Semester

Fall

Date of Graduation

2012

Document Type

Dissertation

Degree Type

PhD

College

School of Pharmacy

Department

Pharmaceutical Systems and Policy

Committee Chair

S. Suresh Madhavan

Committee Co-Chair

Joel Halverson

Committee Member

Xiaoyun Lucy Pan

Committee Member

William P. Petros

Committee Member

Cindy Tworek

Abstract

Diabetes is one of the most common chronic comorbid condition seen in elderly CRC patients. Outcomes of CRC patients with diabetes specifically stage at diagnosis, emergency condition for CRC surgery, survival, and mortality have been insufficiently explored. The aims of the study were to investigate the association between diabetes and stage at diagnosis of CRC in elderly Medicare beneficiaries; to check the association of diabetes with presenting as an emergency condition for CRC surgery in the elderly and; to explore the effect of diabetes on survival of elderly Medicare beneficiaries with CRC. Using the SEER-Medicare data from 2003-2005, patients newly diagnosed with CRC were selected and divided into diabetic and nondiabetic cohorts. The two cohorts were compared in terms of stage at CRC diagnosis, emergency presentation for CRC surgery, and five year survival. Logistic regressions were used to check the association between diabetes and stage at diagnosis and emergency condition for CRC surgery. Survival analysis was employed compare time to death between diabetic and non-diabetic CRC patients. Covariates used in the study included the three most common comorbid conditions besides diabetes: coronary atherosclerosis, congestive heart failure, and chronic obstructive pulmonary disease, age, sex, race, tumor location, region in the country, patient location, and frequency of physician office visits. For survival analysis additional treatment variables -- chemotherapy, radiation, and surgery were included. For stage at diagnosis of CRC, diabetes showed a significant inverse association (OR 0.92; 95% CI 0.85-1.00). On adding quintile of physician office visits this association was not significant. Odds of being diagnosed at a later stage was significantly associated with the least number of office visits (OR 2.13; 95% CI 1.86-2.44) as was having a proximal tumor (OR 1.40, 95% C 1.30-1.51) Although the odds of a diabetic patient being an emergency patient were lower than a non-diabetic, this was not statistically significant (OR 0.89; 95% CI 0.79-1.01) Mortality risk was significantly greater for diabetic CRC patients than nondiabetics (HR 1.15, 95% CI 1.09-1.20). Presenting emergently increased the risk of mortality (HR 1.61, 95% CI 1.54-1.68). Surgery for CRC reduced the risk of mortality (HR 0.41, 95% CI 0.39-0.43) and although in bivariate analyses patients who received chemotherapy were more likely to die, the hazard model showed a significant benefit associated with chemotherapy or radiation (HR 0.70, 95% CI 0.67-0.74). The worse terminal outcomes seen in diabetic CRC patients indicates the need for early and timely screening to prevent the disease or diagnosis at earlier stages.

Download

Share

COinS