By Christopher Guadagnino, Ph.D.

Published August 1997

• React to this article in the Discussion Forum.

Richard Baron, M.D., is Medical Director for Care Integration at Health Partners and Project Director of the Healthier Babies Project.

PND: What is the Healthier Babies Project?

RB: What this project proposes to do is to create a universal prenatal encounter form that all the Medicaid organizations are going to agree to use. Through the HealthChoices program all pregnant women eligible for Medicaid will get that care through one of four managed care organizations: Health Partners, Keystone Mercy, Oxford Health Plan and Health Management Alternatives. Each of these managed care organizations had its own payment policies, its own forms submissions. The information on this universal encounter form will all be input into a database. If a woman on Medicaid changes providers, the information about her prenatal care will continue to be fed to the database even though each provider may not know about care provided by the other providers. The Commonwealth of Pennsylvania Department of Health has agreed to give us access to the birth certificate database. We will know from all women on Medicaid in the Delaware Valley their low birth weight rate, various indices of prematurity, all the information that’s recorded on the birth certificate. That will be merged with the information that’s been collected throughout the pregnancy. In one database we will have process and outcomes in prenatal care.

PND: How would you describe the Healthier Babies project from the standpoint of a collaboration between public health and private medicine?

RB: Both managed care and public health are concerned with the health of populations. We define managed care as improving the health of a defined population within a fixed budget. The major obstacle that faces an individual practitioner, especially in the Medicaid population where there’s so much switching from provider to provider, is how to get a handle on population health. You need a different kind of infrastructure than we have had in traditional individual practice. This project represents an important step in that direction, targeted to a vulnerable population with an identifiable group of providers. That allows us to put in place something we think will be a very important piece of infrastructure to measure and improve population health of pregnant women.

PND: How does that new infrastructure differ from the traditional one?

RB: I think that we providers have thought of ourselves as being responsible for giving somebody who comes to our office the best care that we know how. If they don’t come to our office we don’t think of ourselves as being particularly responsible for them. Some of the reasons that they didn’t come to the office might have been things that we could have fixed, like geography or hours or waiting time. The shift to population health requires that you think about all the folks that you could have interacted with, or in some cases folks you did interact with and then were lost to follow-up in your office.

PND: How will this approach make it less likely that women will fall out of the data tracking?

RB: It’s still going to be a problem in terms of continuity and tracking. In the initial phase of this project, we do not propose to provide real time clinical support for practitioners and patients. It is not the case that, when a pregnant woman drops out of one provider and shows up somewhere else, then that provider is going to be able to call us and find out what’s the story on this woman clinically. There’s a bunch of reasons why that decision has been made. One major is confidentiality issues. Look at what the known risk factors are for low birth babies: things like drug abuse, particularly cocaine abuse, young age, STDs, prior pregnancy outcomes. There are legal issues about the way HIV and information related to drug abuse for mental health gets handled. We need to be in compliance with those issues. We expect there to be very sensitive information on the database and we don’t want to make it available until we understand more about how it’s going to be handled. The specific details of how we will handle information in compliance with those regulatory requirements have not actually been totally worked out. The expectation that by doing aggregate reporting and not by doing reporting on individuals, we should be able to be within the bounds of the administrators’ restrictions.

PND: What data constraints does an aggregate reporting system present?

RB: It’s not a clinical record that supports clinical practice for an individual patient. One of the things people can see when looking at this project is that it’s not very good to me if it doesn’t help me take care of Mrs. Johnson on Thursday afternoon when she comes to my office with no previous history in that office. What good are you if you can’t give me that information? But [we have] confidentiality requirements. We think that the population health piece is very worthwhile. For example, as near as we can tell, no one knows the low birth weight rate of the Medicaid population. We have the low birth weight rate for the county of Philadelphia and each surrounding county, by race, by census tracking— a finer look at a zip code. But we don’t know it by insurance status, and the important thing about insurance status is that Medicaid eligibility is obviously a socioeconomic index. I think a major constraint has to do with how much physicians and health care providers see it as our job to provide this kind of information. If health care providers look at the universal encounter form and say, "This is one more insurance form that the managed care organization is making me fill out for them. They’re making me do their work," then we’re not going to be able to move the project forward.

