Rural Health Policy

Maternal Care Coordination for Migrant Farmworker Women: Program Structure and Evaluation of Effects on Use of Prenatal Care

and Birth Outcome Kim Larson, RN, MPH, Joan McGuire, RN, MSPH, Elizabeth Watkins, DSc,

and Karen Mountain, RN, M S N

ABSTRACT Nearly three fourths of the migrant farmworkers in the US. are Hispanic. Cultural and social barriers, along with constant travel, make coordination of care a significant concern for migrant health centers providing perinatal services to female farmworkers. As part of a demonstration project, a migrant-specific maternal care coordination program was devel- oped that used bilingual staff, outreach services, lay health advisers, and a multistate tracking system. Following the initiation of the project, first-trimester entry into prenatal care and number of prenatal visits increased over a five-year period among the target population. Successful tracking methods provided outcome data on more than 80 percent of participants during the project period.

The results of this study suggest that migrant health centers should focus on employing public health-oriented bilingual or bicultural health professionals and that an outreach strategy must be an integral part of a health care delivey system serving migrant farmworkers. Without these key ingredients, health care services will not be accessible or acceptable for this hard-to- reach population. Collaboration among the National Migrant Resource Program, the Migrant Clinicians Network, and the National Perinatal Association can facilitate development of a regionwide perinatal service system for female migrant farmworkers.

igrant and seasonal farmworkers is available take precedence over seeking primary labor in one of the most hazardous health care. The result is a pattern of incomplete, occupations in the nation, endure fragmented health care received by this population. substandard living conditions, and Several studies have documented the high inci- M have limited access to health and dence of infant mortality among female rnigrant

social services (Johnston, 1985). Problems with trans- portation, child care, and the need to work when work

farmworkers (Chase et al., 1971; Slesinger & Christensen, 1986). Perinatal statistics from Colorado’s

The lournal of Rural Health 128 Vol. 8. No. 2

statewide migrant health program revealed a four- year average low-birthweight rate of 44.6 per 1,000 live births (D. Horton, personal communication, March 1990). Lack of health insurance, a decreasing supply of physicians, and inadequate access to care are frequent problems in rural areas (Hughes & Rosenbaum, 1989). Migrant women face these same problems, and they are compounded by language barriers and lack of a preventive health orientation (Trotter, 1988).

A recent survey of seasonal agricultural workers reported that 71 percent are Hispanic, and Spanish is the primary language for two out of three of these workers (Mines, Gabbard, & Boccalandro, 1991 ). Though early prenatal care is one of the most impor- tant factors in assuring positive pregnancy outcomes, Hispanic women enter prenatal care later than women from other ethnic groups (Singh, Forrest, & Torres, 1989). A recent study reported that only 42 percent of female migrant farmworkers enrolled for prenatal care in the first trimester (Puente, 1989), compared to 76 percent of all women nationally (Singh et al., 1989).

The Federal Migrant Health Program recently reported that migrant and community health centers serve fewer than 15 percent of the estimated 3 million migrant and seasonal farmworkers in the nation (Migrant Health Program, 1989). Migrant Health Program funds are small in comparison to the primary care, environmental, and inpatient care needs of farmworker families. Institutional barriers, including the scarcity of bilingual staff, lack of community outreach programs, and cost of care, further limit access to services for farmworker families.

Project - In an effort to improve perinatal outcomes of

female migrant farmworkers and the health status of their children, the Department of Maternal and Child Health at the University of North Carolina’s School of Public Health began collaboration with Tri-County Community Health Center in 1985 to develop a comprehensive and continuous system of health care delivery to this population (Watkins, Larson, Harlan, & Young, 1990).

