Health Characteristics of Foreign-Born Mothers and their Infants in four PRAMS States
1997 to 2000

Denise Roth Allen
MCH EPI Program
Division of Reproductive Health
Centers for Disease Control and Prevention

February 5, 2003

Tishia Smith: Good afternoon, and welcome to the February 2003 presentation for the Maternal and Child Health Epidemiology Monthly Broadcast. This month's title is Health Characteristics of Foreign-born Mothers and their Infants in four PRAMS States, 1997 to 2000.

Prior to working at CDC, Dr.  Allen conducted two years of ethnographic field work that explored rural Tanzanian women's pregnancy related concerns and recently published a book based on that research.

Dr.  Allen, thank you and welcome to today's broadcast.  

Dr. Allen:  Thank you, Tishia.  Hello, everybody.  This afternoon I'm presenting preliminary results of an analysis that my colleagues and I are conducting on the health characteristics of foreign-born women who give birth in the U.S. and their infants. 

The data were collected through CDC's Pregnancy Risk Assessment Monitoring Surveillance System during 1997 and 2000.  This project is very much a work in progress, so I look forward to your comments at the end.

I'd like to acknowledge first a contribution of my co-authors, Bill Sappenfield, Brian Morrow, Mary Rogers and Cynthia Ferre in the Division of Reproductive Health here at CDC and the PRAMS working group in Colorado, Florida, North Carolina and Washington State.

As an anthropologist who has been working closely with epidemiologists for a year and a half year now, I'm even more convinced than ever that research into the maternal and infant health issues can benefit from an approach that draws upon the strengths of both disciplines.

 I'm going to start off with some background information to put our study into context.  I'll briefly examine some of the relevant research literature and present some census data on changing immigration trends.  I'll also talk briefly about vital records, in particular how birth certificate data is analyzed. I'll then move to the results of two related studies.  The first is a recent MMWR that was coauthored with colleagues working in the vital record section at the National Center for Health Statistics or MCHS.  I'll just be referring to it as MCHS from now on.  I'll follow that up with a discussion of our preliminary findings from the PRAMS data, highlighting new findings regarding the maternal characteristics and behaviors among foreign-born mothers.  I end with a brief discussion of our study implications and possible future directions in research among this important population of women.

 Prior to joining CDC as an EIS officer my research interests were focused on maternal and infant health issues in Sub- Saharan, Africa.  As part of my anthropological training I conducted two years of ethnographic field work in a rural community of Tanzania.  And in that research I explored women's experiences during pregnancy and child birth, including their use of biomedical and nonbiomedical such as prenatal care, their perceptions of maternal health risks and their interactions both positive and negative with the health care system.

After moving to Atlanta and beginning work in the Division of Reproductive Health at CDC, I found myself wondering about maternal and infant health issues among women who immigrate to the U.S. and how one could use epi tools to better understand those issues.

These issues include access to care, birth outcomes, the positive and negative effects of the culturation, the quality of immigrant women social support network and issues surrounding health care coverage.

A subsequent search of the literature revealed two research trends related to my interests.  The first of these focused on racial and ethnic disparities in perinatal health.  Collins, et al. published a lot on black/white disparities and low birth weight and preterm delivery.  Iyasu and his colleagues from CDS explored black/white disparities in infant mortality.  While the preterm delivery team here at CDC focused attention on the social context of black/white disparities in preterm delivery and on how social factors, including stress, may contribute to poor outcomes among African-American women.

The landmark article by Bacerra and his colleagues at CDC focused attention on the heterogeneity within Hispanic sub populations in the U.S. and called for interventions that were tailored to ethno-specific risk factors and outcomes.

Another trend in the literature indicated that foreign-born women in the U.S. have better birth outcomes than their U.S.-born racial and ethnic counterparts, even after adjusting for differences in age, education and marital status.

There are two lines of thinking about why this is so. The first represented on the left side of this screen suggests that women who immigrate to the U.S. have better birth outcomes because they represent a particular group of women with positive attributes.  That is, they're healthy immigrants, i.e., they're healthier more educated and more savvy than those who remain behind.  By savvy here, I mean it's been suggested that these women have that something extra that is needed to navigate the various hurdles associated with obtaining a U.S. visa and eventually arriving in the states.

