SEARCH

SEARCH BY CITATION

Keywords:

  • cancer screening;
  • colorectal cancer;
  • colonoscopy;
  • program evaluation;
  • community health centers;
  • community health workers;
  • rural health

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURE
  9. REFERENCES

BACKGROUND

Colorectal cancer (CRC) is a leading cause of cancer death in the United States. Early detection through recommended screening has been shown to have favorable treatment outcomes, yet screening rates among the medically underserved and uninsured are low, particularly for rural and minority populations. This study evaluated the effectiveness of a patient navigation program that addresses individual and systemic barriers to CRC screening for patients at rural, federally qualified community health centers.

METHODS

This quasiexperimental evaluation compared low-income patients at average risk for CRC (n = 809) from 4 intervention clinics and 9 comparison clinics. We abstracted medical chart data on patient demographics, CRC history and risk factors, and CRC screening referrals and examinations. Outcomes of interest were colonoscopy referral and examination during the study period and being compliant with recommended screening guidelines at the end of the study period. We conducted multilevel logistic analyses to evaluate the program's effectiveness.

RESULTS

Patients at intervention clinics were significantly more likely than patients at comparison clinics to undergo colonoscopy screening (35% versus 7%, odds ratio = 7.9, P < .01) and be guideline-compliant on at least one CRC screening test (43% versus 11%, odds ratio = 5.9, P < .001).

CONCLUSIONS

Patient navigation, delivered through the Community Cancer Screening Program, can be an effective approach to ensure that lifesaving, preventive health screenings are provided to low-income adults in a rural setting. Cancer 2013;119:3059—3066. © 2013 American Cancer Society.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURE
  9. REFERENCES

Colorectal cancers (CRCs) are the third leading site of cancer diagnosis and death among males and females in the United States and the state of Georgia.[1] Because early detection is associated with more successful treatment and better prognosis, several national organizations have issued screening guidelines for people at average risk for developing CRC.[2-5] Screening options include tests that can prevent and detect cancer and those can detect, but not prevent, cancer.[4]

The US Preventive Services Task Force (USPSTF) recommends CRC screening using colonoscopy, flexible sigmoidoscopy, or fecal occult blood test (FOBT) for most adults aged 50 to 75 years.[2] Because detection and removal of precancerous polyps can prevent CRC, organizations including the American Cancer Society and American College of Gastroenterology recommend colonoscopy or other cancer prevention tests as the preferred CRC screening method.[3-5] Colonoscopy is the most sensitive test for detecting CRC, and promoting colonoscopy as the preferred screening method may increase the likelihood of referral and allow for greater detection of adenomatous polyps.[2, 4]

In 2010, only 58.6% of adults aged 50 to 75 years were current on any modality of CRC screening according to USPSTF guidelines.[6] Screening rates are particularly low among minorities, low-income populations, individuals who are uninsured or lack access to quality health care, and rural populations.[7-11] Barriers to colonoscopy screening include not receiving a provider referral for screening, inadequate health insurance, not having a medical home, health systems barriers (eg, scheduling challenges), logistic obstacles (eg, cost, transportation, time commitment), cognitive-emotional factors (eg, fear of procedure, disagreeable preparation, embarrassment), and lack of information about risk factors and the importance of screening.[12-16] These barriers disproportionately affect minorities, people of low socioeconomic status, and residents of rural areas.[16-18]

The patient navigation framework argues that cancer health disparities are fueled largely by a combination of cultural factors, poverty or low socioeconomic status, and social injustice.[19, 20] These powerful social determinants intersect and create barriers to screening among certain groups. Barriers exist at the level of the individual (eg, lack of health insurance), the health care provider (eg, reimbursement costs), and health care systems (eg, lack of screening facilities).[19] Racial bias in cancer care is well documented,[21-23] and the patient navigation framework suggests that it can result in physician biases against recommending particular screening services for some individuals.[19, 20]

