Table of Contents
Michael J. Mugavero, M.D. is associate professor of medicine at the University of Alabama at Birmingham and project director of the University of Alabama at Birmingham 1917 Clinic Cohort.
Dr. Mugavero is one of a handful of HIV clinicians who closely studies discontinuities in HIV care from diagnosis through referral and retention, both in the 1917 Clinic Cohort at the University of Alabama at Birmingham and through analysis of other research. His recent publications include work on temporal trends in presentation for outpatient HIV medical care in the past decade, health care system and policy factors influencing engagement in HIV medical care, and underutilization of the AIDS Drug Assistance Program (ADAP).
Mascolini: To give readers perspective on your insights based on experience at your own clinic, can you outline the demographics of your patient population?
Mugavero: The University of Alabama (UAB) 1917 Clinic Cohort includes patients who are in care at the UAB 1917 Clinic. We currently have over 1800 patients receiving primary HIV medical care, in addition to other supportive and specialty care. Our clinic population is roughly 50% white and 50% African American, although 80% of our women are African American and 20% are white. In terms of transmission risk, slightly more than 40% of our patients are men who have sex with men, with the majority of the rest reporting heterosexual transmission. Injection drug use explains HIV transmission in less than 10% of our clinic population. Roughly 35% of new patients in our clinic have private health insurance. The remaining patients have either public health insurance or are uninsured, and insurance status is roughly split between those other two categories.
Mascolini: The CDC estimates that 20% of HIV-positive people in the United States remain unaware of their infection.1 What's your perspective on the scope of this problem from experience at your clinic and from your understanding of data across the US?
Mugavero: The CDC has wonderful surveillance and epidemiology teams that use sophisticated approaches to track and estimate the number of individuals who are HIV positive and unaware. In our clinic, since 2006 we've observed a dramatic increase in the number of new HIV patients coming in. It's hard to say whether we can attribute that increase to implementation of the CDC's revised HIV testing recommendations2 or to other factors. For example, the Alabama Department of Public Health put into place a new program to enhance referral for treatment services, and we developed a new patient orientation program in our clinic.
It's also hard to say whether the recent increase in new HIV patients in our clinic corresponds to the CDC revising its estimate of infected-but-unaware people from 25% to 20%.1 Regardless of the cause, this influx of new patients is encouraging. At the same time, we've found that our new patients are coming into care earlier in the course of infection in terms of CD4 count and presence of opportunistic infections. Through 2006, roughly half of our patients presenting for HIV care and never seen elsewhere had a presenting CD4 counts below 200, and about one third had an opportunistic infection before or at presentation to care. From 2006 through the end of 2010, we saw a steady and continuous decline in presentations with low CD4 counts or opportunistic diseases. In the most recent years, 31% of new patients are coming in with a CD4 count below 200, and well below 20% are coming in with an opportunistic infection.
Again, it's hard to say what factors are associated with these changes, but the trend over time is encouraging, and the bolus of new patients coming in earlier supports the idea that we're doing a better job identifying infected people and getting them into care more expeditiously.
Mascolini: Are clinicians you know at other HIV centers across the country telling you they're having the same experience?
Mugavero: I've heard mixed anecdotal feedback from different sites. But an NA-ACCORD cohort analysis of US and Canadian patients from 1997 to 2007 also showed a steady increase in median CD4 count among new patients entering care and a decrease in the proportion of patients presenting with a CD4 count below 350,3 which until recently was the prevailing recommendation of when to start therapy. These changes were less dramatic than our single-site findings, but NA-ACCORD appears to be the largest national study that suggests persons newly entering care are doing so at an earlier disease stage.
Mascolini: Your study of late diagnosis at Duke University in North Carolina found associations between late diagnosis and both older age and female gender.4 What explains those associations and what other factors contribute to late diagnosis in the United States?
Mugavero: First it's important to note that this was a fairly small, single-site study.4 We assessed about 100 patients in that study. The findings for older age and late diagnosis have been seen fairly consistently across different settings in recent years.5-8
Several factors may contribute to that association. First, perceived likelihood of becoming infected with HIV probably declines with age among individuals as well as their health care providers. There may well be a misconception that older individuals are less likely to be at risk. The other factor that may contribute is length of time since infection: Some people diagnosed with HIV infection at an older age probably became infected years earlier, so they had had more time to experience CD4 decline and more advanced infection at the time of testing positive.
The gender finding in our study is interesting.4 We found that women were more likely than men to be diagnosed with HIV during hospitalization (adjusted odds ratio 6.74, 95% confidence interval 2.08 to 21.81, P = 0.001). We didn't have enough details to know whether some of those hospitalizations might be related to pregnancy or what specifically caused those hospitalizations. We did not see that women had lower CD4 counts than men at presentation. Actually, a number of studies found that women are often diagnosed at a higher CD4 count than men.5,6,9 I think part of that success can be explained by the long-standing recommendation for routine HIV screening during pregnancy, which was in place long before the CDC's 2006 recommendation for general opt-out screening.2
Mascolini: Turning to linkage to care, a fair amount of research has focused on patient-related factors affecting late entry to care. You published an interesting analysis of health care system and policy factors that influence linkage to care.10 How do you weigh the relative importance of these two sets of factors?
