates LLMs with physician oversight
changes the surgical decision-making process, potentially in terms of increasing or decreasing the number
of patients who opt for surgical intervention, and/or improving their comfort level and understanding of the
risks/benefits.

One final exciting application of AI in surgery is its significant role in enhancing the surgical consent
process in spinal surgery. Recent studies have demonstrated that AI can effectively simplify complex
medical information, improving readability and comprehension for patients. For example, Ali et al. showed
that ChatGPT-4.0 could generate procedure-specific consent forms at an average 6th-grade reading level,
[23]
significantly enhancing patient understanding without sacrificing important medical details . This AI-
human expert collaborative approach not only improves patient education but also addresses medico-legal
concerns by ensuring that consent forms meet both medical and legal standards. Given the litigious nature
of spinal surgery, it is critical for future work to address the medico-legal implications of incorporating AI
into this field. LLMs can still produce errors or “hallucinations”, making it essential to implement strict
[24]
validation processes to ensure that only the most accurate information is conveyed to patients . This is
particularly important as AI-generated content becomes more integrated into surgical decision making,
where the stakes are highest.

Notably, ChatGPT provided the following statement after answering the last question: “Always consult with
your spine surgeon for personalized advice and to address specific concerns regarding your ACDF surgery”.
As LLMs evolve and enhance their precision, clarity, and comprehensiveness, it is important for physicians
and medical researchers to evaluate and study the best way to incorporate these tools into the routine care
of our patients moving forward.

DECLARATIONS
Authors’ contributions
Made substantial contributions to the conception and design of the study: Yoseph ET, Lang S, Zygourakis
CC
Performed data analysis and interpretation: Yoseph ET, Gonzalez-Suarez AD
Supported data acquisition: Yoseph ET, Zygourakis CC, Hu SS, Desai A
Wrote the manuscript: Yoseph ET, Gonzalez-Suarez AD, Zygourakis CC
Critically revised and approved final manuscript: Yoseph ET, Gonzalez-Suarez AD, Lang S, Desai A, Hu SS,
Zygourakis CC

Availability of data and materials
Data supporting the findings of this study are available in Supplementary Tables.

Financial support and sponsorship
None.

Conflicts of interest
Zygourakis CC is a consultant for Stryker and Amgen; Desai A is a consultant for Stryker and Carlsmed.
While the other authors have declared that they have no conflicts of interest.