PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Cancer. Author manuscript; available in PMC Feb 15, 2011.
Published in final edited form as:
PMCID: PMC2819540
NIHMSID: NIHMS158481
A Trial of Three Interventions to Promote Colorectal Cancer Screening in African Americans
Daniel S. Blumenthal, MD, MPH,corresponding author Selina A. Smith, PhD, MDiv, Charlye D. Majett, MBA, and Ernest Alema-Mensah, PhD
Daniel S. Blumenthal, Department of Community Health and Preventive Medicine, Morehouse School of Medicine, 720 Westview Dr. SW, Atlanta GA 30310, Phone: 404/752-1624, Fax: 404/752-1160;
corresponding authorCorresponding author.
Daniel S. Blumenthal: dblumenthal/at/msm.edu
Background
Colorectal cancer (CRC) is the second-leading cause of cancer death in the U.S. CRC incidence and mortality rates are higher in blacks than in whites and screening rates are lower in blacks than in whites. We tested three interventions intended to increase the rate of colorectal cancer screening among African Americans.
Methods
The interventions were chosen to address evidence gaps in the Guide to Community Preventive Services: one-on-one education, group education, and reducing out-of-pocket costs. Three hundred sixty-nine African American men and women aged ≥50 years were enrolled in this randomized controlled community intervention trial. The main outcome measures were post-intervention increase in colorectal cancer knowledge and obtaining a screening test within six months.
Results
There was substantial attrition: 257 participants completed the intervention and were available for follow-up 3–6 months later. Among completers, there were significant increases in knowledge in both educational cohorts but in neither of the other two. By the 6 month follow-up, 17.7% (11/62) of control group members reported having undergone screening, as compared to 33.9% (22/65) of the group education cohort (p = 0.039). Screening rate increases in the other 2 cohorts were not statistically significant.
Conclusions
Group education can increase colorectal cancer screening rates among African Americans. The screening rate of less than 35% in a group of people who participated in an educational program through multiple sessions over a period of several weeks indicates that there are still barriers to overcome.
Keywords: Colorectal cancer, health status disparities, minority health, health education, screening, community-based participatory research
Colorectal cancer (CRC) is the third most commonly diagnosed cancer and is second only to lung cancer as a cause of cancer death in the United States. Each year, nearly 140,000 cases of CRC are diagnosed, and approximately 56,600 people die of the disease.1 Both incidence and mortality rates are higher in blacks than whites (incidence per 100,000: 72.9 v. 61.4 for men; 56.1 v. 44.7 for women.2 Mortality per 100,000: 26.1 vs. 19.3 for men; 22.4 vs. 15.3 for women3).
CRC screening reduces morbidity and mortality, and screening by endoscopy reduces incidence by allowing the removal of pre-cancerous polyps. However, low rates of CRC screening have been documented nationwide. The 2006 Behavioral Risk Factor Surveillance System found that the median screening rates for adults aged 50+ who had had a fecal occult blood test (FOBT) within the past two years were 24.4% for whites and 23.6% for blacks; rates for those ever experiencing sigmoidoscopy or colonoscopy were 59.1% for whites and 53.1% for blacks.4 It is estimated that 50–60% of CRC deaths could be avoided if all people over age 50 had regular screening tests.5
Subramanian et al,6 in a review of 30 articles, found that older age, more education, health insurance, physician recommendation, and a positive attitude toward screening were factors associated with patient adherence to CRC screening guidelines. They also found that “misconceptions and lack of knowledge are present despite efforts to inform the public about cancer screening.”
