ICECAP Tool Developed to Identify Errors in Vascular Procedures

March 1, 2013—A pilot study that sought to design a valid, reproducible tool for observers and teams to identify and categorize errors in the vascular and endovascular theater was published by S.L. Mason, MD, et al in the European Journal of Vascular and Endovascular Surgery (2013;45:248–254). The investigators concluded that the ICECAP (Imperial College Error Capture) tool may be useful for this purpose and may also have a role as an error-recall prompt for self-reporting purposes by vascular surgical teams.

In the background commentary on the study, the investigators stated that accurate evaluation is crucial to improve quality in the unique and complex vascular and endovascular theater environment, which is associated with significant risks of patient harm and procedural inefficiency.

In the study, the investigators analyzed relevant published literature and previously collected ethnographic field notes from over 250 hours of arterial surgery. A comprehensive log of vascular procedural errors was compiled, and 12 vascular experts graded each error for the potential to disrupt procedural flow and cause harm. Using this multimodal approach, the ICECAP tool was developed.

The investigators stated that during 21 consecutive arterial cases (52 hours of operating time), the ICECAP tool was validated as an observer-led error capture record and as a prompt for surgical teams to determine the feasibility of error self-reporting.

As summarized in the European Journal of Vascular and Endovascular Surgery, six primary categories (communication, equipment, procedure-independent pressures, technical, safety awareness, and patient-related) and 20 error subcategories were determined as the most frequent and important vascular procedural errors. Using the ICECAP, the number of errors detected correlated well between two observers (Spearman rho = .984; P < .001). Both observers identified all moderate or severe errors similarly and categorized all but 4/139 (2.9%) of the total errors in an identical fashion. Self-reporting of errors without prompting identified a mean of 24.4% (range, 0–50%) of all recorded errors, whereas surgical teams reported a mean of 69.7% (range, 50–100%) of errors when ICECAP error-category prompts were used.