Health Education Research Advance Access originally published online on July 14, 2004
Health Education Research 2005 20(2):149-162; doi:10.1093/her/cyg108
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Health Education Research Vol.20 no.2, © Oxford University Press 2005; All rights reserved
Intervention to increase screening mammography among women 65 and older
1 Department of Family and Community Medicine and 2 Department of Public Health Sciences, Wake Forest University School of Medicine, Winston-Salem, NC 27157, 3 The National Cancer Institute, Office of Education & Special Initiatives, Bethesda, MD 20892-8334 and 4 Comprehensive Cancer Center, School of Public Health, Ohio State University, Columbus, OH 43210, USA
5 Corresspondence to: R. Michielutte; E-mail: bmichiel{at}wfubmc.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
This paper reports the results of a practice-based intervention program to increase mammography screening among women 65 and older who receive their health care in the private sector. Forty-three primary-care practices and 2147 women in central and western North Carolina were enrolled in the study, and 1911 women completed all phases of the study. The intervention was a three-stage educational and counseling program designed to become progressively more intensive at each stage. The interventions included provider education in the form of current information on issues in mammography for older women, simply written educational materials on breast cancer and screening mailed to women, and a brief telephone counseling session for the women. While the analysis revealed no overall effect across all three stages of the intervention program, tests for interaction indicated a significant program effect for women who were 80 or older, had less than 9 years of education, were black, or had no private insurance to supplement Medicare. The results suggested that providing primary-care physicians with information on screening older women and providing the women with useful educational materials can increase participation in screening mammography among subgroups of women currently least likely to receive mammography screening.
This paper reports the results of a practice-based intervention program to increase mammography screening among women 65 and older who receive their health care in the private sector. Forty-three primary-care practices and 2147 women in central and western North Carolina were enrolled in the study, and 1911 women completed all phases of the study. The intervention was a three-stage educational and counseling program designed to become progressively more intensive at each stage. The interventions included provider education in the form of current information on issues in mammography for older women, simply written educational materials on breast cancer and screening mailed to women, and a brief telephone counseling session for the women. While the analysis revealed no overall effect across all three stages of the intervention program, tests for interaction indicated a significant program effect for women who were 80 or older, had less than 9 years of education, were black, or had no private insurance to supplement Medicare. The results suggested that providing primary-care physicians with information on screening older women and providing the women with useful educational materials can increase participation in screening mammography among subgroups of women currently least likely to receive mammography screening.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Several studies support the value of mammography and clinical breast exam (CBE) for women age 50 and older as effective means for early detection of breast cancer (Kerlikowske et al., 1995
; Demisse, et al., 1998; Overmoyer, 1999). Intervals for frequency of screening are not consistent among the various health organizations, but most recommend that annual CBE should begin no later than 40 years of age and yearly mammography should begin at age 50. While estimates of screening behavior vary considerably, there is evidence that breast cancer screening has increased significantly in the past decade (Anonymous, 1997; American Cancer Society, 1999; Breen et al., 2001; Ostbye et al., 2003). However, despite the fact that 80% of all women over 40 report ever having had a mammogram and a CBE (American Cancer Society, 1999), participation in regular screening for breast cancer is low among many underserved populations (Harris et al., 1991; Breen and Kessler, 1994; Phillips et al., 1998; Breen et al., 2001; Ostbye et al., 2003).
Particularly important is the fact that participation in screening decreases as age increases (Harris et al., 1991; Breen and Kessler, 1994; Makuc et al., 1994; Ruchlin, 1997; Phillips et al., 1998; American Cancer Society, 1999; Ostbye et al., 2003). Ostbye et al. (Ostbye et al., 2003), for example, found that mammography screening during the period 1996–2000 remained stable at 70–80% for women age 50–64 and declined consistently as age increased to 40% for women 85–90. Thus, women at highest risk for incidence and mortality from breast cancer are also the least likely to be screened. The purpose of the present paper is to present the results of a practice-based intervention to increase screening mammography among women age 65 and older who receive medical care in the private sector.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The Program to Increase Screening Mammography (PRISM)
The North Carolina PRISM was a 3-year study (1999–2002) funded by the National Cancer Institute. The study's goal was to increase screening for women 65 and older who receive their health care in the private sector. It tested the value of simple, inexpensive interventions that could be implemented by staff in a primary-care setting. The rationale for the intervention approach was two-fold. Practice-based interventions appear to be particularly appropriate for elderly women. The results of the 1997 National Health Interview Survey indicated that over 83% of all women 65 and older saw or spoke to a primary-care physician in the year preceding the survey, and 91% reported at least one visit to a health-care professional (National Health Interview Survey, 2000). Almost 75% had two or more visits to a health-care professional. Secondly, the most successful practice-based screening interventions will be ones that can be integrated into the existing routine of a primary-care practice, and which are designed to become progressively more intensive and time consuming as necessary (King et al., 1994
; Janz et al., 1999). The intervention program was designed to fit these criteria. The interventions were introduced in a sequential design that allowed examination of the relative effectiveness of each stage of the intervention. The program included provider education in the form of current information on issues in screening mammography for older women, simply written educational materials on breast cancer and screening mailed to women, and brief issue-oriented telephone counseling for women. Forty-three primary-care practices in central and western North Carolina were enrolled in the study, and randomized to intervention and control conditions. The intervention and data collection methods for the study were approved by the Wake Forest University School of Medicine Institutional Review Board.