The other line of thinking represented on the right side of the slide suggests that foreign-born women have better outcomes not because they are a self-selected group of women from the start, but because immigrant status may serve as a proxy for a variety of protective behavioral cultural and psychosocial factors that positively influence pregnancy outcomes.

For example, some researchers suggest that foreign-born women are less likely to smoke and eat diets high in fat and sugar and due to strong social support networks they're less likely to become pregnant at an early age or outside the context of marriage.  They further suggest that the process of acculturation to the American lifestyle may reduce these protective benefits over time.

So just who are the foreign-born?  There are two main definitions.  According to U.S. census Bureau, a foreign-born person is anyone living in the U.S. who was not a U.S. citizen at birth, although she or he may become a naturalized citizen later on.  The population of foreign-born in the U.S. include legal immigrants and legal nonimmigrants, refugees and illegal immigrants.  Legal nonimmigrants here are people who come on student or work visas.

The designation of foreign-born status by birth certificate data is quite different.  A woman is identified as a foreign-born mother if she was born outside the 50 states, the District of Columbia or U.S. territories.  Regardless of the citizenship at the time of birth.  I'll return to this definition later in the presentation when I talk about the U.S. vital record system.  But what I'd like to do now is briefly discuss the changing trends in U.S. immigration.

Prior to the 1970s, the majority of people who made up the population of foreign-born in the U.S. were European.  Since the 1970s, the majority of the foreign-born have been Hispanic and Asian.  Note here there's also a category "Other."  This kind of category drive anthropologists crazy.  I know Africa is included there.  I'm not quite sure who else "Other" is.  I just wanted to point that out.

According to the U.S. census Bureau, this is an outcome of the changing nature of U.S. immigration policy, these changing trends.

Legislation enacted in the 1920s and reaffirmed in 1952 established a national quota system which favored countries in the western hemisphere and northern and western Europe. Subsequent legislation in 1965, 1986 and 1990 which permitted some illegal aliens to obtain lawful permanent residence and increase the annual cap on immigration has contributed to an increase in international migration.

Let's take a look at these trends over the past decade.  According to the U.S. Census Bureau, 7.9 percent of the U.S. population in 1990 were foreign-born.  In 2000, the foreign-born comprised of 11.1 percent of the population.

This slide shows the top ten countries of birth for the foreign-born population in the U.S. in 1990 and 2000.  China here includes Taiwan and Hong Kong.  The direction of the arrow indicates the largest to smallest.

Although the top three countries remain the same, on the right side of the slide we can see a change in the distribution of these, as well as the appearance of new groups.  New population subgroups within the top ten are highlighted in green.  Subgroups that experience an increase in numbers are highlighted in yellow. Notice here that by 2000 no European countries figured among the top ten countries for immigration.  We wanted to compare these trends to changing trends in the U.S. births and so we turn to the vital record system maintained by NCHS.

Although I initially hope to identify births to immigrant women by the country of their birth, the U.S. birth certificate does not record information on citizenship or immigration status.  However, the mother's place of birth and race and ethnicity are recorded.  States differ quite considerably in how they code this information.  Some have very detailed categories for place of birth.  Others have very detailed categories for ethnic origin.  After cleaning their data, all states send their birth certificate information to MCHS which uses a broader coding scheme.

This slide shows MCHS's coding scheme from mother's place of birth.  As you can see, there's a very limited number of possibilities.  A mother's place of birth can be identified as occurring in one of the 50 states and Washington D.C., one of the U.S. territories, Canada, Cuba, Mexico, or the remainder of the world.

I didn't find this latter category particularly helpful for what I was trying to accomplish.  That left me with a category for race and ethnicity.

As with mother's place of birth, MCHS records states' information on race and ethnicity into broader categories.  From these codes a combined code for race and Hispanic ethnicities can be created.  These include non-Hispanic white, non-Hispanic black, American, Indian and Alaska native, Asian Pacific Islander and Hispanic.  And then you see here both Asian Pacific Islanders and Hispanics can be classified further into subcategories.

In order to get at least some sense of where a woman immigrated from, I would have to stratify those born in the remainder of the world by their race and ethnicity.  After discussing this with colleagues at MCHS they suggested I might want to contact individual states as state categories for mother's place of birth and ethnicity are often more detailed.