Mugavero: Both sets of factors are critically important, so you can't necessarily weigh the importance of one against the other. In that paper we explored a socioecological model that looked beyond individual sociodemographic characteristics to relationship characteristics and factors related to the community, the healthcare system, and healthcare policy.10 Most HIV linkage studies have looked at individual-level factors -- often at very fixed sociodemographic variables such as age, race, sex, and insurance status -- as they relate to delayed entry to care. But it's also important to assess how system-level and policy-level variables influence linkage and retention because system-level changes, if successful, can have broad reach. Addressing system-level obstacles can complement individual-level interventions and expand their scope.
For example, re-emergence of ADAP waiting lists in recent years illustrates how a policy factor can influence testing uptake and linkage to care. When people at risk for HIV or recently diagnosed with HIV see news stories about perceived or real lack of availability of medications, those stories can be a disincentive to getting tested or getting into medical care. The impact of news about something like an ADAP waiting list on testing and care decisions is hard to measure, but I think variables like these can have a great impact.
As another example of system-level factors, a number of studies show that when testing and medical services are located at the same site, people are more likely to get linked into care and are more likely to get linked expeditiously.11,12 So, at a system level, having those services co-located can help.
Another system-level variable is the waiting time between calling to make a first appointment in an HIV clinic and the actual date of that appointment.13 HIV is a life-changing diagnosis for everyone. In the weeks after that diagnosis, individuals go through a range of emotions and psychological stresses. Linking that person to care gives them access to so many resources besides medical treatment -- resources that really help them take the next steps in entering care, educating themselves, and assuming responsibility for self-care and disease management. The sooner you can make that bridge from the community to the clinic, the better.
Mascolini: Your study of the "no-show phenomenon" implicated gender, race, lack of insurance, and a long waiting time to the first appointment date as determinants of poor linkage to care.13 Clinicians can't do anything about their patients' gender, race, or insurance status. What have you done to address the waiting-time factor?
Mugavero: Historically, at our clinic the waiting time was up to 4 weeks from someone calling to schedule a first appointment until a health care provider was available. That waiting time actually increased recently to between 4 and 6 weeks. When you think about how much happens in a person's life in 4 to 6 weeks, especially at such a vulnerable time after testing positive for HIV, it's easy to see how readily barriers might arise to taking the next step and getting into care.
After we did that waiting-time analysis,13 we brought together a multidisciplinary team of stakeholders in the clinic and developed a new patient orientation program. First, we agreed that 4 weeks is too long for a first-time patient to wait and probably partly explains why one third of potential patients were not getting into care. Because the clinic didn't have the capacity to shorten the time to the provider visit, we developed a new patient orientation visit. We call it Project CONNECT.14
Figure 1. Project CONNECT at the University of Alabama 1917 Clinic aims to bring people to the clinic within 3 to 5 days of HIV diagnosis, to orient them to HIV care and address immediate concerns, and to begin collecting data the medical provider can use at the first care visit.
Other sites have different types of orientation visits -- a lab-only visit or a brief nursing visit. Project CONNECT is a detailed patient-focused visit led by social work services and including a semistructured interview, a psychosocial questionnaire, and baseline laboratory testing (Figure 1). The interview aims to assess how new patients are adjusting to their diagnosis, while the questionnaire captures information about depression, anxiety, substance use, alcohol use, and other factors that may be impediments to staying in care, starting therapy, and staying on medications. The orientation visit also lets individuals know what resources are available to them at the clinic.
All of this information -- expansive contextual and behavioral information in addition to baseline lab work -- lets the medical provider engage the patient immediately at the first provider visit. As a result, the clinician begins care for a new patient with rich information about social support, stigma, disclosure, depression -- many potentially modifiable factors that may present challenges to staying in care.
The CONNECT visit occurs within 3 to 5 days of someone calling, sometimes even sooner. The idea is that the sooner we can get someone from the community to the clinic, the sooner we can make that personal connection, have someone welcome them, and have someone address their early questions and concerns. If we can make that strong connection up front, it will help get new patients to the medical provider visit, make the first provider visit more meaningful and productive, and provide a foundation for successful partnerships in managing HIV.
Mascolini: You and others have outlined an array of factors that can be addressed to improve linkage to care.14 Besides shortening the time to a new patient's first clinic visit, what other factors can clinicians address to get more newly diagnosed people into care?
Mugavero: Testing is often done in nonmedical settings or at least not in an HIV medical clinic. Many people are diagnosed after testing in outreach settings, emergency departments, or hospitals and are then reported to the health department and referred to an HIV clinic. One thing clinicians can do is to develop relationships -- memoranda of understanding -- for how to work with these community partners on linkage to care. The key factor is determining how to integrate service delivery and work together.
Once someone makes it to clinic, the provider should talk openly about retention in care within the paradigm of adherence. When I give talks about these issues and ask providers how often in their encounters they talk about medication adherence, it's over 90%. Medication adherence has become engrained in our culture of providing care as HIV clinicians. I similarly encourage clinicians to talk to patients about retention in care -- not just adherence to meds, but adherence to visits.