Theoretical framework
The elements of two theoretical models, the Health Belief Model (Becker, 1974; Rosenstock et al., 1988) and the Transtheoretical or Stages of Change model (Prochaska and DiClemente, 1982; Prochaska et al., 1994) were used in development and conduct of the intervention program. The Health Belief Model assumes that individuals fear illness, and that their health actions in response to illness are motivated by the degree of fear and the expected fear-reduction potential of actions, as long as this potential outweighs the practical and psychological obstacles to taking action (Glanz and Rimer, 1995). The constructs of the Health Belief Model in its present form include perceived susceptibility, perceived severity, perceived benefits, perceived barriers, cues to action and self-efficacy. As Glanz and Rimer (Glanz and Rimer, 1995) note, the Health Belief Model can be an effective mechanism for guiding the development of persuasive messages for health actions such as screening mammography.
In the context of the present study, the development of the messages in the printed materials was directed by this model. For example, perceived susceptibility was addressed by emphasizing the importance of age as a risk factor for breast cancer, clearly describing the general lifetime risk of breast cancer, and by providing information on other important risk factors, specifically noting that most women who get breast cancer do not have a family history of the disease. Examples of messages dealing with perceived benefits included graphic illustration of the sizes of tumors that can be detected by mammography as compared to CBE or breast self-exam. In some instances some constructs were addressed simultaneously. Fear of pain is an important barrier for a fairly large proportion of older women. The messages discussed the issue and addressed self-efficacy by providing the women with information on how they could deal with pain and steps that could be taken on how to minimize it.
The Stages of Change model assumes that behavior change is a process rather than an event that includes different levels of motivation (readiness) to change. The model suggests that interventions will be more effective if they take into account where a person is in the process of change. The stages identified in the model and their adaptation for use in the present study are as follows:
- Pre-contemplation, never heard of mammogram; heard of, but has not thought about getting one.
- Contemplation, considered getting a mammogram, but either has made no decision or has decided not to get one.
- Decision/determination (preparation), intends to get a mammogram, but has not made an appointment.
- Action, has scheduled or gotten a mammogram (first mammogram for women who have never had a mammogram or repeat mammogram for women who have had a previous mammogram, but are overdue for another one).
- Maintenance, has had repeat mammogram at recommended interval.
Information obtained on where a woman was in terms of the Stages of Change model with regard to getting screened at the time of the counseling phase of the intervention allowed tailoring of the telephone counseling to address the specific stage of readiness for women who had not yet obtained a mammogram. This information was used in addition to the report of the most important barriers to screening to direct the counseling session. If, for example, a woman indicated that she intended to get a mammogram, but has not made an appointment, the counselor identified the reasons for not making the appointment, provided information on how to address these barriers (e.g. information on transportation assistance available in the county), discussed any other general barriers mentioned by the woman and offered assistance in making the initial appointment with her primary-care physician.
Description of the intervention
The intervention design was sequential, with progressively more intensive interventions introduced at each stage. The three stages were as follows:
Stage 1
The first stage of the intervention began after the women were recruited for the study, completed a baseline interview and the practice was randomized to intervention or control condition. Physicians in the intervention clinics received (1) a fact sheet providing current information on screening mammography for older women, (2) telephone follow-up (on request) of any questions and (3) copies of a simply written pamphlet on mammography for older women that could be distributed to patients. A letter accompanying the materials emphasized key points in the fact sheet such as Medicare and private insurance guidelines for coverage, and encouraged physicians to contact patients 65 and older in need of screening.
Stage 2
Approximately 4 months after the physicians in a participating practice received the educational materials, women enrolled in the study were mailed educational materials dealing with breast cancer and screening for breast cancer. The educational materials included the pamphlet on breast cancer and breast cancer screening for older women sent to the physicians in the first phase of the intervention, as well as a fact sheet that reinforced key points for older women and provided additional information on breast cancer screening. Both the pamphlet and fact sheet were developed by study investigators using the framework of the Health Belief Model. The fact sheet emphasized (1) age is the most important risk factor for cancer, (2) older women benefit from early detection, (3) most insurance companies in North Carolina are now required by law to pay for mammograms and that Medicare helps pay for annual mammograms, and (4) now is the time to seek screening. The sheet also stressed that a complete program of breast cancer screening includes mammography, CBE and breast self-exam. All materials emphasized that women should contact their primary-care physician. These materials were accompanied by a brief letter from the PRISM coordinator describing the materials and stressing the importance of breast cancer screening.
Stage 3
About 4 months after a woman in a participating practice received the educational materials she was contacted by telephone by a trained counselor and asked to complete a second interview. Counselors had a 2-week period in which to contact the woman and at least three attempts were made to reach her. Incorrect telephone numbers were resolved by checking with staff in the woman's primary-care practice. Once a woman was contacted, she was first asked a series of questions identifying her stage of change for mammography screening. If the woman reported not having been screened in the past year, she was asked to identify the most important reason for not being screened as well as other factors that may have affected her decision. She was then offered a brief telephone counseling session that addressed any general issues dealing with the woman's stage of readiness such as precontemplation and discussed the most important barrier mentioned. The counselor also answered any questions or concerns expressed by the woman. The counselor urged the woman to contact her physician and offered to help make this appointment. Women who reported having a mammogram in the past year were not offered the counseling session.
