B. T. Fry2, L. W. Hess3, M. Jain1, J. T. Anger1, R. Avenido1, B. Gewertz1, K. Catchpole1 1Cedars-Sinai Medical Center,Los Angeles, CA, USA 2University Of Michigan Medical School,Ann Arbor, MI, USA 3Pennsylvania State University,Eberly College Of Science,University Park, PA, USA
Introduction:
Robotic surgery offers advantages over conventional operative approaches but may also be associated with higher costs, additional risks, and new challenges. Surgical flow disruptions (FD) are defined as ‘deviations from the natural progression of an operation,’ and have been empirically associated with surgical errors, adverse events, and inefficiency. Understanding the etiology of FD in robotic surgery will help target training techniques and identify opportunities for improvement. This study explored the relationships between surgeon console time (SCT), the number and types of FD, resident involvement, and other contextual parameters.
Methods:
Thirty-two robotic surgery operations were observed over a six-week period at one 900-bed surgical center. Ten cases prior to this sampling were used to train two researchers and ensure high inter-rater reliability.
The researchers observed FD throughout the time the patient was in the operating room. Each FD was designated with the time and a descriptor, and was then classified into one of 11 different categories: communication, coordination, external interruptions, training, equipment, environment, patient factors, surgical decision making, instrument changes, psychomotor error, and robot console switch. SCT, resident involvement, robot model, and procedure type were also recorded.
Multi-variable statistics were used to evaluate the effects of these parameters on SCT and the number of FD.
Results:
Eight sacrocolpopexies, 21 prostatectomies, and 3 nephrectomies were observed. The mean number of FD was 48.2 (95% CI 38.6-54.8), and mean SCT was 163mins (95% CI 148-179). The Da Vinci S robot model was used in 14 cases, and the Si model was used in 18 cases. Nineteen cases involved residents, and 13 did not.
There was a mean of 60.8 FD (95% CI 47.8-73.8) in resident cases and 29.8 FD (95% CI 22.1-37.5) in non-resident cases. Resident cases demonstrated mostly training, equipment, and robot switch FD, whereas non-resident cases demonstrated mostly equipment, instrument change, and external interruption FD. A linear regression (r2=0.34) demonstrated that residents had a significant effect on number of FD (p<0.002), whereas robot model and procedure type demonstrated a non-significant effect.
The mean SCT with residents was 165.8mins (95% 149.7-181.9) and without residents was 160.2mins (95% CI 130.1-190.3). A linear regression model (r2=0.35) found resident involvement and robot model to be non-significant parameters, while procedure type (p<0.001) and total FD (p<0.034) significantly affected SCT.
Conclusion:
Resident involvement significantly increased the number of FD but did not affect SCT. This suggests that the FD encountered in resident training may not significantly affect operating time. Other FD, such as equipment issues or external interruptions, may be more impactful. Limiting these specific FD should be the focus of performance improvement efforts.