Breast Cancer Perceptions, Knowledge and Behavioral Practices among Women Living in a Rural Community

Abstract

We performed this study to assess women's perceptions, knowledge and behavioral practices for breast cancer prevention in a rural setting. A 61-item questionnaire was developed based on Health Belief Model constructs and completed by 185 women age 35 and older. Results showed significant differences in several areas including perceived susceptibility and severity. Overall knowledge was poor. In logistic regression perceived barriers and yearly clinical breast examination appeared to be significant predictors for regular screening behavior (OR=0.02, CI=0.03-0.09 and OR=0.23, CI=0.05-0.99, respectively). Behavioral interventions targeting barriers for rural women need to be designed to include consideration of specific barriers and clear information on the need for regular screening.

Introduction

An abundance of evidence suggests that there are clear disparities in utilization of preventive services by rural populations. A study based on Behavioral Risk Factor Surveillance showed that women residing in nonmetropolitan areas were less likely to receive mammograms or Pap smears in accordance with recommended guidelines than their urban counterparts (Casey, Call, and Klinger, 2001). Similar findings are also shown in studies based on Medicare Current Beneficiaries Survey (Stearns et al., 2000) and the National Health Interview Survey (Zhang, Tao, and Irwin, 2000). Underutilization of preventive health care services may result in a failure of identifying health problems in time and missing opportunities to reduce mortality or morbidity. A review of cancer incidence in rural versus urban populations found that cancer tends to be diagnosed at more advanced stages among rural populations (Monroe, Ricketts, and Savitz, 1992; Liff, Chow, and Greenburg, 1991), suggesting that rural residents are less likely to receive timely cancer screening tests. In addition, rural residence has been found to be a strong predictor of mammography underuse (Casey, Call, K and Klinger, 2001; Rettig, Nelson, and Faulk, 1994). Although preventive health care utilization has increased in recent years and relationship between rural and urban residence has been quantified in previous literature, there is relatively little information available focusing on psychological, social and behavioral factors of rural women in relation to screening mammogram.

The American Cancer Society (ACS) estimates that 182,460 new cases, and 40,480 deaths from breast cancer, will occur among women in the United States in 2008 (ACS, 2008). Overall, the rate of screening for breast cancer in the US is gradually increasing and breast cancer mortality has declined slightly. In spite of increasing screening and decreasing mortality, the mortality rates from breast cancer remain unacceptably high. Despite technological advancement in screening for breast cancer, mammography remains as the single most cost-effective method of screening for breast cancer. If used optimally, mammography could prevent 15-30% of all deaths from breast cancer through early detection (CDC, 2000). Results from several large randomized clinical trials suggest that mammography is associated with reductions of breast cancer mortality up to 39% among the 50-69 age group women (Day, 1991). Improving understanding of psychological, socio-economic and environmental factors that may influence screening behavior is a critical element of developing programs to reduce breast cancer morbidity and mortality.

Given the limited information on factors associated with participation in breast cancer screening by rural women, this project was conducted this study to measure knowledge, perceptions and behavioral practices related to breast cancer prevention among women living in a rural community in Ohio.

Methods

Participants

The study participants were recruited from Wood County, Ohio. According to Ohio Department of Development, Office of Strategic Research data (2006), the total Wood county population is 121,065, out of which 62,461 are women. The 40 years and older age group comprised of 25,740 women who are eligible for yearly clinical breast exam and/or mammography based on the age and family history of breast cancer. Wood County is rural and its inhabitants include 4033 (3.8%) Hispanic, 1,864 (1.6%) African American and 1,514 (1.3%) Asian and Pacific Islanders among the 121,065 population. According to the Lucas and Wood County Chapter of the American Cancer Society, cancer was the second leading cause of death in 2004.

