Validation of a Low-Fidelity Model for Office-Based Hysteroscopy with a Vaginoscopic Approach

Abstract

<h3>Study Objective</h3> Hysteroscopy is an established method for the diagnosis and treatment of intrauterine pathology. A vaginoscopic approach for office-based hysteroscopy confers less pain; however, trainees report lack of confidence with this procedure. We sought to create a low-fidelity simulation model for office-based hysteroscopy with a vaginoscopic approach and to evaluate the validity and reliability of this model. <h3>Design</h3> Prospective cohort study. <h3>Setting</h3> A single site academic medical center. <h3>Patients or Participants</h3> Eligible participants included obstetrics and gynecology residents and attendings who regularly perform hysteroscopy. <h3>Interventions</h3> The vaginoscopy model was created with a female pelvis simulator and an exam glove placed within the vagina. Following two instructional videos, participants performed a hysteroscopy simulation with a vaginoscopic approach. The primary outcome was total score on a modified Global Rating Scale (GRS) and Objective Structured Assessment of Technical Skills (OSATS). The OSATS describes key procedural steps and was created with experts in hysteroscopy for content validity. Time to complete each task was recorded and summed. A post-procedure survey assessed the model and physician experience. <h3>Measurements and Main Results</h3> Ten attending and 20 resident physicians (9 junior and 11 senior) participated. Attending physicians completed the simulation significantly faster than junior residents (197(±31) vs. 290(±107) seconds, Cohen's <i>d</i>=1.2, p=0.022). On the GRS, both attending physicians (26.1(±2.4), <i>d</i>=2.5, p=0.001) and senior residents (22.5(±3.7), <i>d</i>=1.3, p=0.010) scored significantly higher than junior residents (17.4(±4.3)). Ninety-three percent of surveyed participants were satisfied with simulation, 97% found it useful, 80% found it realistic and 93% indicated that they may use this technique in the future. <h3>Conclusion</h3> This study shows our low-fidelity model to have construct, content, and face validity for vaginoscopic approach to hysteroscopy. A sufficiently powered, yet small sample of physician performance data yielded a large effect size in global mean score differences between attendings and junior residents. Physicians reported the technique to be satisfactory, realistic, and useful for future application.