Control practices
In order to control for possible effects of simply receiving attention on screening behavior, physicians and patients in the control practices received a skin cancer educational program that included the same frequency of contacts as those in the intervention practices. During Stage 1, physicians in the control practices received information on skin cancer screening for older patients. All participating women in the control practices received a mailing of educational materials on skin cancer in Stage 2 of the program and were administered the second interview (with no counseling) at the beginning of Stage 3. The skin cancer educational materials included a simply written pamphlet on skin cancer prevention and early detection, as well as a fact sheet dealing with personal risk and prevention measures.
Time to completion
For each practice, the total time between the baseline interview and completion of Stage 3 of the intervention was approximately 9 months. The total time between the baseline interview and completion of the final chart review was about 13 months.
Study sample
Practices
Primary-care practices participating in the study were recruited from 15 counties in central and western North Carolina. Practices were identified through a local HMO and through county medical societies. Letters were sent to the clinic director of each practice describing the study and inviting the practice to participate. One week later the principal investigator or the project coordinator called the clinic director, answered any questions regarding the study and, if the clinic was willing to consider participating, scheduled a visit. Approximately 30% of the practices contacted regarding the study agreed to participate. About 50% of the practices believed that they did not have enough eligible older women in their patient population to meet recruitment goals and about 20% indicated that they either simply were not interested or could not participate due to organizational changes or heavy physician and staff turnover. In the latter case, several practices that initially expressed interest in participating closed or merged with larger practices before the study began.
Forty-three practices staffed by 127 physicians agreed to participate in the project and all completed the study. Twenty-one practices were randomized to the intervention group and 22 to the control group. Practices were not randomized until recruitment of patients and all baseline interviews were completed.
Patients
The initial eligibility criteria for women in the study were that they (1) be age 65 or older, (2) have no history of breast cancer, (3) have not had a mammogram in the past 15 months and (4) have no serious physical or cognitive problem that ruled out screening for breast cancer. With the assistance of staff in each practice a list of potentially eligible women was identified from the medical records of currently enrolled patients. Typically, the practice provided a computerized list of all women 65 and older. This list was randomly ordered and chart reviews conducted to identify a random sample of women that met all the additional eligibility criteria previously described. Two practices were not able to provide a computerized list. In these practices, a count of all records was obtained and randomly ordered. Reviewers then worked through the records discarding all men, women under 65 and women over 65 otherwise ineligible. In order to take into account ineligible women and refusals to participate, at least 100 potentially eligible women were sampled in each practice. If the chart review identified less than 100 women, all potentially eligible women were selected.
The women were sent a letter signed by their primary-care physician briefly describing the project and inviting them to participate. The letter also stated that they would be called by a member of the PRISM staff to discuss participating in the project. The letter also included the basic elements of informed consent and stressed that they were under no obligation to participate.
A week after the letter was sent each woman was contacted by telephone by a trained interviewer who explained the study and answered any questions. If she agreed to participate, the woman was administered a baseline interview that included questions on breast and skin cancer, use of screening tests for these cancers, questions about her current health and health behaviors, and basic social and demographic background questions. The baseline interview was used to verify eligibility for the study by identifying women who had a mammogram more recently than indicated in the practice records and also to identify women ineligible for other reasons such as cognitive problems, illness or language problems not indicated in the practice records.
During the recruitment phase, it became evident that the practice records were not up-to-date for many of the women. Over one-third of the women contacted reported a mammogram within the year preceding the interview and for about 80% of these women the recent mammogram was verified by the mammography facility or the practice. Since the women had already expressed an interest in participating in the study, we included them as a second sample. This sample was designated as the ‘maintenance’ sample and was followed to test the effectiveness of the educational program in encouraging repeat mammograms. The sample of women overdue for a mammogram (15 months or longer) was designated as the ‘primary’ sample. Since all women recruited for the study enrolled with the same expectations (mailed educational materials, reviews of their medical records, second interview after 8 months), conduct of the intervention components and assessments were identical for women in both the primary and maintenance samples.
Interviewers attempted to contact 3522 women. Of these, 300 were ineligible for the reasons previously noted and 181 could never be contacted. Including the 181 never contacted as potentially eligible, 2147 of 3222 women (66.7% response rate) were recruited for the study. The overall response rate was almost identical for the intervention and control practices (P value for difference of proportions test = 0.624). Comparisons of responders and non-responders by race and age revealed no significant difference by race (P = 0.374) and a significant difference by age (P < 0.001). However, the actual difference by age was small. The mean age of responders was 72.4 (SD = 4.9) as compared to 73.3 (SD = 5.2) for non-responders. Complete follow-up data were available for 1911 (89.0 %) of the 2147 women enrolled in the study. The 11% attrition rate was due to (1) a woman's decision to end participation in the study at the time of the second interview, (2) left the practice or (3) death. Attrition rates did not differ significantly for women in the intervention and control groups.