North Carolina ranks fifth in the United States in the number of farmworkers. An estimated 344,944 migrant and seasonal farmworkers, including depen- dents, are employed annually (Migrant Health

Program, 1990). Located in southeastern North Carolina, Tri-County Community Health Center (TCCHC) is the largest of four migrant health centers in the state and has 32,413 farmworkers in its catchment area (Migrant Health Program, 1990). The farmworker population is predominantly Hispanic and black , with small numbers of whites, Haitians, and American Indians. In the past five years, the Hispanic (91% Mexican or Mexican-American and 9% Central American) maternal and child population at this center has nearly doubled.

ing of a project coordinator (a public health nurse), health educator (a public health nurse), nutritionist, and social worker, was located at the migrant health center to assist staff in designing culturally appropri- ate strategies for delivering care to migrant women and children. This article describes a system of maternal care coordination developed by the Migrant Health Project to function within North Carolina and other states along the East Coast migrant stream. The aim of this approach was to enhance the already established medical assessment and intervention for migrant women receiving maternal health services at the TCCHC.

Specific project objectives were to: (1) increase first-trimester enrollment into prenatal care, (2) improve continuity of care, including frequency of visits, and (3) improve perinatal outcomes. Program components included public health-oriented bilin- gual staff, maternal-child focused outreach, migrant lay health advisers, and a multi-state tracking system.

A bilingual Migrant Health Project team, consist-

Program Structure - Bilingual Staff. Despite the large number of

Spanish-speaking migrant farmworkers served by the TCCHC, the only bilingual health care provider at the beginning of the project was the medical director. Many migrant health centers have difficulty recruit-

The authors would like to acknowledge f inding of projects MC] 373415 and M C J 3736003 through Special Projects of Regional and National Significance, Maternal and Child Health Bureau, Department of Health and Human Services, Health Resources and Services Administration. Data analysis was provided by Stuart Gansky and G a y Koch of the University of North Carolina Biornetrics Consulting Laboratory. A special thanks goes to the migrant farmworker women who participated in the projects.

An earlier version of this paper was presented at the National Perinatal Association Conference on Nov. 16,1990, in NEW Orleans, LA.

For additional information, contact Kim Larson, RN, MPH, Route 1 , Box 179, Lumber Bridge, N C 28357.

Larson, McGuire, Watkin, and Mountain 129 Spring 1992

ing and retaining qualified bilingual staff (S.L. Reig, personal communication, Sept. 12,1991). Salaries are not competitive, and the centers are often located in isolated rural areas. Moreover, many of the centers are staffed with National Health Service Corps physicians who have a short-term commitment.

Communication in migrant health centers is, therefore, frequently mediated by interpreters. Caution must be exercised when using interpreters. Men and children are often the only bilingual mem- bers of the family, and issues concerning modesty and respect must be considered. The use of friends and other family members poses problems of confi- dentiality. Most lay interpreters are unskilled in medical translation and are not required to take a language fluency test. Ultimately, the monolingual provider does not know exactly what was translated or whether something was left unsaid.

Staff on this project initially provided on-site Spanish language classes for Center personnel. Unfortunately, the high staff turnover and sporadic attendance (due to clinic workload) did not allow for testing of fluency. However, the TCCHC administra- tors were convinced of the value of bilingual skills in their recruitment and retention of staff. By raising salaries and providing incentives, such as financial assistance for language study in Mexico, the TCCHC had employed four bilingual registered nurses by the end of the project.

Maternal-Child Focused Outreach. Problems of camp isolation and lack of telephone services lead migrant women to delay seeking prenatal care. Center outreach workers routinely visit migrant camps to inform farmworkers of available services and to identify and refer those in need of health care. To introduce a maternal-child health focus, outreach staff were oriented to find pregnant women and children, and a protocol for their referral into the TCCHC was developed. Outreach workers began asking all women in the camps the question: "Do you think you are pregnant?" Pregnant women were offered transportation to the clinic on either an appointment or on a walk-in basis. Following clinic enrollment, prenatal and postpartum home visits were made to most of the women either by project or outreach staff. Documentation of the serious need for transportation enabled the clinic to receive grant support for a van and a bilingual driver for the maternal health program.

Head Start Program, which provides services to In addition to these outreach efforts, the Migrant

migrant infants and children six weeks to four years of age, already had in place a policy of asking all migrant women who enrolled their children if they were pregnant and documenting where they were receiving care. The Head Start nurse worked coopera- tively with project staff to enroll pregnant women in prenatal care at the earliest possible date and to arrange transportation to clinic appointments.

