TissueDB/SELF/VesselBox Vessel Ligation Assessment

VesselBox Vessel Ligation Assessment is the published assessment and adaptive-practice protocol for the VesselBox Vessel Ligation Trainer (Hu). It sets out how a trainee's ligations are scored, how practice length is decided by the trainee's own learning curve rather than by a fixed number of repetitions, and how the resulting skill was tested for transfer.

The build page (VesselBox Vessel Ligation Trainer (Hu)) contains construction and per-session setup only. This page captures how to ASSESS a trainee on the finished simulator.

Procedure focus: Open vessel ligation — gaining proximal and distal control of a vessel, dividing it, and tying both ends with 3-0 silk.^[1]

Target learners: Graduating final-year medical students matching into general surgery or a surgical specialty, and incoming surgical interns. The published cohort was 16 fourth-year medical students at the University of Virginia, taking part in a two-week surgery internship preparedness course.^[1]

Prerequisites: Every participant was given an instructional video demonstrating proper vessel ligation technique on VesselBox, and had to be able to perform a one-handed knot tie before enrolling.^[1]

Two instruments are used together, plus a timing measure. Both were adopted because neither alone is sufficient: the global scale captures how the trainee operates, the checklist captures whether the specific technical steps were done, and neither accounts for speed.^[1]

An eight-item binary checklist created by investigator consensus, capturing the technical components critical to a correct vessel ligation. Each item is scored done or not done, and the metric is the number of items properly performed, to a maximum of 8.^[1]

1. Tie placed directly under clamp (< 2 mm)
2. Passer tip brought under clamp tip
3. Knots pushed down with 1st finger
4. Knots pushed down beyond clamp tip/knot
5. Requests clamp release while holding tension on first knot throw
6. Square knots (alternate hand-crossing)
7. Knots secure (does not slip with light pull)
8. Proper suture length (not impeded by excessively long/short tail)

Pass mark for a single attempt: 7 or higher out of 8. Anything below 7 counts as a failed attempt.^[1]

The Objective Structured Assessment of Technical Skills global rating scale for hemostasis. It scores seven proficiency descriptors, each from 1 to 5; the seven are averaged into a composite score to a maximum of 5.^[1]

1. Respect for tissue
2. Time and motion
3. Instrument handling
4. Knowledge of instruments
5. Use of assistants
6. Flow of operation
7. Knowledge of procedure

Because neither instrument accounts for speed, the time required to control and ligate each vessel is recorded separately.^[1]

Assistant. A surgical assistant is integral to the task and performs only those actions the trainee directly requests. In the published protocol a single postgraduate year-3 surgery resident acted as assistant for every participant, so that the assistant was not a source of variation between trainees. The task therefore assesses operative technique, instrument knowledge and communication together.^[1]

Blinding. Pre-test performance was captured on video with the camera field limited to the simulator and the participant's forearms, so raters could not identify the trainee. Three surgical faculty, blinded to participant identity, scored each recording on both the global rating scale and the checklist.^[1]

Pre-test. Each participant ligated two vessels on VesselBox before any practice, to establish a baseline.^[1]

Scoring across raters. For each participant, checklist and composite global-rating scores were averaged across the three evaluators.^[1]

Cumulative sum (Cusum) is a sequential analysis tool that tracks learning progression in real time. A Cusum learning curve deflects downward on each successful attempt and upward on each failed one; crossing the decision interval with a positive deflection indicates sub-proficient performance. The protocol below follows Bolsin and Colson's description of Cusum for procedural training, as adopted by Hu et al.^[1]

These were assigned in advance, by investigator consensus, to approximate the expected proficiency of a mid-level resident.^[1]

Parameter	Value
Acceptable failure rate	5%
Unacceptable failure rate	20%
Type I error rate (wrongly judging a proficient trainee sub-proficient)	10%
Decision interval (h0), calculated from the three values above	1.41

1. Deliver up to two demonstrations of correct vessel ligation technique, then hand over to the trainee for repeated practice.^[1]
2. Have an assistant deliver and retract instruments on the trainee's direction for every practice attempt, as in the assessed task.^[1]
3. Give feedback after each attempt, and score that attempt on the eight-item checklist.^[1]
4. Record the attempt as a success if it scored 7 or 8, and as a failure below that, then recompute the Cusum after every attempt.^[1]
5. Apply the Cusum over a rolling window of the most recent 8 attempts, updating after each one. This is a simplified derivation of the exponentially weighted moving-average method of building memory-loss into a Cusum function.^[1]
6. End the session when the trainee has completed eight consecutive attempts without crossing a decision interval. Worked example from the source: if a participant has completed 13 attempts, a Cusum value of 0 is assigned at attempt 6; if no decision interval is crossed from attempt 6 through 13, that participant has demonstrated Cusum proficiency at attempt 13.^[1]

