Is gallbladder PoCUS diagnostic accuracy accessible to medical students after PoCUS training exclusively on healthy volunteers? A pilot randomized control trial

Study design and ethics

This prospective pilot study was based on a double-blind randomized controlled model approved by the institutional review board (Comité d’éthique Hospitalo-Facultaire Saint Luc-UCL) of a tertiary university hospital in Belgium. The study did not involve any patients. The approval of the ethics committee was therefore consultative, and no registration number was given.

This paper was written according to the CONSORT guidelines. The study protocol was previously registered under clinical trials.gov (NCT04879459).

Study timeline

The study timeline (Fig. 1) depicts the interaction between the study population and the study members.

Fig. 1figure 1

Study timeline. IG: intervention Group. CG: control group

Study population

Fifth-year medical students enrolled at the Université Catholique de Louvain in Brussels, Belgium, were invited by email to participate in this PoCUS study. The first 20 eligible students who responded were selected. Students were eligible if they were PoCUS naïve and available on the four chosen study days between October 12 and December 10, 2021 (Fig. 1). Prior to inclusion, students signed an informed consent and a non-disclosure agreement written in French. According to the non-disclosure agreement, participants could not disclose the study to any third party for the entire study period; any breach would lead to their exclusion. Subsequently, the 20 students were randomly assigned to the two predefined groups of 10 students using a computed randomization (Randomizer for Clinical Trial Lite, Medsharing©, Fontenay-Sous-Bois, France).

The 10 CG students received gallbladder PoCUS training in accordance with the current recommendations [10, 18, 19] as well as a publication about learning curves [20]. They learned PoCUS on a mixture of volunteers with healthy and pathological (gallstones/polyps) gallbladders. The 10 IG students underwent gallbladder training exclusively on volunteers presenting a normal gallbladder. Students were blind to group allocation and study outcomes for the entire study period.

Study membersTeaching faculty

Five instructors participated in the three teaching days. All instructors were emergency physicians certified in PoCUS use for gallbladders. Two instructors had validated an interuniversity diploma in PoCUS at University Paris V–Descartes (France), one had a Master’s degree in medical ultrasound from Teesside University (UK), one had a certificate in medical ultrasound from Teesside University (UK), and one had an advanced POCUS certificate from Université Libre de Bruxelles (Belgium). Both groups were taught by the same instructors who were not blind to student allocation. All instructors signed a non-disclosure agreement.

Volunteers

A total of 43 volunteers standardized patients (scanning models) were invited to participate in the study. Healthy volunteers were recruited by email, while volunteers with gallstones were recruited during surgical consultation for elective cholecystectomy or when diagnosed with gallstones during gallbladder ultrasound sessions organized for healthy volunteers before the study. Volunteers were eligible if they were aged over 17 years, had no persistent abdominal pain, and signed the informed consent form and non-disclosure agreement. All volunteers underwent a comprehensive gallbladder ultrasound in radiology at most 1 month prior to the start of the study to determine their gallbladder status (normal or gallstones and/or sludge). Volunteers, 17 male (40%) and 26 female (60%), were paired for body mass index (BMI) (+/−1) and age (+/− 5 years) between groups. Their characteristics are described in Table 1. All volunteers fasted for at least 3 h prior to the gallbladder scan. They were not aware of the study objectives or methodology. A total of 38 (88%) volunteers were allocated to the IG or CG for the teaching sessions depending on their gallbladder status and general characteristics. Among the 38 volunteers, 7 (18%) had a pathological gallbladder (gallstones or polyps) and participated exclusively in the CG teaching days. Five other volunteers participated in the evaluation day.

Table 1 Characteristics of volunteersAssessment faculty

Three examiners from different specialties (surgeon, radiologist, and emergency physician) certified in gallbladder ultrasound use and blind to students’ allocation participated in the evaluation. They were independent of the study and did not work in the university conducting the research; they did not contribute to the study protocol or participate in the theoretical and practical PoCUS teaching sessions. Two Belgian examiners spoke French, while the third examiner was from the UK and spoke English.