Study Design

A survey of women age 35 and older was conducted in Wood County, OH, focusing on breast cancer perceptions, knowledge, and behaviors. The survey was conducted between March 2004 and January 2005. A sample of 500 women age 35 and older was randomly selected from the approximate 17,000 county population of women of the same age range. The lists of 35 years and older living in the Wood County, Ohio were generated by a third party vendor. A personalized cover letter, along with the questionnaire, a self-addressed and stamped envelope and an incentive in the form of a crisp, new one dollar bill were mailed. A reminder letter was sent after two weeks to those who did not respond. A second reminder post-card was sent to the non-responders after four weeks. Anonymity was maintained by assigning codes for each questionnaire and envelop. A total of 160 completed surveys were returned in the first wave, and 90 mails were returned as undelivered. Thirty completed surveys were returned in the second wave and another 45 mails were returned as undelivered. Five surveys were eliminated from the analysis because of large numbers of missing responses leaving a final study sample of 185, which translates, into a returned rate of approximately 52%. Prior to initiating data collection, approval was attained from the Institutional Review Board (IRB).

The Questionnaire

Data was collected using a 61-item questionnaire that was developed based on the constructs of Health Belief Model. Four items each were included on perceived susceptibility, severity of breast cancer and benefits of having mammogram. Seventeen items on barriers to obtaining mammograms were also included. The barrier items were developed based on the reported barriers to screening mammogram in a comprehensive literature search. To measure knowledge about breast cancer and screening mammography seventeen items on known risk factors of breast cancer, symptoms of breast cancer, and misconception about breast cancer were included. Demographics and previous behavioral practices such as previous mammogram, clinical breast examination and regular health visits were also included. To ensure content and construct validity, questionnaire was designed based on several published literature on the perception and knowledge related to breast cancer and mammogram (Friedman, Moore, Webb, and Puryear, 1999; Michielutte, Dignan, and Smith, 1999; McDonald, Throne, Pearson, and Adams-Campbell, 1998; Crane, Kaplan, Bastani, and Scrimshaw, 1996).

Theoretical Underpinnings

According to the Health Belief Model Perceived Susceptibility has been defined as individual's subjective perception of his/her risk of contracting a health condition. For cases of medically established illness, such as breast cancer, perceived susceptibility includes acceptance of the diagnosis, personal estimates of susceptibly and susceptibility to illness in general (Strecher and Rosenstock, 1997).

Perceived Severity is an Individual's feelings concerning the seriousness of contracting an illness or leaving it untreated. Perceived severity also includes the evaluation of medical and clinical consequences (death, disability, pain) and possible social consequences (work, family life and social relations) associated with a disease. Individual perceptions of personal susceptibility to specific illnesses or accidents often vary from any realistic appraisal of their statistical probability (Strecher and Rosenstock, 1997). Perceived Benefits is defined as the decision about a course of action taken depends on beliefs regarding the effectiveness of the various available actions in reducing the disease threat, for example having a mammogram

Perceived benefits also include non-health-related benefits, such as, having peace of mind from performing regular screening mammogram The anticipated value of taking the recommended course of action is a final consideration (Strecher and Rosenstock, 1997). Perceived Barriers are the potential negative aspects of a particular health action may act as impediments to undertaking the recommended behavior. Perceived barriers include the costs, time, difficulty, or pain of taking a particular action.

Data Management and Statistical Analysis

Original data was entered into MS Excel and converted into SPSS and SAS data format using Database Management System (DBMS). Data cleaning, coding and recoding was done before performing data analyses. Both elementary and inferential analyses were done using univariate and multivariate methods using SPSS 11.0 and SAS 9.1 version.

Results

The sociodemographic characteristics of the study respondents are summarized in Table 1. The study population was all female with a mean age of 38.23 years (SD=9.19). Almost all of the participants were white (95%) and all non-whites were African American. Fifty eight percent reported high school and/or some college, and another eighteen percent were college graduates. Fifty one percent were working full-time, 28% were retired, and almost 11% reported annual incomes below $10,000. Twenty five percent had income level more than or equal to $50,000. Regarding health insurance, 8.2% reported not having any health insurance, 23.1% had private health insurance, 38.5% had HMO, 28% had Medicare and 11% had other types of health insurance.

Table 1.

