National Cancer InstituteU.S. National Institutes of
Center to Reduce Cancer Health Disparities logo

2010 AACR Science of Cancer Health Disparities Meeting: Abstract Highlights

3rd AACR Conference on the Science of Cancer Health disparities American Association for Cancer Research (AACR)

2010 AACR Science of Cancer Health Disparities Meeting
September 30-October 3, 2010
Loews Miami Beach Hotel
Miami, FL

Session 1
Session 2
Poster Session A
Poster Session B

Many CRCHD grantees presented at the 3rd Annual AACR Science of Cancer Health Disparities Meeting. Below you will find highlights from the poster sessions.

For a complete list of abstracts from the conference, please click here.

PR-2 - Within-group difference sbetween native-born and foreign-born black men on prostate cancer risk reduction and early detection practices
Folakemi Odedina, Getachew Dagne, Margareth Larose-Pierre3, John Scrivens, Frank Emanuel, Angela Adams, Shannon Pressey, Oladapo Odedina. University of Florida, Gainesville, FL; University of South Florida, Tampa, FL; Florida A&M University, Tallahassee, FL; Central Florida Pharmacy Council, Oviedo, FL; Florida Black Living Navigator (FBLN), Tampa, FL.

Background: The Healthy Immigrant Effect (HIE) phenomenon proposes that: (1) when immigrants arrive in the U.S., they are healthier than their locally born counterparts; and that (2) this health advantage dissipates over time.While this phenomenon has been well studied and documented among Asians and Hispanics, few studies have explored the HIE among blacks. Most research studies on prostate cancer (CaP) health disparities have focused on comparing Black men to other ethnic groups within the U.S. A research area that has been overlooked and understudied is the within-group differences among black men.We investigated the ethnic variations among native-born black men, Africanborn black men, and Caribbean-born black men on: (1) personal history of CaP; (2) discussion of CaP risk reduction and early detection with physician; (3) screening by digital rectal examination (DRE); (4) screening by prostate specific antigen (PSA); (5) CaP knowledge; (6) diet; (7) use of chemoprevention products; and (8) physical activity.

Methods: This study was part of the Florida Prostate Cancer Disparity Project referred to as the PIPCaD project. The inclusion criteria were black men (regardless of country of origin) who were 35 years and older. Data collection took place primarily in five Florida counties. Using a previously validated survey, data were collected from over 3,400 black men. A one-way ANOVA analyses were carried out to determine if there were significant differences among the three groups on outcome variables.

Results: Of the 3,410 responses received, 3,040 indicated their ethnicity: 2,405 were native-born black men, 315 were African-born black men, and 320 were Caribbean-born black men.We found significant differences on CaP knowledge level (F{2, 3016} = 13.43, p<.001), vegetable consumption (F{2, 3016} = 13.43, p<.001), use of chemoprevention products (F{2, 3016} = 13.43, p < .001). Post hoc tests revealed that native-born men had significantly higher CaP knowledge compared to African-born and Caribbean-born men (p<.05). Chi-square test results comparing the three groups were significant for frequencies of meat consumption (?2{2} = 45.2845, p <.001), discussion of CaP riskreduction and early detection with a doctor (?2{2} = 3.81, p =.022), and insurance status (?2{2} = 8.5071, p <.014). Relative to diet, significantly more native-born men reported meat products make up the biggest portion of their meal. Over 70% of native-born men reported eating meat at least four to six times a week compared to 64% African-born men and 62% Caribbean-born men. Although African-born men were the least insured of the three groups, more men in this group reported that they discuss CaP risk-reduction and early detection with a physician every chance they get compared to native born and Caribbean-born men. In addition, the frequency of chemoprevention use was highest among African-born men.

Discussion: Given the significant number of foreign-born blacks in the U.S., it is important to disaggregate the data of native-born and foreign-born blacks to identify knowledge gaps and develop effective programs and policies to address disparities.

PR-7 - A church-based hepatitis B intervention for Korean Americans.
Grace X. Ma, Yin Tan,Wanzhen Gao, Trong Tram, Joanne Rhee1, Wang G. Chae. Center for Asian Health, Temple University, Philadelphia, PA; Asian Community Health Coalition, Philadelphia, PA.

