Ninth Annual Maternal and Child Health Epidemiology Conference / December 10-12, 2003
Foreign-Born Uninsured Women in New Mexico
SSU WENG: First of all, with the title slide, I wanted to clarify that my talk is about New Mexican foreign-born women, so it's not a national data set. First of all, I'd like to acknowledge all the people who've made this project possible, especially people in the PRAMS team. What stimulated my interest in this topic was that studies have described an epidemiologic paradox that foreign-born women tended to be at risk in terms of their demographics and other characteristics but their outcomes are either similar or better in many instances than state-born women. And there are a couple of other presentations from (inaudible). One was last year, and one is actually simultaneous with this one, addressing a similar question.
The original purpose of my study was to provide data for the New Mexico Prenatal Care Network Task Force to seek Medicaid coverage for undocumented pregnant women. This has become a moot issue because we don't have funding for increasing Medicaid coverage for anybody. But what I was trying to do was to compare the undocumented women with women who had Medicaid for prenatal care. Unfortunately, my data set does not have any information about legal and citizenship status, so I wasn't able to do this. So I thought it's still interesting to look at how foreign-born uninsured women do compared with U.S.-born uninsured women, and I looked at three groups of so-called outcomes--they're dependent variables--maternal characteristics, access to and utilization of prenatal care, and then actual outcome variables--mother or infant outcomes.
My methods were to combine two sources of data. This is done automatically when you have a PRAMS data set. And just as an aside, for those--well, I'll explain about PRAMS later. Anyway, the birth certificates provide demographics and some other variables. PRAMS provides a wealth of information about maternal behaviors, healthcare, hospital stay, and then for some of the variables, I used data from both data sets. This is a descriptive exploratory study, and I hope to continue this later on. What I'm going to show you are charts with the confidence intervals, and for those of you who want P values, I'll give those on request.
Is everybody here familiar with PRAMS or no? Just a handful? It's a multiyear, multi-state study led by CDC. New Mexico has collected data since 1997. We're population based. We get our frame from live birth certificates, and there're some exclusions. Well, for all states, it's basically a stratified systematic sample. For New Mexico, what we've done is for the first few years we over-sampled low birth weight infants and Native Americans. And then because of our users' request, we changed to equal allocation by region. The data collection's really intensive.
We have a mailed survey, mail it out up to three times, then telephone follow-up, attempts to do telephone interviews for non-responders, and we get a 70-percent overall response rate, somewhat lower for certain subgroups. Now, what I did with this particular study was I made my independent variable based upon two variables. One is the maternal country of birth, and I excluded the territories. I divided it into U.S. states versus any foreign country. And then the second variable I took was the payer of prenatal care, which comes from the PRAMS data set. The country of birth comes from the birth certificate.
And so first of all, I looked at two groups of uninsured women. They're either foreign born or U.S. born. Then for the women who remained, I had a hierarchy according to the payers. And first, I took Indian Health Service, so that group may or may not have another payer in that group because some women go to Indian Health Service and have private insurance and/or Medicaid. Then the next step was Medicaid, which may or may not have private payer. And then the next one was the private pay only. And so I ended up with these five groups. And the focus of this presentation is to compare the foreign-born uninsured women with either the U.S.-born uninsured or with the Medicaid women. And the second group has no foreign-born women; the third group has a very few, and I plan to address this problem later when I redo the study. And these are the sample sizes that came up. And as you can see, we have a relatively small data set.
For the period from 1997 to 2001, we had 7,310 respondents, and I restricted this to singletons. By the time I include the ones with missing data, I have 380 in the group of main interest, and only 216 in the U.S.-born uninsured. Now, the first issue is to look at maternal characteristics. Are the foreign-born uninsured actually disadvantaged? And yes, indeed, in terms of education, income, and potential language barrier, they are at a disadvantage. But in terms of age, they're less likely to have teenage mothers than the Medicaid group, less likely to have mothers who are more than 35 years old than the U.S.-born uninsured group, and less likely than the Medicaid mothers to be unmarried.
Now, I'm going to look at some of their characteristics. In terms of intention of pregnancy, the proportion of foreign-born uninsured who had intended pregnancy was greater than U.S.-born uninsured or Medicaid. Unwanted pregnancy--they were most like the U.S.-born uninsured, but they had an advantage over the Medicaid. For use of alcohol frequently or binging during the three months before pregnancy, they look as if they had an advantage over the U.S.-born uninsured or the Medicaid, either of those groups. Smoking cigarettes, they did much better than all of the groups except for the Indian Health Service group.
For stressful events, we have a group of 12 questions that we ask in PRAMS. If you just add them up and make a score, the percent of mothers with six to 12 stressful events during the 12 months before delivery was similar to the U.S.-born uninsured but lower than the Medicaid group, and this just shows you the stressful events broken down. And you can see that except for divorce and being homeless, the foreign-born uninsured did better than the Medicaid in all of these measures. They did better than the U.S. uninsured except in the divorce and being homeless, and this is in your handouts. Oh, by the way, excuse me, the order of the slides is a little different from the handouts up here, but the data are all in the handouts.