Study settingTeaching days

Teaching was organized on separate but consecutive days for the IG and CG to prevent interaction between groups and thus respect the non-disclosure agreement. On teaching day 1, students from both groups attended a 1-h theoretical session that included a web-based video in French (https://youtu.be/7ZqP2mKNOQg) about gallbladder PoCUS use with images and loops of normal and pathological (gallstones or sludge/polyps) gallbladders. The theoretical session focused on ultrasound machine settings (curvilinear probe, abdominal preset, depth, and gain), gallbladder recognition using POCUS, description of normal gallbladders, identification of gallstones, difference between gallstones and polyps, and description of POCUS pitfalls (duodenum, collapsed gallbladder). Students also participated in three 3-h practical teaching sessions (teaching days 1, 2, and 3) using a bedside Vscan Air™ (GE VINGMED Ultrasound AS, Horten, Norway) portable ultrasound probe connected to an iPad. Gallbladders were assessed using an abdominal setting with the curvilinear probe. The practical teaching days were distributed according to an expanding retrieval practice for effective learning, as supported by a recent review showing that expanding intervals between teaching sessions, as opposed to regular intervals, enhance long-term retention [21]. All students performed a total of 45 gallbladder PoCUS, with 15 being completed in each 3-h teaching session repeated across 3 teaching days. In the IG, all 45 (100%) of the gallbladder PoCUS performed by students were normal. In the CG, 24 (53%) of the gallbladder PoCUS were normal and 21 (47%) were pathological. Among the 21 pathological PoCUS, 18 (86%) showed gallstones and 3 (14%) polyps. The rotation of students between instructors was organized to ensure that all students underwent the same PoCUS training. During the 3-h session, students had access to one-on-one teaching at the bedside of volunteers. Every 10 min, the students changed places to perform the gallbladder PoCUS on another volunteer supervised by another instructor. During each 3-h session, the students performed a gallbladder PoCUS on five different volunteers and rotated three times, thus corresponding to a total of 15 PoCUS performed by each student. For the practical training, instructors used a teaching chart with relevant PoCUS information that was given to students at each practical session (Additional file 1—translated from French). At the end of the third teaching day, students from both groups were randomly assigned for the 30-min evaluation.

Evaluation day

Sensitivity and specificity for gallstone diagnosis by students using PoCUS was assessed on five volunteers: two had gallstones, two had a normal gallbladder, and one had polyps. Every 30 min, in front of the three examiners, a student was scheduled to perform five gallbladder PoCUS on the five volunteers set up in different rooms. Gallbladder ultrasound images are shown in Fig. 2 following the sequence of evaluations. During the entire evaluation, students were asked not to interact with the examiners. They could nonetheless speak with the volunteers and ask them to change position or take a deep breath. Students completed a form with their diagnosis (gallstones and/or sludge) and then provided a final diagnosis (normal/gallstones/polyps) after performing the PoCUS. Students’ confidence about their diagnosis and machine use was evaluated using a five-point Likert scale. The examiners who were aware of the volunteers’ gallbladder status were asked to fill in a form assessing image quality and quality of machine use. Examiners classified the PoCUS images obtained by students as interpretable or uninterpretable and assessed their machine use as satisfactory or unsatisfactory in terms of their ability to handle the probe with one hand and to properly manage gain and depth with the other. The students were unable to share their results until the end of the evaluation. The examiners were blind to the students’ diagnosis and confidence about their diagnosis and machine use.

Fig. 2figure 2

Gallbladder ultrasound images from the evaluation day. WES: wall-echo-shadow

Outcomes

The main outcome of this study was the sensitivity and specificity of gallstone diagnosis by students using PoCUS. Sensitivity and specificity were explored in each group. As secondary outcomes, students’ confidence in their diagnosis and machine use was also reported, as were the examiners’ evaluation of the image quality and PoCUS use. The acquisition time was also evaluated in the IG and CG.

Data management

Corresponding groups and names were kept in a secure database. Except for the study coordinator, no other investigator had access to the data. A number was randomly assigned to each participant and a letter to each group allocation to ensure blind analysis.

Statistical analysis

Statistical analysis was performed with SPSS 26.0® and JMP pro16.0.0®. The primary outcome was evaluated using sensitivity and specificity estimates with a 95% confidence interval (CI) and the area under the curve (AUC) calculated based on receiver operating characteristic (ROC) curve analysis for the IG and CG. Regarding the secondary outcomes, a Mann–Whitney–Wilcoxon test was used to assess between-group differences in students’ confidence about their diagnosis and machine use according to a Likert scale. A Mann–Whitney–Wilcoxon test was also used to assess the between-group difference in the time taken for image acquisition. A nonparametric test was chosen due to the relatively small sample sizes and the potential skewness of the distribution of acquisition times. A proportion test was used to represent the examiner-assessed image quality and machine use of both groups. A p-value of < 5% corrected for multiple tests was considered significant.

As this is a pilot study, sample size calculation could not be performed due to the lack of previous evidence. We included 10 participants in each group due to the expected low attrition rate and the study feasibility [22].

留言 (0)

沒有登入
gif