On Dec. 9, Salil Deshpande, M.D., chief medical officer of United Healthcare Community Plan of Texas, a division of United Healthcare of Texas, was interviewed by HCI Editor-in-Chief Mark Hagland, in the concluding session of the Health IT Summit in Houston, sponsored by the Institute for Health Technology Transformation (iHT2—a sister organization to Healthcare Informatics under the Vendome Group LLC corporate umbrella).
On Wednesday’s final session, held at the Omni Houston Westside, Dr. Deshpande, an internal medicine physician by training, who has practiced clinically, has worked in hospital administration, and has been a senior executive at United Healthcare of Texas for eight-and-a-half years, focuses in his medical administration work on United Healthcare’s Medicaid and dual-eligible populations in Texas.
Below are excerpts from Hagland’s interview with Dr. Deshpande, as conducted in front of the iHT2 audience.
Just now, in introducing yourself, you described a current focus at United Healthcare of Texas on dual-eligible plan members. That is particularly interesting to me, given that dual-eligibles are among the health plan members with the most complex situations, including poverty, sometimes homelessness, and other complications that aren’t purely medical. I continue to say that what we learn about the care management of dual-eligibles will point the way to successful care management across healthcare populations.
Yes, absolutely. And to step back for a moment, that emphasis on dual-eligibles, it is something that is a priority for the health plans. I think the focus and the eye on healthcare will continue for the country as a whole. It is obviously an issue that the payer industry is also very focused on. There’s been increased access to healthcare for the past several years. We expect that additional people will continue to move into the insured category. There are still continued pressures around healthcare costs in general, of course, but of those reasons will underlie why we are continuing to focus on healthcare as a society.
In terms of the cost pressure, that issue is particularly relevant among the dual-eligible population. They are among the sicker people, and are either elderly or have qualified for SSI by virtue of some other disability, and also are likely to be living below the poverty line. And they likely have a history of lack of access to care, that has caused their chronic conditions to be under-diagnosed. And for all those reasons, that patient population tends to be an expensive one, but with opportunities for healthcare improvement.
That is the reality for any population, but more dramatically in the elderly population and among relatively poor populations; in both cases, a relatively small percentage of members are driving the expenses. So to the extent that we can drive care management, we will both improve outcomes and lower costs, and thus, the focus on dual-eligible members makes eminent sense.
Have you and your colleagues been plunging into health risk assessment for that population?
Yes, the social determinants of care really do drive a patient’s clinical condition and overall well-being. So certainly, that patient with diabetes or congestive heart failure, or a congenital anomaly, will have conditions that will drive care management. But also, homelessness is an issue. Historically, healthcare has not focused on non-clinical issues, and yet those people who lack shelter, who do not have a steady job, or the income to buy groceries regularly—they are not paying attention to their prescriptions, because frankly they have other priorities; and so they are ripe for care management.
So those are significant opportunities. And where this dovetails with the data analytics issue that I know we wanted to discuss, is related to Optum, the United HealthGroup division (United Healthcare is the other) through which we provide population health management to providers. Optum’s services are broad, of course, and extend all the way through to fraud, waste and abuse detection solutions.
What are you doing here in Texas, around data analytics for population health?
I’ll start out with something that was originally deployed for our commercially insured population, but which we’re rolling out everywhere. Our AdvocateForMe program is essentially, a customer service program. For instance, let’s say you’ve lost your insurance ID card and need a replacement. Here’s the thing: if you call in, what will show on the customer service representative’s screen will be all sorts of clinical data, which will be generated by your claims history, or may emerge out of pharmacy or laboratory data, or out of information we’ve gathered around any clinical care quality gaps.
And if that customer service rep is actually a nurse who can make sense of a lot of that clinical data, when he or she answers the phone to get you a replacement ID card, maybe you’re a diabetic and haven’t had your hemoglobin a1c measured in six months, for example. So we might be able to make a connection for you and help you to access an endocrinologist, or connect you to one.