PND: The form is voluntary?

RB: No, the form is actually going to be mandated by the different managed care organizations. But realistically speaking, it’s difficult to mandate without a significant piece of voluntary collaboration and expectation on the part of the provider community.

PND: Will there be sanctions for noncompliance?

RB: I think the commitment of the managed care organizations is to be willing to tie use of the universal encounter form to compensation. We’re going to try to see the form as an intrinsic part of the billing record. We would like docs to think of this as part of their responsibility for practice in a world that has population health as an importance index of excellence. And I think this is a real change in the way that we as doctors think about what we do. It’s a paradigm shift. All of us have to start thinking not just about the individual who is in front of us but the whole system of care and how that does or doesn’t meet the individuals needs.

PND: What additional knowledge can be gained from this sort of public health tracking that wouldn’t be available otherwise ?

RB: By having process and outcome data in one database we should be able to get much better risk adjusted outcome data than has been available before. Providers should be able to know how they are doing in terms of providing care to this vulnerable population. Managed care organizations need to run outreach programs. None of us really knows what’s working and not working with great precision because it’s difficult for us to know whether our programs are making a difference or whether our populations are different from some other populations. One of the things we’re working on most intensively is trying to include data on what we call social risk. This has been a very difficult concept to operationalize. A single mom with three kids probably has a higher degree of social risk than a married woman with no kids. She may have a higher risk if she lives alone with no other adults in the house than if she lives with her mother or her sister who can help provide care for the kids. You can argue the other way—if there’s lots of adults living at home, it’s a fragmented family and drains resources. We’re struggling to get a credible social risk measurement. We are sitting down with clinicians and public health people doing literature reviews, talking with CDC, with people who have some interest and expertise in risk adjustment for prenatal care and trying to decide what are the reports we want to produce in the end.

PND: How well do you think four competing HMOs will collaborate?

RB: The managed care organizations don’t have a track record of the four of us sitting down and collaborating on a significant initiative like this. I think that it’s going to be overcome because we all have a pretty high stake in the care of this population. We’re all probably spending 40 percent of our entire health services budgets on maternal child health. The Medicaid population is disproportionally young and female. We have some HEDIS reporting obligations to the state. All of us can struggle to try to get it on our own, but it’s basically a common problem. All of us need the providers to do the same thing, and if each of us tries a different strategy to make that happen, then providers are going to be confronted with four different systems, which is not what they want. The Philadelphia Department of Health is engaged in the delivery of prenatal care directly to the Medicaid population at city health centers. There are a series of advocacy groups involved as well.

PND: What is the role of the advocacy groups?

RB: They are interested in trying to run programs to improve outcomes. They need to know whether their outreach and intervention programs and their care delivery systems are working, and so they have an interest in participating in the database. They know a lot about maternal child health. We’ve appreciated their input and feedback on population aspects of maternal child health.

PND: Some advocacy groups have been very critical of the HealthChoices program. How did such a seemingly unlikely collaboration come about?

RB: The Robert Wood Johnson Foundation was interested in funding collaborative ventures that sought to use Medicaid managed care in ways that had direct positive impact on the health status of the population. Health Partners applied for a grant to do this project and we needed to get support from the competing organizations and from the Department of Health. Everybody sees enough potential opportunities for getting something they need and value that, at least so far, the collaboration is moving forward. It’s a tenuous collaboration. RWJ will fund the research and development costs of the actual construction of the database, but only when they’ve been convinced that the collaboration is firm enough. We are going to produce a business plan, develop more precise models of what the reports and the data collection tools will look like. We expect to submit it to the Foundation in the fall. We would like to be collecting data sometime in 1998.