Cancer

Barriers to Hereditary Breast and Ovarian Cancer Genetic Services at A Large Health Maintenance Organization

Mouchawar J, Rahm AK, Ellis J

Contact:

Judy M. Mouchawar, MD, MSPH, Kaiser Permanente Colorado, 580 Mohawk Dr., Boulder, CO 80302. E-mail: judy.mouchawar@kp.org.

Background: Cancer genetic counseling and testing services for Hereditary Breast and Ovarian Cancer (HBOC) are a covered benefit for Kaiser Permanente Colorado (KPCO) members, and are offered at the University of Colorado Cancer Center's Hereditary Cancer Clinic. Counseling services require a referral from any provider within KPCO, and pretest counseling is required prior to testing. Over the last 3 years, KPCO has identified that consistently only one third of members with approved referrals for HBOC services attend.

Objective: To describe barriers to receiving HBOC services among KPCO persons approved for referral.

Methods: Members who had an approved referral for HBOC services between March 1, 2001 and November 30, 2001 who had not had contact with genetic services within 10 months of their referral were mailed a written survey to identify potential barriers to attending services. Chart review further identified screening behaviors pre- and postgenetic referral as well as referring provider specialty.

Results: Seventy two members were eligible for the survey, which had a 40% response rate. Respondents did not differ from nonrespondents in terms of demographic characteristics, family history, comorbidities, and cancer screening behavior. Respondents identified 4 major barriers related to the precounseling required family history form, confidentiality/discrimination concerns, having to go outside of KPCO, and worry/anxiety about counseling and testing. We are currently examining associations between patient and provider characteristics with barriers to counseling uptake.

Conclusions: Future delivery of HBOC services should focus on reducing barriers to the uptake of counseling services, a course that could potentially dramatically increase patient volume.

Direct-to-Consumer Advertising for Hereditary Breast Cancer Testing at a Managed Care Organization: A Naturally-Occurring Experiment

Mouchawar J, Alford S, Laurion S, et al.

Contact:

Judy M. Mouchawar, MD, MSPH, Kaiser Permanente Colorado, 580 Mohawk Dr., Boulder, CO 80302. E-mail: judy.mouchawar@kp.org.

Background: Shortly after Myriad's direct-to-consumer advertising (DTC-ad) campaign for testing of the BRCA1 and BRCA2 genes began, articles on the overall and ethical appropriateness of DTC-ad for genetic testing were published based on previous print advertisements, with viewpoints emphasizing possible public misconceptions and disproportionate demand leading to overuse of testing.

Objective: To describe the patient, provider, and organizational impact of the DTC-ad at Kaiser Permanente Colorado (KPCO) where the DTC-ad occurred in terms of what previous literature predicted.

Methods: In 2003, written surveys were mailed to a random sample of 800 KPCO women aged 25-54 and to all KPCO primary care physicians. Organizational outcomes were captured from the existing genetic services database and were compared with Henry Ford Health System (HFHS), Detroit, Michigan, where there were no advertisements.

P

Results: Among KPCO women who reported seeing the DTC-ad, we found no undue negative emotions. Overall, the majority reported no anxiety at all (51%), and only 1% of women at low mutation probability (<4%) reported that the DTC-ad made them feel very anxious. Less than 4% of women with low mutation probability reported that the DTC-ad led them to seek counseling/testing. Referrals at KPCO increased 244% during the DTC-ad compared to the same time period a year earlier ( < .001) with the greatest increase seen in moderate probability (5%-7%) women, but with no significant change in pretest mutation probability among those undergoing testing. HFHS reported no significant change between the 2 time periods for numbers or mutation probability of referrals, or for mutation probability of women tested. The provider survey is under analyses.

Conclusion: In the face of potential future DTC-ad, providers and payers need to think now about the delivery of genetic services for persons of all risk levels. This may include establishment or re-evaluation of benefit coverage as well as referral and/or testing criteria, preparation of all healthcare providers, and educational material availability for patients.

Evaluating End of Life Care Within Managed Care for Women Who Die of Ovarian Cancer

Rolnick SJ, Herrinton L, Jackson J, et al.