Many of the factors this framework links to cancer disparities are found in the largely rural southwest Georgia region, which has lower CRC screening rates and higher average annual CRC incidence and death rates than the state as a whole.[24, 25] Southwest Georgia is a medically underserved region: the Health Resources and Services Administration (HRSA) has designated 31 of the region's 32 counties as Medically Underserved Areas/Populations, and 29 of the 32 counties have been designated as Health Professional Shortage Areas.[26] Southwest Georgia is largely rural and has a greater proportion of African American inhabitants than Georgia as a whole and the United States (37%, compared to 31% and 13%, respectively).[27] Educational attainment in the region is lower than the state average, and the region is also poorer than the state as a whole, with 25% of southwest Georgia residents below the poverty level, compared with 16% of Georgians.[27]

The Task Force on Community Preventive Services recommends several interventions that are effective at increasing CRC screening with FOBT: one-on-one education, client reminders, reducing structural barriers, provider assessment and feedback, and provider reminders and recall.[28, 29] However, the task force found insufficient evidence to recommend any interventions for promoting colonoscopy, which many consider to be the preferred CRC screening method.[4, 5, 30, 31] Since 2006, the Cancer Coalition of South Georgia has operated the Community Cancer Screening Program (CCSP) to increase breast, cervical, and CRC screening and informed decision-making for prostate cancer screening among uninsured and underinsured patients of federally qualified community health centers (CHCs) and other clinics. The CCSP is integrated into the clinic and supports a patient-centered medical home model of care. Although this model is widely endorsed and is included in the Patient Protection and Affordable Care Act,[32] there has been limited research about how patient-centered medical homes can best promote cancer screening.[33] The CCSP employs multiple strategies to reduce both patient- and system-level barriers to colonoscopy, including all of the strategies that the Task Force on Community Preventive Services recommends for promoting FOBT.[28, 29] Trained professional health navigators work with health centers, hospitals, and gastroenterology practices to facilitate program activities. Specifically, at each clinic, health navigators: 1) conduct chart audits to identify patients due for screening, 2) manage provider reminder systems to prompt health care providers to refer patients for screening, 3) coordinate screening and follow-up services, 4) provide one-on-one patient education and appointment reminders, 5) assist patients in overcoming barriers to screening (eg, costs, transportation, literacy), 6) ensure that the colonoscopy recall schedule, based on gastroenterologist specialist recommendation, is entered into patient charts, and 6) coordinate provider feedback on screening referral patterns (Fig. 1). Partnering gastroenterology practices provide colonoscopies at a reduced cost, which is paid entirely by the CCSP for all patients.

image

Figure 1. Flow diagram is shown for Community Cancer Screening Program patient navigator activities at community health center clinics.

Download figure to PowerPoint

At the start of the study period, the CCSP operated in 4 CHC clinics. Colonoscopies provided by the CCSP are performed by 9 gastroenterologists at 2 outpatient endoscopy centers and 1 hospital-based endoscopy center, all of which are operated by 2 Disproportionate Share Hospitals that provide follow-up testing and treatment. Since the inception of the CCSP in 2006, the program has facilitated no-cost colonoscopy screening services to more than 1000 uninsured and underinsured patients.

The purpose of this evaluation was to determine whether the CRC screening component of the CCSP is associated with increased rates of colonoscopy referral, colonoscopy examination, and CRC screening guideline compliance at CHCs. Specific hypotheses were:

  1. Intervention clinic patients who are due for CRC screening will be significantly more likely than comparison clinic patients to receive a colonoscopy referral.
  2. Intervention clinic patients who are due for CRC screening will be significantly more likely than comparison clinic patients to receive a colonoscopy examination.
  3. Intervention clinic patients will be significantly more likely than comparison clinic patients to be compliant with guidelines for CRC screening (any modality).

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURE
  9. REFERENCES

Overview

For this quasiexperimental research study, we abstracted medical records of eligible patients at intervention and comparison CHC clinics. Individuals who were aged 50 to 64 years, eligible for sliding-fee–scale services (ie, documented low-income, underinsured, or uninsured), and who visited a clinic at least once during the study period were eligible for inclusion in the study. Patients with a history of CRC, colorectal polyps, ulcerative colitis, Crohn's disease, or a first-degree relative with CRC or adenomatous polyps were considered high risk and were excluded from analysis. The simultaneous nonrandomized design (program versus no program)[34] compares 2 populations that are not randomized to condition. Over the course of the 18-month study period (November 1, 2009, to April 30, 2011), one patient population at the 4 intervention clinics was exposed to the program, and one patient population at the 9 comparison clinics was not exposed to the program. This study was approved by the Emory University Institutional Review Board.