Introduction: Despite the proven efficacy of hepatitis B screening and vaccination in preventing hepatitis B infection, most Korean Americans are neither screened nor vaccinated. Hepatitis Brelated liver cancer incidence in Korean Americans is considerably higher than that of other ethnic groups in the U.S. About 80% of Korean Americans are active church members, which is an important venue for health education and intervention programs. Community-based participatory research (CBPR) approaches can stimulate partnership development and community engagement especially among hard-toreach Asian American communities. The purpose of this study was to design and pilot test a culturally appropriate hepatitis B screening and vaccination program for Korean Americans using CBPR approaches.

Methods: A quasi-experimental design with intervention and control groups was used in the study. Participants (n=330), age 18 and over, were recruited from four Korean churches (two for intervention and two for control group. The intervention group received bilingual community health educator-led hepatitis B education and patient navigation. The control group received general health education. Assessments were obtained at baseline, post-education, and 6-month follow up. Primary outcome measures were hepatitis B screening and vaccination behavior change.

Results: Participants of intervention group reported significantly higher hepatitis B screening and vaccine. The data of intervention group also suggested significant improvement in knowledge of hepatitis B and measures of Health Belief Model.

Conclusions/Discussion: This study suggested that use CBPR principles to engage churches/pastors and cultural appropriate intervention can be an effective approach to address hepatitis B disparities in Korean Americans.

A1 - NCI CRCHD Geographic Management Program (GMaP): Catalyzing inter-institutional collaboration for enhanced cancer health disparity (CHD) research and outreach.
Steven R. Patierno, Peter Shields, Sarah Hijaz, Sarah Reisinger. George Washington University Cancer Institute,Washington, DC; Georgetown University Lombardi Comprehensive Cancer Center,Washington, DC.

In 2009 the NCI Center to Reduce Cancer Health Disparities launched its Geographic Management Program (GMaP) initiative, which groups cancer health disparity (CHD) research/outreach efforts according to geographic region. Region 1 (DC, MD, VA) includes: Bowie State University, Georgetown University, George Washington University, Hampton University, Howard University, Johns Hopkins University, Morgan State University, University of the District of Columbia, University of Maryland-Baltimore, University of Maryland-Eastern Shore, University of Virginia,Washington Hospital Center, Virginia Commonwealth University.

The George Washington University Cancer Institute (GWCI) was selected as the Region 1 coordinating hub. Region 1 also includes additional funding for BMaP which is focused on Biospecimen Banking for CHD research. The Georgetown Lombardi Cancer Center is the BMaP coordinating hub.

The GMaP goal is to create a regional network wherein resources can be pooled to facilitate progress through inter-institutional collaborations with large-scale funding. There are six regions across the country.

A vital aspect of the Region 1 infrastructure are its Core/Elective Groups, which each have the goal of assessing research/outreach potential in the following fields: Administration and Communication, Training, Bioinformatics, Biospecimen Banking (BMaP), Clinical Trials, Community Translation, Emerging Technology, Evaluation, Social and Behavioral Oncology.

To begin cataloging institutional resources, two survey tools were circulated to Region 1 institutions. The first is the NCI's Readiness Assessment Tool (RAT), which collects institutional data on: communication abilities, bioinformatics, biospecimen banking, clinical trials, emerging technology, training programs. Region 1 institutional data has been organized and submitted to the NCI.

The second survey tool, the Resource Assessment Database (RAD), is Region 1-specific.While the RAD aims to obtain a more thorough understanding of institutional resources, including community translation and social and behavioral oncology programs, it crucially is also meant to identify cooperation potential. The RAD is considered a "living document" that is to be modified as GMaP progresses in order to best capture developing research/outreach data interests. The first RAD draft was circulated to all participants, and the resulting data is being organized. All RAT/RAD data will be used to assess and plan collaborative efforts within the Region.

The BMaP Group also surveyed three Region 1 institutions on metabolomics specimens by type and race. Further information on tissue type, disease state, was culled for specimens located at Georgetown University. BMaP also plans to host a meeting to discuss the feasibility and interest of a virtual repository for CHD biospecimens. Specific discussions will focus on shared access, universal consents, logistics, as well as institutional and other barriers.

Leaders from Region 1 participants have met to initiate more detailed discussions of potential inter-institutional collaborative research/outreach projects. The RAT/RAD data submitted thus far indicates a wide variety of resources that can be brought together to meet its goal of addressing CHDs as a whole.

Funded by grant NIH 5U01-CA-116937.

A64 - Review on the application of laser thermal therapy for palliation of recurrent head and neck cancer.
Marcos B. Paiva, Michael Bublik, Marcio Abrahao, Joel Sercarz1. University of California, Los Angeles, CA; University of Miami, Miami, FL; Federal University of Sao Paulo, Sao Paulo, Brazil.

