This article has Open Peer Review reports available.
Recruitment to a physical activity intervention study in women at increased risk of breast cancer
© Korde et al; licensee BioMed Central Ltd. 2009
Received: 02 February 2009
Accepted: 27 April 2009
Published: 27 April 2009
Physical activity is being studied as a breast cancer prevention strategy. Women at risk of breast cancer report interest in lifestyle modification, but recruitment to randomized physical activity intervention studies is challenging.
We conducted an analysis of recruitment techniques used for a prospective, randomized pilot study of physical activity in women at risk of breast cancer. We evaluated differences in proportion of eligible patients, enrolled patients, and successful patients identified by each individual recruitment method. The Fisher-Freeman-Halton test (an extension of Fisher's exact test from 2 × 2 tables to general row by column tables) was used to compare the success of different recruitment strategies.
We received 352 inquiries from women interested in participating, of whom 171 (54%) were eligible. Ninety-nine women completed a baseline activity evaluation, and 58 (34% of eligible; 16% of total inquiries) were randomized. Recruitment methods fell into three broad categories: media techniques, direct contact with potential participants, and contacts with health care providers. Recruitment strategies differed significantly in their ability to identify eligible women (p = 0.01), and women who subsequently enrolled in the study (p = 0.02).
Recruitment techniques had varying success. Our data illustrate the challenges in recruiting to behavior modification studies, and provide useful information for tailoring future recruitment efforts for lifestyle intervention trials.
Trial Registration No(s)
CDR0000393790, NCI-04-C-0276, NCI-NAVY-B05-001
Breast cancer is the most common malignancy among women in the United States and the second most common cause of cancer mortality . Although tamoxifen and raloxifene have both been shown to decrease breast cancer incidence by approximately 50% [2, 3], chemoprevention is not widely accepted by high risk women, mostly due to fear of side effects [4, 5]. Consequently, effective and acceptable prevention strategies can have enormous public health implications. Epidemiologic studies suggest that a lifestyle that includes some degree of physical activity is protective against breast cancer. While the amount of activity needed to impart a benefit is unknown, several data from several studies suggest that even a modest amount and intensity of activity is beneficial [6–9]; in the Women's Health Initiative cohort study, brisk walking for 1.25 – 2.5 hours per week resulted in about a 20% decrease in breast cancer risk. Recent data also suggest that physical activity is associated with both decreased breast cancer mortality and reduced risk of recurrence in breast cancer survivors [10, 11]. The underlying biologic mechanisms for these effects have yet to be elucidated, but several pathways have been proposed, including alterations of estrogen levels, body weight, growth factor pathways, and immune function .
Recruitment to lifestyle modification studies focusing on breast cancer risk reduction is challenging because it requires identifying participants who are not currently exercising, but are, or can be, motivated to participate in a study to increase physical activity. In other exercise intervention studies, only 2 – 20% of potential recruits were successfully randomized [13–16], highlighting the need to identify the most efficacious recruitment strategies. The present recruitment analysis was conducted within a prospective, randomized study of a home-based physical activity intervention comparing the use of a pedometer, an exercise prescription, and a motivational booklet versus a control of stretching exercises.
We employed various recruitment strategies to identify eligible subjects for a randomized physical activity intervention study in women at increased risk of breast cancer and breast cancer survivors. Simple frequencies were calculated to determine the most efficient method of recruiting eligible subjects. We also evaluated differences in proportion of eligible patients, enrolled patients, and successful patients identified by each individual recruitment technique. Statistical significance was assessed using the Fisher-Freeman-Halton test. The Fisher-Freeman-Halton test is the extension of Fisher's exact test from 2 × 2 tables to general row by column tables. Fisher's exact test gives the probability of observing a table that gives at least as much evidence of association as the one actually observed, given that the null hypothesis is true. All analyses were performed in SAS version 9.1.