For these reasons we initially decided to focus our analysis at the state level.  I selected a state that had a very detailed coding scheme for mother's ethnic origin.  However, after a preliminary analysis, I found that the numbers were too small.  There were also some overlapping codes for ethnicity.

Because we didn't find this state's coding scheme very helpful, we decided to focus our analysis at the national level, to try to paint a picture of changing demographic trends for the U.S. as a whole.  That analysis resulted in the publication of an MMWR article.

Because some of the literature we reviewed suggested that foreign-born women and women born in U.S. territories had better birth outcomes than their racial and ethnic state-born counterparts, we decided to analyze our data in terms of two categories of women.  Those born inside the 50 states in DC and those born outside them.

In the next couple of slides I present some of those findings.  In 1990, 15.6 percent of all births in the U.S. were to women who were born outside the 50 states in DC.  In 2000, such births represented 21.4 percent of all births, in both 1990 and 2000, births to Hispanic women comprised the majority of these births. As we can see in this slide, in all categories, births to women born outside the States increased with the largest increase among Hispanics.  Mexican women who were born outside the states accounted for 66 percent of these births in 1990 and 72 percent in 2000.

Although the percent of births to Mexican and Central South American women increased from 1990 and 2000, this was not true for Puerto Ricans, Cubans and other Hispanics.  Although women born in U.S. territories such as Puerto Rico are also U.S. citizens, previous analysis suggested that their outcomes are different than state-born Puerto Ricans.  Because our focus today here is on the health characteristics of foreign-born women, we won't be comparing birth outcomes between Puerto Ricans inside and outside the state, because both groups of women are U.S. citizens.

Health care issues among women born in U.S. territories should be addressed separately.

In 1990, six states accounted for 76 percent of live births to women born outside the 50 states in DC.  These states were California, New York, Texas, Florida, Illinois and New Jersey. These same six states accounted for 69 percent of such births in 2000, an absolute decrease of seven percent.

The U.S. census Bureau has reported a similar pattern of change from 1990 and 2000 with regard to where the foreign-born in the U.S. reside.

From 1990 to 2000, six states experienced more than a 10 percent absolute increase in births to women born outside the states and DC.  These were Oregon, Nevada, Arizona, Colorado, Georgia and North Carolina.  For example, in 1990, the percent of live births to foreign-born or, I'm sorry, to women born outside the states in Georgia were 5.5 percent.  By 2000, 16.1 percent of live births to women in Georgia were to women born outside the United States and DC. 

In sum, approximately one in five live births in the U.S. are to women born outside the United States and DC.  State specific comparisons of the number and distribution of such births from 1990 and 2000 revealed a shift to states in the west and south. Births to Hispanic women accounted for most of this increase.  Among non-Hispanic the greatest increases in women born outside the states were found in the northeast, on the East Coast and in Hawaii.

Overall women born outside the states had higher percentages of education than their state-born counterparts except for Mexican and Central South American women.  They were also less likely to be a teenager or to be unmarried when they gave birth.  Although they had later entry into prenatal care than their state-born counterparts, women born outside the states had better birth outcomes.

Previous research had indicated similar differences even after adjustment for age and education and marital status. Although our MMWR article provided another piece of the puzzle regarding maternal and infant health issues among immigrant women in the U.S., for the kinds of questions we were trying to answer, the analysis of birth certificate data had its limitations. For example, we had limited country of birth specific details.  The birth certificate also provided very few details regarding access to care.  Also the information on maternal characteristics behavior and women's experiences of care are limited.  There is also a question regarding the quality of some reporting, in particular issues surrounding prenatal care during pregnancy.  To get a better sense of some of these issues we decided to analyze data from CDC's Pregnancy Risk and Assessment Monitoring Surveillance System or PRAMS.  Our study was the first time that PRAMS data was used to look at outcomes among foreign-born women. So what is PRAMS?  PRAMS is part of the CDC initiative to reduce infant mortality and low birth weight.  It is a population-based surveillance system that was designed to identify and monitor selected self-reporting maternal behaviors and experiences that occur before, during and after pregnancy among women who deliver a live born infant.

Each participating state uses a standardized data collection method developed by CDC.  Every month the stratified sample of 100 to 250 new mothers is selected from eligible birth certificates.  PRAMS staff in each state collect data through statewide mailings and follow up with non-responders by telephone.  At two to six months after delivery, each sample mother is then (inaudible) (audio difficulty:Static).This slide shows the 32 states who were participating in PRAMS as of 2000 by the year of their entry.

