Assessing reliability and validity of SIPAT and opportunities for improvement: A single-center cohort study

Abstract

Psychosocial assessment is a standard component of patient evaluations for transplant candidacy. The Stanford Integrated Psychosocial Assessment for Transplant (SIPAT) is a widely used measure to assess psychosocial risk for transplant. However, there are questions regarding the SIPAT's reliability and validity. We examined the SIPAT’s psychometric performance and its impact on equitable access to transplant in a diverse cohort of 2825 patients seeking liver transplantation between 2014 and 2021 at an urban transplant center. The SIPAT demonstrated good internal consistency reliability at the overall score [Cronbach’s α = 0.85, 95% CI (0.83, 0.86)] and domain levels (0.80 > α > 0.70). There was mixed support for structural validity, with poor overall model fit in confirmatory factor analysis and 50% of questions achieving the 0.70-factor loadings threshold. Adjusting for sociodemographic variables, the odds of not being waitlisted for psychosocial reasons were three times higher for patients with Medicaid insurance than patients with private insurance [OR 3.24, 95% CI (2.09, 4.99)] or Medicare [OR 2.89, 95% CI (1.84, 4.53)], mediated by higher SIPAT scores. Black patients had nearly twice the odds of White patients [OR 1.88, 95% CI (1.20, 2.91)], partially mediated by higher social support domain scores. Patients with Medicaid, non-White patients, and those without a college degree scored significantly higher on collinear questions, disproportionately contributing to higher SIPAT scores. The SIPAT did not perform equally across insurance type, race/ethnicity, and education groups, with the lowest subgroup validity associated with patient readiness and psychopathology domains. The SIPAT should be interpreted with caution, especially as a composite score. Future studies should examine validity in other populations.