Generic Error Rating Tool (GERT)
Husslein et al.24 were the first to test a
non-procedure-specific error assessment tool. Called GERT, the tool,
which uses a Likert scale with nine anchors, is designed to analyse
technical errors and resulting events during laparoscopy. GERT is based
on the inverse relationship between surgeon and skill, i.e. more skilled
surgeons make fewer errors.
Technical errors are defined as “the failure of planned actions to
achieve their desired goal” and an event as “an action that may
require additional measures to avoid an adverse outcome”.24 The GERT technical error analysis comprises nine
generic surgical tasks during which errors can occur. Each of these
generic task groups is subdivided into four distinct error modes: 1) too
much use of force or distance, 2) too little use of force or distance,
3) inadequate visualisation and 4) wrong orientation of instrument. To
assess error distribution within different operative sub-steps, the
procedures are divided into insertion of trocars, creation of bladder
flap, colpotomy and vault closure.
Two blinded reviewers analysed twenty video recordings of total
laparoscopic hysterectomies, and correlation analyses were performed
between GERT and OSATS. Scores from the latter were used to establish a
measure of technical skills and to divide surgeons into two groups as
either high or low performers. The results showed a significant negative
correlation between OSATS and GERT scores (rater 1: Spearman = -0.76,
p<0.001; rater 2 = -0.88, p< 0.001). Group
comparison showed that high performers made significantly fewer
technical errors than low performers.
Interrater reliability was high (CCI>0.95) for the total
number of errors and events. Within the nine anchors (task groups) ICC
was >0.8 in all groups except for cutting, transection and
stapling; clipping; and use of suction. Intrarater reliability was high
(ICC>0.95) for total number of errors and events. By
analysing the different operative sub-steps, the study was able to
detect procedures more prone to technical errors, e.g. vault closure.