Today I'll be presenting data for four of these PRAM states for the combined years 1997 to 2000.  We selected Florida and Washington because both were among the top ten states for births to foreign-born mothers in both 1990 and 2000.

We selected Colorado and North Carolina because, according to our analysis of MCHS data, they were among the six states during 1990 to 2000 that experienced a greater than 10 percent absolute increase in births to women born outside the states.

As mentioned earlier, we use data for the combined years of 1997 to 2000.  Because mother's place of birth was not one of the standard PRAMS variables, all four states linked back to birth certificates for all women sampled to provide this information. 0.4 percent of these files were missing information on mother's place of birth.  Our analysis is based on the weighted data. As a focus of this presentation it's on the health characteristics of foreign-born mothers and their infants results from women born in U.S. territories will not be presented.  This represented 0.8 percent of the data.

In the remainder of this talk I will be discussing our results in terms of non-Hispanic white, non-Hispanic black, Asian Pacific Islander and Hispanic origin, with a focus on the subcategories of Mexican, Cuban, Central South American and other.

I will also highlight our findings regarding maternal age, education, marital status and prenatal care in the first trimester.  I'll also present results from selected maternal health behaviors and birth outcomes.

Our data were analyzed using SASS and the SUDAAN Statistical Package in order to calculate weighted population estimates.  Percentages, standard errors, odds ratios confidence intervals and P-values were also calculated.  For this presentation I will only be presenting weighted percentages and their statistical significance.

As noted earlier, according to PRAMS standard, only data that achieve a weighted response rate of 70 percent are analyzed.  This slide shows the overall weighted response rates for the selected PRAM states, as well as their individual response rates.  As this was the first time PRAMS data had been stratified by mother's place of birth, we calculated those response rates as well.

Here are those rates overall and for the four individual states.  As you can see, response rates for foreign-born women in Colorado and North Carolina are lower than 70 percent.  For this reason I'm presenting our results for all four states combined.  We were also interested in seeing the response rates when the sample of women were stratified by race and ethnicity.

Here are those numbers for the four states combined.  All are above the 70 percent cut-off level.  When stratified by Hispanic ethnicity, Mexican and other Hispanic mothers fell below the 70 percent level from the four states combined and when stratified by mother's place of birth, given that Hispanics made up the largest group of foreign-born mothers in our study, this lower response rate among these two subpopulations is notable.

In looking at the distribution of state-born and foreign-born mothers overall we found that foreign-born mothers made up the majority of births among Asian Pacific Islanders and Hispanics. Among Hispanic subcategories, foreign-born mothers make up the greater percentage of all births, with births to Mexican and Central South American women predominating.  Let me turn now to some of our results regarding maternal demographics.  Similar to what we found in our analysis of national data, foreign-born mothers were less likely to be teen mothers or to be unmarried.  With the exception of Hispanic mothers, they were also less likely to have less than a high school education.

Because the income variable differed among all four states, we used the ratio of persons per room in household as a proxy.  Foreign-born mothers in all race and ethnic categories were more likely than their state-born counterparts to report living in a household with greater than one person per room ratio.  With the exception of whites these findings were statistically significant.

Statistical significance at this  .05 level is indicated by the yellow star.

Let me turn to some health service data.  In our preliminary analysis of health care service variables, foreign-born women were less likely than their state-born counterparts to begin prenatal care during their first trimester.

This trend was reversed among black women and was statistically significant.  Among Hispanics, this trend was reversed for Cuban women.  For birth outcomes, similar to what we found in our analysis of national level data, foreign-born women were less likely to deliver preterm than their state-born counterparts. This trend was reversed but not statistically significant for Asian Pacific Islander women.  Differences within Hispanic categories were not significant either.

We found a similar trend for low birth weight, which, due to small numbers, only differences between black and Hispanic state-born and foreign-born women were statistically significant.  Among Hispanic women, these differences were significant for Mexican women only.

In the next set of slides I present findings from our analysis of selected maternal behaviors.  Alcohol use during the last three months of pregnancy, six or more stressful events during the year prior to giving birth.  Smoking during the last three months of pregnancy.  Physical abuse prior to pregnancy and breast feeding for at least one month, and putting the infant to sleep on its back.

