Unfragmenting the System: A Q&A with Dr. Amy Young
By Daniel Oppenheimer
Editor, Texas Health Journal
Dr. Amy Young’s ambition for low-income and uninsured pregnant women in Travis County is simple. Their care should be equal to what is typical for those who have private insurance. The mechanism for achieving this goal is “perinatal redesign,” which is a rather abstract term for a very concrete, and often fairly complex, set of reforms.
Young, Chair of Women’s Health at UT Austin Dell Medical School, is leading the effort along with colleagues at Central Health, CommUnityCare Clinics, the Seton Healthcare Family, the Community Care Collaborative and St. David’s HealthCare.
In an op-ed published in the Austin American-Statesman published not long after her arrival in Austin, Young articulated specific objectives for the redesign, including:
- Improving access to the right providers at the right time
- Embedding academic specialists and subspecialists in the community
- Consolidating specialized care in centers of excellence
- Establishing standards of care that all providers must meet
- Providing specialized ultrasound services to all women, improving detection of abnormalities in routine screening.
- Creating opportunity for warm hand-offs between providers
- Reducing the delay in treatment that comes from multiple referrals
- Making care more cost-effective
Texas Health Journal spoke to Young about the reasoning behind the redesign, its components, and how its success (or failure) is being measured.
The phrase “perinatal redesign” is very abstract. I’d like to put some flesh on it. What does that mean? What’s the problem that requires a redesign in the first place? And what are the solutions?
Amy Young: The problem, and this isn’t at all unique to Austin or Travis County, is that care for patients is fragmented, especially if you are talking about under- or uninsured populations. This fragmentation of care can be exacerbated in the case of pregnancy, which is a complex condition that changes rapidly over time. It requires inpatient and outpatient visits, and after delivery also involves a second patient, the baby, who has his or her own set of records and health concerns. All the different care may not be linked, either through electronic medical records or within the same provider group. What you often end up with, then, is an over-arching care pattern that is interrupted, which is then superimposed over an already fragmented or dis-coordinated system.
Before you describe some of the solutions to this fragmentation and interruption, can you elaborate on what problems this can produce for the actual health of women and babies?
This complexity, and lack of coordination, can lead to more tests and visits than are necessary, more stress for the patient (and therefore the fetus), and greater likelihood that a problem can be missed or treated inconsistently. All of that can produce worse outcomes for the mother and baby. It can also lead to unnecessary costs for the system.
So how do you fix that?
The simple answer is care standardization, care integration, and care decentralization (which doesn’t seem like it goes with integration, but it does).
Let’s take those in turn. What do you mean by care standardization?
Providers come to their patients with different training, in different contexts, and with different degrees of experience in seeing various conditions. Those differences can be valuable. We want providers to bring themselves to their care. But we don’t want variability in care when dealing with conditions where there is good evidence about the best ways to screen, diagnose, and treat. In those cases, patients should receive the standard best care every time. If we standardize the care, by implementing protocols, automating reminders, following checklists, and implementing other, similar kinds of measures, care will be better.
I’ll give you an example. We know that mental health can be very adversely affected during and after a pregnancy, and that signs of mental distress can very easily be missed. All patients, as part of standardization, are now screened for depression at least twice during pregnancy, and at least once postpartum. Postpartum depression is present in about 20-25 percent of our patients, so we want to make sure we are giving our patients every opportunity to get referred to appropriate treatment by standardizing that screening.
We’ve standardized care around the issue of postpartum contraception. Nationally, we know that less than 50 percent of underinsured pregnant women are counseled about postpartum contraception. In our system, all patients are now queried during the prenatal period about their postpartum contraception plans. We document that plan in the prenatal record, then follow up at the two-week postpartum visit. The goal is to reach 100 percent of patients.
Another example. One of the ways that we have enhanced care dramatically is by doing an ultrasound at the patient’s first obstetric visit. This is incredibly important, because so many of the things we do in caring for a normal or an abnormal pregnancy are related to the dating of the pregnancy. If the dating is wrong, we are making decisions based on bad information. The earlier you date the pregnancy, the more accurate the dating is going to be, and the less likely we are to make wrong decisions later.
This initial ultrasound also documents the viability of the pregnancy. A quarter of all pregnancies end in miscarriage. There are also a fair number of ectopic pregnancies. When a mom is not going to have a viable pregnancy, the early ultrasound allows us to pro-actively manage her care. If we had waited, and that ultrasound had not taken place, she may be coming back to the clinic or arrive at the ER with spotting or bleeding or needing emergency care. By understanding what is going on with that patient sooner, and making a diagnosis sooner, we may give her more time to grieve, and more time to have input on how she wants that pregnancy to be managed.
Tell me about care integration and decentralization.
It involves better connecting records between inpatient and outpatient visits, at various facilities. It is better communication with patients, particularly when they receive care at different facilities.
We have created care delivery guidelines that risk-stratify patients. When they come in for their initial obstetric visit, they are identified as needing low-risk care, intermediate-risk care, or high-risk care. The high-risk patients, and in some cases the intermediate-risk patients, can be seen at central facilities, where they have the specialists and resources to manage complex pregnancies. The low-risk patients—and this is how integration and decentralization are really flip sides of the same things— can be seen at clinics closer to their home. That is better for them, and more cost effective for the system. It also frees up resources at the more centralized locations, where intermediate and high-risk care can be delivered to the patients who need it most.
We’re more often allowing providers to practice to the top of their training. So, for instance, general OB-GYNs are already trained to take care of women who have intermediate risk conditions, like gestational diabetes or well-controlled hypertension. But we weren’t necessarily allowing our general OB-GYNs to be able to practice to that level of training. In many cases, we had been requiring patients to come down for a high-risk clinic appointment, with a different provider, when their care could have just as well provided in a setting close to home, with a provider they knew and trusted. We’ve changed that.
Another example is the 20-week anatomic survey, which is the big scan that everyone gets excited about. It’s often when you find out about the gender. In the past, this has often been performed by general OB-GYNs, in their offices. We have now redirected those ultrasounds into a specialized antenatal testing unit. There is a fair amount of literature showing that someone who is used to doing that kind of work, all day, every day, reading and interpreting ultrasound images, is much more likely to catch problems. The risk of missing something, or the risk of identifying something that is not really there, goes down when we have that systematic approach to imaging. So this is an example of both care integration and care standardization.
These are just some examples. It’s not just one thing. It’s a series of measures to make the whole system more coherent, and to make sure that resources are distributed in the right way.
How will you know if all this is making a difference?
We want to look at a wide range of things, and at measures on which we can get good data. For example, we are looking at what patients say they want for post-partum conception, while they are pregnant, and what they ask for post-delivery. We are measuring the time to get patients in to an appointment, and whether have we are able to improve access by reducing the time. How many patients’ dates did we change based on that initial ultrasound? What percentage of patients who have had a previous preterm birth have been able to get weekly shots of 17-hydroxyprogesterone, which has been shown to decrease risk of another preterm birth by up to 30 percent, if begun early enough in the pregnancy? How many instances of postpartum depression can we identify and refer for treatment?
It will take a long time to get really good outcome measures, but even in the short term we can measure improvements in these processes and practices, and there is good basis in the evidence to believe that such improvements will lead to better outcomes for the mothers and babies. These kinds of system improvements are not necessarily the sexiest thing in medicine today, but they may be what makes the most impact in the long run.