Cervical Cancer Screening Practice and Knowledge Among Hispanic Migrant and Seasonal Farmworkers of Michigan

Abstract

Background

High incidence of cervical cancer among Hispanics and low utilization of cervical screening among farmworkers led us to examine Pap test use and knowledge among Hispanic farmworkers in Michigan.

Methods

Patients and potential patients of Northwest Michigan Health Services, Inc were surveyed in 2 communities (A and B) about their screening knowledge, attitudes, and practice.

Results

In all, 324 farmworkers participated, including 184 seasonal and 125 migrant farmworkers. Among the entire study population, 87.7% women reported receiving a Pap test recently, with no difference between migrant and seasonal farmworkers (88.0% and 87.4%, respectively, P = .088). More women from community B reported a recent Pap (93%) compared with those from community A (83%, P = .01). Only 35% of the participants had knowledge of any cervical cancer risk factors.

Discussion

Migrant farmworkers may not experience more difficulty in accessing Pap tests than seasonal farmworkers in Michigan. Knowledge about cervical cancer risk factors is low and needs to be addressed in future educational interventions.

Background

Minorities in the United States, especially Hispanics, are disproportionately affected by cervical cancer as evidenced by the SEER incidence rates of 11.1 per 100 000 for Hispanics compared with 7.4 per 100 000 for non-Hispanic whites.¹ Current studies may not be capturing those Hispanics that are most at risk for cervical cancer, such as farmworkers. This is an important, potentially complicating factor in the cervical cancer disparity between Hispanic and non-Hispanic white women. Farmworkers can be considered at risk for cervical cancer for many reasons. The overall low level of income of the population often leads to long working hours and disincentives to seek health care.² This, in turn, can result in poor utilization of health care and an emphasis on acute care rather than preventive care.³ Current data on Hispanics likely underestimate the disparity between Hispanics and non-Hispanics as poor Hispanics and recent immigrants, including farmworkers, are likely underrepresented in cancer registry data and not included in official reports because of the transience required by their occupation.²

The utilization of the Papanicolaou (Pap) test, which can detect cervical abnormalities and lower cervical cancer incidence and mortality via timely intervention, also potentially complicates the cervical cancer disparity between Hispanics and non-Hispanic whites. Hispanic women reported lower Pap test rates in the past 3 years than non-Hispanic white women at 73.4% and 77.7%, respectively.⁴ Furthermore, Hispanic women may experience increased barriers to appropriate follow-up of abnormal Pap test results.⁵ This highlights the necessity of adding to the knowledge generated by a few studies that have examined Pap test utilization in Hispanic farmworkers specifically. One such study in California showed that United Farmworkers of America members (a largely Hispanic population) had a significantly increased risk of cervical cancer diagnosis compared with the general Hispanic population.⁶

Michigan has an estimated farmworker population of 45 800, making the state the fifth largest employer of farmworkers.^7,8 Farmworker employment is often seasonal in nature. The Office of Migrant Health recognizes 2 subpopulations of farmworkers. A migrant farmworker is “an individual whose principal employment is in agriculture on a seasonal basis … and who establishes for the purposes of such employment a temporary abode.” A seasonal farmworker also works in agriculture on a seasonal basis but does not migrate (United States Code: Title 42, 1962). Farmworkers generally live in “camps” consisting of apartment-style housing or communities of mobile homes on land owned by the farmers employing them. The distinction between migrant and seasonal farmworkers is of interest in determining the health needs of this population because of possible differential access to care because of the mobility of the migrant lifestyle.

Only one study has focused on cervical cancer screening utilization in the farmworker population of Michigan.⁹ Michigan implemented a Breast and Cervical Cancer Control Program in 1991 to increase the access of eligible women to screening.¹⁰ The current study aimed to determine whether a disparity exists in screening rates by the Pap test between Hispanic migrant and seasonal farmworkers in northwest Michigan. Secondarily, this study intended to determine demographic characteristics associated with cervical cancer knowledge in this population of women. Migrant farmworkers were predicted to have lower Pap test utilization than seasonal farmworkers. Higher level of acculturation, younger age and seasonal residency status were expected to be associated with identification of cervical cancer risk factors and symptoms.

