Back

Development of a novel healthcare discrimination measure: PreDict

Oladele, C.; Gonzalez-Colaso, R.; Schulze, A.; Rizzo, T.; The PreDict Writing Group, ; Nunez-Smith, M.

2023-10-19 public and global health
10.1101/2023.10.16.23291749 medRxiv
Show abstract

IntroductionPatient reported quality of care measures are widely recognized tools for healthcare system performance assessment. Yet, there are few existing patient reported quality of care measures regarding health equity, and none to specifically collect patient experiences of discrimination in health care. ObjectiveTo develop an item pool to measure patient experiences of healthcare discrimination-the Patient-Reported Experiences of Discrimination in Care Tool (PreDict). MethodsUtilizing a multistage, exploratory sequential mixed methods study design, we conducted qualitative interviews (n=73) and expert panel consensus analysis to develop items to capture patient experiences of discrimination. This process plus systematic literature review identified extant items and informed de novo items for inclusion in the item pool. Items were developed in English and Spanish and were not represented by extant items. Following identification of the initial item pool (n=125), candidate items underwent cognitive interview testing with English (n=113) and Spanish (n=70) speaking participants to evaluate items for clarity and comprehensiveness. English and Spanish items were also evaluated by a bilingual expert panel to recommend pool items for inpatient field testing. ResultsOne hundred and three items underwent cognitive interview testing and fifty-nine items were retained. Lack of clarity was the most cited factor for removal or revision of items. Expert panel review resulted in the removal of one additional item and the revision of seven items. Fifty-eight candidate items were retained for inclusion in field testing and future analyses using item response theory modeling. ConclusionPreDict fills an important gap in measurement of discrimination, which is known to influence patient health outcomes. Development and testing to date demonstrate evidence of validity in characterizing the complex phenomenon of healthcare discrimination.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.