Background: The number of new cases of head and neck cancers in the United States was 40,490 in 2006, accounting for about 3% of adult malignancies. 11,170 patients died of their disease in 2006. The worldwide incidence exceeds half a million cases annually. In North America and Europe, the tumors usually arise from the oral cavity, oropharynx, or larynx, whereas nasopharyngeal cancer is more common in the Mediterranean countries and in the Far East. In the United States, African Americans are disproportionately affected by head and neck cancer, with younger ages of incidence, increased mortality, and more advanced disease at presentation.

Objective: To review the outcomes of a phase II study using laserinduced thermal therapy (LITT) as a palliative treatment for 106 patients with recurrent head and neck tumors.

Study Design: Retrospective study. Setting: Tertiary hospital in the United States. Subjects And Methods: The primary endpoints were tumor response and survival. Prognostic values were assessed by the Kaplan- Meier method.

Results: The best results were seen in oral cavity tumors, in which mean survival was 29.1 months, as compared to neck tumors (mean 14.4 +/- 6.9 months; range 7.5-20.7 months; with a 95% confidence interval). Further analysis showed that clinical factors such as gender, smoking, and alcohol use were not indicators of poor prognosis, whereas neck disease and tumor stage at first treatment were relevant factors.

Conclusion: In this study, 40 out of 106 patients treated by LITT remained alive at the end of our follow-up, and a complete response was seen in 24 (22.6%) patients. The highest response rate was seen in oral cavity tumors, which suggests that tumor location at this site may be a predictor of favorable outcome with LITT.

A96 - Breast tumor subtypes and survival outcomes in African American and Latina patients in South Los Angeles.
Jaydutt V. Vadgama, Yanyuan Wu, Marianna Sarkissyan, Yayha Elshimali, Hezla Mohamed, Sheila Clayton, Dennis Slamon, H Phillip Koeffler. Charles Drew University of Medicine & Science/David Geffen UCLA School of Medicine, Los Angeles, CA; Charles Drew University of Medicine & Science, Los Angeles, CA; David Geffen UCLA School of Medicine, Los Angeles, CA; Cedars Sinai Medical Center/David Geffen UCLA School of Medicine, Los Angeles, CA.

Background: Recent studies have shown that minority women, in particular African American are more likely to express Triple Negative (TN) tumors compared to Non-Hispanic Whites. TN tumors are more difficult to treat this may be one of the factors contributing to disparities in disease outcome and overall survival. In our study, we have studied a cohort of African American and Latina women with equal socio-economic status (SES) and access to care. The purpose of this study was to determine if these two ethnic groups with equal SES and access to care had differences in disease outcome and if this difference was related to tumor subtypes.We have also examined cellular and molecular markers associated with tumor subtypes and disease outcome.

Methods: A total of 300 subjects were retrospectively selected. 156 were African American and 144 Latina. Demographic, pathologic, and clinical follow-up information were collected from existing database. The prevalence of breast cancer subtypes within racial and menopausal subsets was examined and the associations with tumor size, lymph nodal status, staging, and histologic grade were determined by logistic regression. Disease-free survival (DFS) was analyzed by Kaplan-Meier survival curves with log-rank test. Gene and protein signatures associated with PI3kinase/Akt and the CD44/CD24 expression profiles were analyzed by immunohistochemistry analysis (IHC) in different tumor subtypes.

Results: The TN or basal-like breast cancer subtype (ER/PR/HER2 negative) was more prevalent among premenopausal African American women (36%) compared to postmenopausal African Americans (28.1); premenopausal Latina women (31%); and postmenopausal Latinas (17%) (p=0.05). The HER2+/ER- subtype was more prevalent among premenopausal Latina women (17%) than postmenopausal Latinas (14%), postmenopausal African American women (9%), and premenopausal African American women (3%) (p=0.046). Prevalence of Luminal A and Luminal B breast cancer subtypes were similar with race and menopausal status. Compared to Luminal A, basal-like tumors were more poorly differentiated (OR=6.4, p less than 0.001) and more likely to have CD44+/CD24- phenotype, while HER2+/ER- tumors were more associated with large tumor size (OR=3.1, p=0.01) and CD24+ phenotype. The Luminal B tumors were also more likely to be associated with larger tumor size and were poorly differentiated (OR=3.0, p=0.002). Patients with basal-like and HER2+/ER- tumors had significant shorter DFS compared to patients with Luminal A tumors (log rank=11.97, p=0.0005 and log rank=8.95, p=0.002). The DFS was also reduced significantly in patients with Luminal B tumors than Luminal A tumors (log rank=7.09, p=0.007). The poor disease outcome in those patients may due to have increased plasma IGF-I, activated tissue pAkt and loss of functional FOXO1A.