Sample and Setting

There are 4 CHCs in southwest Georgia; each CHC includes multiple clinics, for a total of 13 CHC clinics in the region. All 13 clinics participated in the study, and the 4 clinics where the CCSP was present during the study period comprised the intervention group. Because of changes to their record-keeping system, one comparison clinic was not able to provide information about patients seen during the last 2 months of the study period.

Power calculations were estimated to account for group-level intervention,[35] assuming a baseline CRC screening prevalence of 46%[24] and an intraclass correlation coefficient of between 0.01 and 0.05, as is commonly reported for trials of CRC screening studies where the group is the clinic.[36] With a sample size of just 542 participants, this study would have 80% power to detect a minimum 11% increase in screening prevalence, which is a very conservative estimate of intervention effect, based on results of similar studies.[37-39] The CHCs generated a list of all eligible patients (n = 3773), and we randomly sampled 350 patients from intervention clinics (27.6%) and 625 patients from comparison clinics (24.9%). Seventy patients did not meet eligibility criteria, and 96 were excluded from analysis because they were at high risk for CRC, leaving a total sample of 809.

Data Collection Procedures

Two CHCs, consisting of 8 clinics, had electronic medical records (EMRs); in these clinics, abstractors primarily relied on EMRs and referred to paper charts as needed when the EMR was incomplete. Two CHCs, consisting of 5 clinics, had paper medical charts. All 4 intervention sites had EMRs; 4 of 9 comparison sites had EMRs. A total of 7 trained data abstractors used an abstraction form to record patient demographics, CRC history and risk factors, colonoscopy screening and referral (2 most recent colonoscopies and 5 most recent referrals), sigmoidoscopy screening and referral (4 most recent sigmoidoscopies and 5 most recent referrals), and blood stool screening (4 most recent blood stool tests). For quality control, we double abstracted fields related to our primary outcomes (colonoscopy screening and referral) for all records; 10% of all patient records were fully double abstracted. The data were double entered by 2 research assistants to identify data entry errors.

Measures

Patients who did not have a colonoscopy between May 1, 2001, and October 31, 2009, were considered “due for a colonoscopy” during the study period (n = 767, 94.8%). Among patients due for a colonoscopy, those who had documented evidence of a colonoscopy referral or examination during the study period were considered to have received a “colonoscopy referral” (n = 271, 35.3%), and those who had documented evidence of a colonoscopy examination were considered to have received a “colonoscopy examination” (n = 123, 16.0%). Patients who met at least one of the following conditions at the end of the study period were considered “guideline-compliant” on any screening modality (n = 179, 22.1%): colonoscopy in the prior 10 years, sigmoidoscopy in the prior 5 years, or blood stool test in the prior year.

Statistical Analysis

We initially examined frequency distributions of all relevant categorical variables and performed descriptive statistics on one continuous variable (age). Next, we examined each demographic variable (age, sex, race, and marital status) in relation to study condition (intervention versus comparison) and each of the 3 outcomes (colonoscopy referral, colonoscopy examination, and guideline-compliant). Results indicated that race was associated with study condition such that blacks were more likely to be in the intervention group (Table 1), receive a referral for colonoscopy, undergo colonoscopy screening, and be guideline-compliant (all P values ≤ .01). In addition, older patients tended to be less guideline-compliant as compared with younger patients (P ≤ .05). Thus, we conservatively opted to control for race and age in all 3 multivariate models.

Table 1. Sample Demographic Characteristics
 TotalInterventionComparisonP Value
Characteristicn (%)n (%)n (%)
  1. a

    Sex, n = 799 (10 missing); race, n = 780 (13 missing, 16 other); marital status, n = 794 (15 missing).

