Ninth Annual Maternal and Child Health Epidemiology Conference / December 10-12, 2003

Dynamic Changes in Medicaid Coverage of Pregnant Women

LI YAN:  The title of my talk today is "Dynamic Changes in Medicaid Coverage of Pregnant Women."  I was new to this field.  Actually, I obtained my Ph.D. in physics, so if I make outrageous statements here, please point it to me, and it's tremendous help to me.  But I will say for my short time of in this field, I was absolutely thrilled how we can--the potential we can do to help our children and all the knowledge just pour into my mind, so thrilled I just sometimes I wake up in the night thinking, "What was in there?"  Anyway, the objective here in my talk here is I want to determine when the pregnant woman enters and exits the Medicaid program around the time of delivery.  As I was here, I was follow-up the one year before and after the delivery, and want to evaluate the changes in these coverage.  This is actually very important because we have this Medicaid program designed to help pregnant women and to know when and how they are in and out of this program gave us actual knowledge whether we achieved or designed the goal for these programs.  So the target women in here is women who is in Florida residence who give birth between '95 to 2000 I was linking this (inaudible) studies data to the Medicaid eligibility records because I need one year before and after, so it's studied more in those parts. 

Before I go any further, I want to give some brief introduction or background or definitions what I was talking about because I'm going to use extensive (inaudible) for my talk.  I would say a pregnancy is Medicaid pregnancy if the woman had at least one day eligible for Medicaid during her pregnancies.  In here, I generally separate them to two subgroups.  One is so-called SOBRA group, which is defined in the Sixth Omnibus Budget Reconciliation Acts.  To my knowledge, the program is designed for those women who need help in the pregnancy.  They are relatively poor, is about 25 to 85 percent of federal poverty in state of Florida.  And another subgroup is all other Medicaid coverage is called non-SOBRA, which is generally--most of them is regular Medicaid participant who are with very low-income levels, about 24 percent of federal poverty levels.  The purpose of divide is that to two subgroups is a relatively simple way to estimate the party level of pregnant women without go into too much details. 

And I highlight this (inaudible) in there which is supposedly a SOBRA subgroup because this subgroup is inherited from the research done generally by other people, and how we explain later, this is actually a problematic subgroup in there.  But in general, you will remember that the non-SOBRA group is poorer than SOBRA group, and certainly there's some Medicaid pregnancy people or women is poorer than non-Medicaid people.  And then I want to say how we're going to classify this group because, like I mentioned, that the coverage is quite dynamic during their pregnancy.  In my purpose, I will determine this subgroup by the date of the delivery.  Normally, your thinking is the longest coverage just during the pregnancy just to see the impact of those programs, but in here I want to say day-to-day changes in there.  So I was just choosing what happened at their deliveries.  Here is what a simplified process, how you got into the Medicaid programs.  If you never be in Medicaid programs, January you go into this (inaudible) presumptive coverage, and within 60 days of you in there, they evaluate your income, income validates, and then decided that you are actually going to be into SOBRA type of coverage or non-SOBRA, or you're not just eligible or something like that.  Certainly, also, your coverage could be changing during the pregnancy because your income level is changed; your situation has changed in here. 

So from here, you can see (inaudible) is actually a temporary coverage.  We should not include (inaudible) in the SOBRA group, but I will deal with that issue later.  (Inaudible) briefly talk about my data.  Data is (inaudible) studied from Florida PBS office.  There are about 200,000 peoples each year, and they are unique under the (inaudible) number.  They're including other identifiers like social security number, name, address, date of birth, something like that.  Include births of the mom, dad, if possible, something like that.  Medicaid eligibility, they are (inaudible) is Medicaid ID.  They include social number, name address, date (inaudible), something like that.  But the big problem is you don't have common unique identifier.  So we developed a strategy, a program to try to match as many as possible moms to the case because the standard practice using social security number to link this to is quite unsatisfying to me.  That is used earlier.  But because you're going to miss a big chunk of population who don't have social security number, which could be very different and vulnerable to all the risk factors.  We have another colleague Friday morning talking about that issue.  So because I want to know one year before and after delivery, so (inaudible) three years of Medicaid eligibility file, and like I mentioned, (inaudible) the combination of those triple matches. 

