Background
Physical examination and auscultation can be challenging for medical students. The aim of this study was to investigate whether a brief session of group training in focused cardiac ultrasound (FCU) with a pocket-sized device would allow medical students to improve their ability to detect clinically relevant cardiac lesions at the bedside.
Methods
Twenty-one medical students in their clinical curriculum completed 4 hours of FCU training in groups. The students examined patients referred for echocardiography with emphasis on auscultation, followed by FCU. Findings from physical examination and FCU were compared with those from standard echocardiography performed and analyzed by cardiologists.
Results
In total, 72 patients were included in the study, and 110 examinations were performed. With a stethoscope, sensitivity to detect clinically relevant (moderate or greater) valvular disease was 29% for mitral regurgitation, 33% for aortic regurgitation, and 67% for aortic stenosis. FCU improved sensitivity to detect mitral regurgitation (69%, P < .001). However, sensitivity to detect aortic regurgitation (43%) and aortic stenosis (70%) did not improve significantly. Specificity was ≥89% for all valvular diagnoses by both methods. For nonvalvular diagnoses, FCU’s sensitivity to detect moderate or greater left ventricular dysfunction (90%) was excellent, detection of right ventricular dysfunction (79%) was good, while detection of dilated left atrium (53%), dilated right atrium (49%), pericardial effusion (40%), and dilated aortic root (25%) was less accurate. Specificity varied from 57% to 94%.
Conclusions
After brief group training in FCU, medical students could detect mitral regurgitation significantly better compared with physical examination, whereas detection of aortic regurgitation and aortic stenosis did not improve. Left ventricular dysfunction was detected with high sensitivity. More extensive training is advised.
Highlights
- •
We trained 21 medical students in FCU using a group model.
- •
We examined whether the students improved at bedside diagnosis after 4 hours of training.
- •
LV dysfunction and MR were identified with high sensitivity.
- •
Detection of aortic valve pathology did not improve compared with physical examination.
- •
The group model limited individual scanning time. More extensive training is advised.
Although patient history and physical examination, including auscultation, remain the basis of the initial assessment of a cardiac patient, the limited diagnostic accuracy of the stethoscope is well known. In recent decades, the emerging field of portable ultrasound has challenged the use of the stethoscope. The recent arrival of pocket-sized devices facilitates true bedside routine use and has created a new paradigm for the use of ultrasound. The concept of focused cardiac ultrasound (FCU) has been introduced. Despite their small size and limited features, pocket-sized scanners have proven diagnostic value when used as an adjunct to physical examination by experienced echocardiographers. Low cost and simplified operation have opened their potential use to nontraditional cardiac ultrasound users, and a growing body of evidence suggests that inexperienced operators also improve bedside diagnosis with such scanners. However, the amount of training required to reach a given and standardized level of accomplishment is still a matter of debate. The prospect of educating and training all physicians represents an enormous challenge, and the demand for cost-effective training programs may gain in importance.
Few studies have thus far evaluated the use of the pocket-sized FCU devices in a large group of medical students. Our aim with the present study was to investigate whether a brief, group-based FCU training course would allow medical students to improve their ability to detect clinically relevant cardiac lesions at the bedside.
Methods
Study Population
In this prospective study, 21 medical students from the University of Oslo, all in their second half of medical school and without prior echocardiographic experience, were randomly recruited from 104 applicants to complete a standardized 4-hour FCU training program. A total of 72 elective patients already referred for routine echocardiographic examinations were included in the study, which was conducted in the Department of Cardiology, Oslo University Hospital, Rikshospitalet (Oslo, Norway), between February and May 2012. All patients who were available during the days of inclusion were asked to participate. Exclusion criteria included practical and medical considerations, such as lack of consent, shortage of time between scheduled procedures, and postprocedural or hemodynamically unstable patients. Written informed consent was obtained from all participants, including patients and students. The study was approved by the Regional Committee for Medical Research Ethics and conducted according to the second Declaration of Helsinki.
Echocardiographic Training
Before attending the training course, the students were encouraged to study a selection of echocardiographic loops provided online, demonstrating normal cardiac anatomy and common pathologies. The precourse material also featured a compendium describing the cardiac views in ultrasound and instructions on how to position the transducer to obtain the different views. The course for six trainees at a time (pilot group of three) consisted of a 45-minute introduction to cardiac ultrasound with a review of the same echocardiographic loops that were provided online. The loops demonstrated (which were recorded with an FCU device) are summarized in Table 1 and included both normal cardiac anatomy and common pathologies. In addition, the FCU examination protocol was demonstrated in practice, with emphasis on scanning technique to obtain the cardiac views and the evaluation criteria for each parameter ( Table 2 ). After the initial session, the students were given 60 minutes to practice on one another, to familiarize themselves with the device and to practice obtaining all images according to the protocol. Each student examined two other students. Furthermore, the students had 75 minutes of practice on patients in the cardiology ward. These patients were selected for cardiac pathology. Two students examined each patient together, and each student pair examined two different patients. This was followed by 60 minutes of case reviews, in which the recorded images from the ward patients were discussed and compared with a standard echocardiogram. The students thus were challenged in image interpretation. Training in electrocardiographic interpretation and auscultation was not involved, as these skills had been taught in the clinical curriculum.
