Statistical analysis
We ranked patients by national ADI percentile, and divided them into 4 equal-sized quartiles, where the lowest quartile (Q1) represents least socioeconomically deprived status, and highest quartile (Q4) represents most deprived.10 To compare differences in demographics and comorbidities between ADI quartiles, we used Kruskal-Wallis rank sum test for continuous variables and Pearson chi-squared test or Fisher’s exact test for categorical variables, as appropriate. Pairwise chi-squared test with nominal independence and Bonferroni correction was used for between-group differences. Two-sided Cochrane-Armitage test was used to compare occurrence of clinical and imaging follow-up between ADI quartiles.
Bivariate and multivariate generalized logistic regression were used to determine factors associated with occurrence of surveillance imaging (CT chest or echocardiogram) at or before 24 months following index CT scan. Variables with p-value <0.1 on bivariate analysis were included in the multivariate model.
Competing risks analysis, as described by Fine et al.,17, 18 was used to model the effect of ADI quartile on time from index CT scan to first encounter with a cardiologist or cardiac surgeon for ATAA aneurysm, with the competing risk of death. Time to follow-up was defined as 0 for patients who were already established with either of these specialties. This methodology allows for visual representation of the subset of patients with pre-established follow-up prior to index scan, in which new first-time encounters do not apply. We performed competing risks regression adjusted for ADI quartile, age, sex, race, and ATAA size on index scan, factors which may influence a provider’s decision to refer patients to a cardiovascular specialist. We defined statistical significance asp <0.05. All data analysis was performed in R version 3.6.3 (R Foundation for Statistical Computing).
The Yale University Institutional Review Board approved this study and individual consent was waived (IRB 2000027551). The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines for cohort studies.