The mean number of women recruited in each practice was 50, representing an average of 35 women in the primary sample and 15 in the maintenance sample. We estimated that the final sample represented about 10% of the population of eligible women in the participating practices. The average number of women per practice did not differ significantly between the intervention and control practices for the primary sample (P = 0.322) or for the maintenance sample (P = 0.913).
Examination of characteristics of women in the intervention and control practices indicated no significant differences for social and demographic characteristics among women in either the primary or the maintenance sample (Table I). In both samples, women in the intervention and control practices also were similar with regard to CBE in the past year, ever had a mammogram (primary sample only), ever did a skin self-exam, did a skin self-exam in the past year and ever had a clinical skin exam. The results for breast and skin cancer screening behaviors did show one significant difference for each sample. In the maintenance sample, women in the control group were less likely to have reported a clinical skin exam in the past year than women in the intervention group (P = 0.022). In the primary sample, while there was no difference between the intervention and control groups in the percentage of women who reported performing BSE once a month or more often, women in the intervention group were less likely to perform appropriate BSE only once a month (P = 0.004). Overall, however, comparisons of the intervention and control groups indicate that the randomization procedure was effective in both the primary and maintenance samples.
|
Hypotheses and outcomes
The major hypotheses of the study were that each stage of the intervention program would have an independent effect on screening rates, and that the program would show an overall cumulative effect across all three intervention stages. Operationally defined, we hypothesized that for each stage and overall, the proportion of women receiving screening mammography in the intervention practices would be significantly greater than the proportion screened in the control practices. Secondary research questions asked whether the intervention program was differentially effective by socioeconomic status, race or age.
The outcome evaluated was whether the women had a screening mammogram after each stage of the intervention program. The source of data for mammograms was chart review. Four months after implementation of Stage 1, provider education, a first evaluation chart review was conducted to identify the women who had mammograms during that period. Immediately after the review, Stage 2, mailed educational materials, was implemented. Four months after that a second evaluation chart review was conducted to determine the effectiveness of Stage 2. After the second evaluation review, the women were called for the post-test interview and Stage 3 telephone counseling. Finally, about 4–5 months after Stage 3 was implemented, a third and final evaluation chart review was completed.
As previously noted, delays in recording mammograms in the medical charts resulted in an underestimation of screening behavior among women in the study. This problem was largely eliminated for the first two evaluation reviews, since each could be updated by the subsequent review. Some underestimation is present in the final review, but the degree of error should be the same for both intervention and control practices.
Statistical analysis
Women were randomly selected within practices, and the practices were randomized to intervention and control conditions. Data analysis involved comparisons of mammography screening rates for the intervention and control samples. Linear and logistic regression models were used to assess differences between samples and between groups within samples in screening mammography rates, and for selected characteristics of the women. The generalized estimating equations (GEE) method was used to account for the cluster sampling design used in the study. GEE allows one to account for the possible correlation of patient outcomes within a practice, while assuming that patient outcomes from different practices are uncorrelated. This approach allows one to incorporate both patient and practice-level covariates in the analysis, provides better estimates of the standard error of the intervention effect, and results in more appropriate inference. The correlation within practices was modeled using an exchangeable covariance structure, assuming the responses of individuals within a practice are equally correlated. The criterion for statistical significance for all analyses was P 
0.05, two-tailed test.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The results by stage of intervention and cumulative results across all stages are shown in Table II. Overall, the data indicate no significant differences between the intervention and control groups for the physician education and telephone counseling components of the program. Stage 2 of the program, mailed educational materials, shows a small but significant difference for the total sample (P = 0.020) and the primary sample (P = 0.026) in favor of the intervention group. Also, the percentage point difference in the percent screened by mammography after receiving the educational materials [Intervention (I) – Control (C) = 4.4%] is identical for both the primary and maintenance samples. Cumulative differences across all three stages of the intervention indicate no overall significant intervention effect.
|
Controlling for potentially confounding variables (social, demographic and health behaviors) did not change the original results. Tests for interactions between group (intervention versus control) and indicators of socioeconomic status (education and type of insurance), minority status (race) and age (65–79 versus 80 and older) revealed a differential intervention effect. No significant differences were found for women in the higher socioeconomic status categories, white women or younger women in the sample. However, consistent differences by stage of the intervention and type of sample were found for women in the lowest education and insurance categories, black women, and women 80 and older. For example, the cumulative results across all stages of the intervention for the total sample showed a larger percentage of women in the intervention group receiving screening mammography among (1) women over 80 (I = 33%, C = 17%, P = 0.005), (2) black women (I = 50%, C = 29%, P = 0.049), (3) women with less than 9 years of education (I = 38%, C = 19%, P = 0.026) and (4) women with no private Medicare supplemental insurance (I = 35%, C = 24%, P = 0.057).
Typically, considerable overlap exists among these subgroups in the population, e.g. older women and black women are more likely to have less education and/or less insurance coverage. However, while significant, these associations were not particularly strong in the study sample. Since all the characteristics of the identified subgroups can be considered indicators of underserved status, we created a summary variable that totaled the number of underserved characteristics (80 or older, less than 9 years of education, black race, no private insurance to supplement Medicare) for each woman. A majority of women (N = 1197) had no underserved characteristics, 474 women had only one characteristic, 164 women had two, 30 women had three and only one woman had all four characteristics. Complete data were not available for 45 women. For further analysis this variable was defined as ‘underserved status’ and analyzed as a dichotomy (none versus one or more underserved characteristics).