Parent Study Design
Inclusion and exclusion criteria for study participants
ALL of the following:
Medically fit to exercise
ECOG Performance Status 0–1
If on hormonal therapy, >2 months since starting or changing therapy
and ANY of the following:
Breast cancer survivor:
History of Stage I, II, or III breast cancer
(> 2 mos since completion of treatment)
Increased risk of breast cancer:
Gail model 5 year risk ≥1.67%
Claus model lifetime risk >20%
Previous breast biopsy showing high-risk lesion
(e.g., atypical ductal hyperplasia or lobular neoplasia)
History of appropriately treated ductal carcinoma in situ
Known or suspected BRCA1 or BRCA2 gene mutation
Current or planned pregnancy
Uncontrolled intercurrent illness
Physical conditions that preclude daily walking
(e.g., use of wheelchair, walker, cane, etc)
History of cancer other than breast or non-melanoma skin cancer within 2 years preceding enrollment
Metastatic or recurrent cancer
Proportion of study participants identified by individual sources of recruitment
Number of inquiries
Article discussing physical activity and breast cancer risk
Advertisement in free newspaper in DC metro area
TV/print news story
General information in TV/print media
Referral service linked to NCI website http://www.cancer.gov
Multiple sources, including listservs and posting on group websites
Direct Patient Contact
Recruitment of patients already enrolled in an NCI study
NNMC Breast Care Center
Referral of patient at Nat'l Naval Medical Center
NNMC Patient mailing
Recruitment letter mailed directly to patients (n = 97)
Recruitment of current employees of NIH or NNMC
Contact with Health Care Providers
Recruitment letter sent to area ob-gyn and oncology practices (n = 500)
Community/support group talk
Visit by professional to discuss cancer prevention studies
In an attempt to reach an even broader population, we ran an advertisement in The Express, a free publication distributed on the Washington, DC Metrorail system, which has an average weekday ridership of >600,000 people http://wmata.com. The ad ran on three separate occasions over the course of three weeks, at a cost of $1,272, and solicited a total of 22 inquiries. Of these women, 5/22 (23%) women met pre-screen eligibility criteria. One woman underwent the baseline assessment and was subsequently randomized and completed the study.
Recruitment mailings were a reasonably successful means of generating interest in this study, with direct mailings to potential participants having a higher yield than letters to physicians. We sent a letter describing the study to 97 women receiving annual screening and follow-up care at the NNMC Breast Care Center, and received study inquiries from 10 women (10% of those contacted). This was a low cost recruitment method – the only costs were related to labor time for generating the letters and the cost of postage. By contrast, a mailing to local internal medicine and gynecology practices, which was sent to about 500 physicians, at a cost of about $5,000 for the purchase of physician mailing lists, resulted in only 16 referrals to the study.
An important aspect of targeted recruitment is the ability to reach participants who are sufficiently motivated to complete the study. Of 58 participants randomized, 43 (74%) completed the study, yielding an attrition rate of 26% (see Figure 2). Recruiting study participants who had previously participated in clinical trials was most successful for patient retention; 36% (4/11) of previous NIH study participants were randomized, and 75% (3/4) completed the study (3/11, see Table 2). The patterns of retention by source of recruitment (media techniques, direct physician contact, and contact with healthcare providers) approached, but did not reach, statistical significance (p = 0.06).
The difference in success of recruitment methods (media techniques, direct patient contact and contact with healthcare providers) in identifying eligible subjects was statistically significant (p-value for difference in proportion of patients meeting eligibility criteria by individual recruitment strategies = 0.01), suggesting that these differences were not due to chance alone. Similarly, recruitment strategies were significantly different (p = 0.02) in terms of identifying subjects who went on to enroll in the study.
