[Objective assessment of transfusion-related knowledge of nurses using modern test theory]

Clinical practice requires knowledge of health professionals for the application of problem solving of different tasks in various clinical settings. Therefore, a set of reliable measurement tools is required to assess various components of the professional knowledge including factual knowledge, skills and competence as outcomes of nursing education and training. Traditional questionnaires and test methods do not allow the measurement of these characteristics properly, as these tools typically provide overall scores without relevant insight into areas in which nurses perform well, and those where their knowledge and/or skills should be improved substantially.

The aim of this nationwide survey conducted among nurses between November 2014 and February 2015 was to determine how the teaching/training objectives have been achieved in the nurses' transfusion practice. The authors attempted to exceed the capabilities of classical test theory and acquire a detailed picture about what the nurses know about transfusion therapy and how they are involved and behave in routine clinical practice.

The knowledge and skills of 657 participants were assessed using a validated instrument consisting of a set of questions covering every aspects of transfusion therapy. The answers to these items were evaluated on a binary (good or bad) scale. Recorded answers of the participants were analysed using hierarchical cluster analysis and item response theory tools such as the one-parametric Rasch model suitable for dichotomous data. Data analysis was performed with the SPSS program and the ltm module of the R statistical program.

The paper presents the distribution of correct and incorrect answers to various questions about transfusion therapy along with the corresponding logit values and odds ratios, respectively. The characteristic curves of each item were determined on the basis of the number of correct answers that have been recorded. These curves highlight which questions were answered easily and which items were found harder to answer by the nurses who participated in the survey. In addition to the separate analysis of individual questions, a set of response patterns is also presented which shows how frequently the nurses responded correctly to different combinations (sub-sets) of questions. On the whole, nurses exhibited medium level performance in terms of knowledge and skills required for efficient and safe transfusion practice.

Objective and reliable measurement of the level of acquired knowledge is a key requirement in nursing education. This paper, which demonstrates the use of cluster analysis and item response theory for the assessment of transfusion-related knowledge of nurses, focuses on this issue for the first time in nursing research. The results of this survey have revealed substantial limitations and deficiencies both in knowledge and skills of nurses which need to be addressed by training in order to improve the efficiency and safety of transfusion therapy.