When you partner with providers around accountable care or population health development, are you increasingly able to provide them with analytics, in the context of risk-based contracting?
We don’t necessarily expect that all providers are ready to bear risk, and so whoever providers are on that continuum, we have a variety of things we can provide them. Maybe they’re in a two-person practice in a rural part of the state, and they may not be able to bear risk, but we can incent them to close clinical quality gaps in care. So maybe in addition to their fee-for-service payment, we’re making an incentive payment if they address all the diabetes-related gaps in care for their patient panel. Now in terms of the analytics to support them, we obviously do have a lot of data, and there’s an opportunity to bring that to the table, to let them see how their patients are faring based on our understanding of their claims, etc.
How is that data/analytics delivered?
There are a variety of ways. And frankly, that’s a perennial issue for us as a payer. And it’s different for every physician practice, depending on whether the doctors are practicing in a small rural practice, or a large multispecialty practice. We send things by mail, by fax, etc., but we’ll also have face-to-face meetings with medical directors. We’ll make these things available via provider portals. Our latest version of a population registry incorporates a variety of data elements, including medical claims, pharmacy claims, lab data, we’re also beginning to incorporate information from other providers, whether ADT feeds from other hospitals in our network, or CCD data from other physician practices in our network. Also, states have their own immunization registries, so we might be able to bring in that data, and in states where there are well-established registries, to leverage some of the data from those registries.
Can you provide a practical example in this context?
In the Medicaid space, in the past year, we’ve deployed our accountable care community program. We’re looking for large-volume practices, per United Healthcare Medicaid. With these practices, we see the opportunity to improve care coordination, especially for the patient in their panel who has recently been in the ED or admitted as an inpatient; we’re looking for that practice to follow up within seven to 10 days of an encounter. Or to fill in gaps in encounters. We’d like to make sure that members’ hemoglobin a1c is measured and under control, for example. So we share data back and forth. We can provide it in hard-copy form, but also make it accessible via population registries and portals, to incorporate it into their daily workflows in the physician practices.
How close to real-time have you been able to get, so far?
With that first objective in time, trying to get the patient into see their primary care physician within seven days of an inpatient admission or ER visit, our policy is that hospitals are required to notify the payers at the time of an elective or urgent admission. With an inpatient admission, it’s easy to put that into a portal for a PCP. With regard to emergency department visits, we don’t have that information, so we’d have to rely on a claim. The reality is that most hospitals are dropping their claims pretty quickly, but that’s where ADT feeds or alerts are valuable to us. And to the extent we can’t get those feeds for political reasons, rankly, for the most part, that is a clinical opportunity.
In Maine, the statewide HIE has advanced to the point at which they’re able to provide real-time alerts to hospitals and primary care physicians of ED visits.
Yes, we’d like to be able to do that.
In the next couple of years, what plans do you have to use data and share it with providers?
To touch on that example of the HIE, a lot of our focus is on building our connections with those HIEs. In the Houston area as an example, we’ve been driving the information exchanges in certain directions. For example here in Houston, it’s primarily been around beefing up their security protocols—that’s a bit beyond the scope of my work personally. But as a result of some of the data breaches, particularly the large national payers, there’s a substantial emphasis on the security protocols being in place before we really continue to participate going forward with some of these regional and statewide HIEs. So here in Houston, we’ve played a big role in pushing the local HIE to adopt additional certification processes, and that came with an expense. But we believe that that avenue is a means of obtaining data that we can share with the rest of our provider network.
Another thing around external connections to data is that we’re also working with some of the larger-by-volume EHR [electronic health record] vendors around the country, initially to obtain certain data elements, but in the future, to be able to exchange data elements with the provider community across the country. We strongly believe in that connectivity. Otherwise, much of what we’ll be working on here in Texas will be broadening the scope of those value-based contracts, with a significant emphasis on really having a partnership where we’re sharing data, and making it available to hospitals and doctors going forward.