Contact:

Sharon J. Rolnick, PhD, MPH, HealthPartners Research Foundation, 3105 East 80th Street, Mod. C, Suite 299, Bloomington, MN 55425. E-mail: cheri.j.rolnick@healthpartners.com.

Background: One in four deaths is caused by cancer, making it the second leading cause of death in the United States. Thus, end of life care is a critical component of the cancer care continuum. Ovarian cancer is the leading cause of death in women with gynecological cancer.

Objective: To describe end of life care (or lack of it) for women who die of ovarian cancer. In the pilot phase of this study, we focused on the quantity and quality of information we could obtain from the medical records.

Methods: This observational, pilot study was conducted by 5 HMOs (Kaiser Permanente Northern California, Kaiser Permanente Northwest, Group Health Cooperative, Henry Ford Health Systems, HealthPartners Research Foundation). Through medical record abstraction, data was collected on clinical signs and symptoms within the last 6 months of life, therapies used to manage symptoms, factors associated with a determination of end of life (EOL), change in care following EOL determination, comorbidities, and demographics. To be eligible, women must have died with epithelial ovarian cancer (primary or secondary cause of death) between 1995 and 2000. Abstraction focused on retrospective data for each subject's last 6 months of life.

Results: The initial abstraction form was extremely cumbersome. Recording data from each encounter to identify signs, symptoms, and procedures was extremely time intensive. The most time consuming aspect related to details of complications. Data not readily obtained from charts included practice guidelines, results of clinical tests (ie, we could not tell if tests were not done or not documented), evaluation of pain management, and alternative and complimentary therapies.

Conclusions: The data available in the paper medical record is good both in quantity and quality. However, some desired data may be too time intensive to obtain. Further, additional sources are needed to augment the information found. In the larger study, computerized databases will be used to obtain comorbidity information. Charts from hospice and hospitals may be needed to obtain a more comprehensive picture of end of life care.

Hospice Admission and Length of Stay Among Men Dying of Prostate Cancer

Bohlke K, Neslund-Dudas C, Schultz L, et al.

Contact:

Kari J. Bohlke, ScD, Group Health Cooperative-Center for Health Studies, 1730 Minor Ave, Suite 1600, Seattle, WA 98101. E-mail: bohlke.k@ghc.org.

Background: In spite of increasing participation in hospice, lengths of stay remain short.

Objective: To explore predictors of hospice admission and length of stay among men dying of prostate cancer.

Methods: Using the Seattle-Puget Sound Surveillance, Epidemiology, and End Results registry, we identified prostate cancer deaths occurring between 1993 and 1999 among members of a managed care organization. For 250 of these men, we collected information about the prostate cancer diagnosis and the last 6 months of life from medical records. We used logistic regression to analyze predictors of hospice admission, and Cox regression to analyze hospice length of stay (LOS).

Results: Among the 240 men with known hospice admission status, 84% received hospice care. Men aged &#8805;80 years were less likely to be admitted to hospice than younger men (odds ratio (OR) = 0.20, 95% confidence interval (CI), 0.09-0.44). The 11 African-American men in the study were less likely to be admitted than non-African-Americans (OR = 0.25, 95% CI, 0.07-0.94). Median LOS among the 194 men with hospice admission date information was 39.5 days. Twenty-six men (13%) were admitted during the last week of life. Another 26 men were admitted more than 180 days before death. Analyses of LOS are still underway, but men with localized disease at diagnosis appear to have shorter lengths of stay in hospice than men with distant disease at diagnosis (age- and race-adjusted hazard ratio = 1.62, 95% CI, 1.11-2.35).

Conclusions: Hospice participation was high, but there were disparities by age and race. Understanding the basis for variability in LOS by stage at diagnosis may suggest ways to increase LOS.

New Frontiers in Cancer Prevention: The Center for Health Communications Research

Greene SM, Strecher V, Little R, Resnicow K, Saunders E, Ubel P

Contact:

Sarah Greene, MPH, Group Health Cooperative-Center for Health Studies, 1730 Minor Ave, Suite 1600, Seattle, WA 98105.E-mail: greene.sm@ghc.org.