For physical partner abuse during the 12 months before pregnancy, the foreign-born uninsured were less likely than the Medicaid women to report this, but I believe that they were statistically significant and significantly more likely than the U.S.-born uninsured. Now, for social support, all these three groups were about the same in the proportion of mothers who reported support from a husband or partner, but the foreign-born uninsured were less likely to have social support from a family member or friend than either of the U.S.-born uninsured or the Medicaid women.
This is social support in terms of help raising the baby. And then I created a variable for medical risks. Oh, no, actually, the medical risk hypertension, I took information from both the birth certificate and PRAMS, and the foreign-born uninsured were less likely than the Medicaid women but about the same as the U.S.-born uninsured to be treated for hypertension during pregnancy and similarly for being hospitalized during pregnancy. But if you look at the risk of being overweight, the foreign-born uninsured don't have such a definite advantage, and they're actually significantly more likely than the U.S.-born uninsured to have been overweight before pregnancy. And then if you look at diabetes, which you would think would be related to being overweight, there's really no definite difference in the three groups of interest.
And finally, the second group of variables that I looked at was related to access to and utilization of care. If you look at late entry to prenatal care, the foreign-born uninsured were more likely than the Medicaid women, as you would expect since they lacked financial access, to be late entering prenatal care and about the same as the U.S.-born uninsured women. And what I thought was interesting was that we always speculate that there may be cultural factors that affect the entry to prenatal care, and if you look at the foreign-born uninsured, they were more likely than any group except for the private insured to say--if you look at the women who started prenatal care late across the board and you'd compared the foreign-born with the others, they're more likely to say that they started as early as they wanted, and so starting late may not be an issue; just the financial access is the motivation to start early.
When you look at the Kotelchuck Index, the APNCU Index, and look at the inadequate level, the foreign-born uninsured were about the same as the U.S.-born uninsured and more likely than the Medicaid women to have inadequate level. And I can give you data on the other levels if you're interested. Now, we are very much interested in our state in services, and so I looked at the safety-net sources of prenatal care, at the public health departments and community health centers, and the foreign-born uninsured women were more likely than the other two groups to use public health departments or community health centers, so they were making use of the safety net.
Unfortunately, for delivery, we don't have a real safety net, and this slide only looks at the foreign-born uninsured women, and I created a variable to see whether or not they had either a third-party payer or indigent fund, yes or no. And almost 60 percent did not have a third-party payer or indigent fund. So in other words, they're going to have to use their own income or just not pay the bill. And the proportion who used their personal income was over 35 percent. The proportion who eventually got Medicaid payer for delivery was a little bit over 25 percent. An indigent fund helps about 12 or 13 percent of them.
Now, I'm coming to some true outcomes. For infants' birth weight, in our state, it looks as if the foreign-born uninsured women are less likely than Medicaid women to have very low birth weight, less than 1,500 grams, or babies 1,500 to 2,500 grams, and more likely to have high birth weight babies. And for gestational age, you see a similar pattern of the two extremes. The foreign-born uninsured women are less likely than the Medicaid women to have babies who are extremely premature, less than 32 weeks; they're a little bit more likely than the Medicaid babies to have babies who are term.
In terms of hospital stay, the median number of days was the same for all groups. The mean number of days was similar for all groups, although there's a suggestion that it's lower for foreign-born women, uninsured women, than the Medicaid women. And for the infant, admission to ICU was similar for the three groups that I'm most interested in: the foreign-born insured, U.S.-born uninsured, and Medicaid women. And there was no difference for staying six days or more.
So when you look at the paradox with New Mexico data, indeed, the foreign-born uninsured women are challenged in terms of their education, income, homelessness, divorce, or partner abuse before pregnancy, and they're certainly challenged in terms of access to and utilization of prenatal care. And when you compare them with Medicaid women, they do seem to do better with low birth weight or preterm delivery, but there's no difference in terms of their newborn stay in ICU or nights in hospital. And some of the factors contributing to these outcomes may be their behavior, smoking cigarettes, use of alcohol, social issues, stressful events, their attitudes towards the intention of pregnancy, and the fact that there are healthcare resources for them.
And some of the implications in our state for practice and policy were that we need to continue supporting community health centers and public health clinics, that we still need to find a pair of delivery for these foreign-born uninsured women, and that we need to motivate women to get prenatal care in order to get adequate prenatal care. Some of the implications of these exploratory findings for future studies are that foreign-born uninsured women quite often resemble the U.S.-born uninsured, and in comparing foreign-born and U.S.-born women, one needs to consider their financial access as well as country of birth.
I don't feel as if I can just lump all the women who are U.S.-born women. And one thing that I would be very much interested in studying later on is high birth weight and large for gestational age infants and to do a little bit more refined study. The other thing is I'm hoping some day we'll be able to link our data sets so that we can do more outcome studies, and these were some topics that I'm interested in exploring eventually. Certainly, the birth defects and neural tube defects, they've shown that women on the border are more likely to have neural tube defects, and this has to do with use of prenatal diagnosis perhaps. And there are many limitations to this study, and I won't dwell on them, but I'm quite aware of them.