Overall, foreign-born mothers were less likely to smoke during pregnancy.  This was also true for Hispanic and ethnic categories.  Although the differences for Cuban and south American women were not significant.

Overall, foreign-born mothers were less likely to drink alcohol during pregnancy.  This trend was reversed for white and black women, although the differences were not statistically significant.

With the exception of central south American women, a similar pattern was found among Hispanic women.  Although none of these differences were significant at the  .05 level but the differences between Cuban women were borderline significant at  .06.  With the exception of white women, foreign-born women were less likely to report ever being physically abused before pregnancy, although these differences were only statistically significant among black women.

According to the PRAMS definition, physically abused meant being pushed, hit, slapped, kicked, choked or physically hurt.

A similar pattern was found regarding abuse during pregnancy.  Again, none of the findings were statistically significant. Foreign-born women were less likely to report six or more stressful events during the year prior to their giving birth.  These findings were all statistically significant.

With the exception of Cubans, a similar pattern was found among Hispanic women but only statistically significant for Mexican women.

So what is a stressful event?  Women were asked their experiences of these 12 stressful events, which were adapted from the 18 item modified life events inventory.

It's also quite possible that this list doesn't capture stressful events as perceived by immigrant women.  For example, immigrant women who don't speak English may find it very stressful to go to a prenatal clinic and not be able to communicate.  But this type of stress is not  -- this is very American focused stressful events.

Foreign-born mothers were less likely to breast feed their infant for at least one month.  With the exception of Central South American mothers, a similar pattern was found among Hispanic mothers.

Note that the difference between state-born and foreign-born  -- note the difference here between state-born and foreign-born black mothers in terms of breast feeding.  Quite a large difference.

We found very little difference between foreign-born, state-born mothers with regard to how they laid their babies down to sleep.  As putting the baby on its back to sleep is associated with lower rates of Sudden Infant Death Syndrome, the overall low rates between and within race and ethnic categories is notable. In sum, foreign-born mothers were less likely than their state-born counterparts to be a teenager when they gave birth and less likely to be unmarried.  They were also less likely to report having six or more stressful events in the year prior to giving birth.

Foreign-born mothers were also less likely to smoke or drink alcohol during the last three months of their pregnancy.  They were more likely to breast feed their infants for at least one month.

Foreign-born mothers were more likely to live in a household with greater than one person per room ratio.  They were also more likely to start prenatal care later.  Although the person to room ratio is used as a proxy for lower income, I think it's also possible that in some cases this ratio could be an indication of greater social support.  Anthropological research methods may be one way of exploring this further.

To recap some of our specific findings with regard to differences between foreign-born women and state-born women of the same race and ethnicity.  Our analysis found that foreign-born black mothers were more likely to begin prenatal care during their first trimester while foreign-born next can mothers were more likely than state-born Mexican women to have less than a high school education.

Foreign-born Cuban mothers were less likely to report alcohol use during pregnancy but more likely to experience greater than six or more stressful events during the year prior to birth.  So I want to point out again, this is based just on four states.  So what we're hoping to do is get more states in the study and we may find this ethnic differences vary when we get different states in the analysis.

We anticipate expanding our analysis of access to health care issues by looking at PRAMS variables for Medicaid and health care coverage, as well as barriers to prenatal care, content of prenatal care and prenatal care education.  We're also beginning to look at some of the qualitative information.  At the end of each survey women are able to write in their comments.  And I've begun looking at some of these comments for the year, just for the year 2000, and overall approximately six percent of state-born and six percent of foreign-born women gave some comments.  And so there's some really interesting  -- it's a real rich data said that I'm hoping to spend a lot of time on. So I'd like point out some of our limitations.  Although we hoped to analyze our data in terms of foreign-born women's ethnic origin, the coding scheme for vital records data eliminates our ability to do so.  For example, what does a category foreign-born black woman mean exactly?  It's very likely that black women from Carribbean countries have different health care issues than women who come from Subsaharan Africa.  Our analysis was also limited by our inability to cover unique health issues from the perspective of foreign-born mothers.