Conclusions: Our study suggests a highest prevalence of basallike tumors (triple negative) in premenopausal African American women, followed by premenopausal Latina women. Premenopausal Latina women also had a higher prevalence of HER2+/ER- tumors. DFS was poor in both ethnic groups with either basal like tumors or those with HER2+/ER- tumors. Activation of pAkt in the tumor tissue was a significant contributor to the poor disease outcome.

A100 - Alcohol promotes breast cancer cell invasion by regulating the Nm23-ITGA5 pathway.
Nomeli P. Nunez. The University of Texas at Austin, Austin, TX.

Introduction: Alcohol consumption increases breast cancer risk in women. Additionally, alcohol increases the metastatic ability of cancer cells. Metastasis suppressor genes encode proteins that have the ability to slow down or prevent the invasiveness of cancer cells, thus inhibiting the establishment of metastases.We analyzed the expression level of metastasis suppressor genes to determine the mechanisms by which alcohol increases the invasive ability of breast cancer cells.

Methods: The invasiveness, or metastatic ability, of breast cancer cells was determined by the Boyden assay. To determine if alcohol increases invasion of breast cancer cells via the metastasis suppressor gene Nm23, we over-expressed or down-regulated Nm23 using siRNAmediated knockdown; subsequently, we determined the effects of alcohol on the invasiveness of these cancer cells.

Results: Results show that alcohol increased breast cancer cell invasion in a dose-dependent manner. Alcohol significantly decreased the expression of Nm23. Overexpression of Nm23 in breast cancer cells suppressed the effects of alcohol on the invasive ability of the cancer cells. Also, exposure of cancer cells to alcohol led to an increase in the expression level of the fibronectin receptor subunit ITGA5, and overexpression of Nm23 blocked the effects of alcohol. Further, the invasiveness of the breast cancer cells was inhibited when ITGA5 was knocked down by siRNA, and alcohol was unable to increase the invasiveness of the cancer cells when ITGA5 was knocked down.

Conclusions: Results suggest that alcohol increases the invasive ability of breast cancer cells by down-regulating the metastatic suppressor Nm23 gene, which leads to increased expression of ITGA5. Consequently, up-regulated ITGA5 may increase the invasive ability of the cancer cells.

B5 - Mexican-American and Puerto Rican breast cancer survivors' perspectives on exercise.
Rose A. Trevino, Liliana Vallejo, Daniel C. Hughes, Velda Gonzalez, Maribel Tirado-Gomez, Karen Basen-Engquist. The University of Texas Health Science Center San Antonio, Harlingen, TX; UT M. D. Anderson Cancer Center, Houston, TX; The University of Puerto Rico Cancer Center, San Juan, PR.

Purpose: To provide formative data for culturally tailoring an exercise intervention based on social cognitive theory, six focus groups of Mexican-American and Puerto Rican breast cancer survivors were formed to gain insight from breast cancer survivors' perspectives on the issues surrounding cancer survivorship and exercise.

Methods: Six focus groups with 4-7 breast cancer survivors were conducted; 3 in San Juan Puerto Rico metropolitan area and 3 in different cities in Texas (Houston, San Antonio, and Harlingen in the Lower Rio Grande Valley). A total of 29 breast cancer survivors participated in the sessions. This is the first out of four focus group sessions projected to be held over a year's time. Using a social cognitive theory-based behavioral model as a guide, six questions were consistently presented to each focus group. The focus groups explored the meaning of exercise, previous experience with exercise, relationship between exercise and cancer, safety issues, barriers to exercise, and ideas on a possible exercise intervention. Based on discussion, follow-up questions were added as relevant to the overall six questions. The focus groups met for approximately an hour and a half. Each session was recorded, transcribed verbatim and then the responses were coded with Atlas.ti software using the same social cognitive theory behavioral model as a guide. Frequency and details of coded responses were compared between the six focus groups and compared between the Mexican- American and Puerto Rican groups.

