Background: Using hospital-based cancer registry data, we examined the predictors of late-stage CRC diagnosis and identified factors contributing to CRC treatment patterns in Virginia.
Methods: We conducted a retrospective cohort study using the Sentara Cancer Registry for patients diagnosed with CRC from 2008 to 2016. Patients' demographics and clinical characteristics were summarized and compared between Caucasian and African American in a bivariate analysis using the Chi-square test for categorical variables, and the Student's t test for continuous variables. A hierarchical logistic regression model adjusting for age, sex, race, marital status, primary payer, cancer site, initial year of diagnosis, and the Charlson Comorbidity Index (CCI) was used to model the odds of late-stage CRC disease, and to determine factors associated with treatment patterns, i.e., surgery, chemotherapy, and radiation. Adjusted odds ratios with their 95% confidence intervals were reported.
Results: The data consisted of 4,505 cases of CRC reported to Sentara hospitals during 2008 to 2016. Among these patients, 2,371 (54.34%) were diagnosed late and 1,992 (45.66%) were early-stage diagnosis. About three quarters of the patients were Caucasian (74.93%). On average, patients were 65.95 years old (SD=13.22), and the majority were males (52.01%), married (55.8%), and had primarily Medicare (57.02%), followed by private insurance (31.88%). During the study period, the overall rate of CRC diagnosis decreased from 12.32% in 2008 to 10.36% in 2016. About 47.26% of the patients had a Charlson Comorbidity Index (CCI) score ≤ 2. The bivariate analysis of racial disparity revealed that African American with CRC are younger than their counterpart Caucasian (p < 0.0001), and they are generally more likely to be late-stage diagnosis (59.41% vs. 52.65%; P < 0.0001). Time to treatment initiation is statistically higher among Black than White (p = 0.0307). The fully adjusted multilevel logistic regression model showed a nonexistent racial disparity in late stage diagnosis between White and Black. The odds of late-stage CRC diagnosis were higher among younger patients (<50), those with a CCI score greater than 5, and among the uninsured and self-pay (All P<0.05). For treatment patterns, younger patients (<50) had higher odds of having surgery, performing chemotherapy and radiation. The odds of surgery are less likely among Black and more likely among those with a private insurance, and a CCI score between 3 and 5 (All P<0.05). Finally, the likelihood of performing radiation is statistically significantly less among patients with a CCI score between 3 and 5 and above 5, [OR=0.563 (95%CI: 0.398 - 0.795]. Race, sex, and marital status were not significant predictors of late stage CRC, chemotherapy, and radiation.
Conclusion: The impact of patient demographics and clinical factors on late stage and treatment patterns is striking in Virginia. These findings suggest the need for policy to try to delineate those factors associated with those disparities.
Citation Format: Hadiza Galadima, Georges Adunlin, Marybeth Hughes, and Jennifer May. Racial disparity, stage at diagnosis, and treatment patterns of colorectal cancer: Analysis of data from a hospital-based cancer registry, Virginia 2008-2016 [abstract]. In: Proceedings of the Eleventh AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2018 Nov 2-5; New Orleans, LA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(6 Suppl):Abstract nr C005.