Background: The Center for Health Communications Research (CHCR) was funded by NCI in 2003 as a collaboration between 3 HMO Research Network sites (GHC, Henry Ford, KP Georgia) and the University of Michigan. It is one of four Centers of Excellence in Cancer Communications Research funded nationally by the National Cancer Institute. The overall goal of this initiative is to understand, apply, and disseminate effective communication approaches that maximize access to and appropriate use of cancer information by all who need it.

Objective: To report on efforts by the CHCR to develop an efficient model for generating tailored health behavior interventions that is generalizable across health behaviors and populations.

Methods: The CHCR is comprised of 3 research projects and 5 cores. The cores provide an infrastructure to support these initial projects. All projects utilize a fractional factorial design to determine the potential active ingredients of tailoring, including communication factors such as message content, message framing, message source, and graphical presentation; individual factors such as culture and sociodemographics; and other psychosocial factors such as motivation and self-efficacy. This design allows efficient incorporation of many experimental factors and multiple interactions, without the need for a large sample. Each project will consist of 2 phases. Phase I will screen for and identify factors that influence the target behavior (either smoking cessation, fruit and vegetable intake, or decisions about breast cancer prevention). Phase II will optimize factors determined to be active influences on behavior change. Each project varies with respect to delivery of the intervention, collaborating sites, and data collection strategies. Two projects will create Internet-based tailored interventions, an innovative approach in communications research.

Discussion: Results are not available due to the recent inception of this project. This endeavor represents yet another successfully-funded collaboration between HMO Research Network sites and colleagues in academia, as well as another outgrowth of the Cancer Research Network. Effective cancer communication strategies have enormous potential to improve health, and it is anticipated that this interdisciplinary multisite collaboration will result in exciting new models and interventions that can be rapidly disseminated in healthcare settings and beyond.

Qualitative Analysis of Patient Reactions to Genetic Testing for Hereditary Cancer Risk

Rahm AK, Mouchawar J, Hensley-Alford S, Ellis J, Laurion S

Contact:

Alanna Kulchak Rahm, MS, Kaiser Permanente Colorado, Clinical Research Unit, PO Box 378066, Denver, CO 80237-8066. E-mail: alanna.k.rahm@kp.org.

Background: In September, 2002, Myriad Genetics, Inc. initiated an intensive 5-month direct-to-consumer advertising (DTC-ad) campaign for genetic testing for the BRCA1 and BRCA2 genes, projected to reach more than 90% of women in the Denver and Atlanta Metropolitan Areas aged 25-54 an average of 16.5 times each.

Objective: To determine if perceptions or concerns about genetic testing varied by exposure to the DTC-ad.

Methods: A survey was sent to 1600 randomly selected female members of Kaiser Permanente Colorado (KPCO&#150;test site) and Henry Ford Health System (HFHS—control site) aged 25-54 years. The survey instrument began with a brief description of genes and hereditary risk for cancer, and women were asked to write a word or phrase that came to mind when thinking about the phrase "genetic testing for breast or ovarian cancer risk." A codebook was developed to classify themes and perceptions between those exposed to the DTC-ad and those not exposed.

Results: The survey had a response rate of 40%, with 64% reporting being exposed to the ad and 36% not exposed. Of the unexposed, 53% were from HFHS. Overall, there was an even distribution of women reporting a positive feeling about testing ("this is great!") vs. those who thought it was "scary," with an additional 14% stating simply the word "cancer." Many reported the phrase made them think about prevention or cure for cancer. A small number also mentioned specific fears of insurance or job discrimination. Significantly more women at KPCO had questions about testing such as, "is it for me?" "can I test?" or "am I at risk?" or simply mentioned their family history ("my mother had cancer") vs. women from HFHS.

Conclusions: The idea of genetic testing for breast cancer risk elicits both positive and negative feelings, including misconceptions ("curing cancer"). Such may be barriers to uptake of genetic services or negatively impact a woman's counseling experience. These concerns could be incorporated into educational materials as well as elicited and addressed in counseling sessions.

Receipt of Routine Surveillance Care Following Cancer Treatment with Curative Intent

Lafata JE, Simpkins J, Schultz L, et al.

