Survival Analysis of Demographic Factors Associated With 5+ Year Survival of Pancreatic Carcinoma

Background Although pancreatic cancer incidence is low at 13.1 per 100,000 people, this cancer is difficult to treat and carries a poor 5-year survival rate. Additionally, pancreatic cancer survival rates vary disproportionately based on age and race. The objective of this study was to evaluate the association between 5-year survival of pancreatic cancer and the basic demographic factors age, race, and sex. Methods Data were retrieved from the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) 18 database, spanning from 2000-2017, using SEER*Stat. SPSS was used to calculate descriptive statistics for vital status, age, race, and sex. Odds ratios with confidence intervals were calculated using Epi Info. Case data were used to conduct survival analysis by age, race, and sex using OriginPro. Results Out of a total of 118,581 cases, 79.3% were White (n = 106,887), 12.5% were Black (n = 16,866), 7.4% were Asian or Pacific Islander (n = 9,960), 0.6% were American Indian/Alaskan Native (n = 792), and 0.2% were unknown race (n = 321). The odds ratio (OR) of dying before reaching 5+ survival was lowest for the Asian or Pacific Islander group (OR = 0.70, 95% CI: 0.66 - 0.74), followed by the group of Black patients (OR = 1.07, 95% CI: 1.02 - 1.13), the White patients group (OR = 1.12, 95% CI: 1.08 - 1.17), and the American Indian/Alaskan Native group (OR = 1.12, 95% CI: 0.89 - 1.40). The largest age group was 65-69 years old, comprising 14.7% (n = 19,866) of the dataset. Probability of 5+ year survival for pancreatic cancer patients was highest for the age group 15-19 years (n = 74). In general, 5+ year survival probability declined with age. Risk of death before reaching 5+ year pancreatic cancer survival was slightly higher in men (OR = 1.03, 95% CI: 1.00 - 1.07), who comprised 50.9% (n = 68,628) of the dataset. Discussion Findings from this study corroborate differences by age, race, and sex discussed in the literature. Differences in survival rates by race depart from some findings in literature documenting no significant differences in treatment outcome by race. Controlling for age in a future study in both race and sex survival probability analyses may be helpful. Further, stratifying by sex in survival probability analysis by race would be illuminating. In addition to survival analysis, regression modeling would be a useful next step.