Methods

This study was a cross-sectional survey of patients of Northwest Michigan Health Services, Inc (NMHSI) in 2 communities of northern Michigan (A and B). There is one NMHSI clinic in each community. NMHSI is a federally qualified health center that provides culturally sensitive care to farmworkers and the general public on a sliding fee scale. NMHSI patients are primarily of Mexican ancestry.^11,12 Institutional review board approval from the University of Michigan was obtained for this study.

Participants

Participants were recruited and data collected in community A from May to June 2011 and in community B from July to August 2011. This roughly corresponded to the areas’ respective principal harvests. Participants were recruited in the waiting rooms of clinics and at their residences in farmworker camps. Recruitment differed between clinics and camps. In clinics, when possible, every woman that entered was approached for participation. When clinics became busy, every other woman that entered was approached. Recruitment in camps relied on lists of area camps provided by the clinics. Each of the 70 camps in the study area with at least 4 units was visited and every unit in these camps was approached. If multiple eligible women were present in a unit, all were approached for participation. Camp visits occurred primarily in the early evening (3–7 pm).

Data Collection

Participants were interviewed in private rooms within each clinic, or in or around their residences according to their preference. Interviews were conducted orally, lasted approximately 15 minutes, and did not interfere with health care. Inclusion criteria were as follows: aged 21 years or older, female, able to speak and understand English or Spanish, and able to provide oral consent. The survey instrument was written in English. The Language Services Office of NMHSI performed translation into Spanish and assurance of cultural sensitivity.

Measures

Sociodemographic variables measured included age, ethnicity, marital status, and preferred language. Because of the sensitive nature of legal residency status, migrant and seasonal farmworker status was assessed by the question “In the last year, have you left Michigan to work in other states?” Those who had left to work in other states were considered migrants and those who had not were considered seasonal workers. Variables included the community (A or B), interview location (NMHSI clinic or at a camp) and whether the participant had or had not previously received care from an NMHSI clinic. Acculturation was measured by the modified Balcazar General Acculturation Index.¹³

The main outcomes of interest were ever having a Pap test, having had a recent Pap test, and having knowledge of cervical cancer risk factors. Women were categorized as ever receiving a Pap test versus having not heard of the Pap test/never received a Pap test. Recent Pap tests were defined, based on the recommendations of the American College of Obstetricians and Gynecologists, as tests received within the past 2 years for women aged 21 to 29 years and within the past 3 years for women 30 years and older. Eleven questions regarding knowledge of cervical cancer risk factors, signs, and symptoms were taken from the Special Cancer Behavioral Risk Factor Survey (SCBRFS) to facilitate comparisons with the general Hispanic population of Michigan.¹⁴ Risk factor knowledge was defined as the identification of any correct, clinically significant risk factor of 18 possible responses from SCBRFS.

Analyses

Analyses were limited to those participants that self-identified as Hispanic/Latina. Frequency distributions were calculated for all study variables. Descriptive statistics of sociodemographic variables, classified by key exposures, were calculated, and associations were examined using χ² analysis. Relationships between sociodemographic variables and each outcome were examined using simple logistic regression. Variables included in the multivariable logistic regression were chosen to maximize the fit of the model. All statistical analysis was performed using SAS software version 9.2 (SAS Institute Inc, Cary, NC).

Results

In all, a total of 479 women were approached for participation in the study; 398 were eligible for participation and 324 (81.4%) participated. The specific response rates for the clinics and camps were comparable at 82.2% and 80.0%, respectively. Of 324 women, 309 self-identified as Hispanic/Latina and were included in these analyses.

The study population included 184 seasonal and 125 migrant farmworkers with a mean age of 38.4 years (SD = 12.4). Table 1 presents the sociodemographic characteristics of the study population. Approximately 88% of the participants had received a recent Pap test, and 65% did not correctly identify any correct cervical cancer risk factor. Participants varied significantly in terms of interview city, acculturation level, and recent Pap test utilization across categories of previous care from an NMHSI clinic. The distribution of participants varied significantly by interview city across categories of Michigan residency status and the distribution varied significantly by risk factor knowledge across categories of interview location.

Table 1.

Sociodemographic Characteristics of the Hispanic Study Population (N = 309).