| Cardiac pathology | Imaging mode | Cardiac views | Severity levels demonstrated |
|---|---|---|---|
| LV dysfunction | Grayscale | PLAX, A4C | Normal (EF, 60%), moderately impaired (EF, 45%), and severely impaired (EF, 30%) |
| Pericardial effusion | Grayscale | PLAX, A4C | Mild, moderate, and large pericardial effusion |
| MR | Color Doppler | PLAX, A4C | None, mild, moderate, and severe regurgitation |
| AR | Color Doppler | PLAX, A5C | None, mild, moderate, and severe regurgitation |
| AS | Grayscale and color Doppler | PLAX, A5C | None, mild aortic sclerosis, moderate stenosis, and severe stenosis |
| TR | Color Doppler | PLAX, A4C | None, mild, moderate, and severe regurgitation |
| RV dysfunction | Grayscale | PLAX, A4C | Normal, moderately impaired, and severely impaired |
| Left atrial dilation | Grayscale | PLAX, A4C | None, dilated |
| Right atrial dilation | Grayscale | PLAX, A4C | None, dilated |
| Aortic root dilation | Grayscale | PLAX, A4C | None, dilated |
| Landmark | Evaluation method | |
|---|---|---|
| FCU | Standard echocardiography | |
| LV systolic function ∗ | Visual assessment by observing endocardial motion, overall LV cavity size, and mitral valve excursion, reported as normal, moderately impaired, or severely impaired | LV EF by volume calculations using the biplane method of disks (modified Simpson’s rule); normal (≥55%), moderately impaired (54%–36%), severely impaired (≤35%) |
| Pericardial effusion ∗ | By caliper in end-diastole, reported as none, moderate (≥0.5 cm), or severe (≥1.0 cm) | |
| MR ∗ AR ∗ | Visual assessment of the color flow regurgitant jet, reported as none, mild, moderate, or severe | Valve morphology, color flow regurgitant jet, vena cava width, PISA method, CW regurgitant profile, pulmonary vein flow (MR), peak E velocity (MR), pressure half-time (AR), diastolic flow reversal in descending aorta (AR) |
| AS ∗ | Visual assessment of the color flow jet and presence of calcified aortic ring or aortic cusps with reduced opening, reported as none, mild, moderate, or severe | CW Doppler measurements of maximal velocity and mean pressure gradient, including the aortic valve area estimated by the continuity equation |
| TR ∗ | As described for MR | |
| RV systolic function ∗ | Visual assessment, reported as normal, moderately impaired, or severely impaired | RV fractional area change, considered normal (≥32%), moderately impaired (31%–18%), or severely impaired (≤17%) |
| Left atrial dilation Right atrial dilation | Visual assessment, reported as normal or dilated (if greater than half the size of the LV longitudinal dimension in A4C) | Atrial area; normal (<20 cm 2 ) or dilated (≥20 cm 2 ) |
| Aortic root dilation | By caliper, reported as normal or dilated (if ≥3.5 cm at the sinuses of Valsalva) | |
Study Protocol
The students were blinded to medical and drug history, prior echocardiograms, and other interventions. They completed a sequential assessment of each patient, consisting of a brief medical history, a physical examination, and finally an FCU examination. History was limited to exploring the presenting symptoms and New York Heart Association class. Physical examination included inspection (detection of peripheral pitting edema and jugular venous distension), cardiac and lung auscultation, and interpretation of the provided electrocardiogram. Cardiac murmurs were reported in terms of location (parasternal or apical), intensity (I–VI), and timing (systolic or diastolic). Furthermore, the students were asked to suggest if their auscultatory findings represented mitral regurgitation (MR), aortic regurgitation (AR), or aortic stenosis (AS) and to classify the lesion as either mild, moderate, or severe. The cardiac landmarks assessed by FCU are shown in Table 2 . The report sheets for physical examination are in Appendix 1 and 2 (available at www.onlinejase.com ) and FCU examination in Appendix 3 (available at www.onlinejase.com ). Findings from the students’ examinations were compared with a standard echocardiogram as the reference method.
