Evidence for Validity for the Cognitive Load Inventory for Handoffs.

OBJECTIVE Patient handoffs remain a significant patient safety challenge. Cognitive Load Theory (CLT) can be used to identify the cognitive mechanisms of handoff errors. The ability to measure cognitive load types during handoffs could drive the development of effective protocols and educational strategies. No such measure currently exists. METHOD The authors developed the Cognitive Load Inventory for Handoffs (CLIH) using a multi-step process, including expert interview to enhance content validity and talk-alouds to optimize response process validity. The final versions contained 28 items. From January to March, 2019, we administered a cross-sectional survey to 1,807 residents and fellows from a large health system in the United States. Participants completed the CLIH following a handoff. Exploratory factor analysis from one-third of respondents identified high performing items; confirmatory factor analysis with the remaining sample assessed model fit. Model fit was evaluated using the comparative fit index (>0.90), Tucker-Lewis Index (>0.80), standardized root mean square residual (<0.08), and root mean square of approximation (<0.08). RESULTS Participants included 693 trainees (38.4%) (231 in the exploratory study and 462 in the confirmatory study). Eleven items were removed during exploratory factor analysis. Confirmatory factor analysis of the 16 remaining items (5 intrinsic load, 7 extraneous load, and 4 germane load) supported a three-factor model and met criteria for good model fit: CFI was 0.93, RMSEA was 0.074, and SRMR was 0.07. The fit was comparable for gender and role. Intrinsic, extraneous and germane scales had high internal consistency. With one exception, scale scores associated, as hypothesized, with postgraduate level and clinical setting. CONCLUSION The results provide evidence for validity: content, response process, internal structure and association with other variables. This instrument can be used to determine the relative drivers of cognitive load during handoffs as well as the relative effectiveness of handoff instruction and protocols.