Development of novel training for magnet-assisted capsule endoscopy (MACE) using cumulative sum analysis

Aims Magnet-assisted capsule endoscopy (MACE) is a novel, non-invasive alternative to upper gastrointestinal endoscopy for the diagnosis of gastric diseases. Operating the MACE system requires a different skillset from conventional endoscopy. However, dedicated training for MACE has not yet been developed. We established a training programme in which novices developed their MACE skills on ex-vivo models prior to examining human subjects. The learning curve was evaluated, and competencies were established using CUSUM analysis.

Methods Two ex-vivo models were developed and used to train participants in MACE. Ex-vivo exercise 1: the four points of a compass were drawn on a flat surface, each 7.5cm from the centre point. The novice was required to use the two joysticks and magnet robot to move the capsule from, and returning to, the centre to each point in sequence, passing through the centre each time within one minute. Ex-vivo exercise 2: novices were required to identify numbers (1-8), on the inside of a water-filled, opaque, plastic gastric model within four minutes. The learning curve of trainees was evaluated, and competency was defined using CUSUM analysis. CUSUM values were calculated from a mixture of increments which rise with each failed attempt and decrements which fall with each successful attempt. Competence was declared when the CUSUM curve crossed two consecutive boundary lines. Having become competent at these exercises, trainees performed examinations in human subjects. This was done in a standard fashion [1] in which they were required to save distant and close images of each gastric landmark.

Results Six trainees (median age 27 years, 83.3% male) completed the ex-vivo training. Half the trainees did not have previous endoscopy experience. The number of attempts to gain competency varied between trainees, ranging from 53 to 171 in exercise 1 and 44 to 74 in exercise 2. Overall, the trainees required a median of 70.5 attempts (IQR 53.7 – 116.3) over a median of 72 minutes (IQR 39.3 – 115) to achieve competency in exercise 1. Achieving competency in exercise 2 required fewer attempts, with a median of 47.5 attempts (IQR 41.2 – 64.2), but a longer time, with a median of 150.9 minutes (IQR 110.9 – 214.2). Data regarding examinations in human subjects were obtained from 22 trainees (median age 33 years, 9% male) who completed the same ex-vivo training programme, of whom 59% did not have previous endoscopy experience. All trainees successfully completed 8 – 12 examinations on human subjects with median time ranging between 28 – 35 minutes.

Conclusions All trainees who completed the ex-vivo training exercises demonstrated competence in examining human subjects. Competency in performing MACE can be achieved much more quickly than in conventional endoscopy. CUSUM analysis could be used to guide progress in training and clinical practice as part of a quality assurance programme.

Publication History

Article published online:
15 April 2024

• References

• 1 Ching HL. et al. "Diagnostic yield of magnetically assisted capsule endoscopy versus gastroscopy in recurrent and refractory iron deficiency anemia.". Endoscopy 2019; 51 (05) 409-418