Contact:

Jennifer Elston-Lafata, PhD, Center for Health Services Research, Henry Ford Health System, 1 Ford Place, 3A, Detroit, MI 48202. E-mail: jlafata1@hfhs.org.

Background: Medical societies often recommend routine surveillance to detect recurrent disease among cancer survivors, yet such care is rarely studied.

Objective: To compare surveillance care receipt among cancer survivors to guideline recommendations.

Methods: Cohorts of patients aged &#8805;30 years diagnosed with breast, colorectal, endometrial, lung, or prostate cancer between 1990-1995 and treated with curative intent were identified via tumor registry (n = 100 per site). Receipt of procedures and indication were abstracted from medical records for up to 5 years following treatment. Kaplan-Meier product estimates were used to estimate median time to service receipt (routine visits, and recurrence and metastatic disease testing) and care that met and exceeded guideline recommendations.

Results: Except among colorectal patients, more than 88% of patients had 2 routine office visit(s) and more than 71% had testing for new primaries within 18 months of treatment. Overall, median time to receipt of first-year recommended care was <10 months. More than 88% of lung, endometrial, and prostate patients received first-year care by 18 months; rates were 78% and 69% for colorectal and breast patients, respectively. Care exceeding guideline recommendations was also common. More than 2/3 of patients received &#8805; 4 visits in 18 months. By 30 months post-treatment, 74% of lung and 99% of prostate patients, respectively, received metastatic testing. Approximately 1/3 of colorectal and prostate patients exceeded recommendations for routine laboratory testing and almost 1/4 of breast patients exceeded the recommended number of mammograms in 18 months. By 30 months, these numbers had risen to 51% and 61%, respectively. Conclusions: Receipt of routine follow-up care is common among these cancer survivors. Notably, care in the first year often exceeded published clinical guidelines. Also of note is that a surprisingly high number of patients fail to receive recommended care. Compared to established guidelines, both under- and over-use of surveillance care is evident among cancer survivors. Clinical reasons for these variations and cost and health implications deserve further study.

Screening Clinical Breast Examination Sensitivity, Specificity, and Predictors of Accuracy

Rolnick SJ, Fenton JJ, Elmore JG, on behalf of the CRN PROTECTS Group

Contact:

Sharon J. Rolnick, PhD, MPH, HealthPartners Research Foundation, 8100 34th Avenue South, 11th Floor, Bloomington MN 55425. E-mail: cheri.j.rolnick@healthpartners.com.

Background: Although many American women receive regular screening clinical breast examination (CBE), the accuracy of CBE in the community setting remains uncertain.

Objective: To determine the accuracy of CBE screening in the community setting.

Methods: We determined the accuracy of CBE among asymptomatic female health plan enrollees in 5 states (Washington, Oregon, California, Massachusetts, and Minnesota). Among women who received a screening CBE within 1 year of breast cancer diagnosis and who subsequently died of breast cancer (n = 485), sensitivity was estimated as the proportion of women whose most recent CBE was abnormal. Among women without a breast cancer diagnosis in the year following a screening CBE (n = 1427), specificity was estimated as the proportion whose screening CBE was normal. Bivariate and logistic regression analyses identified patient characteristics associated with CBE accuracy.

Results: Among women who subsequently died of breast cancer, the sensitivity of screening CBE was 21.6% (95% CI, 18.1%-25.6%). Decreased sensitivity was associated with estrogen use at the time of CBE (OR 0.23; 95% CI, 0.07-0.80) and concurrent receiving a Pap smear with CBE (OR 0.45; 95% CI: 0.27-0.72). There were nonsignificant trends toward decreased sensitivity among women with a family history of breast cancer and increasing chronic disease comorbidity. Specificity of screening CBE was 98.6% (95% CI, 97.8%-99.0%). Both a family history of breast cancer (OR: 0.31; 95% CI, 0.13-0.78) and a history of breast biopsy (OR 0.22; 95% CI, 0.09-0.55) were independently associated with decreased specificity.

Conclusions: Screening CBE provided in the community is less sensitive but more specific than in clinical trials of breast cancer screening.