Age, y809289520.842
50-54348 (43.0)124 (42.9)224 (43.1) 
55-59282 (34.9)98 (33.9)184 (35.4) 
60-64179 (22.1)67 (23.2)112 (21.5) 
Sexa799288511.363
Male263 (32.9)89 (30.9)174 (34.1) 
Female536 (67.1)199 (69.1)337 (65.9) 
Racea780282498<.001
Black491 (62.9)213 (75.5)278 (55.8) 
White289 (37.1)69 (24.5)220 (44.2) 
Marital statusa794282512.733
Not married267 (33.6)100 (35.5)167 (32.6) 
Married266 (33.5)93 (33.0)173 (33.8) 
Divorced/separated204 (25.7)72 (25.5)132 (25.8) 
Widowed57 (7.2)17 (6.0)40 (7.8) 
Insurance status8092895201.000
Sliding-fee–scale eligible809 (100)289 (100)520 (100) 
Colonoscopy screening status    
Due for colonoscopy during the study period767 (94.8)257 (88.9)510 (98.1)<.001
Had colonoscopy referral during study271 (35.3)149 (58.0)122 (23.9)<.001
Had colonoscopy exam during study123 (16.0)90 (35.0)33 (6.5)<.001
Guideline-compliant at end of study179 (22.1)123 (42.6)56 (10.8)<.001

To investigate the effectiveness of the intervention while accounting for the clustering of patients within the 13 clinics, we conducted multilevel analyses using logistic models.[40, 41] As stated above, colonoscopy referral, colonoscopy examination, and guideline-compliance were the outcome variables of interest. For each analysis, we first investigated the unconditional model to calculate the intraclass correlation coefficient. Then, we added the intervention variable to the model (Model 2). Finally, we controlled for age and race (Model 3). Level-2 predictors beyond the intervention condition were not included due to the small number of clusters. The full model tested was:

  • display math

The mixed equation was:

  • display math

where i indicates the individual and j the clinic. Residuals of the final model were normally distributed.[42] To address the possibility of differences between clinics with EMRs and paper charts (eg, that EMRs may be set up to prompt providers to make screening referrals or may make it easier to systematically record and locate screening data), we repeated the multilevel analyses with the subset of patients who have EMRs (n = 615). Descriptive analyses were conducted using SPSS version 19.0. All multilevel analyses were conducted using HLM7.[43] An alpha of .05 was used to establish statistical significance.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURE
  9. REFERENCES

Table 1 presents the sample demographic characteristics for the 809 study patients determined to be at average CRC risk. They ranged in age from 50 to 64 years (mean= 55.8, median = 55, standard deviation = 4.1). Patients tended to be black and female. There were no significant differences in age, sex, marital status, or insurance status between intervention and comparison clinic patients. Intervention clinics had a higher percentage of black patients than did comparison clinics (76%, as compared to 56%, P < .001). Analyses of the colonoscopy referral and colonoscopy examination outcomes are restricted to the 767 patients who were due for a colonoscopy during the study period. Analyses of the guideline-compliance outcome include the full sample of 809 patients. Among intervention clinic patients due for a colonoscopy, 58% received a referral and 35% received an examination; among comparison clinic patients, 24% received a referral and 7% received an examination (all P values < .001). At the end of the study, 43% of intervention clinic patients and 11% of comparison clinic patients were guideline-compliant (P < .001).

The multilevel analyses indicate that intervention patients were almost 5 times more likely to receive a colonoscopy referral, 8 times more likely to undergo a colonoscopy examination, and 6 times more likely to be guideline-compliant than comparison patients while controlling for age and race (Table 2). For the subset of patients with EMRs, intervention patients were not significantly more likely to receive a colonoscopy referral than comparison clinic patients. However, intervention patients were approximately 5 times more likely to receive a colonoscopy examination and to be guideline-compliant than comparison patients while controlling for age and race (Table 3).