Echocardiographic Equipment and Methods
The FCU examination was performed with the Vscan (GE Vingmed Ultrasound AS, Horten, Norway). The device is handheld, fits in the pocket, and consists of a display unit (135 × 73 × 28 mm) and a broad-bandwidth phased-array probe (120 × 33 × 26 mm; frequency, 1.7–3.8 MHz). Other specifications include a 3.5-inch flip-up display (resolution, 240 × 320 pixels), a total weight of 390 g, and approximately 60 minutes of scanning time. The scanner provides grayscale two-dimensional imaging and color Doppler imaging, automatic gain adjustment, and automatic detection of a full heart cycle for storage without the need for electrocardiography. Basic measurements can be performed using the provided caliper tool.
Standard echocardiography was performed in the hospital’s echocardiography laboratory by experienced cardiologists, with the high-end Vivid E9 or Vivid 7 scanner (GE Vingmed Ultrasound AS). The investigators were blinded to the result of the FCU examinations. Data were digitally stored for offline analysis using dedicated software (EchoPAC; GE Vingmed Ultrasound AS). The evaluation criteria for the two echocardiographic methods are shown in Table 2 .
Data Analysis
As suggested by the Standards for Reporting of Diagnostic Accuracy statement, diagnostic accuracy was calculated in terms of sensitivity, specificity, positive and negative predictive value, and κ values.
A cardiac lesion was defined as clinically relevant when it was moderate or greater in severity, which routinely leads to an additional evaluation by standard echocardiography. In addition to presenting accuracy data with the cutoff for clinically relevant lesions at moderate or greater, results from valvular diagnoses are also presented with the cutoff at mild or greater pathology to show the total number of detected lesions.
To compare accuracy among the students, we developed a diagnostic scoring system modified from Decara et al . and Panoulas et al . For each true-positive finding of a significant cardiac lesion, two points were given. For each true-positive finding of a mild lesion, one point was given. For each true-negative or normal finding, 0.5 points were given. For each false-negative finding, zero points were given. For each false-positive finding, 0.5 points were deducted. In cases of underestimation of a lesion (student reports mild when truly moderate or severe), 0.5 points were still given. In cases of overestimation (student reports moderate or severe when truly mild), 0.5 points were still given. To calculate the diagnostic score, the total score obtained by the student was divided by the maximum score possible.
Quality Assessment
For the purpose of assessing feasibility and image quality, two experienced echocardiographers retrospectively reviewed all recorded FCU images and standard echocardiograms in a blinded fashion and categorized each examination as good, fair, or poor. Note that examinations already deemed inadequate by the students were omitted ( n = 2). To be classified as good, all four chambers along with the mitral, aortic, and tricuspid valves had to be visualized from both apical four-chamber and parasternal long-axis views. For valvular assessment, loops of all three valves both with and without color Doppler had to be available. For the grading of fair, a prerequisite was satisfactory imaging of the entire left ventricle and left-sided heart valves, even if only apical or parasternal views were recorded. In case of inadequate or blurred image quality of both or only one view, the examination was classified as poor. A criterion for including poor quality recordings was that the left ventricle could be visualized for functional assessment.
In addition, two experienced echocardiographers interpreted the FCU images recorded by students to separately assess the students’ interpreting abilities. The accuracy of expert interpretations was compared with student accuracy.
Statistical Analysis
Numeric data are presented as mean ± SD or, as appropriate, median (interquartile range) and categorical data as numbers and percentages. Sensitivity, specificity, and positive and negative predictive value were calculated using binary variables and are presented as percentages with 95% confidence intervals. The following statistical analyses were performed using SPSS version 20 (IBM, Armonk, NY): χ 2 tests were used to assess differences in categorical variables. McNemar’ tests were used to compare the sensitivities of the two diagnostic methods. Two-tailed P values < .05 were considered significant. Cohen κ coefficients were used to measure interrater agreement for categorical variables. Kappa values < 0.2 were interpreted as indicating slight, 0.21 to 0.4 fair, 0.41 to 0.6 moderate, 0.61 to 0.8 good, and 0.81 to 1.00 very good agreement.
Results
In 72 patients, 110 FCU examinations were conducted. In 25 of the patients, repeat evaluations were performed by two to five students. The mean age of the patient population was 65 ± 16 years, with 52 (72%) men. The mean time required for physical examination was 7 ± 2 minutes. Cardiac murmurs were present in 63 patients (57%), as reported by the students. Of these, 50 (79%) were systolic, eight (13%) were diastolic, and five (8%) were combined systolic and diastolic murmurs. The mean FCU examination duration was 17 ± 6 minutes. The median time period between standard echocardiography and FCU was 0 days (interquartile range, 0–4 days).
