664 - Measuring response stability of a novel assessment tool for promoting decoloniality within global health

Publication Number: 3643.664

Jeffrey G. Edwards, St. Jude Children's Research Hospital, Memphis, TN, United States; Christiana M. Russ, Boston Children's Hospital, Boston, MA, United States; Priscilla Gonzalez, Boston Children's Hospital, Boston, MA, United States; Elisha Siegel, BCH Global Health Program, Boston, MA, United States; Jared Gabriel L.. Dela Rosa, College of Medicine, University of the Philippines-Manila, Manila City, National Capital Region, Philippines; Tamara N. Fitzgerald, Duke University School of Medicine, Durham, NC, United States; Barnabas Tobi. Tlayande, University of Global HEalth Equity, Kigali, Kigali, Rwanda

Background: Global health is rooted in colonial practices though academia has recently more actively engaged in historical and ongoing community-led efforts to decolonize global health work. Despite more global health curricula developed with the aim of promoting decoloniality, there are not any formal assessments or self-reflection tools that have been published. An assessment tool was created by the author team via literature review, with the tool components refined to achieved consensus via a modified Delphi containing eleven decoloniality and global health experts (Tables 1 and 2). The reliability of that tool has not yet been assessed.

Objective: This study aimed to assess the internal validity of both the self-assessment and educator assessment of a learner tool by measuring each tool's response stability using paired samples.

Design/Methods: The study design included a sociodemographic survey followed by two administrations of the training level-specific version of the survey (self-assessment for learners, educator assessment for educators), separated by one month's time. The study consisted of a convenience sample of learners or educators engaged in global health work. Data were captured via REDCap and analyzed utilizing R software. This study received ethical approval from the Boston Children's Hospital IRB. A Wilcoxon signed-rank test was used to compare response differences across survey administration time point for each domain since paired pre-post differences were non-normal in several domains. A Holm correction was applied across the five domains within each assessment instrument to control the family-wise error rate.

Results: 62 individuals (36 learners, 26 educators) completed the study. Four of five self-assessment domains and three of five educator-assessment domains showed no statistically significant shift after Holm correction; significant shifts were observed in the research domain for the self-assessment, and in the structural humility and leadership & development domains for the educator assessment (Table 3).

Conclusion(s): These results indicate that an assessment tool in decoloniality for learners in global health can demonstrate response stability in both self-assessment and educator assessment of a learner formats. The three domains that did not demonstrate reliability likely require adjustments before being utilized in a meaningful tool. Follow-up studies should be conducted to assess external validity in different contexts. These tools might aid learners and educator in self-reflection and overall promotion of decoloniality within global health.

Table 1
Complete description of each learner level across five domains as a self-assessment, after reaching consensus (>75% agree) in a modified Delphi study

Table 2
Complete description of each learner level across five domains as an educator assessment of a learner, after reaching consensus (>75% agree) in a modified Delphi study

Table 3
Results of the Wilcoxon signed-rank tests for paired survey administrations, with median responses of the learner level and interquartertile ranges for each domain