Table III summarizes the tests for interaction between group and underserved status in their association with the outcome variable of screening mammography. Although not entirely consistent, the interaction tests suggest that the program was differentially effective by underserved status for Stages 1 and 2 of the intervention program, as well as for the cumulative program effects across all three stages of the intervention. Only Stage 3 showed no significant interaction between group and underserved status.
|
The results by underserved status are shown in Table IV. These data clearly indicate that (1) significant differences between the intervention and control groups in screening mammography exist only among women with one or more underserved characteristics, and (2) almost all of the differences occurred in Stages 1 and 2 of the intervention program. The cumulative results for women in the primary sample with one or more underserved characteristics indicate that screening mammography was reported for 33% of women in the intervention group and 19.8% of women in the control group. This 13.2 percentage point difference was due to a 6.6 percentage point difference in Stage 1, 5.9 percentage points in Stage 2 and less than 1 percentage point in Stage 3. Among women in the maintenance sample, an overall 25.5 percentage point advantage in favor of the intervention group (I = 58.5, C = 33.0) was due to a 6.4 percentage point difference in Stage 1, 17.6 percentage points in Stage 2 and 1.5 percentage points in Stage 3.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This paper addressed the issue of screening mammography for older women and tested the efficacy of a practice-based intervention to increase screening among women 65 and older. The analysis revealed no overall effect across all three stages of the intervention program. However, tests for interaction and subsequent subgroup analyses suggested a significant program effect for women who were 80 or older, had less than 9 years of education, were black, or had no private insurance to supplement Medicare. In general, women in these subgroups are among those currently least likely to receive screening mammography. Thus, women with one or more of these characteristics were defined as underserved in the present study.
Among underserved women in the primary sample, the 13 percentage point higher screening rate for women in the intervention group was about equally split between Stage 1, physician education, and Stage 2, mailed educational materials (Table IV). The 26 percentage point advantage for underserved women in the maintenance sample who received the intervention showed about the same effect of physician education as for women in the primary sample and a much larger effect of mailed educational materials.
Although the increase in screening following the provision of educational materials to physicians was small for underserved women in both the primary and maintenance samples, the data support earlier work that suggests physician recommendation for screening mammography may be particularly important for older women (Fox et al., 1991; Grady et al., 1992; Costanza, 1994; Ruchlin, 1997). Our data also support the conclusion that simply written mailed educational materials can be useful in increasing participation in screening mammography among underserved older women. Similar results were reported in a study of women 65 and older in southern California (Fox et al., 2001). In that study, mailed educational materials informing women of the Medicare benefit for screening mammography, low-cost screening opportunities and mammography facilities significantly increased screening among black and Hispanic women, but not among white women.
The stronger effect of mailed educational materials for underserved women in the maintenance sample also suggests that these materials can be particularly useful as a reminder, or perhaps an extra motivator, to promote maintenance among women coming due for a mammogram. Previous research suggests that postcard reminders, form letters and tailored letters all can be effective in promoting screening among patient populations [e.g. (Bodiya et al., 1999; Lauver and Kane, 1999; Janz et al., 1999; Saywell et al., 1999; Lipkus et al., 2000)].
One of the most interesting negative results of the present study is the fact that telephone counseling was not an effective component of the intervention program. Telephone counseling is considered one of the most promising approaches to delivering interventions to improve health behavior and health care (McBride and Rimer, 1999) and has been found to be effective in several studies [e.g. (King et al., 1995; Lauver and Kane, 1999; Saywell et al., 1999; Costanza et al., 2000; Preston et al., 2000; Taplin et al., 2000)]. However, a recent large community-based study of telephone counseling to increase screening mammography among women 50 and older found that it was not effective (Stoddard et al., 2002). The authors concluded that telephone counseling has potential, but that further modifications are needed to reach this potential.
In the case of the present study, the telephone counseling sessions were designed to be brief, and the mean session length was only three minutes (range = 1–15 min). In addition, the counselor focused on a woman's stage of readiness regarding mammography screening and the main reason for not getting a mammogram in the past year. While it is clear that the counseling must be individualized for each women, additional research is needed to determine if an optimal session length exists, whether the discussion should routinely include some basic information on screening and the extent to which barriers to screening should be pursued. Furthermore, if the intent is to incorporate telephone counseling into the office routine of a primary-care practice, all of the aforementioned issues must take into account the time and skills of office staff. It may be that truly effective telephone counseling for screening is not possible in the context of a busy primary-care practice.
The results of the present study must be viewed in light of certain design and practical limitations. Randomizing practices to intervention and control conditions instead of women reduced the possibility of contamination through exchange of information between women in the intervention and control groups, but did not eliminate all sources of contamination. All women were told that they would be receiving educational materials on breast and skin cancer and all participating physicians knew that the study was one designed to increase screening mammography. Thus, all participants were sensitized to the issue of screening for breast cancer. To the extent that this information increased the likelihood of women in the control group seeking screening, differences between the intervention and control groups would be reduced, resulting in an underestimate of the effectiveness of the intervention program.