Here is what actually my meat is.  You can say overall the SOBRA woman began to enroll at the beginning of their pregnancy.  My arrow is about 217 days before their delivery.  So from here is quite encouraging because you can see many people enrolled in the SOBRA Medicaid during the first trimester, which is what we intended to do.  So I think that is a success.  They all gradually increased during the second trimester, both SOBRA and non-SOBRA women.  But this (inaudible) when they realize they have needs.  But what really began to worry me, is this non-SOBRA women would know they are really poor, they need help the most, is responding quite slow during their pregnancy.  They are getting help too late to me because we all know the first trimester is critical time to help their outcomes of children, the outcomes of women's health, and yet, you can see they're only coming here around date of delivery.  Actually, almost half of them are coming in there at a last trimester, where basically you don't have much to do to affect the outcomes.  Another thing is worrying me on data side because I'm a data person is we have a huge spike at date of delivery, which meaning some (inaudible).  I want to know why.  But, anyway, you can say really poor women got help late.  And then the last point is SOBRA women got their coverage decreased within two months of delivery, which is actually natural because they are going to lose their eligibility after 60-some day delivery.  So that is expected. 

In here, I have *FP, and there is a new program, a Family Planning Waiver Program.  It's designed to help women who not qualify for regular Medicaid after they have Medicaid funded delivery, we gave them this eligibility so they can have those family-planning-related coverage, but that is not the central topic today we have.  Here I will show over years--'95, '96, '97--they're all basically the same.  From '98, the Family Planning Waiver Program kicked in, so you have something in there.  Over years, the change is the same.  And all of them basically really worry me is non-SOBRA women got coverage late and another one is like I mentioned, that we have a spike around the date of delivery, which meaning the data of non-SOBRA have problems, so we talked, talked, talked with state department of health.  People, finally we began to learn what is really happening here.  Actually, this is happening for the past two, three weeks. 

So this is a new data.  I was working on this the last Sunday and Monday, so some of them is really--anyway, what we learned is we have--we should--excuse me.  The Medicaid coverage is actually a little bit more complex than the SOBRA and non-SOBRA.  They have emergency coverage for alien (inaudible) which by being alien, they are not eligible for Medicaid, but they could be paid for their delivery.  And they may have special needs for the Medicaid, this SSI, because the women may have health or behavioral problems, such as retarded, something like that.  And they may be medically needy.  This is actually not really poor women.  They have their own policy insurance, but Medicaid as a secondary coverage kick in, so they are actually richer than SOBRA women.  And then we have really a small amount of bad data in here.  So during this regrouping, we kind of make this SOBRA and non-SOBRA women more homogeneous.  And in here we can say basically the behavior has not changed, so my worries not go away.  The SOBRA women have got their help after this, more or less, as we expected, although I would like they got help even earlier, like the leading (inaudible) would be better.  But the non-SOBRA woman is really not getting help until really late.  The majority of this spike in here we suspect is due to alien coverage. 

So by further grouping like this, we can have a better understanding on the outcome analysis, and we can better targeting what the problem they may have instead of the general population.  So in summary, we have created a program to look at these dynamic changes on Medicaid coverage during pregnancy.  During this process, it's aroused my concern how well we're actually helping this population.  And I think I've run out of time, so I'll run through this faster.  To me, I was thinking what should we do differently to help these really poor women.  Is their (inaudible) for example, their lack of knowledge of this coverage, or is it their motive because they are just, for example, we further are going to start whether teenage women in there is more and something--what caused this slow increase of enrollment in non-SOBRA women.  And as I mentioned, our outcome after considering all other factors--the social, racial, all the other factors--the non-SOBRA woman is always worse than in other groups.  So and recently we began to *tap the costs effect on there.  We found out the early intervention is always big cost saver in there, basically saying if you do it at the beginning of your trimester, you always save the money.  So in this aspect, we still need to do a little bit more solely with that, but I think in this respect, I think for policy what target population should we put more effort to improve our health outcome in our state and the United States in general will be a question to everybody.  Thanks.