Purpose: Quality improvement (QI) has been part of medical education for over a decade. Assessment of QI learning remains challenging. The Quality Improvement Knowledge Application Tool (QIKAT), developed a decade ago, is widely used despite its subjective nature and inconsistent reliability. From 2009 to 2012, the authors developed and assessed the validation of a revised QIKAT, the "QIKAT-R."
Method: Phase 1: Using an iterative, consensus-building process, a national group of QI educators developed a scoring rubric with defined language and elements. Phase 2: Five scorers pilot tested the QIKAT-R to assess validity and inter- and intrarater reliability using responses to four scenarios, each with three different levels of response quality: "excellent," "fair," and "poor." Phase 3: Eighteen scorers from three countries used the QIKAT-R to assess the same sets of student responses.
Results: Phase 1: The QI educators developed a nine-point scale that uses dichotomous answers (yes/no) for each of three QIKAT-R subsections: Aim, Measure, and Change. Phase 2: The QIKAT-R showed strong discrimination between "poor" and "excellent" responses, and the intra- and interrater reliability were strong. Phase 3: The discriminative validity of the instrument remained strong between excellent and poor responses. The intraclass correlation was 0.66 for the total nine-point scale.
Conclusions: The QIKAT-R is a user-friendly instrument that maintains the content and construct validity of the original QIKAT but provides greatly improved interrater reliability. The clarity within the key subsections aligns the assessment closely with QI knowledge application for students and residents.