Conversely, deterrents or barriers to screening include pessimism and fear of cancer,6 lack of knowledge,7 lack of a provider recommendation,8 attitudinal factors such as fear or embarrassment,8 lack of insurance,9 and “life difficulty.”10 Fatalism – the belief that cancer inevitably leads to death – is a common deterrent among African-Americans.11 Many investigators have recommended education for the general population12, 13 and specifically for African-Americans14, 15 as an approach to promoting CRC screening. They have also recommended further research to determine which educational programs are effective.6
This report describes the Colorectal Cancer Screening Intervention Trial (CCSIT), the goal of which was to test three interventions to increase CRC screening rates among African Americans. Two of the interventions were educational; the third responded to financial barriers. The educational interventions were consistent with the contention that education could increase screening. These interventions also addressed gaps in the evidence as summarized in the Guide to Community Preventive Services; each approach is considered by the Guide to have “insufficient evidence” on which to base a recommendation.16
The CCSIT study was a randomized community intervention trial to assess the efficacy of two different behavioral and educational counseling interventions, and a financial support intervention, in promoting CRC screening. The project utilized community-based participatory research (CBPR) methods. The study protocol was reviewed by the Community Coalition Board17 of the Morehouse School of Medicine Prevention Research Center and was approved after the Board’s suggestions were incorporated. The study was also approved by the CDC and Morehouse School of Medicine Institutional Review Boards. The Metropolitan Atlanta Coalition on Cancer Awareness (MACCA, part of the Atlanta-based National Black Leadership Initiative on Cancer) participated in recruitment as did members of the Community Coalition Board. MACCA consisted of representatives of community organizations and agencies concerned with cancer as well as unaffiliated cancer prevention advocates. Other community organizations, churches, and clinics also participated. We have shared the results of the project with these groups, and the Community Coalition Board approved this manuscript.
Recruitment
The eligibility criteria for this project were: African American, over age 49, no history of CRC, and no previous CRC screening test within the recommended time interval. Two full-time and one part-time community health worker (CHW) made contacts in-person at senior centers, churches, community centers, and public health clinics. The CHWs worked with MACCA and with members of the Community Coalition Board.
Eligible volunteers were invited to an introductory meeting, where they received logo gift bags containing a cookbook, message fan, pamphlets, and CRC screening information, including a National Cancer Institute (NCI)-developed pamphlet. Also included was a “Resource Directory” that included CRC screening guidelines; methods of referral for screening tests; Medicare, Medicaid, and selected insurance coverage policies; locations of sites (including low-cost sites) in metro Atlanta where screening could be obtained; transportation routes to screening sites; and free or low-cost transportation services. At this meeting, participants completed informed consent forms and pre-test questionnaires. At a second introductory meeting, participants were assigned to one of the four cohorts and, for the two educational interventions, were notified of the time and place of the next meeting. Participants were randomized by the site – church, community/senior center, or clinic at which they volunteered. That is, all participants at a given site were assigned to the same arm of the study. This prevented the “contamination” that could result from participants in different arms of the study sharing information with each other. At this second introductory meeting, the participants were also shown a slide presentation on healthy living (“Down Home Healthy Living,” tailored for African Americans).
Interventions
We selected for testing three of the interventions for which the Community Guide stated that there was “Insufficient Evidence”: one-on-one education, group education, and reducing out-of-pocket costs.
One-on-One Education Cohort
A health educator met individually with participants in this group for three 45-minute sessions over three weeks to review educational materials on CRC risk and screening. All meetings took place at the recruitment site. The educational material included descriptions of CRC symptoms, diagnosis and treatment, screening tests, and screening recommendations.
Group Education Cohort
For this approach, a health educator met with participants in groups of 4 – 14 (average = 5) to review the educational material. Although the material was identical to that covered in the one-on-one counseling, there were more people at each meeting, and four sessions over four weeks were required to cover the material. Again, all meetings took place at the recruitment site.
Reduced Out-of-Pocket Expense Cohort (Financial Support)
In this group, participants were offered financial reimbursement up to $500 for out-of-pocket expenses incurred for CRC screening, including transportation and other non-medical expenses. A health educator was available to assist with negotiating direct payment and arranging transportation to the doctor’s office or medical clinic for the screening test. The health educator also provided letters of introduction and guarantee of payment for the patient to assist in scheduling the test visit.
The amount offered was more than sufficient to pay all costs involved in obtaining an FOBT, but less than half the cost of a colonoscopy. For patients seeking an endoscopic screening, the amount could cover the co-payment and/or deductable for those with private insurance or Medicare or could pay the full cost of screening at a reduced-cost public facility. Of note, the waiting list for reduced-cost screening was generally long.
Control Cohort
Participants in this group attended the introductory session but received no intervention other than accepting the contents of the gift bag with the educational pamphlets. They received pretesting (at the introductory session), post-testing, and follow-up on a schedule identical to that of the participants in the other cohorts.
Measures
We requested medical records from the participants’ physicians or health care facilities and abstracted the records to determine whether or not each participant had, in fact, been screened for CRC prior to enrollment in the study.
