Graduation Date

Spring 5-5-2018

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Programs

Epidemiology

First Advisor

Shinobu Watanabe-Galloway

Second Advisor

Jane Meza

Third Advisor

KM Monirul Islam

Abstract

Colorectal Cancer (CRC) is the third most common and leading cause of cancer death in the United States. Although CRC screening can prevent and detect CRC at an early stage, about 35% of Americans are not screened. Despite the recent increase in screening, people with lower SES and those who live in rural areas have lowest screening. In rural areas, a common obstacle for screening is the long trips for health services which is associated with advanced CRC.

Moreover, surgery is a substantial part of CRC treatment since stages I-III and some metastatic CRC (mCRC) patients are treated with surgery. Up to 25% of patients who undergo surgery get readmitted to the hospital due to several factors which costs $300 million annually. Prior studies showed some variations in CRC treatment between rural and urban patients.

The purpose of this study was to assess the association between rural-urban status and CRC screening, stage at diagnosis and the receipt of CRC surgery. There were three specific aims: 1) To assess the impact of rurality on CRC screening, 2) To assess the impact of travel time on the stage of CRC diagnosis, and 3) To evaluate rural-urban differences in healthcare utilization.

We conducted analyses using data from Blue Cross Blue Shield of Nebraska (BCBSNE) between 2012 and 2016. For Aim 1, the study population included BCBSNE members aged 50-64 years with average-risk CRC. For Aim2, the study population included BCBSNE members aged 50-64 years with average-risk CRC. For Aim 3, the study population consisted of CRC patients between the ages of 19-65 years old who had CRC surgery during the study period.

Claims data were used to ascertain the CRC screening, diagnosis, receipt of surgery and hospital readmission using ICD and CPT codes. Rural-urban status was based on the Rural-Urban Commuting Area Codes and travel time between the residence and the provider facility was calculated using Google Map. For Aim 1, prevalence rates for FOBT and colonoscopy were calculated and compared using X2-test. Univariate and multivariate logistic regression analyses were performed to assess the relationship between the independent variables and CRC screening test. For Aim 2, we used Wilcoxon rank-sum tests for continuous variables and X2-tests for categorical variables and we adjusted for covariates using logistic regression. For Aim 3, Readmission and surgery status were estimated using multivariate logistic regression.

There was no significant difference between rural and rural residents in colonoscopy use. However, after adjustment, rural residents were 47% more likely to use FOBT. Patients who do not use preventive services were 2.80 more likely to present with mCRC and urban residents were 3.50 times more likely to receive mCRC. The fact that 12% of our population presents with mCRC suggests some non-compliance with screening guidelines. Therefore, we recommend removing barriers that prevent rural patients from receiving screening colonoscopy and thus increase early detection of CRC. Until these obstacles have been lessened, screening with more convenient tests is encouraged. The use of mailed FOBT test is easy and more accessible.

Included in

Epidemiology Commons

Share

COinS