Overall, we were able to recruit and randomize 34% (58/171) of eligible subjects and 16% (58/352) of those who expressed interest. We found that media techniques were most successful for recruiting patients to this lifestyle modification study. The strategies that generated the most patient inquiries were a news article in a local newspaper, the recruitment of patients from the NNMC Breast Care Center, and the NCI clinical trials website. Differences in success of recruitment methods were statistically significant in terms of identifying participants who were eligible and identifying participants who went on to enroll in the study. Differences in retention in the study by recruitment source approached significance.
Our overall recruitment rate of 16%, defined as those who were randomized, was in the range of those reported in other studies [13, 14, 16], and is similar to that seen for self-referred women in a recent randomized physical activity study for breast cancer survivors . Close to half of the women who expressed interest in participating in our study were not eligible; other investigators have noted similar results . One possible reason is that women tend to overestimate their risk of breast cancer , and therefore many women who perceive themselves at high risk do not meet the established study entry criteria. We also had numerous inquiries from women who were already physically active, and had a higher than expected percentage of women who met our pre-screen criteria for being sedentary by self-report, but who were too active for the study when assessed with an objective measure (i.e., a pedometer-determined step count for one week). Increasing the baseline step count cut-off from 5000 to 6000, to include women who were not sedentary by strict criteria, but who could be considered "low-active," partially alleviated this issue. Prior to this modification, 39% of women undergoing the baseline evaluation were too active; after increasing the baseline step count criteria to 6000, only 9% exceeded the baseline activity criteria (data not shown). We initially felt that women with lower baseline step counts would be most likely to benefit from increased physical activity, and would present the greatest opportunity to see measurable changes in both physical activity levels and secondary endpoints. Recently reported studies have allowed higher baseline step counts, and have also shown success in increasing physical activity [16, 28]. However, the level of activity needed to impact breast cancer risk, and the incremental benefit of additional activity in already active women, have yet to be determined.
The data provided here can inform the design of future attempts to reach a target population for lifestyle modification studies. An article describing the study that was published in a local newspaper proved to be very successful in generating interest in the study. In contrast, a short newspaper advertisement, which lacked detailed study information, was far less productive. Others have had similar experiences with media techniques [14, 29], and data suggest that women who are provided clear, concise information are more likely to participate in research studies than those who feel they are provided a less thorough explanation . In a randomized exercise intervention study for breast cancer survivors, Daley and colleagues had about equal success in generating interest with media techniques and with invitation letters generated by physicians, but media based strategies were more successful in terms of proportion of participants enrolled.
Recruitment mailings have been utilized and evaluated in a number of studies with varying success [14, 29, 31, 32]. In the study described above by Daley and colleagues, in which mailings were particularly effective, eligible women were identified by community oncologists and surgeons, who sent personalized letters to potential study participants. In other studies, a less tailored approach was used; Tworoger and colleagues  utilized mailing lists obtained from the Washington State Department of Motor Vehicles, and sent a packet containing an invitation letter, study brochure, eligibility survey and return envelope to over 100,000 women. Seven percent of women responded to the mailing, and 2.5% were enrolled and randomized. In our study, a targeted mailing sent to 97 patients seen in a breast health clinic had a reasonable success rate. This mailing resulted in 10 study inquiries and 4 women subsequently enrolled.
Striking a balance between reaching a large audience and targeting appropriate (and eligible) potential participants is a significant challenge. More broadly based recruitment efforts can reach a larger number of interested participants, but may also create interest among those who do not meet eligibility criteria. The cost of mailings must also be considered, and the more tailored approach may be relatively time intensive.
This study should be interpreted within the context of certain limitations. Recruitment to this study may have been complicated by the fact that women who were interested in a physical activity study were highly motivated to increase their level of physical activity, and therefore potentially less interested in accepting assignment to the control arm of an exercise intervention. Randomizing motivated women to a control arm may discourage participation by those whose original motive was to decrease their risk of breast cancer, an unknown proportion of whom likely assumed that physical activity was going to be effective, despite the absence of proof for this belief. This bias would discourage those most likely to participate. This issue may be alleviated by clear explanation of the opportunity for cross-over. This study did include an informal opportunity for crossover (participants on the stretching arm were offered a pedometer, exercise prescription and educational book at the end of the 3 month study, but were not formally followed for an additional three months). Those randomized to the control arm initially may have felt frustrated with the extra time they must put forth to complete all six months of a study that includes three months of control followed by three months of intervention. By designing the study as a six month cross-over study, all women would have entered the study with the expectation of providing six months of data to the researchers, although with an additional design flaw of asking women to stop exercising.