For example, do birth outcomes from foreign-born women differ depending on whether a woman come to the U.S. as an immigrant or as a refugee from a war-torn setting?  Do foreign-born women who speak English have an easier time than those who don't?  How do cultural practices between and within these groups of women differ?  Our analysis is also limited in this sense that we were not able to differentiate between recent and longer term immigrants.  Some research suggests that immigrants who live in the U.S. for 10 to 12 more years have health outcomes similar to native-born Americas or U.S.-born women.

The smaller size of our sample for some race and ethnic groups was also a limitation.  Although this was a first analysis of PRAMS data by mother's place of birth, and it's yielded some interesting results.  It's still very much a first step and doesn't capture their full experiences.  That's where I think anthropological methods would be useful here at the state level, as states differ with the foreign-born make-up, qualitative inquiries into the different issues within states could yield a lot of interesting information.

In addition to the limitations of our analysis we also have some strengths.  Our preliminary findings suggest additional areas of inquiry into why foreign-born women have better outcomes than their state-born counterparts.

For example, although we found that foreign-born women were more likely to be better educated than those born in the U.S., we did not know if they acquired that education before or after they moved to the states.  However, from the housing perspective, in terms of women per room ratio, they were probably poorer women. Regarding lower rates of teen birth, we have no way of knowing if the foreign-born women in our study came at an older age.  We also had no way of determining whether or not they were healthier either.  Our analysis does suggest, however, that foreign-born women engage in healthier behaviors during pregnancy or at least less harmful ones.  For example, they were less likely to smoke during pregnancy.

A larger sample might have found that the differences in alcohol use were statistically significant.  We also saw that foreign-born mothers, black and Cubans, enter prenatal care earlier.  Foreign-born mothers also reported less stressful events and were more likely to breast feed their infants. This study highlights the need for further research into the economic cultural and language barriers foreign-born mothers living in the U.S. face.  And the impact those barriers have on the health of their infants.  Research that draws upon the strengths of both anthropological and epidemiological methods would be a perfect place to start.

And that's the end of my presentation.

Webcast Manager:   Okay.  There are no questions on the Internet site.

Susan Nalder:  This is Susan Nalder in New Mexico. Denise, thank you very much.  We are hoping that we'll be able to share our data with you.

Dr. Allen:  Oh, great.

Susan Nalder:  We have to get vital records permission.  But I wanted to comment that we are now doing an analysis of U.S.-born, foreign-born for the U.S. mexico border health community.  And we are going to conduct what we call a partnership consultation on March 7th in Los Crusas.  This means we sit down with promontories and local folks and ask them to tell us what does this data suggest, what does it mean.  And so if you're interested to participate, just follow back and we'll, we're inviting you now.

Dr. Allen:  Okay.  That's great.  That's great.  I'm very interested.  So these are health care workers on both side of the border will be at this conference?

Susan Nalder:  These would be the U.S. mexico border team from the U.S. side, including the promontories, who are familiar with what goes on on both sides.  I would say the teams are familiar.  But it gives us an opportunity to get interpretation.

Dr. Allen:  Right that's great.  The U.S. border commission on health just came and recently gave a presentation at CDC.  I'm assuming that's part of that.  That might be connected somehow.

Susan Nalder:  They were not connected.

Dr. Allen:  They're not.  Okay.  Well, it sounds great.  Would I really like to be involved in that.  I'm hearing static again.  I'm not sure if  --

Webcast Manager: -- It's gone.  We actually have a question from one of our Internet participants and she's wondering, I think one of the questions in PRAMS relates to whether mom's work outside of their homes or not.  Did your study include what percentage of foreign-born mom's worked outside of the home compared to U.S.-born?

Dr. Allen:  We're doing an ongoing analysis now.  That's one of the variables we're going to be looking at.  It is there.  We can look at sources of income.  There's a question that asks where all your sources of income were.  I just haven't looked at that yet.  But we will be looking at that.  We'll also be looking at Medicaid and also coverage and also the type of health care provider that a woman went to, is there a difference between U.S. and state-born and foreign-born women in terms of the type of health care provider they seek out.

So that's a great question.  And, yes, we will be looking at that but I don't have the information right now.

Webcast Manager: We also have another question, what kind of outcomes would you expect to find if you were able to differentiate longer term versus shorter term immigration?

Dr. Allen:  You know, that's a really good question.  I think that because some of the literature is suggesting that women who have lived in the states for ten years or more have similar outcomes to U.S. born women, I would suggest that the longer term, the longer term immigrant have their birth outcomes are similar to those in the U.S.