Table 2. Hierarchical Linear Modeling of the Association Between Study Condition and Each of 3 Outcomes (Colonoscopy Referral, Colonoscopy Exam, and Guideline-Compliant on at Least One Screening Modality)
 Colonoscopy ReferralColonoscopy ExaminationGuideline-Compliant
  1. a

    P < .05.

  2. b

    P < .01.

  3. c

    P < .001.

  4. d

    The intraclass correlation coefficient (ICC) is the variance between the clusters (clinics) divided by the total variance, ie, it is the proportion of variance that can be explained by differences between clusters. For example, an ICC of 20% indicates that 20% of the variance can be explained by differences between clinics.

Fixed effects   
Level 1   
Intercept0.2820.0241a0.0097b
Age (continuous)0.9961.02791.0431
Race (African American = 1)1.3071.15621.2281
Level 2   
Treatment condition (intervention = 1)4.807a7.9079b5.9045c
Random effects   
τ00 (Intercept)0.738c0.7949c0.0823
Reduction in variance and goodness of fit
Reduction in τ0036.5%61.5%90.9%
Deviance2030.92226.32277.7
ICC (unconditional model)d26.1%68.6%21.5%
Table 3. Hierarchical Linear Modeling of the Association Between Study Condition and Each of 3 Outcomes (Colonoscopy Referral, Colonoscopy Examination, and Guideline Compliant on at Least One Screening Modality) for Patients With Electronic Medical Records
 Colonoscopy ReferralColonoscopy ExaminationGuideline-Compliant
  1. a

    P < .05.

  2. b

    P < .01.

  3. c

    P< .001.

Fixed effects   
Level 1   
Intercept0.3420.0390.016a
Age (continuous)1.0061.0121.037
Race (African American = 1)1.2751.1861.223
Level 2   
Treatment condition (Intervention = 1)2.3044.751a5.060b
Random effects   
τ00 (Intercept)0.248c0.429c0.110a
Reduction in variance and goodness of fit
Reduction in τ0038.0%61.0%88.5%
Deviance1628.11589.61706.3
Intraclass correlation coefficient (unconditional model)10.9%10.9%10.9%

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURE
  9. REFERENCES

The CCSP is a novel intervention that uses professional patient navigators to implement multiple strategies to address structural and individual barriers to CRC screening for uninsured and underinsured rural populations. We conducted a quasiexperimental evaluation of the program to determine whether the program is associated with increased rates of colonoscopy referral, colonoscopy examination, and CRC screening guideline compliance. Intervention patients were significantly more likely than comparison patients to undergo a colonoscopy examination and to be guideline-compliant at the end of the study. Among all average-risk patients who were due for CRC screening, intervention patients were significantly more likely than comparison patients to receive a colonoscopy referral; however, this difference was not significant when analyses were limited to the subset of patients with EMRs. It is noteworthy that blacks in our study were more likely to be in the intervention group and to have positive screening outcomes. The CCSP is intended to reach low-income individuals, which in Georgia includes a disproportionate share of blacks.[44]

Research suggests that both individual and structural barriers influence CRC screening rates.[12-17] Patient navigation offers an approach for addressing the multiple barriers that contribute to health disparities by preventing some individuals from accessing high-quality cancer services.[19, 20] With the notable exception of the National Cancer Institute's Patient Navigation Research Program (PNRP),[45] there has been a lack of systematic evaluation of patient navigation programs, especially in rural areas. Despite well-documented urban/rural cancer disparities,[11, 46] the vast majority of the patient navigation programs described in the literature that promoted CRC screening were in urban settings. The 9 sites funded through the PNRP are mostly urban, and only one includes CRC screening promotion for a rural population.[45]

There is a clear and compelling need to increase the evidence base of interventions that are effective at increasing colonoscopy screening, particularly in rural settings.[28, 29] Further research about how patient-centered medical homes can promote cancer screening is also needed.[33] The CCSP is a multicomponent patient navigation intervention that promotes colonoscopy in rural CHCs and supports a patient-centered medical home model. The CCSP includes strategies to address both system-level and patient-level barriers to screening. The program offers an innovative way of dealing with financial barriers to screening by partnering with gastroenterology practices to provide low-cost colonoscopies, paid for by the CCSP. The program also covers the cost of preparation materials and transportation, as needed. These types of financial supports are rarely used in programs described in the research literature.