The questionnaire survey (pre-test/post-test) included an assessment of knowledge, attitudes, beliefs and practices regarding general health issues and, in particular, knowledge about CRC risk and screening. Screening status for CRC was assessed by phone and/or mail at the follow-up assessment periods at 3 and, if necessary, at 6 months after the intervention.
Educational Materials
Educational resources for the small groups and one-on-one participants included materials from the National Cancer Institute (NCI) and the American Cancer Society as well as original materials prepared for the CCSIT project. The curriculum for the educational sessions described, in lay terms, the genetics, incidence and mortality rates for CRC; disparities in detection and treatment; and the screening methods used to detect CRC. It also covered the effectiveness of diet and exercise in the prevention of CRC. Time was devoted to explaining the screening methods and relating screening to early detection and prevention of CRC. We pilot-tested all measures and procedures at a senior citizens’ center in Atlanta. Based on the experience and feedback from nine participants, the materials were refined.
Training for Personnel
One of the CCSIT investigators (S.S.) conducted training for the staff responsible for recruiting participants and conducting the interventions.
Follow-up
We administered the post-test questionnaires to participants in the two educational cohorts at their final intervention session; those in the control and financial groups were given their post-test questionnaires two weeks after the second introductory session. All participants were contacted by telephone approximately three months after completing their post-test questionnaires to determine whether they had been screened for CRC; if they had not been screened, they were called again at six months.
If the telephone was not answered, or if the participant was not available, at least two repeat attempts were made either by telephone or through mail. If the phone had been disconnected or the number changed, an attempt to identify a new number was made through Directory Assistance.
Study Population
From January 2003 through April 2005, we recruited participants from 34 churches, 13 senior residences, 11 community senior centers, 3 medical clinics and 4 men’s groups based at social organizations. All recruitment sites were in the Atlanta Metropolitan Area counties of DeKalb, Fulton, Cobb and Clayton. Initially, 803 persons expressed interest in participating in the study. Of these, 158 were ineligible because they had previously been screened, could not verify their screening status, or did not meet one or more of the other eligibility criteria. The remaining 645 eligible participants were randomized to an intervention group or the control cohort. Of these randomized participants, 276 withdrew by refusing consent, failing to attend an intervention session, or otherwise dropping out. The remaining 369 completed an educational intervention or were in the control or reduced-out-of-pocket expense cohorts. Of these, 259 (69.6%) completed the post-test survey and were successfully contacted at follow-up by 6 months to assess their CRC screening status. These participants were considered to be “completers.” The 110 who did not complete the post-test survey or could not be contacted at follow-up were “non-completers.” In the 3–6-month follow-up, we were unable to contact non-completers because of their death (3 persons); because their telephone had been disconnected and they did not respond to the mailed requests for them to call for follow-up; or because they refused to participate in the follow-up phone interview.
. We compared completers with non-completers among the 369 enrolled participants to identify demographic factors associated with completion. Participants who completed the study were similar to those who did not with respect to major demographic variables including age, sex, marital status, and insurance coverage. However, completers were more educated than non-completers, with more than a 50% greater rate of college graduation (p=0.026)
Establishing Group Equivalence
Participants were mostly non-married females, over age 65, with a high school diploma or some technical college. For all groups, Medicare and/or Medicaid covered most of the participants. The four groups were similar with respect to mean age, proportion of males and females, educational background, and marital status (Table 1). Approximately one-third had attended at least some college. However, insurance coverage was not distributed evenly among the groups. Persons in the Financial Support group were more likely to have private health insurance than participants in any other group, while participants in the One-on-One group were more likely to have no health insurance than any other group (p<.0003).
Table 1
Table 1
Demographics by Study Arm N= 369 Participants at Baseline
Effect of Group Intervention on Cancer Knowledge and Screening Behavior
Change in Cancer Knowledge
Cancer knowledge was determined by administering a seven-item pre-test and post-test. We performed a pair-wise analysis of data within groups by use of the McNemar chi square test for categorical data (Table 2). Knowledge increased significantly in the two educational cohorts on six of the seven items. On the seventh, (“Can you name an examination that can be done to find colorectal cancer in its very early stages?”), over 90% of respondents knew the correct answer on the pre-test; hence, there was little room for improvement on the post-test. The financial support cohort did not show improvement on any question; the control cohort showed improvement on only one.
