TY - GEN
T1 - Evidence-based education in robot-assisted surgery
AU - Hertz, Peter
PY - 2023/11/20
Y1 - 2023/11/20
N2 - Robotic-assisted surgery has progressed enormously since the first operations onpatients more than twenty years ago. More than 66,000 surgeons have been trained inrobotic-assisted surgery. The current training varies in quality and often relies on thetraining offered by the industry. There is a great variation in the introduction and use ofthe robotic system among the intra-abdominal specialities (urology, gastrointestinalsurgery and gynaecology). For gastrointestinal surgery, robot-assisted surgery to resectthe right colon due to cancer has become more common in recent years. A vital part ofthe procedure is the removal of the lymph nodes associated with the affected part of thecolon by dissection in the fetal planes. This technique requires anatomicalunderstanding, skills in fine dissection, and the ability to work within limited space.This technique is called complete mesocolic excision (CME).This thesis aims to close some of the gaps in the literature on basic robot-assistedtraining and develop new ways to achieve competencies in the advanced procedure ofrobot-assisted CME.Study IWe sought to identify common learning goals for basic skills for the three specialities(urology, gastrointestinal surgery and gynaecology) to help guide evidence-based crossspeciality education of robot-assisted surgery.Method: We invited every department in urology, gastrointestinal surgery andgynaecology in Denmark that performs robot-assisted surgery to participate in a threeround Delphi survey with two surgeons from each. We asked them to list relevantcontent for a cross-speciality robot-assisted surgery course, and after the third round, theDelphi panel was asked who should learn the specific learning goals and how it istaught best, e.g. e-learning, virtual reality, and team training.Results: fifty-six robot surgeons participated, and we identified 40 learning goals thatreached a consensus on the degree of relevance to include in a curriculum. And for eachidentified learning goal, we ranked the best way to train/learn.Conclusions: We identified the content for a cross-speciality robotic curriculum andinvestigated the surgeons' preferred ways to train and learn each identified learninggoal.Study IIWe conducted a systematic review to identify and evaluate current surgical training andcompetency assessment tools. We included all studies on the anatomical colorectal areawhen using laparoscopic, open, or robot-assisted approaches.Method: The Preferred Reporting Items for Systematic Reviews and Meta-Analysesguidelines were followed, and two authors individually screened and did dataextraction. The identified studies were assessed for quality using the Medical EducationResearch Study Quality Instrument.Results: Overall, the findings were of varying quality. Fifty-one studies were identified,and we identified many laparoscopic assessment tools but very few for open and robotassisted surgery. In general, the consequences of the assessments were not discussed,nor were relevant pass/fail standards established.Conclusions: We have established an overview of current training and assessment toolsand evaluated the validity of the evidence. Based on that, we have suggested areas ofinterest for further research.Study IIIaSimulation-based training has become one of the cornerstones of basic surgical training.It is very well implemented in laparoscopic surgery through inanimate, virtual reality,and real tissue models (animal or cadavers). The same modalities are available forrobotic surgery but need further research. Robotic-assisted surgery on tissue is difficultand expensive because of the need for a dedicated training robotic system. Developing amodel for advanced procedure training that could be used in a patient-dedicated roboticsystem would enable easy access and lower training expenses. The aim of this studywas to build a phantom for procedure-specific training of complete mesocolic excisionthat could be used in a patient-dedicated robotic system.Method: The items identified as relevant when assessing the surgical competency ofcomplete mesocolic excision were translated into engineering tasks. Each task wasdesigned in regard to assessment and realism. The phantom was developed with the aidof patient scans and 3D printing, and experienced surgeons evaluated the developedphantom.Results: We developed a model from patient scans and 3D printing for completemesocolic excision. It is possible to assess 35 out of the 48 items from the usedassessment tool.Conclusions: The model can be used in the robotic system to practice completemesocolic. In order to evaluate the training value, validity data is needed for reliableassessments.Study IIIbWhen assessing surgical competencies, an assessment tool ensures relevant and equalassessment of the trainees.To ensure that the tool measures what it is supposed to measure validity data should bereported with the tool. A validity framework suggested Messick is considered themodern framework and is the most used among medical educators. Performance dataare needed