I think that there might be differences though in terms of access to care issues.  They may have less barriers to care.  But I also think that in terms of alcohol use during pregnancy and smoking during pregnancy among longer term immigrants may mirror those of U.S.-born.  And that's why we wish we really could look at that, because this term foreign-born may be masking the heterogeneity within even that group.  And also recent immigrants have a lot of different, I mentioned that, they have issues in terms of access to care language barriers, but also I'm not an expert on Medicaid and funding, but I think that there's a five-year, you have to be in the states, I think five years before you can qualify I'm not sure maybe somebody out there can correct me.

Webcast Manager:  You have to be in the United States five years to qualify for Medicare or else you have to reimburse them if they catch you on Medicaid and you've not been there for five years.

Dr. Allen:  Right.  So that, if we were able to differentiate between longer term and shorter term immigrants, we would find differences right there.  What we would probably find is that there's a lot of, there's a question on PRAMS that asks were you on Medicaid during pregnancy and then were you on Medicaid during delivery.  And I think what we would find is among recent immigrants that Medicaid coverage during delivery would be higher because they can get it for emergency, somebody comes in and is giving, in labor and giving birth, they'd be covered.  So I think that's where we'd find the differences.

But so I mean that might be a question that PRAMS could include at some point down the line, is how long have you lived in the states.  But maybe some people might see that as too much information.  So I'm not sure that can be asked. I hope that answered the question.

Any more questions?

Webcast Manager: Thank you.  There are no more questions on the Internet.

Ken Rosenberg:  This is Ken Rosenberg in Oregon.  We're doing something not with PRAMS data but might be of interest.  One of the issues that comes up in a lot of states is whether Medicaid should cover the prenatal care of pregnant women who are undocumented.  We're going to play with that by looking at the Medicaid enrollment data and linking that with the birth certificate and looking at the prenatal care, the initiation and number of visits for (inaudible) women, undocumented women and comparing them to the general population and to the other Medicaid women, interested in looking at birth outcomes because obviously foreign-born women will have better birth outcomes but just looking at number and type of prenatal care visits.  That's certainly links with what you're describing.  I'm not sure it has anything to do with PRAMS.

Dr. Allen:  Yeah.  I loved your article, by the way.  I thought it was a great one.  When you looked at in New York City these issues in New York City, I thought it was a great article.

Ken Rosenberg:  Thank you.

Dr. Allen:  I'd recommend it to everyone out there.

Ken Rosenberg:  If people want the reference, they can send me a note and I'll give it to you.

Webcast Manager: Okay.  We have another question from the Internet site.  Did you look at the question of whether the women received prenatal care as soon as they wanted?  If so, was there any indication of foreign-born women not wanting prenatal care early in pregnancy.

Dr. Allen:  That's a great question.  I actually just looked at it last week.  And what's really interesting is U.S. born women were more likely to  -- there were two sets of questions.  There's a set of questions that says did you get prenatal care as early as you wanted and then there's a series of questions that asks what were the reasons you didn't get prenatal care.  And so when we asked for the first question, actually I don't have the numbers right in front of me, but what was interesting, what came out of it is that foreign-born women said that they got prenatal care as early as they wanted.  But when I looked at that by when they actually got prenatal care, even women who got care later, like in their third trimester said they got it as early as they wanted.  What would be interesting to do and what we'll be doing is looking at that also by parity, women who are pregnant for the first time, are they more likely to want it early or not.  But that's a great question.  I don't have the data in front of me but we just started to look at that.

What was also interesting about that, I don't have the numbers, but U.S.-born women or state-born women who didn't get prenatal care as early as they wanted were more likely to say it was because they didn't know they were pregnant.

So there's a lot of analysis that we're doing along those lines.

Webcast Manager: Okay.  Thank you.  That's all the questions on the Internet site again.

Are there any other questions on the phone side?

Mary Ellen Simpson:  This is Mary Ellen Simpson in Illinois.  I really enjoyed your presentation.

Dr. Allen:  It's so weird giving a presentation and not being able to see faces.

