Development and evaluation of a new survey instrument to measure the quality of colorectal cancer screening decisions
1 Division of General Internal Medicine, Health Decision Sciences Center, Massachusetts General Hospital, 50 Staniford Street, 8th Floor, Boston, MA 02114, USA
2 Harvard Medical School, Boston, MA 02115, USA
3 Center for Survey Research, University of Massachusetts Boston, 100 Morrissey Boulevard, Boston, MA 02125, USA
4 Informed Medical Decisions Foundation, 40 Court Street, 3rd Floor, Boston, MA 02108, USA
5 UNC Division of General Internal Medicine and UNC Institute for Healthcare Quality Improvement, University of North Carolina, 5045 Old Clinic Building UNC Hospital, Chapel Hill, NC 27599-7110, USA
BMC Medical Informatics and Decision Making 2014, 14:72 doi:10.1186/1472-6947-14-72Published: 20 August 2014
Guidelines for colorectal cancer screening recommend that patients be informed about options and be able to select preferred method of screening; however, there are no existing measures available to assess whether this happens.
Colorectal Cancer Screening Decision Quality Instrument (CRC-DQI) includes knowledge items and patients' goals and concerns. Items were generated through literature review and qualitative work with patients and providers. Hypotheses relating to the acceptability, feasibility, discriminant validity and retest reliability of the survey were examined using data from three studies: (1) 2X2 randomized study of participants recruited online, (2) cross-sectional sample of patients recruited in community health clinics, and (3) cross-sectional sample of providers recruited from American Medical Association Master file.
338 participants were recruited online, 94 participants were recruited from community health centers, and 115 physicians were recruited. The CRC-DQI was feasible and acceptable with low missing data and high response rates for both online and paper-based administrations. The knowledge score was able to discriminate between those who had seen a decision aid or not (84% vs. 64%, p < 0.001) and between providers, online patients and clinic patients (89% vs. 74% vs. 41%, p < 0.001 for all comparisons). The knowledge score and most of the goals had adequate retest reliability. About half of the participants received a test that matched their goals (47% and 51% in online and clinic samples respectively). Many respondents who had never been screened had goals that indicated a preference for colonoscopy. A minority of respondents in the online (21%) and in clinic (2%) samples were both well informed and received a test that matched their goals.
The CRC-DQI demonstrated good psychometric properties in diverse samples, and across different modes of administration. Few respondents made high quality decisions about colon cancer screening.