Sociodemographic Characteristics s of the Study Respondents

	N (N= 185)	%
Age
<40	9	5.0
40-49	58	31.4
50-59	44	23.8
60-69	30	16.0
70 and above	44	23.8
Race
White	176	95.2
African American	5	2.7
Asian and PI	1	0.5
Hispanic	2	1.1
Other	1	0.5
Educational attainment
< High school	9	4.9
High school grad.	56	30.4
Some College or technical school	50	27.2
College Grad.	33	17.9
> Grad. Or professional	36	19.6
Marital status
Married	37	20.0
Single	31	16.8
Separated	3	1.6
Divorced	63	34.0
Widowed	51	27.6
Employment status
Working full-time	94	51.4
Working part-time	11	6.0
Retired	52	28.4
Never worked	16	8.7
Not working now	10	5.5
Income level
< $ 10,000	18	10.8
$ 10,000-$ 19,999	34	20.5
$ 20,000-$ 29,999	26	15.7
$ 30,000-$ 39,000	27	16.3
$ 40,000-$ 49,000	18	10.8
> $ 50,000	43	25.9
Insurance status
Yes	169	91.8
No	15	8.2

Prior breast cancer screening behavior, health care utilization behavior and availability of health insurance were examined. Over 80% of the respondents reported ever having a mammogram, and of those, 76.4% had the mammogram within the past 12 months and 23.6 had a mammogram 2 years back. The remaining respondents (14%) reported never having had a mammogram. For clinical breast examination, 76.4% reported having had a clinical breast examination; among which 83.5% within the past year and 16.5% within the past 2 years. Over 90% of the respondents reported having regular health care visits; among which 98.1% within the past year, 1.3% between 1 and 2 years back and only 0.6% more than 2 years back. Age-specific screening and health care utilization has been summarized in table 2.

Table 2.

Age-Specific Screening History and Health Care Utilization

Percentage of respondents
Age (years)	MM*Ever	MM 1 year ago	CBE MM*Ever	CBE 1 year ago	HCV	HCV 1 years ago
<40	22.2	100.0	100.0	100.0	88.9	100.0
40-49	82.8	76.3	93.1	85.1	96.6	100.0
50-59	90.9	77.4	79.6	88.0	93.2	97.5
60-69	93.3	78.3	93.3	71.4	86.7	100.0
=> 70	93.2	73.5	68.3	80.0	83.7	100.0

^*Significantly different, P < 0.05, by χ² analysis

Breast Cancer Knowledge

Breast cancer and screening related knowledge was assessed using 17 true-false items. The items asked about early signs and symptoms of breast cancer, breast cancer risk factors, misconception about breast cancer and screening, and general knowledge of management of breast cancer. Only 1.08% women had correct answers on all items. The mean number of items answered correctly among the study population was 13 out of total 17 items (SD=1.99). Overall the knowledge on risk factors of breast cancer was low and misconception about breast cancer was prevalent. Almost 35% women believed that being hit in the breast may cause a woman to get breast cancer later in life. About 47% women knew swelling in the breast as possible sign of breast cancer. Seventy four percent women did not think that breast cancer is more common in 65-year-old women than in 40-year-old women. About one third of the women did not know that one out of every nine women in the United Stats would develop breast cancer by the age of 85. About 41% women believed that fibrocystic breast disease (breast lumps that are not cancer) increase a woman's risk of breast cancer, and 34% women did not know that breast cancer is the most common cancer in women.

Breast Cancer Perceptions

Perceived Susceptibility: A significant differences in perceptions of susceptibility were found. In four questions on susceptibility majority of the women disagreed- developing breast cancer in few years (61%), chance higher compared to other women (74%), and life-time risk (78%) with the p value 0.003, 0.001 and <0.001 respectively (Table 3).

Table 3.