The absence of bilingual staff in other public health and social service agencies necessitated project involvement in farmworker advocacy in these agencies. To facilitate communication between migrant women and local hospitals, project staff initiated an on-call birth coach service that provided the obstetrical unit with a list of volunteer translators to be called to assist migrant women in labor. Many migrant women were hesitant to seek assistance from the Department of Social Services because of their non-resident status. Assisting women in obtaining resources for which they were eligible, such as Medicaid, emergency financial aid, and food stamps, was a frequent service. Subsequently, the Center arranged for a social worker from the Department of Social Services to be based at the migrant health center to assist with Medicaid applications.

Migrant Lay Health Advisers. Migrant women within the farmworker community to whom people would naturally turn for advice were recruited by project staff for a lay health adviser training program. The program added basic knowledge about maternal and child health practices to their natural. helping ability (Watkins et al., 1988). Following training, lay health advisers shared this information with other migrant families as they traveled. In North Carolina, the lay health advisers were active in identifying and counseling pregnant farmworkers and in referring them for prenatal care.

and those with low educational levels made it necessary to develop an appropriate mix of audio- visual and clearly illustrated written materials. Individual health education at the clinic and camps, as well as the lay health adviser program, became part of the project's health promotion strategy.

The large proportion of Spanish-spea king clients

Multistate tracking system. Because migrant women may leave an area between one a,ppointment and the next, a system was created to provide conti- nuity of care both within the state and in other states through tracking and follow-up.

All pregnant migrant farmworkers signed a

The ]ourrial of Rural Health 130 Vol. 8, No. 2

consent form to participate in the project during their first visit to the TCCHC. The consent form served as a release-of-information form for tracking. Also during the first prenatal visit, women were given a plastic “MCH record pouch” in which to carry a copy of their medical record, which was updated at each visit. Migrant women with their records could enter

Table 1. Characteristics of Female Migrant Farmworkers Enrolled in Prenatal Care at Tri-County Community Health Center, North Carolina: 1985 to 1989 (N = 599).

prenatalcare more easily at a new location and avoid unnecessary repetition of laboratory or other proce- dures. They were given instructions to present the pouch to each center that provided follow-up care. A letter was sent to other migrant health centers on the East Coast explaining the project and the tracking system. Each center was requested to ask migrant women for their records and the plastic record pouch and to provide information to TCCHC on pregnancy outcome on a self-addressed postcard, included in the pouch. Project staff also made site visits twice each year to selected Florida migrant health centers, to which a majority of the migrant women and children returned. The site visits were made to describe the project and facilitate information ex- change.

migrant woman on her first prenatal visit to the center. Tracking began at the initial contact and continued until the postpartum visit. Missed appoint- ments resulted in a letter or telephone message with a new appointment. A second missed appointment resulted in a home visit. To facilitate tracking of outcomes, a permanent address and anticipated location of delivery were documented on this form. When delivery occurred outside North Carolina, these locations were matched with migrant health centers through the National Migrant Resource Program’s Referral Directory. Outcome data were mailed to the project in North Carolina following written or telephone communication with the identi- fied hospitals and migrant health centers. This process was often expedited by personal contracts made during the site visits to Florida.

A prenatal tracking form was established for each

Outcomes - Data were abstacted from the medical records at

migrant health centers and hospitals which provided care to project participants between 1985 and 1989. A total of 599 migrant farmworker women participated in the project, representing 98 percent of all the women receiving prenatal care at TCCHC. Only one woman refused to participate, due to her husband’s

19851 1986 1987 1988 1989 Ethnic Group (n=109) (n=125) (n=99) (n=120) (n=146)

White 132 12 9 6 5 Black 20 19 23 19 15

80 Hispanic 55 66 66 72 Haitian/other 12 2 2 3 0

P = 0.0002

Mean maternal age 24.3 22.9 22.4 23.3 22.9 Mean gravida 3.3 2.9 2.6 3.1 2.6 Mean parity 2.0 1.6 1.3 1.9 1.4

1.

2. Numbers are percentages.

Note: Percentages may not total 100 due to rounding.