Results: A total of 29 codes were identified and used to catalogue 983 responses. Some responses were catalogued in more than one code. Both groups had similar descriptions of what they considered exercise and similar responses for barriers to exercise. Some of these barriers included time and fatigue.Walking, dancing, and using a gym facility were described throughout as exercise. However, the Mexican- American and the Puerto Rican groups greatly differed in their responses to perceived safety in their community and accessibility to exercise equipment/facilities. For example Puerto Ricans mostly felt secure exercising around their neighborhoods while Mexican-Americans did not; Mexican-Americans had more accessibility to parks or trails while Puerto Ricans did not. Both ethnic groups also expressed the absence of information from health care providers regarding their exercise capabilities and exercise limitations during and after their cancer treatments. In regards to delivering an intervention, the Mexican- American group was open to having a group intervention, while the Puerto Rican group preferred more of an individual intervention.

Conclusions: Using a social cognitive theory-based behavioral model as a guide for coding, we found differences and similarities in factors of exercise behaviors and cancer survivorship exercise adherence between Mexican-American and Puerto Rican breast cancer survivor focus groups. This information can be used to help design culturally competent interventions for these specific ethnic groups.

B44 - GMaP/BMaP Region 3: Coming together to eliminate cancer disparities in the South and Puerto Rico.
Cathy Meade, Isabel Scarinci, Teresita Muņoz-Antonia. H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL; University of Alabama, Birmingham, AL.

Background: The Transdisciplinary Geographic Management Program (GMaP) and Minority Biospecimen/Biobanking Geographic Management Program (BMaP) for Region 3 is designed to develop a systematic and comprehensive strategy for building a state-of-the-art network for the support and efficient management of cancer health disparities research, training, and outreach in the five Southeastern states and Puerto Rico. The overall goal of the GMaP/BMaP-3 Network is to eliminate racial/ethnic cancer health disparities by identifying factors associated with such disparities, developing culturally-relevant approaches to address these factors, and disseminating evidence-based interventions across the cancer control continuum and across the continuum of the socio-ecological model. Specific to biobanking the ultimate goal of this network is to create unique, centralized, public resources for multi-ethnic biobanking/biospecimens. These goals are being accomplished in three phases: 1) network establishment, 3) needs/assets assessment, and 3) implementation plan.

Methods: We first established a working infrastructure for a GMaP/BMaP hub in the southeast United States and Puerto Rico as a collaborative partnership among the following institutions: H. L. Moffitt Cancer Center and Research Institute, University of Alabama at Birmingham Comprehensive Cancer Center, Tulane University, Xavier University of Louisiana, University of Mississippi, Tuskegee University, Morehouse School of Medicine, Emory University - Winship Cancer Center, and Ponce School of Medicine. Although the partners have worked closely together in all aspects of the project, we established specific cores in order to obtain an in-depth examination of specific areas in cancer disparities: clinical trials, advanced/emerging technologies, biomedical informatics, communication/dissemination, and evaluation. Given the importance of biospecimen/biobanking (BMaP), additional cores were established specific to this topic: collection, processing, storage and analysis of biospecimen; community and sociocultural beliefs; ethical, legal, and policy. Once the network was established, we designed a comprehensive needs/assets assessment of current strengths, weaknesses, capabilities, needs and interest in cancer health disparities in our participating institutions, CRCHD programs, and targeted communities with regard to research, training, and infrastructure development.

Results: A draft of the Comprehensive Assessment Tool (CAT) specific to our region has being developed through a participatory and methodic process with active involvement of all partnering institutions.

Conclusions: The data obtained from the needs assessment will inform the planning, nature, and extent of activities associated with the development of an operational plan to prioritize cancer research and biobanking activities in the Southeast US and PR.

B95 - Disparities in cancer drug use in the geriatric patient.
June M. McKoy, Denys T. Lau. Northwestern University Feinberg School of Medicine, Chicago, IL.

Background: Many older patients with cancer have multiple comorbidities requiring multiple medications in addition to their anticancer therapies. Ensuring appropriate medication use in this population, including adequate treatment dosage and adhering to treatment guidelines, is therefore a public health priority.We reviewed the epidemiological literature regarding anticancer treatment administration to older cancer patients.

Methods: We conducted a systematic literature review of population-based, epidemiological studies published between 1983 and 2006 on medication use involving chemotherapy and hormonal therapy among cancer patients aged 65 and older. Articles were classified into content areas based on processes of care and adverse drug reactions. Searches were conducted using PubMed, CancerLit, and Ageline. Medical subheadings were used to supplement searches. Only population-based, epidemiological studies published in English between 1983 and 2006 that examined medication use involving anticancer therapies (chemotherapies and hormonal therapies) and included cancer patients 65 years and older were included. Articles were systematically reviewed and classified based on Chassin's paradigm of quality problems in health services use: underuse, overuse, and misuse. For this study, underuse reflected the use of substandard, low-intensity anticancer drug therapies or the omission of standard drug treatments. Overuse included greater than appropriate use of pharmaceuticals, such as polypharmacy or over-dosage. Misuse included avoidable improper use of anticancer therapies, such as inappropriate drug choices, and adverse interactions among drugs or between pharmacotherapies and co-morbidities.