Table 2
Table 2
Cancer Knowledge Change over Time by Study Arm N= 257
Colorectal Cancer Screening Behaviors
For purposes of reporting the findings on cancer screening, the 3- and 6-month follow-up assessments were combined and referred to as “follow-up.” Participants who were contacted at 3 months and indicated that they had been screened were considered to have completed the study and were not contacted again. Participants who had not been screened were encouraged to be screened and were contacted again at 6 months. At follow-up, 257 of the 369 enrolled participants were contacted for a telephone interview about whether or not they had had any type of CRC screening since the start of the project. We were unable to contact 112 participants.
We conducted two types of analyses. The first considered only those participants who were successfully contacted to determine their screening status post-intervention. The second used the more conservative “intention to treat” approach, in which those who could not be contacted were considered not to have been screened.
Follow-up results for each of the intervention groups are shown in Tables 3 and and4.4. Considering only the contacted group (Table 3; n=237), the Group Education intervention showed significantly more participants screened (33.9%) compared to the Control Group (17.7%, p=.039). The Financial Support Group and the One-on-One Group were similar in percent of participants screened (22.2% and 25.4% respectively), rates that exceeded that of the Control Group but did not differ from it statistically in an individual chi square comparison.
Table 3
Table 3
Screening Status by 6-Month Follow-up for all Contacted Participants (N=257)
Table 4
Table 4
Screening Status by 6-Month Follow-up for all Enrolled Participants (N=369)
For the intention-to-treat analysis (overall chi square 4-group comparison with all participants), there was a similar trend but with only borderline statistical significance (Table 4).
Only two “client-based” interventions, client reminders and small media, are recommended by the Community Guide to promote CRC screening. The other six are rated as “Insufficient Evidence.”1 Moreover, the interventions listed were primarily tested on participants who were not African-American, so that their potential for reducing health disparities is not clear.
To add to the evidence base, we tested three CRC screening interventions that fell into the “Insufficient Evidence” category. By testing them on African-American participants, we hoped to demonstrate an approach that would help reduce CRC disparities.
The results suggest that all three interventions had some efficacy. Knowledge was increased in the two cohorts that participated in an educational intervention, as contrasted with the two cohorts that did not. Each of the interventions appeared to be efficacious, although statistical significance was borderline. Because of shortfalls in recruitment and retention, the study was underpowered.
The least successful of the interventions was Financial Support. This might be attributed to the fact that randomization was not successful with respect to the distribution of health insurance among our participants. The Financial Support cohort, the most heavily insured group, probably needed the support the least. Still, the prospect of reimbursement of out-of-pocket costs should have been attractive since, even with insurance, co-payments, deductibles, transportation costs etc, can be substantial.
The most successful intervention was group education, which was modeled on social support theories such as Social Ecological Theory18 and Social Cognitive Theory19. These theories suggest that the informational and emotional support received by participants in a support network may buffer barriers related to cancer screening. The involvement of family members, friends, volunteers and others can promote cancer early detection and augment treatment.20
Limitations and Challenges
Recruitment
The recruitment of participants was labor-intensive, with more than 2000 individuals being approached over 3 years (Figure). However, we were unable to reach our enrollment and retention target of 500 individuals. This was not a result of reluctance on the part of potential participants: over 800 persons volunteered to participate, but about 20% were ineligible, mostly because of age or previous screening exclusion criteria. This still left over 600 eligible participants.
Our success in this area can be attributed to the active participation of our community partners. The Prevention Research Center Community Coalition Board and the Metropolitan Atlanta Coalition on Cancer Awareness participated in the design of the interventions and, along with other community partners, played a major role in recruitment.
Retention
Retaining participants in the study was our greatest challenge: over 40% failed to complete the intervention. Three of these died, but the remainder dropped out.
Retaining participants in a community-based prevention initiative such as this is a common problem, especially low-income and poorly-educated participants. Continued participation represents a low priority for participants, especially since the benefits may not be apparent. We have observed this in previous projects, 21 as have others; in particular, low income, less education, and African-American race have been associated with attrition from research projects.22 Brown and colleagues23 identified lack of awareness, lack of transportation, interference with work/family responsibilities, financial costs, negative side effects, and burdensome procedures as barriers, and it is likely that all of these played a role in our project. Financial incentives are sometimes effective in retaining participants24, but we did not offer significant financial incentives to participants in our educational interventions, since the comparison intervention was based on a financial incentive. We offered only minor incentives to participate, such as key rings and supermarket gift cards.