Represents a nine-month period due to project start up; all other years are full calendar years.

general distrust of the program. Hispanic representation at this migrant health

center increased significantly from 55 percent in 1985 to 80 percent in 1989 (Table 1). Also over the five years, a younger prenatal population having fewer pregnancies was noted. In 1985 the mean age was 24.3 years with an average of 3.3 pregnancies; in 1989 the mean age was 22.9 years with an average of 2.6 pregnancies.

Improvement in the use of prenatal services at TCCHC during the five-year period was also found. First trimester entry into prenatal care changed signifi- cantly, increasing from 35 percent in 1985 to 51 percent in 1989 (Table 2). Also, the percentage of women receiving nine or more prenatal visits rose significantly from 24 percent in 1985 to 50 percent in 1989.

the proportion of low birthweight infants during the project period (Table 3). However, caution must be used when evaluating the program’s effect on birth outcomes. In fact, adjusting for race, no significant difference was noted between year and birthweight. Black women had the highest proportion (19%) of low- birthweight infants and Hispanic women had the lowest (4%) during the five-year period.

Pregnancy outcome data were tracked for 84

A nonsignificant declining trend was observed in

Larson, McGuire, Watkin, and Mountain 131 Spring 1992

Table 2. Characteristics of Use of Prenatal Care by Female Migrant Farmworkers: 1985 to 1989 (N = 599)l.

Table 3. Distribution of Low and Adequate Birthweight Infants: 1985 to 1989 (N = 468)l.

1985 1986 1987 1988 1989 Trimester entry (n=109) (n=125) (n=99) (n=120) (n=146)

1985 1986 1987 1988 1989 Birthweight (n=82) (n=102) (n=83) (n=94) (n=107)

First (< 15 weeks) 352 41 52 45 51 Second (15-27 weeks) 41 30 38 38 30 Third (> 27 weeks) 24 30 10 18 19

S 2,500 grams 122 7 7 9 6 > 2,500 grams 88 93 93 91 94

P=0.009

Number of prenatal (n=83) (n=102) (n=85) (n=95) (n=110) visits

1-4 39 30 22 20 19 5-8 37 34 21 26 28 9+ 24 35 56 54 53

P = 0.0002

1. The sample size in the number of prenatal visits is 475;women who had spontaneous or therapeutic abortions were excluded from this analysis. Numbers for first, second, and third trimesters and number of prenatal visits are percentages.

2.

Note: Percentages may not total 100 due to rounding.

percent (n=500) of the 599 female farmworkers who participated in this study. These data provide a far clearer picture of the birth outcomes of female farmworkers attending this migrant health center than previous record reviews have shown (Watkins, Peoples, & Gates, 1985). One of the most common reasons for tracking failure was that the consent form for release of information expired before the delivery date. Hospitals would not honor a consent form that had been signed more than six months before the request for information. Although 16 percent of the women were lost to our tracking efforts, it was determined that they were equally representative of the women whose outcome was known.

Discussion - The results of this project suggest that the elimi-

nation of barriers to care, particularly language and

P = 0.5457; P=0.487, adjusting for race.

1. 2. Numbers are percentages.

Only live births were included in this sample.

transportation barriers, combined with close tracking and follow-up can have a positive effect on prenatal care and pregnancy outcome for a traditionally hard- to-reach migrant maternity population.

grant-specific, culturally tailored health care. Special projects, such as the University of North Carolina (UNC) Migrant Health Project, demonstrate the effectiveness of maternal care coordination in address- ing the problems of incomplete, fragmented care.