Results: A total of 78 articles were identified. Sixty five articles about overuse, underuse, and misuse of medications involving anticancer therapies among older cancer patients were identified: 64 (98%) evaluated underuse, 1 (2%) overuse, and none evaluated misuse. An additional 13 retrieved articles focused on ADRs associated with chemotherapy or hormonal therapy administration to older patients with cancer. One article reported overuse of hormonal treatments in elderly cancer patients with localized prostate cancer. The 64 articles on underuse were further subcategorized into undertreatment (n=47), deviation from treatment criteria (n=16), and patient non-adherence (n=1). Forty-seven studies described under-treatment, operationally defined as: no therapy, less frequent administration, and other non-standard anticancer treatment regimens. Seven studies analyzed medical records and primarily included data from one geographical location or healthcare system. Breast cancer was the most frequently studied cancer diagnosis, being the focus of 22 studies on under-treatment (47%). Thirty-nine of these studies reported that advanced age, regardless of physiologic condition, was a significant risk factor for under-treatment in anticancer therapy.

Conclusions: 98% of the epidemiological studies on processes of care associated with use of anticancer therapies among older patients with cancer evaluated underuse. Since cancer cases are expected to rise as the population ages and a tandem increase in the complexity of drug regimens is expected, additional research needs to focus how decisions are made relative to the utilization of cancer therapies older patients.

B105 - Addressing colorectal cancer disparities in rural Appalachia.
Mark Dignan, Nancy Schoenberg, Brent Shelton, Kevin Pearce, Cheri Tolle. University of Kentucky, Lexington, KY.

Colorectal cancer, a disease that is both preventable and detectable in early stages, is currently the third leading cause of cancer death in the U.S. In Kentucky, this disease is the second leading cause of cancer death. The elevated rates of colorectal cancer are a concern because they have increased in recent decades and because of low rates of use of available screening modalities. Low screening rates are particularly concerning for rural populations because of limited access to colonoscopy as well as diagnostic and treatment facilities for colorectal cancer. This concern is particularly acute in Appalachia, a region with a long history of pronounced cancer health disparities.

This NCI-funded study is designed to increase colorectal cancer screening by providing outreach and education to primary care physicians practicing in rural, medically underserved areas in Appalachian Kentucky. The intervention is delivered using academic detailing in partnership with Area Health Education Centers (AHEC). General practice, family medicine, and general internal medicine practices in Appalachian Kentucky were identified and invited to participate. A practice representative, usually the lead health care provider, completed a survey describing the practice characteristics and the colorectal cancer screening testing that they provided and/or recommended. Medical record reviews of a sample of 60 asymptomatic patients age 50 and older were conducted to establish baseline screening rates. Practices were then randomized to early or delayed intervention groups. The intervention was delivered at the practices. After an introductory visit, the academic detailer made up to four subsequent visits to the practices to present information on screening efficacy, patient counseling, reimbursement, and practice management. Evaluation data were collected through medical record reviews at 6 and 18 months after the final intervention visit. Medical record reviews collected data on screening recommended and/or provided (fecal occult blood testing (FOBT) in the past year, flexible sigmoidoscopy in the past 5 years, and colonoscopy in the past 10 years).

A total of 66 practices we enrolled. Baseline data from 3906 patient records indicated that health care providers had recommended screening by FOBT in the past year for 18.5% of the eligible patients seen during the period covered by the medical record review. The review also showed that 10.5% of records included documentation of results of FOBT. For flexible sigmoidoscopy in the past 5 years 0.4% had a recommendation and 0.3% results. For colonoscopy in the past 10 years, rates of recommendation and results documentation were 43.7% and 29.7%, respectively. Data from medical record review conducted 6 months after the intervention revealed small changes in FOBT, and a substantial increase in documentation of results from colonoscopy, (p<.05). Medical record review at 18 months post intervention revealed a sharp decline in FOBT and slight increases in colonoscopy. Qualitative interviews with project personnel and participants suggest that a pattern of rapid abandonment of FOBT as a screening modality and increased emphasis on colonoscopy may be competing explanations for intervention effects.

Updated: 12/07/10