Predictive modelling of vascular surgery trends using machine learning: a comparative study of Irish public and private tertiary referral centres

TL;DR

This study uses machine learning to predict future trends in vascular surgery in Ireland, highlighting a potential shortage of surgeons and the need for strategic planning.

Contribution

The study introduces a predictive model using AI/ML to forecast vascular surgery workforce needs in both public and private sectors.

Findings

The private sector saw a 73-fold increase in abdominal aortic aneurysm interventions compared to a 1.25-fold increase in the public sector.

The model predicts a shortage of vascular surgeons, with the workforce potentially meeting demand by 2050.

Surgeons may need to increase yearly wRVU production by 22%–31% by 2030 to accommodate workload.

Abstract

Vascular diseases are increasing in Ireland as well as worldwide alongside an ageing society, posing a growing demand for trained and qualified healthcare professionals. In this study, we have analysed current practices of vascular interventions by using the data from the vascular tertiary centre to predict the future size and capacity of the vascular surgery workforce through artificial intelligence (AI)-powered predictive models.

We employed supervised machine learning (ML) regression model to predict trends in the landscape of complex vascular and endovascular surgery over the next 22 years, utilising data from a high-volume public and private tertiary referral vascular centre spanning two decades (2002 to 2023) in the West of Ireland.

We conducted 1,653 aortic interventions, 1,185 carotid interventions, and 3,069 peripheral vascular interventions, with conversion rates from referral to surgery of 5%, 7.4%, and 9.8%, respectively. The private sector experienced a dramatic 73-fold increase in abdominal aortic aneurysm interventions, contrasted with a modest 1.25-fold increase in the public sector. Our model predicts a shortage of vascular surgeons, with the workforce potentially meeting demand by 2050. By 2030, each surgeon would need to increase yearly wRVU production by 22%–31% and by 2040 by 8%–11% to accommodate the workload.

Our model predicts a shortage of the vascular surgery workforce over the next two decades. We can speculate that addressing future needs in vascular surgery requires either training more specialists or increasing the efficiency and wRUV through strategic planning and integration of AI/ML systems to ensure adequate compensation and the sustainability of the workforce. By focusing on these areas, we can navigate the evolving landscape of vascular surgery and continue providing high-quality patient care.

Figures