Construct validity and reliability of the Concussion Knowledge Assessment Tool (CKAT).

Objective To evaluate the test-retest reliability and construct validity of the concussion knowledge assessment tool (CKAT) as a measure of knowledge of concussion and its management among chiropractic subgroups and to compare these properties for two scoring strategies for the CKAT. Methods Three chiropractic subgroups (first year students, interns and sports chiropractors) completed the CKAT via SurveyMonkey with as second administration two to six weeks later for a subset of respondents. Scatter plots and Intraclass Correlation Coefficients (ICC) were used for test-retest reliability. A priori hypotheses regarding the relationship of CKAT scores across known subgroups, and with concussion knowledge self-rankings were established prior to data collection. Distributions of CKAT scores were compared across the subgroups using boxplots and ANOVA for known groups validity, and correlation of CKAT scores with concussion knowledge self-ranking was examined. Results Test-retest ICC for the revised scoring was 0.68 (95%CI 0.51-0.80). First year students had a mean revised CKAT (out of 49) of 36.9 (SD= 4.7), interns 39.9 (SD=3.0) and sports chiropractors 41.8 (SD=3.2) which are significantly different (F2,125=17.54; p<0.0001). Conclusions The CKAT distinguished between chiropractic subgroups expected to have different levels of knowledge, supporting construct validity, however, it did not achieve adequate test-retest reliability.