Findings from this study should be considered in light of several important limitations. Neither clinics nor patients were randomized to condition, and it is possible that the program was implemented in higher capacity sites, which would have had higher screening rates in the absence of the program. CCSP patient navigators seek to improve how clinics record screening information in the medical record, so it is possible that more colonoscopies were documented at intervention sites relative to comparison sites because this aligns with the purpose of the CCSP. All CHCs in southwest Georgia have multiple clinics, and 2 of the 4 CHCs in our sample had at least one clinic receive the intervention and one clinic in the comparison group. It is possible that the presence of the CCSP within their CHC may have influenced provider referrals at some comparison clinics. Some providers referred patients to intervention clinics specifically for a free colonoscopy through the CCSP. At least 17 comparison patients in the study were sent to intervention clinics where they received colonoscopy referral and/or screening. For these patients, we conservatively included only data from their home clinic: all referrals and examinations noted in the medical record at the comparison clinic were attached to that clinic, even if the CCSP provided the screening. These providers' efforts point to the need for increased availability of screening services for low-income patients in rural settings. Furthermore, we did not distinguish between screening and diagnostic examinations. Many of the screenings done at the comparison clinics appear to have been diagnostic; our effect size may have been larger had we limited the study to screening examinations. Because the program was in place at some intervention clinics before the study period began, it is possible that increases in screenings resulting from the program may have identified people who were positive for CRC risk factors, thus causing those individuals to be excluded from our study.

An evaluation of the cost-effectiveness of the CCSP was beyond the scope of this study; however, such information is of critical importance. Future research should explore the effectiveness of specific patient navigation intervention components for promoting colonoscopy. The CCSP is a multicomponent intervention with multiple strategies. It may be possible to achieve a similar level of impact, or a more cost-effective impact, using a subset of the program's strategies, or the program's impact may be the synergistic result of all its components. Information about how such programs work is particularly timely because the nation's health care system is in a “teachable moment,” moving forward with implementation of health care reform under the Patient Protection and Affordable Care Act.[47]

Patients who received navigation services through CCSP were significantly more likely than those who did not to have a colonoscopy examination and be current on any recommended CRC screening at end of the study. Patient navigation, delivered through the CCSP, can be an effective approach to promote adherence to screening referrals and to ensure that lifesaving, preventive health screenings (colonoscopies) are provided to low-income adults at average risk for CRC.

FUNDING SOURCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURE
  9. REFERENCES