The effectiveness of the intervention program also is somewhat underestimated for women in the maintenance sample, especially for Stage 1. When Stage 1 was completed, just before educational materials were mailed in Stage 2, we estimate that only about 40% of women in the maintenance sample were due for a mammogram. About 80% were due for screening at the beginning of Stage 3. By the end of the program (final record review), all women in the maintenance sample were at least 1 month due for screening. Regardless of screening status, all women in the intervention group could have received the benefit of the physician intervention (Stage 1) and all received the mailed educational materials. The most important consequence regarding Stage 3 is that about 20% of women in the maintenance sample who may have benefited from the counseling component of the program were not yet due for a mammogram and therefore not eligible to receive it. However, the similarity of results for both the primary and maintenance samples suggests that the general conclusions regarding the effectiveness of the program for women defined as underserved were not significantly distorted by this limitation.
It is especially important to emphasize that the summary measure of underserved status used in this study cannot be considered a comprehensive index of underserved status. The measure does not include other important variables such as distance from medical facilities, income or, more importantly for older individuals, total economic assets. The only variables examined in the present study were education, insurance status, minority status and age. Examining differences in screening mammography between the intervention and control groups when controlling for these variables revealed virtually identical results for each variable. The analysis revealed significant intervention effects for underserved subgroups defined as women with low education, those with no insurance other than Medicare/Medicaid, black women and very old women. The measure of underserved status reported here is simply an efficient way of summarizing the results.
The decision to focus on education, insurance status, minority status, and age as the primary indicators of underserved status was based on two factors that guided the entire intervention. First, despite recognition that low literacy is a major barrier to preventive health care, much health care information is not accessible to persons with poor literacy skills (Michielutte et al., 1999). A large proportion of individuals with low education, no insurance and members of minority groups have low literacy skills. Older individuals in general (65 or older) are more likely to have poor literacy skills and the proportion becomes larger with advancing age. A major goal of the intervention program was to develop both written and counseling approaches that would be readily understood by all women regardless of literacy level.
Secondly, considerable confusion exists regarding the appropriateness of mammography for women over 70 and even more so for women over 80. While this issue is not resolved, some evidence suggests that healthy women 70 and older can benefit as much from mammography as women 50–69 [e.g. (Kopans, 1992; Mandelblatt et al., 1992; Rodin and Blesch, 1994/95; Gabriel et al., 1997; Hwang and Cody, 1998; Smith-Bindman et al., 2000; Randolph et al., 2002)]. The intervention components for both physicians and women emphasized that (1) healthy older women can benefit from mammography, and (2) the decision to be screened should be made on an individual basis by both the woman and her physician.
Thus, while the overall lack of intervention effect is disappointing, we are encouraged that the program appears to have been at least modestly successful among those groups of women most likely to benefit from educational materials designed for low-literacy audiences and from materials emphasizing the potential value of mammography for older women. Few preventive health-care initiatives have received the attention given mammography screening in the past two decades. It may be that most women who understand the messages and who have ready access to health care have made their decision regarding screening. This would help explain the overall lack of effect of our program. This interpretation of the intensive effort to date to increase mammography screening also suggests that future programs to increase screening, if they are warranted, should focus on underserved subgroups in the population.
Finally, potential selection bias and the geographic characteristics of the practices enrolled in the study impact the generalizability of the results. Less than one-third of the practices contacted eventually participated in the study. As noted earlier, the practices that declined to participate in the study did not feel they had enough eligible women, had extensive physician and staff turnover, went out of business before the study began or were not interested in the study. It may be that the results can be generalized only to stable primary-care practices with a fairly large geriatric patient population and an interest in prevention. One way to test this possibility would be to test the intervention program through an HMO that could enlist the participation of all enrolled primary-care practices.
One of the most important reasons for the initial response rate of 67% and subsequent 11% attrition rate of women enrolled in the study was the time frame of the study. Each practice was on a unique schedule. In order to maintain a consistent schedule for recruitment of women in each clinic and implementation of the intervention, both baseline interviews and follow-up contact for the second interview and counseling session had to be conducted in a limited amount of time (2 weeks). Women who could not be contacted in that time period and many of those who agreed to participate, but wanted to be contacted at a later time, were lost to the study. Limited comparisons from medical records of the characteristics of women who declined initial participation in the study or who could not be contacted, and all comparisons of those who were lost to follow-up (from the baseline interview) indicated no differences between the intervention and control groups. Overall, we found no evidence of bias due to response rate.
A more significant limitation of the study is the focus on the private sector of primary care and the geographic location of the practices. The participating practices did not include public health clinics and were located in rural areas, small towns and moderate-size cities. Large metropolitan areas were not represented in the study.
The characteristics of the participating clinics, absence of metropolitan areas, together with the fact that the intervention program was only effective for women defined as underserved in the present study, clearly indicate that the results have limited generalizability.