	Total Population		Seasonal		Migrant		Pa	Clinic		Camp		Pa	No PC NMHSIb		PC NMHSI		Pa
	n	%	n	%	n	%		n	%	n	%		n	%	n	%
Recent Pap test
No	38	12.34	23	12.57	15	12.00	.882	19	9.79	19	16.67	.077	17	23.94	21	8.94	.001c
Yes	270	87.66	160	87.43	110	88.00		175	90.21	95	83.33		54	76.06	214	91.06
Correct RF knowledge
No	198	64.92	110	61.11	88	70.40	.095	111	58.12	87	76.32	.001c	50	70.42	148	63.25	.267
Yes	107	35.08	70	38.89	37	29.60		80	41.88	27	23.68		21	29.58	86	36.75
Michigan residency
Seasonal	184	59.55	—	—	—	—		122	62.56	62	54.39	.158	35	49.30	146	62.13	.054
Migrant	125	40.45	—	—	—	—		73	37.44	52	45.61		36	50.70	89	37.87
Acculturation
Low	244	72.49	132	71.74	92	73.60	.719	134	68.72	90	78.95	.052	41	57.75	181	77.02	.001c
High	85	27.51	52	28.26	33	26.40		61	31.28	24	21.05		30	42.25	54	22.98
Community
A	166	53.72	113	61.41	53	42.40	.001c	111	56.92	55	48.25	.140	24	33.80	140	59.57	.000c
B	143	46.28	71	38.59	72	57.60		84	43.08	59	51.75		47	66.20	95	40.43
Location of interview
Clinic	195	63.11	122	66.30	73	58.40	.158	—	—	—	—		43	60.56	149	63.40	.664
Camp	114	36.89	62	33.70	52	41.60		—	—	—	—		28	39.44	86	36.60
PC from NMHSI
No	71	23.20	35	19.34	36	28.80	.054	43	22.40	28	24.56	0.664	—	—	—	—
Yes	233	76. 80	146	80.66	89	71.20		149	77.60	86	75.44		—	—	—	—

Abbreviations: NMHSI, Northwest Michigan Health Services, Inc; PC, previous care; RF, risk factor.

^aP value based on χ² test or t test.

^bN = 306.

^cStatistically significant (P < .05).

Of all Hispanics surveyed, 94.2% reported having ever received a Pap test and less than 6% (n = 18) of the study participants had not heard of or never received a Pap test. Comparatively, 87.7% had received a Pap test recently, and 72.8% had received one in the past year. Reporting ever having or having a recent Pap test did not differ between migrant and seasonal farmworkers (P = .88 for both comparisons).

Bivariate analysis indicated that, being 21 to 29 years old, a seasonal farmworker, having a high level of acculturation, living in community A, being interviewed in a camp, and not having previous care from NMHSI were associated with reporting not having a recent Pap test (Table 2). In the adjusted statistical model, only interview city and having previously received care from an NMHSI clinic were associated with reporting not receiving a recent Pap test.

Table 2.

Crude and Adjusted Odds Ratios of Not Having a Recent^a Pap Test (N = 308).

Characteristic	Not Screened Recently		ORb	95% CI	ORc	95% CI
	n	%
Age (years)
21–29	14	14.89	1.00	—	—	—
30–39	8	9.88	0.63	0.25–1.58	—	—
40–49	8	12.12	0.79	0.31–2.00	—	—
≥50	8	11.94	0.78	0.30–1.97	—	—
Michigan residency
Seasonal	23	12.57	1.00	—	—	—
Migrant	15	12.00	0.95	0.47–1.90	—	—
Acculturation
High	11	12.94	1.00	—	1.00	—
Low	27	12.11	0.93	0.44–1.96	1.09	0.48–2.50
Community
A	28	16.97	1.00	—	1.00	—
B	10	6.99	0.37d	0.17–0.79	0.23d	0.10–0.52
Location of interview
Clinic	19	9.79	1.00	—	1.00	—
Camp	19	16.67	1.84	0.93–3.65	1.93	0.93–4.01
Previous care from NMHSI
No	17	23.94	1.00	—	1.00	—
Yes	21	8.94	0.31d	0.15–0.63	0.20d	0.09–0.45

Abbreviations: OR, odds ratio; 95% CI, 95% confidence interval; NMHSI, Northwest Michigan Health Services, Inc.

^aIn accordance with the guidelines of the American College of Obstetricians and Gynecologists.

^bBivariate analysis.

