V T Nguyen1, S W Edmonds2,3, Y Lou4, D W Roblin5,6, K G Saag7, P Cram8,9, F D Wolinsky4,2,10. 1. College of Public Health, University of Iowa, 5233 Westlawn, Iowa City, IA, 52242, USA. thuy-v-nguyen@uiowa.edu. 2. College of Nursing, University of Iowa, Iowa City, IA, USA. 3. CADRE, Iowa City VA Health System, Iowa City, IA, USA. 4. College of Public Health, University of Iowa, 5233 Westlawn, Iowa City, IA, 52242, USA. 5. Kaiser Permanente of Atlanta, Atlanta, GA, USA. 6. School of Public Health, Georgia State University, Atlanta, GA, USA. 7. School of Medicine, University of Alabama at Birmingham, Birmingham, AL, USA. 8. Faculty of Medicine, University of Toronto, Toronto, ONT, Canada. 9. University Health Network and Mount Sinai Hospital, Toronto, ONT, Canada. 10. College of Medicine, University of Iowa, Iowa City, IA, USA.
Abstract
We studied the Osteoporosis and You knowledge scale in 7749 participants enrolled in a clinical trial. Results confirmed its psychometric properties in a diverse audience. Baseline scores were associated with better recall of bone mineral density test results at follow-up; however, the scale was not responsive to knowledge change. INTRODUCTION: The goal of this study was to confirm the measurement properties of the Osteoporosis and You (O&Y) knowledge scale using classic test theory methods in the 7749 men and women participating in the Patient Activation After DXA Result Notification (PAADRN) randomized controlled trial. We hypothesized a simple factor structure that would reflect the four-factor model previously published. METHODS: We conducted psychometric analyses which included item analysis, internal consistency reliability, construct validity using exploratory and confirmatory factor analysis (EFA and CFA), comparing knowledge levels across pre-specified groups, and responsiveness to change. RESULTS:PAADRN participants were predominantly college educated, White females with low bone density, and a moderate level of 10-year fracture risk. EFA revealed four domains closely matching those in two previous reports. While overall scale reliability was minimally acceptable at 0.68, the reliabilities of the domain subscales were unacceptably low (0.59, 0.64, 0.45, and 0.36 for the Biological, Lifestyle, Consequences, and Prevention and Treatment subscales). CFA revealed the data fit the hypothesized model reasonably well with the items loading on their expected latent variable. The scale was not responsive to change, but although not significant, improved knowledge indicated better DXA result recall at 12 and 52 weeks. CONCLUSIONS: In the PAADRN population, the O&Y knowledge scale had psychometric properties similar to those previously reported. Over 12 and 52 weeks, participants did not demonstrate significant changes in knowledge, but those with higher knowledge at baseline were more likely to accurately recall their baseline DXA result.
RCT Entities:
We studied the Osteoporosis and You knowledge scale in 7749 participants enrolled in a clinical trial. Results confirmed its psychometric properties in a diverse audience. Baseline scores were associated with better recall of bone mineral density test results at follow-up; however, the scale was not responsive to knowledge change. INTRODUCTION: The goal of this study was to confirm the measurement properties of the Osteoporosis and You (O&Y) knowledge scale using classic test theory methods in the 7749 men and women participating in the Patient Activation After DXA Result Notification (PAADRN) randomized controlled trial. We hypothesized a simple factor structure that would reflect the four-factor model previously published. METHODS: We conducted psychometric analyses which included item analysis, internal consistency reliability, construct validity using exploratory and confirmatory factor analysis (EFA and CFA), comparing knowledge levels across pre-specified groups, and responsiveness to change. RESULTS: PAADRN participants were predominantly college educated, White females with low bone density, and a moderate level of 10-year fracture risk. EFA revealed four domains closely matching those in two previous reports. While overall scale reliability was minimally acceptable at 0.68, the reliabilities of the domain subscales were unacceptably low (0.59, 0.64, 0.45, and 0.36 for the Biological, Lifestyle, Consequences, and Prevention and Treatment subscales). CFA revealed the data fit the hypothesized model reasonably well with the items loading on their expected latent variable. The scale was not responsive to change, but although not significant, improved knowledge indicated better DXA result recall at 12 and 52 weeks. CONCLUSIONS: In the PAADRN population, the O&Y knowledge scale had psychometric properties similar to those previously reported. Over 12 and 52 weeks, participants did not demonstrate significant changes in knowledge, but those with higher knowledge at baseline were more likely to accurately recall their baseline DXA result.
Authors: Paul A Harris; Robert Taylor; Robert Thielke; Jonathon Payne; Nathaniel Gonzalez; Jose G Conde Journal: J Biomed Inform Date: 2008-09-30 Impact factor: 6.317
Authors: Santiago Palacios; Rafael Sánchez-Borrego; José Luis Neyro; Francisco Quereda; Francisco Vázquez; Maite Pérez; Montserrat Pérez Journal: Menopause Int Date: 2009-09
Authors: Stephanie W Edmonds; Fredric D Wolinsky; Alan J Christensen; Xin Lu; Michael P Jones; Douglas W Roblin; Kenneth G Saag; Peter Cram Journal: Contemp Clin Trials Date: 2012-10-17 Impact factor: 2.226