Background: We sought to determine whether the video self Instruction kit (CPR anytime
®
) could be a useful tool for the mass CPR training, and to evaluate its effects on the CPR quality compared to the traditional method of learning.
Methods: The CPR training courses were designed to 3 different types by the learning method: 45 munites self learning with CPR anytime
®
(group A), 1 hour lecture + 45 munites self learning with CPR anytime
®
(group B), 1 hour lecture + 1 hour practice with instructors (group C). From July 2007 to August 2007, 405 (34%) participants out of the 1,178 applied trainees were seleted by stratified sampling according to their sex, age, the date of learning, and the learning method. The trainees with any respiratory, skeletal disease or incomplete informed consents were excluded. At the end of the courses, the skill test was done by the checklist and their skill performance was tested using the Laerdal PC SkillReporting System
®
. Statistical analysis was performed using One Way ANOVA in continuous variables and
x
2
analysis in categorical variables.
Results: Two hundred twenty subjects were enrolled properly with 79(35.9%), 66(30.0%), 75(34.1%) belonging to group A, B, and C, respectively. The mean age was 36.3 years and females accounted for 77.3%. We found no significant differences in their willingness to participate, sex, weight, height, and educational status. The skill test indicated no differences among the three groups. In the skill performance, there were no significant differences between group A and B. When compared to group C, group A showed a higher rate of forceful ventilation (58.5 ±38.4% vs 41.3 ±37.9%,
p
=0.03), a larger amount of mean tidal volume (950.0±416.5 mL vs 745.6±343.8 mL,
p
=0.00), a lower rate of shallow compression (47.1±40.3% vs 65.4±40.6%,
p
=0.01), and a deeper depth of compression (38.6±9.6 mm vs 34.1±9.8 mm,
p
=0.01).
Conclusion: The trainees of self leraning with CPR anytime
®
showed similar results of the skill test with the trainees of other two longer leraning method, and presented relatively adequate chest compression with the tendency of frequent hyperventilation. On the condition of adding a solution to prevent hyperventilation, the CPR anytime
®
can be a helpful tool for the mass CPR education.
An Adversarial Self-Learning Method for Cross-City Adaptation in Semantic Segmentation