Perceived susceptibility, Severity and Benefit Constructs

Perceived Susceptibility	Agree	Disagree	χ 2	P
My chances of getting breast cancer in next few years is great	39.01	60.99	8.79	0.003
Your chances of getting breast cancer in relation to average women is higher	25.82	74.18	42.55	0.001
I feel I will get breast cancer sometime during my life	21.67	78.33	57.8	< 0.001
I believe all women are equally likely to develop breast cancer	32.61	67.39	22.26	< 0.001
Perceived Severity
If I had breast cancer, I would be worried and depressed	87.03	12.97	101.45	<0.001
If I had breast cancer, I would have to have my breast taken off by surgery	27.72	72.28	36.54	<0.001
If I had breast cancer, it would cause me to die	12.43	87.57	104.44	<0.001
Perceived Benefits
I believe breast cancer can be cured easily	29.89	70.11	29.76	<0.001
If I get a mammogram and nothing is found, I would not worry about breast cancer	44.57	55.43	2.17	0.14
If I get a mammogram and nothing is found, I would find peace of mind.	82.61	17.39	78.26	<0.001
Regular mammogram will help finding breast lumps early and can help save my breast	93.51	6.49	140.1	<0.001
Having a regular mammogram would help my doctor save my life.	92.43	7.57	133.23	<0.001

Perceived Severity: Most of the women responded that they considered breast cancer as a serious disease because it will cause them worried and depressed (87%), however they disagreed on the question on surgery (72%) and breast cancer as cause of death (88%) with the p value <0.001.

Perceived Benefits of mammograms: Nearly all of the respondents reported substantial perceived benefits of mammograms and early detection. Almost 93% reported believing that regular mammograms will help finding breast lumps early and thus help save their lives. Over 80% agreed that if nothing is found in mammography, it would bring peace of mind. However, 70% women disagreed that breast cancer can be cured easily. Almost equal number of women split up with the thought that if nothing was found in mammogram they would not worry about breast cancer (agree= 45% and disagree=55%, p=0.14).

Perceived Barriers to getting mammograms: About half (52%) of the respondents felt that gender of the provider is a barriers to mammograms, and same percentage of respondents indicated that they preferred to be examined by a female physician. Thirty one percent of women considered not having enough money as a barrier. However, 86% women agreed on the statement that even if mammograms are expensive, if doctors suggested that they should get it, they would get it (p=0.001).

In this study the reliability of the questionnaire was assessed by item correlation. Both raw and standardized Cronbach Alpha value were over 80% (α = 0.80).

Logistic Regression

A logistic regression analysis was done to examine combinations of factors associated with regular screening mammography behavior. Previous mammogram within a year (yes versus no) was regressed with age, race, educational attainment, employment status, household income, marital status, perceptions of susceptibility, and severity of breast cancer, perception of benefits and barriers related to mammograms, recent health care visit, and clinical breast examination in the past year. Perceptions were measured in Likert scale. We created an average score on perception scale of all questions dealing with that specific perception. For example, four separate questions were asked for perceived susceptibility. In the logistic regression, average score of these four questions was used as a continuous variable of perceived susceptibility. Same procedure was followed to create continuous variable on perceived severity, benefit, barrier and also for knowledge. Both univariate and multivariate regression analyses were carried out and results are summarized in table 5.

Table 5.

Crude and Adjusted Odds Ratios for Factors Associated with Screening Behavior

Factors	Crude OR	95% CI	Adjusted OR	95% CI
Age
40-49	1.00		1.00
<40	0.13	0.02-1.06	0.17	0.01-2.25
50-59	1.20	0.55-2.63	0.83	0.25-2.76
60-69	1.50	0.61-3.67	1.46	0.24-8.81
>=70	1.32	0.60-2.89	0.86	0.08-9.75
Race/Ethnicity
White	1.00		1.00
Non-White	0.44	0.11-1.80	0.09	0.01-1.52
Education
< High school graduate	1.00		1.00
High school graduate	1.00	0.26-3.84	3.35	0.29-38.60
Some college and technical	0.85	0.22-3.31	1.78	0.14-22.78
College and Professional	1.46	0.39-5.53	4.80	0.37-62.74
Employment Status
Full or part-time working	1.00		1.00
Retired	1.40	0.71-2.76	1.23	0.16-9.53
Never worked or not working now	0.39	0.16-0.97*	0.34	0.06-1.87
Household Income
< $10,000	1.00		1.00
$10,000-$19,999	1.31	0.52-3.35
$20,000-$29,999	2.10	0.76-5.84	3.62	0.67-19.63
$30,000-$39,999	1.41	0.52-3.83	1.31	0.29-6.00
$40,000-$49,999	1.31	0.42-4.01	0.77	0.11-5.26
>=$50,000	1.82	0.75-4.43
Marital Status
Married	1.00		1.00
Divorced	0.75	0.33-1.71	0.56	0.14-2.31
Single	0.43	0.16-1.1*	0.48	0.10-2.41
Windowed	0.89	0.38-2.12	2.67	0.46-15.64
Separated	0.34	0.03-4.10	2.88	0.14-58.00
Perceptions
Perceived Susceptibility	1.79	0.79-4.09	2.01	0.71-5.66
Perceived Severity	0.54	0.22-1.28	0.53	0.18-1.59
Perceived Benefits	1.55	0.39-6.13	2.60	0.36-18.91
Perceived Barriers	0.04	0.01-0.14*	0.02	0.03-0.09*
Recent Health Care Visit
Yes	1.00
No	0.17	0.05-0.61*	0.13	0.02-1.00*
Clinical Breast Examination in a Year
Yes	1.00
No	0.38	0.14-1.04	0.23	0.05-0.99*