Even brief statements can be validating: "It's great to see you today. I'm glad you made it in." If someone misses a visit, the provider should acknowledge that: "I missed seeing you last time. What's going on? What was happening?" Just paying attention to the fact that you as a provider notice whether or not a patient comes to visits and that you care sends a strong message. It's a relatively simple thing to do, and anecdotally patients respond to it.
I was encouraged to see that one of the new recommendations in the 2009 HIVMA guidelines is expanding the notion of HIV adherence beyond medication adherence to include retention in care.15 That was really encouraging because, when you look at the grand scheme of things, among all persons with HIV in the US, only a small proportion is on antiretroviral therapy.16 A lot of infected people are not linked or not retained, so we've got to pay more attention to those aspects if we're going to have maximal reach.
Mascolini: A study at your clinic determined that missing visits in the first year of HIV care more than doubled the rate of long-term mortality.17 Other studies yielded similar findings.18-20 What are the variables that raise the risk of poor retention in your clinic, and what have you done to address those factors?
Mugavero: When you look at a lot of the literature, we're focusing on factors like age and gender and race, which might help identify risk groups, but these are not modifiable factors. As the field moves forward, it's critical that we start looking at modifiable factors that might influence linkage and retention in care.
Factors that we and others have noticed include common co-occurring disorders: Depression, anxiety, substance use, and alcohol use are associated with worse retention in our clinic. By asking new patients about some of these factors during our initial CONNECT interview, we can identify individuals up front who are at high risk for worse retention in care and worse medication adherence once therapy begins. By identifying those factors early, we can refer people to appropriate services and make primary providers aware of these factors at the first clinical visit. I think screening, identifying some of those factors, and referring to appropriate treatment can be critical.
This screening and assessment should not happen once, but on an ongoing basis. These are things that can be done systematically within a clinic.
Mascolini: When I was doing research for these articles on testing, linkage, and retention, I saw that many studies are done by people on the policy side or in epidemiology or other non-clinical fields. You're one of a handful of HIV clinicians who have studied these issues closely. What was your motivation for getting into this area of research?
Mugavero: From 2003 to 2006, when I was an infectious diseases fellow working on expanding HIV testing, I began to realize the challenges beyond HIV testing and started thinking about the obstacles to ensuring a continuum of care. Much of my research focused on testing, and then in the clinic I was caring for a lot of people with HIV, but I realized that many barriers and challenges lay between those two points. It became clear that the folks I was seeing in clinic were just a subset of those I was trying to reach with the testing efforts.
To me this continuum seemed a natural progression of the focus on medication adherence. If large numbers of individuals are unaware of their infection, or not linked to care or retained in care, the impact of successful treatment and adherence interventions on a population is greatly diminished.
In the last 3 to 5 years we've seen a dramatic expansion in emphasis on linkage and retention in care. A lot of that is driven by the test-and-treat paradigm or TLC-Plus: test, link to care, plus treatment.21 This approach -- expanded HIV testing plus quick linkage to care and treatment -- should have benefits both for the individual and at the population level because identifying infected people and treating them quickly will lower "community viral load" and limit HIV transmission. We saw this in HPTN 052, an international randomized trial that ended early when antiretroviral therapy begun at a higher CD4 count for the positive partner in a discordant couple lowered the risk that the negative partner would become infected by 96%.22 Those results should stimulate even more interest in the test-and-treat concept.
Figure 2. This simplified scheme of HIV care from diagnosis through antiretroviral therapy and retention in care indicates that care should be viewed as a seamless continuum, not a series of discrete steps.
I always stress that we need integrated approaches not just from a research perspective, but also from a practical perspective across this continuum of care (Figure 2). A newly diagnosed person doesn't think, "Now I'm going to go to my outreach intervention. Now I have to hop over to linkage to care. Now that I've made it to clinic, I'm going to focus on risk reduction, then on starting therapy, then on adherence."
Our approaches to these interventions often focus on one piece at a time, and we do need to focus on each element. But ultimately, if we really want to have impact, we need integrated approaches that map to the lived experience. Someone should go seamlessly from testing positive to getting into care, developing early behaviors around starting therapy, having good adherence to visits and to medicines -- all in a matter of months. Although test-and-treat and TLC-Plus focus on secondary prevention and population health, I think they have incredible potential to improve individual health outcomes.
Mascolini: Before we close, would you like to make any other points related to any of these issues?
Mugavero: I think the key message is that improved HIV care is going to require integration at multiple levels -- both scientifically and in practice. And over the past few years we have seen more and more integration, at the funding level and in local service delivery between medical providers, AIDS service organizations, community-based organizations, and health departments. The overriding idea is that we need approaches from testing through outcomes that match the individual's experience. People with HIV don't seek services piecemeal in discrete steps; there should be a seamless transition from one facet of care to the next.
With TLC-Plus21 and other studies, we need to determine the best way to integrate efforts and meet individuals' needs. Right now we're losing too many HIV-positive people at each step along that cascade who are not diagnosed, not linked, not retained. We must redouble our efforts to focus on helping people navigate this continuum of care as seamlessly as possible.