As suggested by the experience of this project, programs that target ethnic groups need to make recruiting and retaining bilingual and bi-cultural health professionals a priority. If interpreters must be used, they should complete a training course that includes medical terminology and issues of confidenti- ality. Likewise, monolingual staff members should receive training in cross-cultural, interpreter-mediated communication. To continue to pursue the UNC Migrant Health Project objectives, TCCHC received funding for and employed a full-time bilingual mater- nal care coordinator in 1988.

female migrant farmworkers remain essentially uncovered because of citizenship requirements and inflexibility in the Medicaid application process. In 1989 fewer than 25 percent of migrant women actually received Medicaid. Migrant health clinicians should be aware of the Medicaid Child Health Amendment of 1991 (HR 1392), which would provide Medicaid reciprocity between states for migrant women and children.

outcomes out-of-state. In fact, most public health

Migrant and seasonal farmworkers require mi-

Despite new laws expanding Medicaid coverage,

Few maternal health programs track pregnancy

The loitrnal of Rural Health 132 Vol. 8, No. 2

agencies generally consider a woman who has left the area a ”closed case.” For the at-risk farmworker who is unfamiliar with health care services, care coordina- tion should ensure that the woman has continued care in her next place of residence. Most farmworker families know when and where they will be moving. This coordination effort can be accomplished by using the National Migrant Resource Program’s Referral Directory. The farmworker can be given her prenatal record, the address of the center in her next location, and the knowledge that someone is expect- ing her. If health professionals convey the importance of continuity of care, women are more likely to continue seeking care.

Based on the experience of the UNC Migrant Health Project, the National Migrant Resource Program, Migrant Clinicians Network, and National Perinatal Association recently began a collaborative effort to promote coordination of perinatal services for migrant women. The National Migrant Resource Program provides resources to migrant care provid- ers in the development of collaborative relationships between agencies serving farmworkers. The Migrant Clinicians Network, an organization for health professionals working with migrant and seasonal farmworkers, has identified the problem of inad- equate prenatal care as a priority focus for the 1990s. Input from the National Perinatal Association is important in creating alliances at the state level that will enhance perinatal services for female farmworkers. These organizations are committed to increasing outreach efforts, reducing barriers to services for migrants, and working directly toward expanding information and service exchange.

References - Chase, H.P., Kumar, V., Dodds, J.M., Sauberlich, H.E., Hunter,

R.M., Burton, R.S., & Spaulding, V. (1971). Nutritional status of preschool Mexican-American migrant farm children. American Journal of Disabled Children, 122,316-324.

Hughes, D., & Rosenbaum, S. (1989). An overview of maternal and infant health services in rural America. Journal of Rural Health, 5,299-319.

Johnston, H.L. (1985). Health for the nation’s harvesters. Farmington

Migrant Health Program. (1989). Migrant Health Program three-year Hills, MI: National Migrant Worker Council, Inc.

strategic work plan, 1989-1991. Washington, DC: Department of Health and Human Services.

estimate number of migrant and seasonal farmworkers and members of their families. Washington, DC: Department of Health and Human Services.

the National Agricultural Workers S u r ~ e y fNAWS) 1990 Washington, DC: US. Department of Labor, Office of Program Economics.

for 1988. Migrant Health Newsline, 7(5), 2.

United States. Washington, DC: The Alan Guttmacher Institute.

Slesinger, D., & Christensen, B. (1986). Health and mortality of migrant farm children. Social Science & Medicine, 23, 65-74.

Trotter, R. (1988). Orientation to multicultural health care in migrant health programs. Austin, TX National Migrant Resource Program.

Ramos-Nunez, M., Gilbertson, S., & Ramirez-Garza, C. (1988). Migrant lay health advisors: A strategyfor health promotion. Chapel Hill, NC: University of North Carolina.

Watkins, E.L., Larson, K., Harlan, C. & Young, S. (1990). A model program for providing health services for migrant farmworker mothers and children. Public Health Reports, 105,

Watkins, E.L., Peoples, M.D., &Gates, C. (1985). Health and social

Migrant Health Program. (1990). An atlas ofstate profiles which

Mines, R., Gabbard, S., & Boccalandro, B. (1991, July). Findingsfrom

Puente, A.M. (1990). Comprehensive Perinatal Care Program Data

Singh, S., Forrest, J.D., & Torres, A. (1989). Prenatal care in the

Watkins, E.L., Larson, K., Harlan, C., Young, S., Wenrich, S.L.,

567-575.

needs of women farmworkers. Migration Today, XIII, 39-42.

Larson, McGuire, Watkin, and Mountain 133 Spring 2992