Note. OR = Odds ratio; CI = Confidence interval

^*= statistically significant.

According to the univariate analysis employment status was important to getting mammograms. Women who had never been employed or who were currently unemployed were 61% less likely to have a yearly mammogram than those working part-time or fulltime (Crude OR=0.39, CI=0.16-0.97). Unmarried women were 57% less likely to have a yearly mammogram compared to married women (Crude OR=0.43, CI=0.16-1.1). Perceived barriers appeared to be a significant predictor of regular mammogram both in univariate and in adjusted model. Women were 96% less likely to have yearly mammogram for one unit increase in barrier scale (Crude OR=0.04, CI=0.01-0.14). Recent health care visit was also a significant predictor both in univariate and multivariate analyses. Women who did not visit for heath care within a year were 83% less likely to have a yearly mammogram (Crude OR=0.17, CI=0.05-0.61). Other variables were not statically significant in univariate analysis.

After controlling for all other variables perceived barriers, recent health care visit and clinical breast examination within previous one year appeared to be significant predictors of regular screening behavior. Women were 98% less likely to have yearly mammogram for one unit increase in barrier scale (Adjusted OR=0.02, CI=0.00-0.09).

Women who did not visit for heath care within past one year were 87% less likely to have a yearly mammogram, though the significance level was at borderline (Adjusted OR=0.13, CI=0.02-1.00). Women who did not have clinical breast examination within past one year were 77% less likely to have a yearly mammogram compared to those who had clinical breast examination within a year (Adjusted OR=0.23, CI=0.05-0.99). We examined the goodness of fit of this model by Hosmer and Lemeshow Goodness of Fit test which indicates model's goodness of fit was acceptable (Chi-sq=11.19, p=0.19).

Discussion

The study population was selected from a rural community of Ohio from a randomly generated database provided by a third party vendor, which provides a basis for generalized conclusion; however, a large number of returned mails due to wrong addresses may have lead to a selection bias. In this study population a small group of women (n=9) were less than 40 years of age. We have included them in the analysis since there was no significant difference excluding them from the analysis. The age-specific screening history and health care utilization suggests that the study population had more than state average screening rate 61.4% (CDC, BRFSS 2006) in all age groups, 76.3%, 77.4%, 78.3% and 73.5% in 40-49, 50-59, 60-69 and =>70 years respectively. Clinical breast examination within past one year was also similar except for the age group 60-69, who are below the state average (71.4% versus 78.7%).

Perceived susceptibility did not appear to be a significant predictor of having had a mammogram in this study population, however, the perception on susceptibility is noteworthy. In all four questions majority of the women did not perceive themselves as susceptible to develop breast cancer. Women in this study had high severity perception and they also perceived mammography as beneficial for cancer prevention. However, 55% women did not think that if nothing was found in screening mammogram they would not worry about breast cancer. Approximately 72% women did not think that if they were diagnosed with breast cancer surgical removal of breast will be done. In one hand these women were concern about being worried and depressed if they were diagnosed with breast cancer, on the other hand, they do not believe that surgery will be done as treatment; fear or denial might have played a role in this situation. About fear or denial in other population similar findings were found (Rahman, Dignan, and Shelton, 2003; Rahman, Mohamed, and Dignan, 2003; Rahman, Dignan, and Shelton, 2005).