Despite these limitations, however, the results of the study suggest that provision of primary-care physicians with information about screening older women and providing the women with simply written educational materials, can increase participation in screening mammography among women in subgroups at highest risk of not being screened. Telephone counseling was not found to be an effective component of the intervention program, but merits further study. Additional refinement and testing of this three-stage intervention for screening mammography as well as for other health behaviors could result in an effective program for health education that can be readily implemented in the primary-care setting.
|
Acknowledgments |
|---|
The research reported in this paper was supported by a grant from the National Cancer Institute (R01-CA79746).
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
American Cancer Society (1999) Breast Cancer Facts and Figures: 1999–2000. American Cancer Society, Atlanta, GA.
Anonymous (1997) Self-reported use of mammography among women aged 40 years and older—United States, 1989 and 1995. Morbidity and Mortality Weekly Report, 46, 937–941.
Becker, M.H. (ed.) (1974) The Health Belief Model and Personal Health Behavior. Slack, Thorofare, NJ.
Bodiya, A., Vorias, D. and Dickson, H.A. (1999) Does telephone contact with a physician's office staff improve mammography screening rates? Family Medicine, 31, 324–326.[Medline]
Breen, N. and Kessler, L. (1994) Changes in the use of screening mammography: evidence from the 1987 and 1990 National Health Interview Surveys. American Journal of Public Health, 84, 62–67.
Breen, N., Wagener, D.K., Brown, M.L., Davis, W.W. and Ballard-Barbash, R. (2001) Progress in cancer screening over a decade: results of cancer screening from the 1987, 1992 and 1998 National Health Interview Surveys. Journal of the National Cancer Institute, 93, 1704–1713.
Costanza, M.E. (1994) The extent of breast cancer screening in older women. Cancer, 74, 2046–2050.[CrossRef][Web of Science][Medline]
Costanza, M.E., Stoddard, A.M., Luckmann, R., White, M.J., Avrunin, J.S. and Clemow, L. (2000) Promoting mammography: results of a randomized trial of telephone counseling and a medical practice intervention. American Journal of Preventive Medicine, 19, 39–46.[CrossRef][Web of Science][Medline]
Demisse, K., Mills, O.F. and Rhoads, G.G. (1998) Empirical comparison of the results of randomized controlled trials and case-control studies in evaluating the effectiveness of screening mammography. Journal of Clinical Epidemiology, 51, 81–91.[CrossRef][Web of Science][Medline]
Fox, S.A., Murata, P.J. and Stein, J.A. (1991) The impact of physician compliance on screening mammography for older women. Archives of Internal Medicine, 151, 50–56.
Fox, S.A., Stein, J.A., Sockloskie, R.J. and Ory, M.G. (2001) Targeted mailed materials and the Medicare beneficiary: increasing mammogram screening among the elderly. American Journal of Public Health, 91, 55–61.[Abstract]
Gabriel, H., Wilson, T.E. and Helvie, M.A. (1997) Breast cancer in women 65–74 years old: earlier detection by mammographic screening. American Journal of Roentgenology, 168, 23–27.
Glanz, K. and Rimer, B.K. (1995) Theory at a Glance: A Guide for Health Promotion Practice. NIH publ. no. 95-3896. US Department of Health and Human Services, Public Health Service, National Institutes of Health, Washington, DC.
Grady, K.E., Lemkau, J.P., McVay, J.M. and Reisine, S.T. (1992) The importance of physician encouragement in breast cancer screening of older women. Preventive Medicine, 21, 766–780.[CrossRef][Web of Science][Medline]
Harris, R.P., Fletcher, S.W., Gonzalez, J.J., Lannin, D.R., Degnan, D. and Earp, J.A. (1991) Mammography and age: are we targeting the wrong women? Cancer, 67, 2010–2014.[CrossRef][Web of Science][Medline]
Hwang, E. and Cody, H.S. (1998) Does the proven benefit of mammography extend to breast cancer patients over age 70? Southern Medical Journal, 91, 522–526.[Web of Science][Medline]
Janz, N.K., Schottenfield, D., Doerr, K.M., Selig, S.M., Dunn, R.L., Strawderman, M. and Levine, P.A. (1999) A two-step intervention to increase mammography among women aged 65 and older. American Journal of Public Health, 87, 1683–1686.
Kerlikowske, K., Grady, D, Rubin, S.M., Sandrock, C. and Ernster, V.L. (1995) Efficacy of screening mammography: a meta-analysis. Journal of the American Medical Association, 273, 149–154.
King, E.S., Balshem, A., Ross, E., Rimer, B. and Seay, J. (1995) Mammography interventions for 65-to-74-year old HMO women. Journal of Aging and Health, 7, 529–551.
King, E.S., Rimer, B.K., Seay, J., Balshem, A. and Engstrom, P.F. (1994) Promoting mammography use through progressive interventions: is it effective? American Journal of Public Health, 84, 104–106.
Kopans, D.B. (1992) Screening mammography in women over age 65. Journal of Gerontology, 47, 59–62.
Lauver, D.R. and Kane, J. (1999) A motivational message, external barriers and mammography utilization. Cancer Detection and Prevention, 223, 254–264.
