Table 3.

Included studies examining the effect of ultrasound education

First author, year	n	Study type	Aim	Education program	Main results	Rating
Short course
Chin, 2012	20 paramedics	Cohort	To determine if paramedics can acquire and interpret US for pneumothorax, pericardial effusion and cardiac activity	2-h session – 1 h lecture and 1 h hands-on session	After-test only: All subjects could identify the pleural line and 19/20 could obtain a cardiac view suitable for interpretation. Test score results were 9.1 out of a possible 10 (95% CI 8.6–9.6).	0
West, 2014	10 paramedics	Diagnostic accuracy	Not specified, but tested diagnostic accuracy for free fluid in abdominal trauma ultrasound	4 h course with lectures and hands-on training	Detecting of free fluid after course (peritoneal dialysis patients). Sensitivity 67%, specificity 56%. Higher false-positive rate than false-negative rate (59% vs 41%, p < 0.01)	0
Bhat, 2015	57 EMTs, paramedics and students	Controlled (before-and-after)	To assess the ability of EMS providers and students to accurately interpret heart and lung US images	1 h lecture on PTX, pericardial effusion and cardiac standstill	Theoretical test before and after: Test score 62.7% vs 91.1%. 95% CI for change 22–30%, p < 0.001). New post test in 19 subjects after one week: 93.1%.	+
Rooney, 2016	4 paramedics, 19 patients	Cohort	To determine if paramedics could perform cardiac ultrasound in the field and correctly identify cardiac activity/standstill	3 h course with 2 h theory and 1 h hands-on training	A total of 17/19 (89, 95% CI 67–99) exams were adequate for clinical decision-making. Correct identification of 17/17 cases of cardiac activity and 2/2 cases of cardiac standstill.	+
1- or 2-day course
Charron, 2015	100 exams	Diagnostic accuracy	To assess the ability of emergency physicians to obtain and interpret heart and inferior vena cava views using portable US	2-day course	Parasternal short axis, long axis and subcostal views were adequate in 44, 46 and 46%, respectively. Apical 4-chamber was adequate in 67%. Agreement with experts was weak for LVF, RV size and pericardial effusion and very weak for IVC.	+
Paddock, 2015	36 paramedics, nurses and physicians	Randomized controlled study	To compare the effectiveness of training using an ultrasound simulator to traditional trauma ultrasound training	Group A: Traditional training. Group B: US simulator training. Group C: Both	No difference between groups on neither image acquisition skills nor theoretical knowledge scores.	+
Booth, 2015	11 paramedics (4 long-term)	Controlled (before-and-after)	To determine if paramedics can be trained to perform and interpret US of the heart in cardiac arrest	1-day course with 2 h theory and 4 h hands-on training.	Theoretical test before and after: Improved theoretical knowledge (test score 54% before vs 89% after, p < 0.001). Practical test only after: 88% success in image acquisition during 10-min pulse-check window. Reduced to 75% (3/4) after 10 weeks.	–
Krogh, 2016	40 physicians	Controlled (before-and-after)	To evaluate the effect of e-learning and a hands-on US course of the lungs, heart, and abdomen	1-day course with 120 min e-learning + 4 h hands-on course	Improvement in theoretical knowledge after e-learning compared to before (51.3 (SD 5.9) vs 37.5 (SD 10.0), p < 0.001). Improvement in practical US performance and image interpretation after hands-on compared to before (p < 0.001).	+
Longer program
Press, 2013	33 paramedics and nurses	Controlled (before-and after)	To evaluate the effectiveness of a trauma US training curriculum and to determine if demographic factors predicted successful completion	1-day course with 2 h lectures, 4 h hands-on training + proctored session (4 exams) during 6 weeks + 60–120 min e-learning + unsupervised real-life exams	Theoretical test: none passed pre-test, 28/33 passed post-test with 78% score (p > 0.001 for difference). 27/33 passed structured clinical examination – only demographic factor predicting passing structured clinical exam was passing theoretical post-test.	+
Bobbia, 2015	14 physicians, 85 patients	Controlled (on experience-level)	To evaluate the interpretability of prehospital heart US based on physician experience	Experienced and non-experienced physicians defined by more or less than 50 exams after initial training (theory, 25 supervised exams)	Eight (57%) experienced physicians performed 51 (60%) exams and 6 (43%) novice physicians performed 34 (40%) exams. In multivariate analysis, only physicians experience was associated with the number of interpretable items (96% vs 56% for LVF, 98% vs 29% for PE, 92% vs 26% for RVD, and 67% vs 21% for IVC)	+
Botker, 2017	24 physicians	Controlled (before-and-after)	To evaluate the effect of a systematical education program in US of the heart and pleura on image acquisition skills, use and barriers	4 h e-learning + 1-day hands-on course + 10 supervised examinations + 3 months unsupervised exams	Proportion of images useful for interpretation increased from 0.70 (95% CI 0.65–0.75) to 0.98 (95% CI 0.95–0.99), p < 0.001. Used by 21/21 (100%) of prehospital providers after 4 years. Barriers for prehospital use comprised image quality in difficult patients and equipment	+

Abbreviations: US ultrasound, CI confidence interval, EMT emergency medical technician, EMS emergency medical services, PTX pneumothorax, M-mode motion mode, 2D-mode 2-dimensional mode, LVF left ventricular function, RV right ventricle, IVC inferior vena cava, SD standard deviation, PE pericardial effusion, RVD right ventricular dilation

Rating scale: ++ High quality, + Acceptable, − Low quality/unacceptable, 0 Rejected