An additional barrier to the recruitment of eligible subjects is that women who feel compelled to actively reduce their risk of breast cancer are typically more active than those who are not so motivated. One study performed in the context of lifestyle interventions for reduction of diabetes risk suggested that those who are knowledgeable regarding their risk and determined to minimize it are usually already taking preventive measures . By measuring daily step counts at baseline we likely eliminated women who have an active job or who perform daily activities that require more than the routine amount of walking but do not engage in a defined exercise regimen.
Although we faced multiple challenges, this study was an important first step in launching a lifestyle modification initiative, and our analysis of recruitment and retention addresses the difficulties in mounting this type of clinical trial. In this analysis of our study methods, we have already gained a better understanding of the best way to target future physical activity intervention studies for recruitment.
The identification of practical yet efficacious interventions for increasing physical activity for women at risk of breast cancer and survivors is crucial, as there is mounting evidence that increased physical activity can reap numerous benefits in these populations. Planned analyses of our primary study outcomes will provide important data on the feasibility of a simple, cost-effective, low burden physical activity intervention in breast cancer survivors and women at high risk for the disease. Future studies will be necessary to determine the impact of such an intervention on breast cancer incidence and markers of breast cancer risk.
This research was supported by the Intramural Research Program of the NIH, National Cancer Institute, Center for Cancer Research and Division of Cancer Epidemiology and Genetics.
- Jemal A, Siegel R, Ward E, Hao Y, Xu J, Murray T, et al: Cancer statistics, 2008. CA Cancer J Clin. 2008, 58 (2): 71-96. 10.3322/CA.2007.0010.View ArticlePubMedGoogle Scholar
- Fisher B, Costantino JP, Wickerham DL, Redmond CK, Kavanah M, Cronin WM, et al: Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst. 1998, 90 (18): 1371-88. 10.1093/jnci/90.18.1371.View ArticlePubMedGoogle Scholar
- Vogel VG, Costantino JP, Wickerham DL, Cronin WM, Cecchini RS, Atkins JN, et al: Effects of tamoxifen vs raloxifene on the risk of developing invasive breast cancer and other disease outcomes: the NSABP Study of Tamoxifen and Raloxifene (STAR) P-2 trial. Jama. 2006, 295 (23): 2727-41. 10.1001/jama.295.23.joc60074.View ArticlePubMedGoogle Scholar
- Port ER, Montgomery LL, Heerdt AS, Borgen PI: Patient reluctance toward tamoxifen use for breast cancer primary prevention. Ann Surg Oncol. 2001, 8 (7): 580-5. 10.1007/s10434-001-0580-9.View ArticlePubMedGoogle Scholar
- Tchou J, Hou N, Rademaker A, Jordan VC, Morrow M: Acceptance of tamoxifen chemoprevention by physicians and women at risk. Cancer. 2004, 100 (9): 1800-6. 10.1002/cncr.20205.View ArticlePubMedGoogle Scholar
- Albanes D, Blair A, Taylor PR: Physical activity and risk of cancer in the NHANES I population. Am J Public Health. 1989, 79 (6): 744-50. 10.2105/AJPH.79.6.744.View ArticlePubMedPubMed CentralGoogle Scholar