Lipkus, I.M., Rimer, B.K., Halabi, S. and Strigo, T.S. (2000) Can tailored interventions increase mammography use among HMO women? American Journal of Preventive Medicine, 18, 1–10.[Web of Science][Medline]
Makuc, D.M., Fried, V.M. and Parsons, P.E. (1994) Health Insurance and Cancer Screening among Women. Advance Data from Vital and Health Statistics 254. National Center for Health Statistics, Hyattsville, MD.
Mandelblatt, J.S., Wheat, M.E., Monane, M., Moshief, R.D., Hollenberg, J.P. and Tang, J. (1992) Breast cancer screen for older women with and without comorbid conditions. Annals of Internal Medicine, 116, 722–730.
McBride, C.M. and Rimer, B.K. (1999) Using the telephone to improve health behavior and health service delivery. Patient Education and Counseling, 37, 3–18.[CrossRef][Web of Science][Medline]
Michielutte, R., Alciati, M.H. and el Arculli, R. (1999) Cancer control and literacy. Journal of Health Care for the Poor and Underserved, 10, 281–297.[Web of Science][Medline]
National Center for Health Statistics (2000) Data File Documentation, National Health Interview Survey, 1997. Machine readable data file and documentation, CD-ROM series 10, no. 12A. National Center For Health Statistics. Hyattsville, MD.
Ostbye, T., Greenberg, G.N., Taylor, D.H. and Lee, A.M.M. (2003) Screening mammography and Pap tests among older women 1996–2000: results from the Health and Retirement Study (HRS) and Asset and Health Dynamics Among the Oldest Old (AHEAD). Annals of Family Medicine, 1, 209–217.
Overmoyer, B. (1999) Breast cancer screening. Medical Clinics of North America, 83, 1443–1466.
Phillips, K.A., Kerlikoske, K., Baker, L.C., Chang, S.W. and Brown, M.L. (1998) Factors associated with women's adherence to mammography screening guidelines. Health Services Research, 33, 29–53.[Web of Science][Medline]
Preston, J.A., Scinto, J.D., Grady, J.N., Schulz, A.F. and Petrillo, M.K. (2000) The effect of a multifaceted physician office-based intervention on older women's mammography use. Journal of the American Geriatrics Society, 48, 1–7.[Web of Science][Medline]
Prochaska, J.O. and DiClemente, C.C. (1982) Transtheoretical therapy: toward a more integrative model of change. Psychotherapy: Theory, Research and Practice, 20, 161–173.[Web of Science]
Prochaska, J.O., Velicer, W.F., Rossi, JS., Goldstein, M.G., Marcus, B.H., Rakowski, W., Fiore, C., Harlow, L.L., Redding, C.A., Rosenbloom, D. and Rossi, S.R. (1994) Stages of change and decisional balance for 12 problem behaviors. Health Psychology, 13, 39–46.[CrossRef][Web of Science][Medline]
Randolph, W.M., Goodwin, J.S., Mahnken, J.D. and Freeman, J.L. (2002) Regular mammography use is associated with elimination of age-related disparities in size and stage of breast cancer at diagnosis. Annals of Internal Medicine, 137, 783–790.
Rosenstock, I.M., Strecher, V.J. and Becker, M.H. (1988) Social learning theory and the health belief model. Health Education Quarterly, 15, 175–183.[Web of Science][Medline]
Rodin, M.B. and Blesch, K.S. (1994/95) Why perform mammograms on the oldest women? Making decisions for screening women beyond age 75. Journal of the Robert H. Lurie Cancer Center, 4, 16–22.
Ruchlin, H.S. (1997) Prevalence and correlates of breast and cervical cancer screening among older women. Obstetrics and Gynecology, 90, 16–21.[CrossRef][Web of Science][Medline]
Saywell, R.M., Champion, V.L., Skinner, C.S., McQuillen, D., Martin, D. and Maraj, M. (1999) Cost-effectiveness comparison of five interventions to increase mammography screening. Preventive Medicine, 29, 374–382.[CrossRef][Web of Science][Medline]
Smith-Bindman, R., Kerlikowske, K., Gebretsadik, T. and Newman, J. (2000) Is screening mammography effective in older women? American Journal of Medicine, 108, 112–119.[Web of Science][Medline]
Stoddard A.M., Fox S.A., Costanza M.E., Lane D.S., Andersen M.R. and Urban N. (2002) Effectiveness of telephone counseling for mammography: results from five randomized trials. Preventive Medicine, 34, 90–99.[CrossRef][Web of Science][Medline]
Taplin, S.H., Barlow, W.E., Ludman, E., MacLehos, R., Meyer, D.M., Seger, D., Herta, D., Chin, C. and Curry, S. (2000) Testing reminder and motivational telephone calls to increase screening mammography: a randomized study. Journal of the National Cancer Institute, 92, 233–242.
Received on September 2, 2003; accepted on April 27, 2004
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. M. Murray, S. L. Pals, J. L. Blitstein, C. M. Alfano, and J. Lehman Design and Analysis of Group-Randomized Trials in Cancer: A Review of Current Practices J Natl Cancer Inst, April 2, 2008; 100(7): 483 - 491. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Montes, L. M. Seijo, A. Campo, A. B. Alcaide, G. Bastarrika, and J. J. Zulueta Factors determining early adherence to a lung cancer screening protocol Eur. Respir. J., September 1, 2007; 30(3): 532 - 537. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||