Recent health care visit appeared to be a significant predictor associated with yearly mammography behavior both in univariate and multivariate analysis, though after adjusting for all other variables it was at the borderline significance level. In previous studies, women who visited a gynecologist as usual care physician had highest rate of mammography (Finison, Wellins, Wennberg, and Lucas, 1999). Physician's recommendation or motivational suggestion from health care professional was found to be effective in promoting mammography (Fox, Klos, and Tsou, 1988; Fox, Murata, and Stein, 1991; Burns, Freund, Ash, Shwartz, and Antab, 1995). In this study 86% women agreed that irrespective of expenses, they would have had a mammogram if it was recommended by their physicians. Clinical breast examination within past one year also appeared as a significant factor associated with yearly screening mammogram. This finding is consistent with the previous studies where regular health care visit or other screening behavior influenced screening mammography.

This study has several limitations. Though the questionnaire was sent to a total of 500 randomly selected women, a large number of returned mails due to wrong address may have caused selection bias. We followed the empirical findings on survey cash incentives between no cash versus one dollar incentives (James, and Bolstein,1992; Lesser, Dillman, Lorenz, Carlson, and Brown, 1999) where they found 12 percent point increase return, and only an additional two to seven percent point increase for five and ten dollar incentives. In our study most of the women who returned the completed survey also returned the one dollar bill. Another limitation is we did not have any objective assessment of breast cancer screening behavior. However, several studies have found high validity of self-reported breast cancer screening behavior and considered self-report as useful information (Paskett et al., 1996; Zapka et al., 1996). Apart from these limitations the study findings clearly show that the rural population are lacking in knowledge about breast cancer, screening mammograms and have low perceived susceptibility of breast cancer. Future behavioral interventions should focus on novel approaches to counteract the perceived barriers to increase unitization of screening and to prevent breast cancer in this population.

Figure 1.

Health Belief Model Components and Linkages (Hochbaum, 1956, Rosenstock, 1966, Becker, 1974 & 1977)

Table 4.

Perceived Barriers to Screening Mammograms

Perceived Barriers	Agree	Disagree	χ 2	P
I don't want to know if I have breast cancer or not.	5.43	94.57	146.17	0.001
Not having enough money would keep me from having a mammogram.	31.15	68.85	26.01	<0.001
I do not know where a woman can go to get a regular mammogram.	5.43	94.57	146.17	0.001
I think having a regular mammogram is too embarrassing.	8.65	91.35	126.5	<0.001
I think having a regular mammogram takes too much time.	4.32	95.68	154.38	<0.001
I would not have regular mammograms because it is likely to be painful	6.56	93.44	138.15	<0.001
Even if mammograms are expensive, if my doctor told me I should get one I would get it.	85.87	14.13	94.69	<0.001
I have trouble with transportation and that would keep me from having regular mammograms.	9.78	90.22	119.04	<0.001
I have other problems more important than having a regular mammogram.	14.13	85.87	94.70	<0.001
I think the people who give the mammograms are not careful.	3.83	96.17	156.07	<0.001
I do not have anyone to take care of my kids while I go to have mammogram.	3.14	96.86	139.63	<0.001
I would not agree to have a regular mammogram, as I do not trust mammograms.	4.32	95.68	154.39	<0.001
If I find I have breast cancer, people will treat me differently, so I don't want to have mammogram.	8.11	91.89	129.86	<0.001
I would not go for a regular mammogram, as I dislike being examined by a male physician.	3.24	96.76	161.78	<0.001
I would not go for a regular mammogram, as the result may not be kept confidential.	4.35	96.65	153.39	<0.001
I never heard or read anything encouraging having regular mammogram.	4.40	95.60	151.41	<0.001
I will prefer to have my breast examined by a female physician rather than by a male physician.	51.91	48.09	0.26	0.68