^cAdjusted for all other variables included in the model.

^dStatistically significant (P < .05).

Participants reporting never having a Pap test (n = 18) tended to be seasonal farmworkers, younger than 40 years, interviewed in Spanish and living in community A. Reasons cited for never having a Pap test included never having sex, not going to the doctor, and embarrassment of screening.

Correct identification of any risk factor was reported by 35.1% of the study population. Knowledge of cervical cancer risk factors was significantly associated with being interviewed in the clinic (P = .001; Table 1). The bivariate analysis (Table 3) indicated that women aged 21 to 39 years, aged 50 years and older, classified as seasonal farmworkers, having high acculturation, living in community A, interviewed in the clinic, and having previously received care from NMHSI were more likely to identify at least one correct risk factor for cervical cancer. The associations between identifying a risk factor, acculturation, and interview location were statistically significant. In the adjusted model, identifying at least one correct risk factor was associated with the age-group of 50 years and older, high acculturation level, and being interviewed in the clinic.

Table 3.

Crude and Adjusted Odds Ratios of Having Cervical Cancer Risk Factor Knowledge (N = 305).

Characteristic	RF Knowledge		ORa	95% CI	ORb	95% CI
	n	%
Age (years)
21–29	31	32.98	1.00	—	1.00	—
30–39	28	35.00	1.09	0.78–2.05	1.44	0.71–2.91
40–49	21	32.31	0.97	0.49–1.90	1.37	0.65–2.89
≥50	27	40.91	1.41	0.73–2.70	2.60c	1.12–5.53
Michigan residency
Seasonal	70	22.95	1.00	—	1.00	—
Migrant	37	12.13	0.66	0.41–1.08	0.72	0.43–1.21
Acculturation
High	41	48.81	1.00	—	1.00	—
Low	66	29.86	0.45c	0.27–0.75	0.33c	0.18–0.61
Community
A	61	20.00	1.00	—	1.00	—
B	46	15.08	0.80	0.50–1.29	0.92	0.55–1.54
Location of interview
Clinic	80	26.23	1.00	—	1.00	—
Camp	27	8.85	0.43c	0.26–0.72	0.46c	0.27–0.78
Previous care from NMHSI
No	21	6.89	1.00	—	1.00	—
Yes	86	28.20	1.38	0.78–2.46	1.72	0.90–2.28

Abbreviations: RF, risk factor; OR, odds ratio; 95% CI, 95% confidence interval; NMHSI, Northwest Michigan Health Services, Inc.

^aBivariate analysis.

^bAdjusted for all other variables included in the model.

^cStatistically significant (P < .05).

Discussion

This study examined and documented cervical cancer screening behaviors and risk factor knowledge among Hispanic farmworkers in northwest Michigan. Our findings reveal that cervical cancer screening rates were high in the farmworker population of northwest Michigan with nearly 88% of participants reporting a recent Pap test. The Pap test screening rate observed in this population is higher than those reported for migrant farmworker populations in other states but is similar to the rate of 90.1% reported for western Michigan in 2002–2004.^9,15,16 Furthermore, this rate is similar to that of 2002–2008 aggregate Michigan SCBRFS data with 83% of Hispanic women reporting a recent Pap test, which may be indicative of the recent success of interventions and campaigns for cervical cancer screening in Michigan (P. Campbell, personal communication). Of note, the American Cancer Society has set the most recent recommendations for cervical cancer screening as of March 2012, calling for cytological testing to begin at age 21 years and occur every 3 years through age 29 years and for cytological and HPV co-testing to occur every 5 years from ages 30 to 65 years.¹⁷

In this study, migrant and seasonal farmworkers reported statistically similar rates of screening by the Pap test. The following factors may account for these unexpected results. First, the cervical cancer screening disparity outlined above has been widely recognized for some years. Our results may reflect recent efforts to increase Pap test utilization in response to the previously noted disparity. Recent efforts include the Michigan Breast and Cervical Cancer Control Program that was implemented in 1991, as well as initiatives led by the Michigan Cancer Consortium (formed in 1998) that have always included goals for increasing Pap testing in minority women. Second, although migrant farmworkers change residences regularly, women may be able to return to the same health care providers with some consistency. Finally, it is possible that the high screening rate detected could be attributed, in part, to the association of the study with NMHSI.