to evaluate the validity of the assessment. The aim of the study is to gathervalidity data on the inanimate model developed in Study IIIa.Methods: Surgeons from across Europe were included in three groups based on roboticcomplete mesocolic excision experience (novice = 0 procedures, intermediate = 1-29procedures, and experienced = >29 procedures). Each participant performed the CMEprocedure on one model as a warm-up and, afterwards, two procedures that were videorecorded. Four trained and blinded raters were used: Two complete mesocolic excisionexperienced surgeons for video rating, and two experienced pathologists evaluated theresected specimens. The raters used published assessment tools for robotic, surgicalCME and pathology assessment of CME specimens.Results: videos and specimens from 33 surgeons were evaluated, and surgeons fromfour different countries in Europe participated. The internal consistency of the surgicalassessment tools was high, whereas the tool used for pathology assessment wasinconsistent with a spearman-brown Coefficient of 0.15 and 0.54 for rater one and two,respectively. The test-retest reliability varied between the raters, and only the raters withmoderate/strong test-retest reliability were able to discriminate between the pre-definedexperience levels. The inter-rater reliability was low/moderate.Conclusion: we did find interesting correlations, but the inconsistency in the ratingreliability excludes the possibility of summative assessment. More data are needed toestablish pass/fail standards for implementation in mastery learning programs.PerspectivesThis thesis has explored different aspects of training in robot-assisted surgery. Thefindings from Study I can be used to develop and guide the training of basic skills andcompetencies in multiple specialities. Study II identified a gap in assessment andtraining tools in robotic-assisted colorectal surgery. We tried to improve on this bydeveloping an advanced phantom for CME based on patient scans. More data areneeded for reliable assessments of CME skills in the simulated setting.We suggest that future training curricula for robot-assisted surgery should focus on boththe patient-side assistant and console surgeon. Training in basic skills could be initiatedas cross-speciality training. For procedural training, our model development methodfrom patient scans enables the possibility to create patient-specific models, which couldpotentially also be used for patient-specific “warm-up surgery” and assessment ofsurgical skills in a patient-risk-free environment.
AB - Robotic-assisted surgery has progressed enormously since the first operations onpatients more than twenty years ago. More than 66,000 surgeons have been trained inrobotic-assisted surgery. The current training varies in quality and often relies on thetraining offered by the industry. There is a great variation in the introduction and use ofthe robotic system among the intra-abdominal specialities (urology, gastrointestinalsurgery and gynaecology). For gastrointestinal surgery, robot-assisted surgery to resectthe right colon due to cancer has become more common in recent years. A vital part ofthe procedure is the removal of the lymph nodes associated with the affected part of thecolon by dissection in the fetal planes. This technique requires anatomicalunderstanding, skills in fine dissection, and the ability to work within limited space.This technique is called complete mesocolic excision (CME).This thesis aims to close some of the gaps in the literature on basic robot-assistedtraining and develop new ways to achieve competencies in the advanced procedure ofrobot-assisted CME.Study IWe sought to identify common learning goals for basic skills for the three specialities(urology, gastrointestinal surgery and gynaecology) to help guide evidence-based crossspeciality education of robot-assisted surgery.Method: We invited every department in urology, gastrointestinal surgery andgynaecology in Denmark that performs robot-assisted surgery to participate in a threeround Delphi survey with two surgeons from each. We asked them to list relevantcontent for a cross-speciality robot-assisted surgery course, and after the third round, theDelphi panel was asked who should learn the specific learning goals and how it istaught best, e.g. e-learning, virtual reality, and team training.Results: fifty-six robot surgeons participated, and we identified 40 learning goals thatreached a consensus on the degree of relevance to include in a curriculum. And for eachidentified learning goal, we ranked the best way to train/learn.Conclusions: We identified the content for a cross-speciality robotic curriculum andinvestigated the surgeons' preferred ways to train and learn each identified learninggoal.Study IIWe conducted a systematic review to identify and evaluate current surgical training andcompetency assessment tools. We included all studies on the anatomical colorectal areawhen using laparoscopic, open, or robot-assisted approaches.Method: The Preferred Reporting Items for Systematic Reviews and Meta-Analysesguidelines were followed, and two authors individually screened