Table 1 Three step screening and eligibility process

First step entailed:

1. Identification of NIH funded grants through a search that used two keywords that were free terms, racism and healthcare, resulted in a total of 558 funded proposals representing all types of grants (e.g., R-, P-series).

2. Non-R-series grants were removed (278 grants).

3. The remaining 280 abstracts and public health relevance statements were exported into excel and then into Covidence. Although Covidence [37] is a cloud-based collaborative platform designed specifically for the management of reviews, it was used in this study by the research team.

4. After uploading them into Covidence, duplicates were identified and removed; and 90 duplicates were eliminated, resulting in a total of 190 R-series grants.

Second step included:

1. The inclusion and exclusion criteria were used during the search and screening phase. The inclusion criteria were: 1) research based R-series grants; 2) focused on racism in healthcare; 3) conducted by US-based entities and within the US territories; and 4) conducted during the five fiscal years of 2017 to 2022. The exclusion criteria were grants: 1) without a focus on racism in healthcare; and 2) non-US-based. Given that research-based grants were the emphasis of this study, R-series were the only types of grants focused on in this portfolio analysis.

2. Both the inclusion and exclusion criteria were entered into Covidence and used independently by each of the two reviewers to appraise the 190 R-series abstracts and public health relevance statements. Each reviewer took notes in the platform and answered the following question: Does the purpose/aim state that the study plans to address racism among healthcare users and within healthcare?

3. After the independent reviews were completed, the first author saw the recommendations (those with consensus or disagreement) in Covidence, and for any discrepancies, the reviewers met via ZOOM to discuss and resolve them by consensus; and 97 were excluded for not meeting the inclusion criteria. The reviewers had a 90% agreement rate.

During the third step:

1. Data were extracted from the remaining 93 abstracts and public health relevance statements.

2. In preparation for data extraction, the quality assessment template in Covidence was used to rate for risk of bias. The data extraction template feature in Covidence was also used by each of the two reviewers, but they met via ZOOM to discuss and customize the headings in Table 3.

3. Other data extracted and analyzed from the NIH RePORTER project sections was the NIH Research, Condition, and Disease Categorization (RCDC) spending categories, number of publications from each corresponding grant, funding history and funding opportunity announcements, during the fiscal years 2017 to 2022 (Table 3).