Although reported screening rates were high, a low level of knowledge of cervical cancer risk factors was detected in the Hispanic farmworker study population. This finding is likely applicable to many Hispanic farmworker populations throughout Midwestern United States as previous studies have similarly detected low levels of cancer knowledge in vulnerable populations.^18–20 A cancer control study in Phoenix found that Hispanic women with low acculturation had especially low levels of cervical cancer screening knowledge with only 43% identifying the Pap test as able to detect cervical cancer.¹⁹ It appears that it may be most beneficial to further educate women in Michigan about the lack of early stage symptoms given that this population has a tendency to delay seeking treatment, resulting in more frequently diagnosed advanced disease.^21,22

There are limitations to this study that should be considered when interpreting results. This study is limited by self-reports of Pap tests. It has been specifically noted in Hispanic farmworker populations that “self-report could be inaccurate as a result of temporal displacement.”¹² Ingratiation is also a potential source of bias in self-report, especially within marginalized populations such as Hispanic farmworkers. However, the Pap test utilization rate detected in this study is comparable to that detected in a Michigan farmworker population in 2002–2004 by reviewing medical records.⁹ Additionally, knowledge about cervical cancer was assessed using English terminology, much of which had to be literally translated into Spanish, in order to facilitate comparisons with SCBRFS data. Since medical terminology was used, as opposed to colloquial language, it is likely that much of the study population had difficulty understanding some survey questions. Therefore, it is possible that cervical cancer and risk factor knowledge was underestimated. Additionally, the questionnaire failed to address the behavior of men on cervical cancer risk. This is a critical question, considering that the promiscuity and infections of a woman’s partner affect her risk of cervical cancer independently of her own behaviors. Finally, due to the necessity of protecting the farmworkers’ legal status, the classification of migrant and seasonal farmworkers was potentially inadequate. If a participant worked only in Michigan during the summer, but spent the rest of the year in Texas without working, this participant would be categorized as seasonal based on the questionnaire, but would likely be considered migrant based on the United States Office of Migrant Health’s definition of “migrant.”³

This study has several notable strengths despite its limitations. This study was among the first of its kind to examine the use of the Pap test to detect cervical cancer among Hispanic farmworkers in Michigan. This population is especially important to capture in health prevention studies considering their vulnerability and the likelihood that they will not be captured by population-level studies. Furthermore, this study highlights the complexity of the cervical cancer disparity among Hispanics and identifies critical aspects that must be addressed in future interventions and campaigns regarding cervical cancer.

Future studies may include reviews of medical records for more accurate information on whether a woman received a Pap test as opposed to a pelvic exam and determination of the date of the test. Future studies could also investigate patterns of care related to the specific geographic location where Pap tests were obtained and corresponding geographic preferences. To obtain more accurate information, it may be beneficial to use simpler language to convey complex terms and concepts related to cervical cancer screening and etiology.

In conclusion, these findings suggest that, in Michigan, migrant farmworkers may not experience more difficulty in receiving Pap tests than seasonal farmworkers. Farmworkers in general recognize the importance of receiving Pap tests even though women have low levels of knowledge about cervical cancer risk factors. Furthermore, intervention efforts directed toward farmworkers should focus on education regarding cervical cancer risk factor and symptom knowledge. Considering that it appears this population is receiving timely Pap tests, health care practitioners could potentially provide appropriate cancer-related education at the time of screening. These interventions may have the most benefit for high-risk groups of women such as those who have not previously accessed related health care and those with a low acculturation levels.

Biographies

Jayne S. Knoff is a cervical cancer research program coordinator at Johns Hopkins University. Her research interests lie in cervical cancer epidemiology and health disparities.

Siobán D. Harlow is a Professor of Epidemiology and the Director of the Center for Integrated Approaches to Complex Diseases at the University of Michigan School of Public Health. Her research interests include women’s health and capacity building in low-resource settings.

May Yassine is the program director of the Cancer Control Services Program at Michigan Public Health Institute. Her expertise in developing and integrating surveillance data to portray disease burden in populations contributes to state-level public health planning. She leads data support and evaluation studies for initiatives such as the Michigan Cancer Consortium and Comprehensive Cancer Control Program.

Amr S. Soliman is Professor of Epidemiology. His research training program is centered around cancer epidemiology in international and minority settings in the United States.