What to expect. Practice sessions averaged 21.8 minutes (range 16 to 60 minutes) and about 10 practice attempts. Individual learning rates varied widely: participants needed anywhere from 8 to 16 attempts to satisfy the proficiency criteria. Attempts to proficiency did not correlate with pre-test global rating score or pre-test checklist score, so a trainee's starting score does not predict how much practice they will need.^[1]

Once a trainee met the Cusum proficiency criteria, practice stopped and post-testing was performed on a high-fidelity fresh human cadaver model, using the same technique practised on VesselBox.^[1]

• Target vessel: a medium-sized vessel — short gastric artery, mesenteric artery branch, or extremity venous branch.^[1]
• Substitute when no suitable vessel is available: friable strands of breast tissue used as surrogate vessels.^[1]
• Rating: each post-test was evaluated in person by the same three surgical faculty who scored the pre-tests, blinded to the participants' pre-test results and practice-session data.^[1]

Sixteen students enrolled and all completed the protocol. Baseline performance averaged 2.29 on the global rating scale (IQR 2.01–2.50) and 4.83 on the checklist (IQR 4.17–6.33).^[1]

Measure	Pre-test	Post-test	p
OSATS global rating scale (max 5)	2.29	3.23	< 0.001
Task-specific checklist (max 8)	4.83	7.33	< 0.001
Median time per vessel ligated	128.2 s	97.5 s	0.001

Practice volume was not associated with post-test performance: there was no significant correlation between the number of practice attempts and post-test global rating score, post-test checklist score, or the size of the improvement in either. This is the finding that justifies the adaptive design — stopping at Cusum proficiency rather than at a fixed repetition count did not cost the trainees who needed fewer attempts.^[1]

1. Set up the trainer and mount a fresh glove finger as described on the build page, which also gives how many attempts each mounted finger allows before it is replaced.
2. Brief the trainee, confirm they can perform a one-handed knot tie, and show the instructional demonstration.
3. Record a baseline: two ligations, filmed with the camera framed on the simulator and the trainee's forearms only.
4. Score the baseline on both the checklist and the global rating scale, ideally with more than one rater blinded to identity.
5. Run the adaptive practice session using the Cusum protocol above, scoring and giving feedback after every attempt.
6. Stop when eight consecutive attempts pass without crossing the decision interval, and record how many attempts it took.
7. If a cadaver or other high-fidelity model is available, run the transfer test and score it with the same instruments.

• The published cohort was 16 graduating medical students at a single institution, so the parameter values were tuned to that population's expected proficiency; a different learner group may warrant different acceptable and unacceptable failure rates.^[1]
• The protocol was run with a purpose-built automated Cusum calculator; the source describes its input form and output but does not distribute the tool, so a site adopting this protocol needs its own means of computing the running Cusum.^[1]
• The transfer test requires a fresh human cadaver, which many sites cannot obtain. The source gives no validated lower-fidelity substitute for the post-test.^[1]
• The eight-item checklist was created by investigator consensus for this construct, not adapted from an externally validated instrument.^[1]

• VesselBox Vessel Ligation Trainer — built as described on its build page
• Disposable latex surgical glove fingers as the vessel substitute — replacement interval on the build page
• 3-0 silk suture ties, for both cut ends of each divided vessel
• The per-station instrument tray listed on the build page
• A surgical assistant for every attempt, practice and assessed alike
• Video camera and a stand allowing a field limited to the simulator and the trainee's forearms
• A stopwatch or the camera's timecode, for the per-vessel timing measure
• A means of computing a running Cusum over the most recent 8 attempts
• Printed or on-screen copies of the eight-item checklist and the seven-descriptor global rating scale

For construction, materials, costs and per-session reset, see: VesselBox Vessel Ligation Trainer (Hu)

The build page contains:

• Materials list and costs
• Step-by-step construction of the four-sided pine box and the vessel mount
• Glove mounting and per-session reset
• Known limitations of the substitute mounting design