Do surgical skills affect outcomes?

Artificial intelligence meets medical robotics

Artificial intelligence (AI) applications in medical robots are bringing a new era to medicine. Advanced medical robots can perform diagnostic and surgical procedures, aid rehabilitation, and provide symbiotic prosthetics to replace limbs. The technology used in these devices, including computer vision, medical image analysis, haptics, navigation, precise manipulation, and machine learning (ML) , could allow autonomous robots to carry out diagnostic imaging, remote surgery, surgical subtasks, or even entire surgical procedures. Moreover, AI in rehabilitation devices and advanced prosthetics can provide individualized support, as well as improved functionality and mobility (see the figure). The combination of extraordinary advances in robotics, medicine, materials science, and computing could bring safer, more efficient, and more widely available patient care in the future. -Gemma K. Alderton.

OUP accepted manuscript

Background

This study aimed to evaluate the use of binary metric-based (proficiency-based progression; PBP) performance assessments and global evaluative assessment of robotic skills (GEARS) of a robotic-assisted low anterior rectal resection (RA-LAR) procedure.

Method

A prospective study of video analysis of RA-LAR procedures was carried out using the PBP metrics with binary parameters previously developed, and GEARS. Recordings were collected from five novice surgeons (≤30 RA-LAR previously performed) and seven experienced surgeons (>30 RA-LAR previously performed). Two consultant colorectal surgeons were trained to be assessors in the use of PBP binary parameters to evaluate the procedure phases, surgical steps, errors, and critical errors in male and female patients and GEARS scores. Novice and experienced surgeons were categorized and assessed using PBP metrics and GEARS; mean scores obtained were compared for statistical purpose. Also, the inter-rater reliability (IRR) of these assessment tools was evaluated.

Results

Twenty unedited recordings of RA-LAR procedures were blindly assessed. Overall, using PBP metric-based assessment, a subgroup of experienced surgeons made more errors (20 versus 16, P = 0.158) and critical errors (9.2 versus 7.8, P = 0.417) than the novice group, although not significantly. However, during the critical phase of RA-LAR, experienced surgeons made significantly fewer errors than the novice group (95% CI of the difference, Lower = 0.104 – Upper = 5.155, df = 11.9, t = 2.23, p = 0.042), and a similar pattern was observed for critical errors. The PBP metric and GEARS assessment tools distinguished between the objectively assessed performance of experienced and novice colorectal surgeons performing RA-LAR (total error scores with PBP metrics, P = 0.019–0.008; GEARS scores, P = 0.029–0.025). GEARS demonstrated poor IRR (mean IRR 0.49) and weaker discrimination between groups (15–41 per cent difference). PBP binary metrics demonstrated good IRR (mean 0.94) and robust discrimination particularly for total error scores (58–64 per cent).

Conclusions

PBP binary metrics seem to be useful for metric-based training for surgeons learning RA-LAR procedures.