Development and Validation of the Rating of CKD Knowledge Among Older Adults (Know-CKD) With Kidney Failure

Rationale & Objective: Few older adults with kidney failure engage in shared decision making (SDM) for kidney replacement therapy. The lack of instruments to assess SDM-relevant knowledge domains may contribute to this. We assessed the reliability and validity of a new instrument, the Rating of CKD Knowledge Older Adults (Know-CKD). Study Design: Multistage process, including a stakeholder-engaged development phase, pilot testing, and validation of a knowledge instrument using a cross-sectional survey of older adults with CKD. Setting & Participants: 363 patients aged 70 + years with nondialysis advanced chronic kidney disease (CKD) (estimated glomerular filtration rate [eGFR] < 30 mL/min/1.73 m(2)) in Boston, Chicago, Portland, ME, and San Diego from June 2018 and January 2020. Exposure: Educational level, higher literacy (Single Item Literacy Screener [SILS]) and numeracy (Subjective Numeracy Scale [SNS]), having participated in clinic-sponsored dialysis education, and self-reported "feeling informed" about options for treatment. Outcome: Validity and reliability of the KnowCKD instrument. Analytical Approach: Reliability was assessed with the Kuder-Richardson-20 coefficient. Construct validity was demonstrated by testing a priori hypotheses using t test, analysis of variance (ANOVA) tests, and linear regression analyses. Results: The mean ( +/- SD) participant age was 77.6 +/- 5.9 years, and mean eGFR was 22.7 +/- 7.2 mL/min/1.73 m(2) ; 281 participants (78%) self-reported as White. The 12-item KnowCKD assessment had good reliability (KuderRichardson-20 reliability coefficient = 0.75), and a mean score of 58.2% +/- 22.3 SD. The subscales did not attain acceptable reliability. The proportion answering correctly on each item ranged from 20.1% to 91.7%. In examining construct validity, the hypothesized associations held; Know-CKD significantly associated with higher education (beta = 6.98 [95% CI, 1.3 412.61], P = 0.02), health literacy (beta = - 12.67 [95% CI, -19.49 to -5.8 6], P <= 0.001), numeracy per 10% higher (beta = 1.85 [95% CI, 1.02-2.6 9], P <= 0.001), and attendance at dialysis class (beta = 18.28 [95% CI, 13.30-23.27], P <= 0.001). These associations were also observed for the subscales except for prognosis (not associated with literacy or numeracy). Limitations: Know-CKD is only available in English and has been used only in research settings. Conclusions: For older adults facing dialysis initiation decisions, Know-CKD is a valid, reliable, and easy to administer measure of knowledge. Further research should examine the relationship of kidney disease knowledge and SDM, patient satisfaction, and clinical outcomes.