Randomization
Randomization of our sample was not entirely successful: members of the financial cohort were more likely to be insured than members of the other cohorts. Insured individuals could have taken advantage of our offer to reimburse out-of-pocket expenses to cover copayments, deductibles, transportation costs, etc., but relatively few did. It would be interesting to do a subset analysis of this cohort to learn if there were a significant difference in screening rates between insured and uninsured persons; but since there were only three uninsured persons in the group, this was not possible.
Screening Rates
Only a modest percentage of participants, regardless of intervention group, were screened even after completing the intervention. In the most responsive arm – Group Education – participants attended six meetings (the two introductory sessions and the four educational sessions) and continued participation in the program through the six-month follow-up. Still, only about a third was screened. Education and social support together doubled the rate at which participants were screened compared to controls, but that rate was relatively low. The remaining barriers could be fear of or aversion to endoscopy, revulsion at handling feces, or other cultural or behavioral obstacles.
A factor possibly contributing to the low screening rates is the likelihood that the target population was resistant to CRC screening. Approximately half of the African-American population over 49 years old has already been screened (and was hence ineligible for our study). The remaining population, from which we recruited participants, was not generally ignorant of the fact that screening could prevent CRC mortality; about two-thirds of completers and non-completers alike knew at baseline that CRC could be prevented and that CRC screening tests could be used for this purpose. They had nonetheless remained unscreened. Regardless of the reason (financial, fear, distrust, or denial), it was more difficult to promote screening to this group than would be the case if we were recruiting from a naive population.
Summary
In this project, we tested three interventions to promote CRC screening in African Americans, the population with the highest CRC incidence and mortality rates. Compared to controls, the group education intervention doubled the rate at which participants were screened, at least in the cohort of participants who completed the study. Hence, this intervention offers potential for reducing CRC disparities. The other two interventions (reduced out-of-pocket expense and one-on-one education) showed promise, but their level of efficacy as compared to the control group did not reach statistical significance. Only about a third of the individuals that participated in the group education intervention obtained screening afterwards, indicating that more effective interventions are needed.
Acknowledgments
This study was supported by grant U57CCU42068 (Community Cancer Control) from the Centers for Disease Control and Prevention (CDC) with additional support from grant 5U48DP000049 (Prevention Research Center) from CDC, grant 1U01CA1146520 (Community Networks Program) from the National Cancer Institute (NCI), grant 2U54CA118638 (Minority Institution/Cancer Center Partnership) from the NCI, and grant 1UL1RR025008 (Clinical Research Center) from the National Center for Research Resources and grant UL1 RR025008 (CTSA) from NCRR. Thanks are owed Latoshia S. Wells, MPH; Lisa Hinton, MPH; Myron Johnson, Jr.; Janet Gross, PhD; Lee Caplan, MD, MPH, PhD; the National Black Leadership Initiative on Cancer (an NCI-funded Community Network Program); the Metropolitan Atlanta Coalition on Cancer Awareness; and the Morehouse School of Medicine Prevention Research Center Community Coalition Board.
Footnotes
None of the authors have any financial disclosures
Contributor Information
Daniel S. Blumenthal, Department of Community Health and Preventive Medicine, Morehouse School of Medicine, 720 Westview Dr. SW, Atlanta GA 30310, Phone: 404/752-1624, Fax: 404/752-1160.
Selina A. Smith, Department of Community Health and Preventive Medicine, Morehouse School of Medicine, Atlanta GA.
Charlye D. Majett, Cardiovascular Research Institute, Morehouse School of Medicine, Atlanta GA.
Ernest Alema-Mensah, Clinical Research Center, Morehouse School of Medicine, Atlanta GA.
1. MMWR Morb Mortal Wkly Rep. 2006. pp. 308–311. www.cdc.gov/mmwr/preview/mmwrhtml/mm551a4.htm. [PubMed]
2. American Cancer Society. 2007. www.CAFF2007AAacspdf2007.pdf.
3. National Cancer Institute, Division of Cancer Control and Population Sciences. Surveillance Research. http://surveillance.cancer.gov/statistics/types/mortality.html.