This research was supported by Cooperative Agreement Number 1U48DP0010909-01-1 from the Centers for Disease Control and Prevention (CDC) and the National Cancer Institute (NCI). The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the CDC or NCI.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURE
  9. REFERENCES
  • 1
    American Cancer Society. Cancer Facts & Figures 2011. Atlanta, GA: American Cancer Society; 2011.
  • 2
    US Preventive Services Task Force. The Guide to Clinical Preventive Services: Recommendations of the US Preventive Services Task Force. AHRQ Publication 10-05145. Rockville, MD: Agency for Healthcare Research and Quality; 2010.
  • 3
    American Cancer Society. Colorectal Cancer Early Detection. http://www.cancer.org/acs/groups/cid/documents/webcontent/003170-pdf.pdf. Accessed October 16, 2012. Last revised January 24, 2013.
  • 4
    Rex DK, Johnson DA, Anderson JC, Schoenfeld PS, Burke CA, Inadomi JM; American College of Gastroenterology. American College of Gastroenterology guidelines for colorectal cancer screening 2009 [corrected]. Am J Gastroenterol. 2009;104:739-750.
  • 5
    National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology: Colorectal Cancer Screening. 2012; version 2.2012. http://www.nccn.org/professionals/physician_gls/f_guidelines.asp#detection. Accessed October 16, 2012.
  • 6
    Centers for Disease Control and Prevention (CDC). Cancer screening - United States, 2010. MMWR Morb Mortal Wkly Rep. 2012;61:41-45.
  • 7
    O'Malley AS, Forrest CB, Feng S, Mandelblatt J. Disparities despite coverage: gaps in colorectal cancer screening among Medicare beneficiaries. Arch Intern Med. 2005;165:2129-2135.
  • 8
    James TM, Greiner KA, Ellerbeck EF, Feng C, Ahluwalia JS. Disparities in colorectal cancer screening: a guideline-based analysis of adherence. Ethn Dis. 2006;16:228-233.
  • 9
    Jerant AF, Fenton JJ, Franks P. Determinants of racial/ethnic colorectal cancer screening disparities. Arch Intern Med. 2008;168:1317-1324.
  • 10
    Rim SH, Joseph DA, Steele CB, Thompson TD, Seeff LC; Centers for Disease Control and Prevention (CDC). Colorectal cancer screening - United States, 2002, 2004, 2006, and 2008. MMWR Surv Summ. 2011;60(suppl):42-46.
  • 11
    Palmer RC, Schneider EC. Social disparities across the continuum of colorectal cancer: a systematic review. Cancer Causes Control. 2005;16:55-61.
  • 12
    Robinson CM, Cassells AN, Greene MA, Beach ML, Tobin JN, Dietrich AJ. Barriers to colorectal cancer screening among publicly insured urban women: no knowledge of tests and no clinician recommendation. J Natl Med Assoc. 2011;103:746-753.
  • 13
    Robinson JM, Shavers V. The role of health insurance coverage in cancer screening utilization. J Health Care Poor Underserved. 2008;19:842-856.
  • 14
    Griffith KA, Passmore SR, Smith D, Wenzel J. African Americans with a family history of colorectal cancer: barriers and facilitators to screening. Oncol Nurs Forum. 2012;39:299-306.
  • 15
    Denberg TD, Melhado TV, Coombes JM, et al. Predictors of nonadherence to screening colonoscopy. J Gen Intern Med. 2005;20:989-995.
  • 16
    Green AR, Peters-Lewis A, Percac-Lima S, et al. Barriers to screening colonoscopy for low-income Latino and white patients in an urban community health center. J Gen Intern Med. 2008;23:834-840.
  • 17
    Gimeno García AZ. Factors influencing colorectal cancer screening participation. Gastroenterol Res Pract. 2012;2012:483417.
  • 18
    McAlearney AS, Reeves KW, Dickinson SL, et al. Racial differences in colorectal cancer screening practices and knowledge within a low-income population. Cancer. 2008;112:391-398.
  • 19
    Freeman HP, Chu KC. Determinants of cancer disparities: barriers to cancer screening, diagnosis, and treatment. Surg Oncol Clin N Am. 2005;14:655-669.
  • 20
    Freeman HP. Commentary on the meaning of race in science and society. Cancer Epidemiol Biomarkers Prev. 2003;12:232s-236s.
  • 21
    Mayberry RM, Mili F, Ofili E. Racial and ethnic differences in access to medical care. Med Care Res Rev. 2000;57(suppl 1):108-145.
  • 22
    Doubeni CA, Laiyemo AO, Klabunde CN, Young AC, Field TS, Fletcher RH. Racial and ethnic trends of colorectal cancer screening among Medicare enrollees. Am J Prev Med. 2010;38:184-191.
  • 23
