Development and assessment of an eLearning course on senology for radiographers: a stratified randomised controlled trial

Breast cancer is a serious public health problem, and screening is a critical tool to combat the mortality rate of the disease. Mammography technique is considered the best imaging examination for this screening and the radiographer plays a crucial role in this process. For this, the radiographer must be properly prepared and should therefore be subject to specific training and continuing education. Therefore, there should be training programmes, and eLearning has been revealed a useful tool and an alternative to the traditional education method. The objectives of this work include the knowledge promotion and understanding of all aspects related to the breast illness and patient care required to radiographers. To achieve this, we proposed to develop an easy-touse course in eLearning environment and assess about its efficacy, effectiveness and the users’ satisfaction, according to evaluation levels 1 and 2 from Kirkpatrick’s framework. A randomised controlled trial was performed with radiographers and radiology students, using as instruments individual satisfaction questionnaires and pre and post-knowledge tests for further comparisons, respectively. Intention-to-treat and per-protocol analysis were considered in the statistical analysis. As result, a total of 54 participants were included in the intervention group (20 students and 34 radiographers) and 53 in the control group (19 students and 34 radiographers). The intention-to-treat analysis showed a mean improvement of 21 percentual points (pp) in the intervention group, compared with 4 pp in the control group (p<0.001). Radiographers clearly improved with the intervention (23pp vs 4pp.; p=0.004). Although the effect was not clear in students (18pp vs 5pp; p=0.098), we found differences in post-test results between intervention and control (88% vs 63%; p=0.003) whereas such difference was absent in pre-test (63% vs 63%; p=0.106). Therefore, an additional per-protocol analysis was applied to reduce the influence of lost to follow-up individuals, which resulted in a higher effect (26pp vs 2pp; p<0.001),