5. Pignone M, Rich M, Teutsch SM, Berg AO, Lohr KN. Screening for colorectal cancer in adults at average risk: a summary of the evidence for the U.S. Preventive Services Task Force. Ann Intern Med. 2002;137:132–41. [PubMed]
6. Subramanian S, Klosterman M, Amonkar M, Hunt T. Adherence with colorectal cancer screening guidelines: a review. Prev Med. 2004;38:536–550. [PubMed]
7. Berkowitz Z, Hawkins NA, Peipins LA, White MC, Nadel MR. Beliefs, risk perceptions, and gaps in knowledge as barriers to colorectal cancer screening in older adults. J Am Geriatr Soc. 2008;56:307–14. [PubMed]
8. Farmer MM, Bastani R, Kwan L, Belman M, Ganz PA. Predictors of colorectal cancer screening from patients enrolled in a managed care health plan. Cancer. 2008;112:1230–8. [PubMed]
9. Beydoun HA, Beydoun MA. Predictors of colorectal cancer screening behaviors among average-risk older adults in the United States. Cancer Causes Control. 2008;19:339–59. [PubMed]
10. Power E, Van Jaarsveld CH, McCaffery K, Miles A, Atkin W, Wardle J. Understanding intentions and action in colorectal cancer screening. Ann Behav Med. 2008;35:285–94. [PubMed]
11. Powe BD, Hamilton J, Brooks P. Perceptions of cancer fatalism and cancer knowledge: a comparison of older and younger African American women. J Psychosoc Oncol. 2006;24:1–13. [PubMed]
12. Lieberman D. Screening, surveillance, and prevention of colorectal cancer. Gastrointest Endosc Clin N Am. 2008;18:595–605. xi. [PubMed]
13. Hoffman-Goetz L, Thomson MD, Donelle L. Reasons for declining colorectal cancer screening by older Canadians: a pilot study. J Cancer Educ. 2008;23:32–6. [PubMed]
14. Hamlyn S. Reducing the incidence of colorectal cancer in African Americans. Gastroenterol Nurs. 2008;31:39–42. [PubMed]
15. Shokar NK, Carlson CA, Weller SC. Factors associated with racial/ethnic differences in colorectal cancer screening. J Am Board Fam Med. 2008;21:414–26. [PubMed]
17. Blumenthal DS. A Community Coalition Board Creates a Set of Values for Community-Based Research. Preventing Chronic Disease. 2006. [serial online] Available from: URL: http://www.cdc.gov/pcd/issues/2006/jan/05_0068.htm. [PMC free article] [PubMed]
18. Breslow L. Social ecological strategies for promoting health lifestyles. Am J Health Promotion. 1996;10:253–257. [PubMed]
19. Bandura A. Social foundations of thought and action. A social cognitive theory. Upper Saddle River, JH: Prentice Hall; 1986.
20. Shoou-Yih DL, Arozullah AM, Cho YI. Health literacy, social support, and health: a research agenda. Soc Sci Med. 2004;58:1309–1321. [PubMed]
21. Blumenthal DS, Sung J, Williams J, Liff J, Coates R. Recruitment and Retention of Subjects for a Longitudinal Cancer Prevention Study in an Inner-city Black Community. Health Serv Res. 1995;30:197–205. [PMC free article] [PubMed]
22. DeVita DA, White MC, Zhao X, Kaidarova Z, Murphy EL. Determinants of subject visit participation in a prospective cohort study of HTLV infection. BMC Med Res Methodol. 2009 Mar 10;9:19. [PMC free article] [PubMed]
23. Brown DR, Fouad MN, Basen-Engquist K, Tortolero-Luna G. Recruitment and retention of minority women in cancer screening, prevention, and treatment trials. Ann Epidemiol. 2000;10(8 Suppl):S13–21. [PubMed]
24. El-Khorazaty MN, Johnson AA, Kiely M, El-Mohandes AA, Subramanian S, Laryea HA, Murray KB, Thornberry JS, Joseph JG. Recruitment and retention of low-income minority women in a behavioral intervention to reduce smoking, depression, and intimate partner violence during pregnancy. BMC Public Health. 2007;7:233. [PMC free article] [PubMed]