and did dataextraction. The identified studies were assessed for quality using the Medical EducationResearch Study Quality Instrument.Results: Overall, the findings were of varying quality. Fifty-one studies were identified,and we identified many laparoscopic assessment tools but very few for open and robotassisted surgery. In general, the consequences of the assessments were not discussed,nor were relevant pass/fail standards established.Conclusions: We have established an overview of current training and assessment toolsand evaluated the validity of the evidence. Based on that, we have suggested areas ofinterest for further research.Study IIIaSimulation-based training has become one of the cornerstones of basic surgical training.It is very well implemented in laparoscopic surgery through inanimate, virtual reality,and real tissue models (animal or cadavers). The same modalities are available forrobotic surgery but need further research. Robotic-assisted surgery on tissue is difficultand expensive because of the need for a dedicated training robotic system. Developing amodel for advanced procedure training that could be used in a patient-dedicated roboticsystem would enable easy access and lower training expenses. The aim of this studywas to build a phantom for procedure-specific training of complete mesocolic excisionthat could be used in a patient-dedicated robotic system.Method: The items identified as relevant when assessing the surgical competency ofcomplete mesocolic excision were translated into engineering tasks. Each task wasdesigned in regard to assessment and realism. The phantom was developed with the aidof patient scans and 3D printing, and experienced surgeons evaluated the developedphantom.Results: We developed a model from patient scans and 3D printing for completemesocolic excision. It is possible to assess 35 out of the 48 items from the usedassessment tool.Conclusions: The model can be used in the robotic system to practice completemesocolic. In order to evaluate the training value, validity data is needed for reliableassessments.Study IIIbWhen assessing surgical competencies, an assessment tool ensures relevant and equalassessment of the trainees.To ensure that the tool measures what it is supposed to measure validity data should bereported with the tool. A validity framework suggested Messick is considered themodern framework and is the most used among medical educators. Performance dataare needed to evaluate the validity of the assessment. The aim of the study is to gathervalidity data on the inanimate model developed in Study IIIa.Methods: Surgeons from across Europe were included in three groups based on roboticcomplete mesocolic excision experience (novice = 0 procedures, intermediate = 1-29procedures, and experienced = >29 procedures). Each participant performed the CMEprocedure on one model as a warm-up and, afterwards, two procedures that were videorecorded. Four trained and blinded raters were used: Two complete mesocolic excisionexperienced surgeons for video rating, and two experienced pathologists evaluated theresected specimens. The raters used published assessment tools for robotic, surgicalCME and pathology assessment of CME specimens.Results: videos and specimens from 33 surgeons were evaluated, and surgeons fromfour different countries in Europe participated. The internal consistency of the surgicalassessment tools was high, whereas the tool used for pathology assessment wasinconsistent with a spearman-brown Coefficient of 0.15 and 0.54 for rater one and two,respectively. The test-retest reliability varied between the raters, and only the raters withmoderate/strong test-retest reliability were able to discriminate between the pre-definedexperience levels. The inter-rater reliability was low/moderate.Conclusion: we did find interesting correlations, but the inconsistency in the ratingreliability excludes the possibility of summative assessment. More data are needed toestablish pass/fail standards for implementation in mastery learning programs.PerspectivesThis thesis has explored different aspects of training in robot-assisted surgery. Thefindings from Study I can be used to develop and guide the training of basic skills andcompetencies in multiple specialities. Study II identified a gap in assessment andtraining tools in robotic-assisted colorectal surgery. We tried to improve on this bydeveloping an advanced phantom for CME based on patient scans. More data areneeded for reliable assessments of CME skills in the simulated setting.We suggest that future training curricula for robot-assisted surgery should focus on boththe patient-side assistant and console surgeon. Training in basic skills could be initiatedas cross-speciality training. For procedural training, our model development methodfrom patient scans enables the possibility to create patient-specific models, which couldpotentially also be used for patient-specific “warm-up surgery” and assessment ofsurgical skills in a patient-risk-free environment.
U2 - 10.21996/xpvg-js43
DO - 10.21996/xpvg-js43
M3 - Ph.D. thesis
PB - Syddansk Universitet. Det Sundhedsvidenskabelige Fakultet
ER -