- McTiernan A, Kooperberg C, White E, Wilcox S, Coates R, Adams-Campbell LL, et al: Recreational physical activity and the risk of breast cancer in postmenopausal women: the Women's Health Initiative Cohort Study. Jama. 2003, 290 (10): 1331-6. 10.1001/jama.290.10.1331.View ArticlePubMedGoogle Scholar
- Rockhill B, Willett WC, Hunter DJ, Manson JE, Hankinson SE, Colditz GA: A prospective study of recreational physical activity and breast cancer risk. Arch Intern Med. 1999, 159 (19): 2290-6. 10.1001/archinte.159.19.2290.View ArticlePubMedGoogle Scholar
- Thune I, Brenn T, Lund E, Gaard M: Physical activity and the risk of breast cancer. N Engl J Med. 1997, 336 (18): 1269-75. 10.1056/NEJM199705013361801.View ArticlePubMedGoogle Scholar
- Holmes MD, Chen WY, Feskanich D, Kroenke CH, Colditz GA: Physical activity and survival after breast cancer diagnosis. Jama. 2005, 293 (20): 2479-86. 10.1001/jama.293.20.2479.View ArticlePubMedGoogle Scholar
- Pierce JP, Stefanick ML, Flatt SW, Natarajan L, Sternfeld B, Madlensky L, et al: Greater survival after breast cancer in physically active women with high vegetable-fruit intake regardless of obesity. J Clin Oncol. 2007, 25 (17): 2345-51. 10.1200/JCO.2006.08.6819.View ArticlePubMedPubMed CentralGoogle Scholar
- Rundle A: Molecular epidemiology of physical activity and cancer. Cancer Epidemiol Biomarkers Prev. 2005, 14 (1): 227-36.PubMedGoogle Scholar
- Church TS, Earnest CP, Skinner JS, Blair SN: Effects of different doses of physical activity on cardiorespiratory fitness among sedentary, overweight or obese postmenopausal women with elevated blood pressure: a randomized controlled trial. Jama. 2007, 297 (19): 2081-91. 10.1001/jama.297.19.2081.View ArticlePubMedGoogle Scholar
- Daley AJ, Crank H, Mutrie N, Saxton JM, Coleman R: Patient recruitment into a randomised controlled trial of supervised exercise therapy in sedentary women treated for breast cancer. Contemp Clin Trials. 2007, 28 (5): 603-13. 10.1016/j.cct.2007.02.009.View ArticlePubMedGoogle Scholar
- Irwin ML, Cadmus L, Alvarez-Reeves M, O'Neil M, Mierzejewski E, Latka R, et al: Recruiting and retaining breast cancer survivors into a randomized controlled exercise trial: the Yale Exercise and Survivorship Study. Cancer. 2008, 112 (11 Suppl): 2593-606. 10.1002/cncr.23446.View ArticlePubMedGoogle Scholar
- Rhew I, Yasui Y, Sorensen B, Ulrich CM, Neuhouser ML, Tworoger SS, et al: Effects of an exercise intervention on other health behaviors in overweight/obese post-menopausal women. Contemp Clin Trials. 2007, 28 (4): 472-81. 10.1016/j.cct.2007.01.002.View ArticlePubMedGoogle Scholar
- Tudor-Locke C, Bassett DR: How many steps/day are enough? Preliminary pedometer indices for public health. Sports Med. 2004, 34 (1): 1-8. 10.2165/00007256-200434010-00001.View ArticlePubMedGoogle Scholar
- Welk GJ, Differding JA, Thompson RW, Blair SN, Dziura J, Hart P: The utility of the Digi-walker step counter to assess daily physical activity patterns. Med Sci Sports Exerc. 2000, 32 (9 Suppl): S481-8.View ArticlePubMedGoogle Scholar
- Iwane M, Arita M, Tomimoto S, Satani O, Matsumoto M, Miyashita K, et al: Walking 10,000 steps/day or more reduces blood pressure and sympathetic nerve activity in mild essential hypertension. Hypertens Res. 2000, 23 (6): 573-80. 10.1291/hypres.23.573.View ArticlePubMedGoogle Scholar
