Competence in transthoracic echocardiography (TTE) is unrelated to traditional measures of TTE competence, such as duration of training and number of examinations performed. This study aims to explore aspects of validity of an instrument for structured assessment of echocardiographic technical skills.
The study included 45 physicians with three different clinical levels of echocardiography competence who all scanned the same healthy male following national guidelines. An expert in echocardiography (OG) evaluated all the recorded, de-identified TTE images blindly using the developed instrument for assessment of TTE technical skills. The instrument consisted of both a global rating scale and a procedure specific checklist. Two scores were calculated for each examination: A global rating score and a total checklist score. OG rated ten examinations twice for intra-rater reliability, and another expert rated the same ten examinations for inter-rater reliability. A small pilot study was then performed with focus on content validity. This pilot study included nine physicians who scanned three patients with different pathologies as well as different technical difficulties.
Validity of the TTE technical skills assessment instrument was supported by a significant correlation found between level of expertise and both the global score (Spearman 0.76, p<0.0001) and the checklist score (Spearman 0.74, p<0.001). Both scores were able to distinguish between the three levels of competence that were represented in the physician group. Reliability was supported by acceptable inter- and intra-rater values. The pilot study showed a tendency to improved scores with increasing expertise levels, suggesting that the instrument could also be used when pathologies were present.
We designed and developed a structured assessment instrument of echocardiographic technical skills that showed evidence of validity in terms of high correlations between test scores on a normal person and the level of physician competence, as well as acceptable inter- and intra-rater reliability scores. Further studies should, however, be performed to determine the adequate number of assessments needed to ensure high content validity and reliability in a clinical setting.