Mary Ellen Simpson:  I have a question and maybe you addressed this and I didn't catch it.  But I think you said there were like, you sort of categorized women that were foreign-born into four categories.  And the latter two are being refugees or illegal immigrants.  Do you think there might have been a bias response bias in using the PRAMS data, especially of those two groups within the foreign-born women of maybe not responding and that possibly there might have been some effect that maybe you were getting the legal immigrants that, it might be a different strata than, economic than perhaps an illegal immigrant or refugee would have been.  Could you discuss that?

Dr. Allen:  That's a great question when I was talking about the types of foreign-born, I was talking those four definitions.  It's according to the U.S. census bureau.  You can't get at, when you're looking at birth certificate data you can't get at whether or not a woman is an immigrant or a refugee or whether or not she's illegal because citizenship status isn't on the birth certificate.  So we have no way of knowing if the women who were responding were illegal or not.

What we were trying to look at that might give some kind of indication of that is to go back to the Medicaid data and look at the percentage of foreign-born women who were not on prenatal, Medicaid during prenatal care but were on Medicaid during delivery, which may indicate that they were not here legally or at least didn't qualify.  But since the birth certificate doesn't say anything about citizenship or immigration status, there's just no way to know that, at least from the way we're analyzing the data.  Does that answer your question?

Mary Ellen Simpson:  Yes, it does.  Thank you.

Webcast Manager: Okay.  Denise, we have another question from the Internet.  And this is from Tina and she's asking:  What about intendedness of pregnancy?

Dr. Allen:  That's another variable that we didn't look at for this presentation.  Definitely we'll be looking at it.  But I haven't looked at that yet.  So I will be looking at it, but I just don't have those numbers right now.

Webcast Manager: Okay.  Thank you.  There's no more questions on the Internet site.  Are there any more questions on the phone side?

Susan Nalder:  It's Susan Nalder.  A couple of quick ones.  Denise, I'm looking at your weighted response rates by state, the very last page.  And you show Florida and then Washington, those response rates are pretty decent. But in Colorado and North Carolina they're not.

Dr. Allen:  Yeah.

Susan Nalder:  Did you talk to the states and ask them  -- we see similar stuff I haven't looked at it recently for Mexican-born, we are told that Mexicans return home for a certain period of time after a new baby is born.  These would be undocumented women more likely than any other group.

Dr. Allen:  Another issue  -- I haven't at this point talked specifically to states about this.  I've talked to the PRAMS people here at CDC, and they pointed out that states stratify or sample, stratify their sample according to different variables, like Colorado stratifies, if I'm right here, they stratify by urban  -- stratify by Denver and by urban and rural.  And then also by low birth weight status.  And then also I think North Carolina stratifies by low birth weight.

So since foreign-born are less likely to give birth to a low birth weight baby than U.S. born, that also might be a reason why their numbers are lower.  That could be.

Susan Nalder:  Why would you think so?  Low birth weight is where we see poor response compared to normal birth weight.

Dr. Allen:  Okay.  Just those that are low birth weight anyway are less likely to respond?

Susan Nalder:  Yeah.

Dr. Allen:  Oh.

Susan Nalder:  Ladies are less likely to have low birth weight but also less likely to respond.

Dr. Allen:  Okay.  I'm not sure what those numbers mean.  I mean we just did this first analysis and I have sent it out to the states and are waiting to start a dialog on that.  I'm not exactly sure what that means.  It could mean a variety of different things.  It could mean  -- it's also true that both Colorado and North Carolina have been in PRAMS a less amount of time, that they came in I think it was in  -- there's a slide there that I show what years these came in.  So they have less time working with PRAMS.  And Florida and Washington have been in PRAMS longer and as the years go by you get better and better at getting response rates.  So that's also another issue in terms of when they came into PRAMS.

Amy Lansky:  This is Amy Lansky.  I wanted to clarify the issue you were talking with Susan about.  With the stratification variables and the low birth weight, you may have fewer foreign-born women who make it into the sample.  But that doesn't necessarily mean they'll have a lower response rate.

Dr. Allen:  Okay.  Sorry I misspoke on that comment.

Webcast Manager: All right.  Are there any further questions? All right.  There are no further questions on the Internet side.  So we'll turn it over to Tishia.

Tishia Smith: Thanks.  Thanks everyone for joining us for this month's broadcast.  Thank you to Denise for a very nice presentation.  I'll also encourage everyone to join us on March 5th for our next broadcast.  You can join us for an educational session on linking data.  Thank you and I hope everyone has a great afternoon.

 ###