- Moreau KL, Degarmo R, Langley J, McMahon C, Howley ET, Bassett DR, et al: Increasing daily walking lowers blood pressure in postmenopausal women. Med Sci Sports Exerc. 2001, 33 (11): 1825-31. 10.1097/00005768-200111000-00005.View ArticlePubMedGoogle Scholar
- Swartz AM, Strath SJ, Bassett DR, Moore JB, Redwine BA, Groer M, et al: Increasing daily walking improves glucose tolerance in overweight women. Prev Med. 2003, 37 (4): 356-62. 10.1016/S0091-7435(03)00144-0.View ArticlePubMedGoogle Scholar
- Croteau KA: A preliminary study on the impact of a pedometer-based intervention on daily steps. Am J Health Promot. 2004, 18 (3): 217-20.View ArticlePubMedGoogle Scholar
- Lindberg R: Active living: on the road with the 10,000 Steps program. J Am Diet Assoc. 2000, 100 (8): 878-9. 10.1016/S0002-8223(00)00251-0.View ArticlePubMedGoogle Scholar
- Rooney B, Smalley K, Larson J, Havens S: Is knowing enough? Increasing physical activity by wearing a pedometer. Wmj. 2003, 102 (4): 31-6.PubMedGoogle Scholar
- Saunders D, Saunders R: Evaluation, Treatment and Prevention of Musculoskeletal Disorders. Chaska MN. 1993, 3Google Scholar
- Program DP: Lifestyle Manuals of Operations. 1996, University of Pittsburgh, Pittsburgh, PAGoogle Scholar
- Buxton JA, Bottorff JL, Balneaves LG, Richardson C, McCullum M, Ratner PA, et al: Women's perceptions of breast cancer risk: are they accurate?. Can J Public Health. 2003, 94 (6): 422-6.PubMedGoogle Scholar
- Vallance JK, Courneya KS, Plotnikoff RC, Yasui Y, Mackey JR: Randomized controlled trial of the effects of print materials and step pedometers on physical activity and quality of life in breast cancer survivors. J Clin Oncol. 2007, 25 (17): 2352-9. 10.1200/JCO.2006.07.9988.View ArticlePubMedGoogle Scholar
- Ott CD, Twiss JJ, Waltman NL, Gross GJ, Lindsey AM: Challenges of recruitment of breast cancer survivors to a randomized clinical trial for osteoporosis prevention. Cancer Nurs. 2006, 29 (1): 21-31. 10.1097/00002820-200601000-00004.View ArticlePubMedGoogle Scholar
- Sharp L, Cotton SC, Alexander L, Williams E, Gray NM, Reid JM: Reasons for participation and non-participation in a randomized controlled trial: postal questionnaire surveys of women eligible for TOMBOLA (Trial Of Management of Borderline and Other Low-grade Abnormal smears). Clin Trials. 2006, 3 (5): 431-42. 10.1177/1740774506070812.View ArticlePubMedGoogle Scholar
- Rubin RR, Fujimoto WY, Marrero DG, Brenneman T, Charleston JB, Edelstein SL, et al: The Diabetes Prevention Program: recruitment methods and results. Control Clin Trials. 2002, 23 (2): 157-71. 10.1016/S0197-2456(01)00184-2.View ArticlePubMedGoogle Scholar
- Tworoger SS, Yasui Y, Ulrich CM, Nakamura H, LaCroix K, Johnston R, et al: Mailing strategies and recruitment into an intervention trial of the exercise effect on breast cancer biomarkers. Cancer Epidemiol Biomarkers Prev. 2002, 11 (1): 73-7.PubMedGoogle Scholar
- Thoolen B, de Ridder D, Bensing J, Gorter K, Rutten G: Who participates in diabetes self-management interventions?: Issues of recruitment and retainment. Diabetes Educ. 2007, 33 (3): 465-74. 10.1177/0145721707301491.View ArticlePubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2288/9/27/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.