J Shannon Swan1, Vanessa F Furtado2, Lisa A Keller3, Judith Borsody Lotti2, Catherine A Saltalamacchia2, Inga T Lennes4, Gloria M Salazar5. 1. Massachusetts General Hospital (MGH) Institute for Technology Assessment, Boston, Massachusetts; Harvard Medical School, Boston, Massachusetts. Electronic address: jsswan@mgh.harvard.edu. 2. MGH Imaging, Boston, Massachusetts. 3. Department of Educational Policy, Research & Administration, University of Massachusetts, Amherst, Massachusetts. 4. Harvard Medical School, Boston, Massachusetts; MGH Cancer Center, Boston, Massachusetts. 5. Harvard Medical School, Boston, Massachusetts; MGH Imaging, Boston, Massachusetts.
Abstract
PURPOSE: The Radiology Process Model (RPM) was previously described in terms of its conceptual basis and proposed survey items. The current study describes the first pilot application of the RPM in the field and the results of initial psychometric analysis. METHODS: We used an Institutional Review Board-approved pilot RPM survey in 100 patients having outpatient interventional radiology procedures. The 24 survey items had 4 or 5 levels of severity. We assessed for missing data, items that patients found confusing, any suggestions by patients for additional items and clarity of items from patient feedback. Factor analysis was performed and internal consistency measured. Construct validity was assessed by correlation of patient responses to the items as a summated scale with a visual analog scale (VAS) they completed indicating their interventional radiology experience. RESULTS: The visual analog scale and the RPM summated scale were strongly correlated (r = 0.7). Factor analysis showed four factors: interactions with facility and doctors/staff, time-sensitive aspects, pain, and anxiety. The items showed high internal consistency (alpha: 0.86) as a group and approximately 0.7 to 0.9 by the factors. Analysis shows that two items could be deleted (cost and communication between radiologist and referrers). Revision of two items and potential addition of others are discussed. CONCLUSIONS: The RPM shows initial evidence of psychometric validity and internal consistency reliability. Minor changes are anticipated before wider use.
PURPOSE: The Radiology Process Model (RPM) was previously described in terms of its conceptual basis and proposed survey items. The current study describes the first pilot application of the RPM in the field and the results of initial psychometric analysis. METHODS: We used an Institutional Review Board-approved pilot RPM survey in 100 patients having outpatient interventional radiology procedures. The 24 survey items had 4 or 5 levels of severity. We assessed for missing data, items that patients found confusing, any suggestions by patients for additional items and clarity of items from patient feedback. Factor analysis was performed and internal consistency measured. Construct validity was assessed by correlation of patient responses to the items as a summated scale with a visual analog scale (VAS) they completed indicating their interventional radiology experience. RESULTS: The visual analog scale and the RPM summated scale were strongly correlated (r = 0.7). Factor analysis showed four factors: interactions with facility and doctors/staff, time-sensitive aspects, pain, and anxiety. The items showed high internal consistency (alpha: 0.86) as a group and approximately 0.7 to 0.9 by the factors. Analysis shows that two items could be deleted (cost and communication between radiologist and referrers). Revision of two items and potential addition of others are discussed. CONCLUSIONS: The RPM shows initial evidence of psychometric validity and internal consistency reliability. Minor changes are anticipated before wider use.
Authors: Monica L Zigman Suchsland; Elizabeth Witwer; Anjali R Truitt; Danielle C Lavallee; Ying Zhang; Philip Posner; Brian Do; Patrick M Bossuyt; Victoria Hardy; Matthew J Thompson Journal: J Am Coll Radiol Date: 2018-10-25 Impact factor: 5.532