VP57 Test-Retest Reliability Analysis Of The Patient Reported Outcomes Burdens And Experiences (PROBE) Study Questionnaire Test-Retest Reliability Analysis Of The PROBE Study Questionnaire

Abstract

INTRODUCTION:

The Patient Reported Outcomes Burdens and Experiences (PROBE) questionnaire was developed with direct patient involvement in questionnaire design, conduct and analysis using patient-centered outcomes to assess health status in patients with hemophilia (PWH). Phase 1 confirmed robustness of the methodology and feasibility. Phase 2a investigated individual test-retest reliability. Phase 2b will explore population level reproducibility.

METHODS:

PWH and non-PWH individuals who attended a hemophilia-related workshop were asked to complete the PROBE questionnaire 3 times (paper-based survey on 2 consecutive days and then a web-based version). Test-retest reliability was analyzed using the percentage agreement and Kappa statistic. Kappa coefficient interpretation .81-1.00 almost perfect, .61- .80 substantial; .41- .60 moderate; .21 -.40 fair; .00 -.20, slight; and < .00 poor agreement.

RESULTS:

Sixty-three participants from twenty-one countries were enrolled with a median age of 50 (range 14–76) years. Of these, thirty (47.6 percent) were PWH or carriers, thirty-three (52.5 percent) were participants with no known bleeding disorders. On general health domain, Kappa coefficients ranged from .69 to .92, indicating substantial to almost perfect agreement, for all items. Reliability of the web-based questionnaire showed moderate to substantial agreement for all except one item. For the hemophilia-related domain, Kappa coefficients ranged from .5-1.0. Of these, five of eleven items were in perfect agreement (Kappa = 1.0). Reliability of web-based questionnaire items were in substantial to almost perfect agreement. For overall health related quality of life, the EuroQol five dimensions questionnaire (EQ-5D) had Kappa coefficients of .62 to .92. Intraclass correlation coefficient of visual analog scale (VAS) was .90 (95 percent Confidence Interval, CI; .83-.94). Test-retest reliability was comparable between hemophilia patients and participants with no known bleed.

CONCLUSIONS:

Phase 2a demonstrated individual test-retest reliability and suggests PROBE is a reliable tool to assess Patient Reported Outcomes in PWH. The Web-based questionnaire has an acceptable agreement with the standard paper-based version in all domains. PROBE Phase 2b, to demonstrate reproducibility at the population level, is on-going. To date, 1,039 participants have been recruited from 10 countries.