Diagnostic Accuracy
Figure 1 presents the number of patients who were diagnosed with any form of clinically relevant valvular lesion by the different screening methods. Significantly more cases were detected by standard echocardiography than by stethoscope and FCU ( P < .001 and P = .004, respectively), but there was no difference between stethoscope and FCU ( P = .274).
Table 3 presents diagnostic accuracy for the stethoscope and FCU examinations. The overall pooled sensitivity to detect clinically relevant valvular disease with a stethoscope was 40%, which improved to 64% after FCU examination ( P < .001). With a stethoscope, sensitivity was 29% to identify clinically relevant (moderate or greater) MR, 33% for AR, and 67% for AS. Adding the FCU examination greatly improved sensitivity to detect MR (69%, P < .001). However, sensitivity to discover AR (43%) and AS (70%) did not improve significantly. Sensitivity to detect tricuspid regurgitation was 33% but was assessed only with FCU. Specificity was ≥89% for all valvular diagnoses by both screening methods.
| Screening method | Cardiac pathology | SE Pos (total [ n ]) | SM Pos ( n ) | Sensitivity (%) (95% CI) | Specificity (%) (95% CI) | PPV (%) (95% CI) | NPV (%) (95% CI) | κ |
|---|---|---|---|---|---|---|---|---|
| Stethoscope | At least moderate MR | 42 (104) | 18 | 29 (16–45) | 90 (79–96) | 67 (41–86) | 65 (54–75) | 0.21 |
| At least mild MR | 89 (104) | 21 | 22 (15–33) | 93 (66–100) | 95 (74–100) | 17 (10–27) | 0.06 | |
| At least moderate AR | 21 (93) | 7 | 33 (15–57) | 100 (94–100) | 100 (56–100) | 84 (74–90) | 0.44 | |
| At least mild AR | 57 (93) | 12 | 19 (10–32) | 97 (84–100) | 92 (60–100) | 43 (32–55) | 0.13 | |
| At least moderate AS | 27 (86) | 19 | 67 (46–83) | 98 (90–100) | 95 (72–100) | 87 (76–93) | 0.71 | |
| At least mild AS | 29 (86) | 25 | 72 (52–87) | 93 (82–98) | 84 (63–95) | 87 (75–95) | 0.68 | |
| Overall ∗ | 89 (283) | 44 | 40 (30–37) | 96 (92–98) | 82 (67–91) | 78 (72–83) | 0.42 | |
| FCU | At least moderate MR | 42 (104) | 36 | 69 (53–82) | 89 (78–95) | 81 (63–91) | 81 (69–89) | 0.59 |
| At least mild MR | 89 (104) | 81 | 90 (81–95) | 93 (66–100) | 99 (92–100) | 61 (39–80) | 0.68 | |
| At least moderate AR | 21 (93) | 16 | 43 (23–66) | 90 (80–96) | 56 (31–79) | 84 (74–91) | 0.36 | |
| At least mild AR | 57 (93) | 49 | 77 (64–87) | 86 (70–95) | 90 (77–96) | 70 (55–83) | 0.61 | |
| At least moderate AS | 27 (86) | 21 | 70 (50–86) | 97 (87–99) | 90 (68–98) | 88 (77–94) | 0.71 | |
| At least mild AS | 29 (86) | 35 | 97 (80–100) | 88 (76–95) | 80 (63–91) | 98 (88–100) | 0.80 | |
| At least moderate TR | 12 (57) | 6 | 33 (11–65) | 96 (84–99) | 67 (24–94) | 84 (71–93) | 0.35 | |
| At least mild TR | 40 (57) | 28 | 70 (56–81) | 100 (77–100) | 100 (85–100) | 59 (39–76) | 0.58 | |
| LV dysfunction ∗ | 61 (105) | 74 | 90 (79–96) | 57 (41–71) | 74 (63–83) | 81 (62–92) | 0.49 | |
| RV dysfunction ∗ | 19 (85) | 36 | 79 (54–93) | 69 (55–78) | 42 (26–59) | 92 (80–97) | 0.36 | |
| Pericardial effusion ∗ | 15 (101) | 12 | 40 (17–67) | 93 (85–97) | 50 (22–78) | 90 (81–95) | 0.36 | |
| Left atrial dilation | 85 (101) | 46 | 53 (42–64) | 94 (68–100) | 98 (87–100) | 27 (17–41) | 0.24 | |
| Right atrial dilation | 47 (89) | 35 | 49 (34–63) | 71 (55–84) | 66 (48–80) | 56 (41–69) | 0.20 | |
| Aortic root dilation | 36 (77) | 14 | 25 (13–43) | 88 (73–95) | 64 (36–86) | 57 (44–69) | 0.13 | |
| Overall ∗ | 364 (903) | 296 | 59 (54–64) | 85 (81–88) | 72 (67–78) | 75 (72–79) | 0.45 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