    Smedley BD, Stith AY, Nelson AR, eds. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, DC: The National Academies Press; 2003.
  • 24
    Georgia Division of Public Health. OASIS-Online Analytical Statistical Information System. Page version 3.0. http://oasis.state.ga.us/. Accessed May 12, 2011.
  • 25
    National Cancer Institute. State Cancer Profiles. Incidence Rates Report. 2011; http://statecancerprofiles.cancer.gov/incidencerates/. Accessed August 13, 2012.
  • 26
    US Department of Health and Human Services. Shortage Designation: Health Professional Shortage Areas & Medically Underserved Areas/Populations. Health Resources and Service Administration Web site. http://www.hrsa.gov/shortage/. Accessed September 12, 2012.
  • 27
    US Census Bureau. State & County QuickFacts. http://quickfacts.census.gov/qfd/index.html. Accessed September 9, 2012.
  • 28
    Community Preventive Services Task Force. Updated recommendations for client- and provider-oriented interventions to increase breast, cervical, and colorectal cancer screening. Am J Prev Med. 2012;43:92-96.
  • 29
    Sabatino SA, Lawrence B, Elder R, et al; Community Preventive Services Task Force. Effectiveness of interventions to increase screening for breast, cervical, and colorectal cancers: nine updated systematic reviews for the guide to community preventive services. Am J Prev Med. 2012;43:97-118.
  • 30
    Christie J, Itzkowitz S, Lihau-Nkanza I, Castillo A, Redd W, Jandorf L. A randomized controlled trial using patient navigation to increase colonoscopy screening among low-income minorities. J Natl Medical Assoc. 2008;100:278-284.
  • 31
    Sonnenberg A, Delco F, Inadomi JM. Cost-effectiveness of colonoscopy in screening for colorectal cancer. Ann Intern Med. 2000;133:573-584.
  • 32
    Kocher R, Emanuel EJ, DeParle NA. The Affordable Care Act and the future of clinical medicine: the opportunities and challenges. Ann Intern Med. 2010;153:536-539.
  • 33
    Sarfaty M, Wender R, Smith R. Promoting cancer screening within the patient centered medical home. CA Cancer J Clin. 2011;61:397-408.
  • 34
    Gordis L. Epidemiology. Philadelphia, PA: W.B. Saunders Company; 1996.
  • 35
    Campbell MK, Thomson S, Ramsay CR, MacLennan GS, Grimshaw JM. Sample size calculator for cluster randomized trials. Comput Biol Med. 2004;34:113-125.
  • 36
    Hade EM, Murray DM, Pennell ML, et al. Intraclass correlation estimates for cancer screening outcomes: estimates and applications in the design of group-randomized cancer screening studies. J Natl Cancer Inst Monogr. 2010;2010:97-103.
  • 37
    Lane DS, Messina CR, Cavanagh MF, Chen JJ. A provider intervention to improve colorectal cancer screening in county health centers. Med Care. 2008;46(9 suppl 1):S109-S116.
  • 38
    Lasser KE, Murillo J, Medlin E, et al. A multilevel intervention to promote colorectal cancer screening among community health center patients: results of a pilot study. BMC Fam Pract. 2009;10:37.
  • 39
    Chen LA, Santos S, Jandorf L, et al. A program to enhance completion of screening colonoscopy among urban minorities. Clin Gastroenterol Hepatol. 2008;6:443-450.
  • 40
    Raudenbush SW, Bryk AS. Hierarchical Linear Models: Applications and Data Analysis Methods. 2nd ed. Thousand Oaks, CA: Sage; 2002.
  • 41
    Luke DA. Multilevel Modeling. Thousand Oaks, CA: Sage; 2004.
  • 42
    Singer JD, Willet JB. Applied Longitudinal Data Analysis: Modeling Change and Event Occurrences. New York, NY: Oxford University; 2003.
  • 43
    Raudenbush SW, Bryk AS, Cheong YF, Congdon RT, du Toit M. HLM 7: Linear and Nonlinear Modeling. Lincolnwood, IL: Scientific Software international, Inc.; 2011.
  • 44
    Kaiser Family Foundation. Poverty Rate by Race/Ethnicity, states (2010-2011), U.S. (2011). http://www.statehealthfacts.org/comparebar.jsp?ind=14&cat=1. Accessed October 16, 2012.
  • 45
    National Cancer Institute. Patient Navigation Research Program (PNRP), Center to Reduce Cancer Health Disparities Web site. http://crchd.cancer.gov/pnp/pnrp-index.html. Accessed August 27, 2012.
  • 46
    Liff JM, Chow WH, Greenberg RS. Rural-urban differences in stage at diagnosis. Possible relationship to cancer screening. Cancer. 1991;67:1454-1459.
  • 47
    Larson EB, Reid R. The patient-centered medical home movement: why now? JAMA. 2010;303:1644-1645.