A conversation with Edward Machtinger, M.D., about the inaccuracies of the death certificates of women with HIV and the importance of trauma-informed care

Women with HIV are more likely to die from trauma complications than from HIV, according to a study of death certificates of patients receiving care at the Women’s HIV Program (WHP) at the University of California, San Francisco (UCSF). Results were recently published in the Journal of Acquired Immune Deficiency Syndromes.

The study cross-referenced the death certificates of 40 women with HIV and their long-term care teams. Results showed that HIV was listed as the primary cause of death in 68% of cases, when in reality, active infection played a role in just 15% of cases. Instead, the providers indicated that mental illness or substance abuse was the main cause of death in 58% of cases.

Senior author Edward Machtinger, M.D., is a professor of medicine at the University of California, San Francisco, director of the Center to Advance Trauma-Informed Healthcare and the co-director of the Women’s HIV Program. He recently sat down with Managed Healthcare Executive to discuss the results of this study and shared why they are important for women living with HIV today.

This conversation has been edited for quality and length.

MHE: Could you provide an overview of the study, including its design and key findings?

Machtinger: For many years, we observed a profound disconnect between what we were hearing about the HIV epidemic in the community, reading about in academic journals and seeing in frontline clinics. We were hearing that people were doing much better with HIV. That is, of course, after the antiretrovirals, HIV went from a chronic killer to a chronic, manageable condition. But there was a belief in the community that the life expectancy of people with HIV had achieved near normal values.

The first disconnect we saw was that those patients in our clinic, which is a typical Ryan White-funded HIV clinic, did not have normal life expectancies. They were dying very frequently, much more frequently than their age-matched controls. The second thing we noticed was that death certificate data described the causes of deaths of patients with HIV as being primarily related to HIV. What we were seeing in clinic was profoundly different than that. Most of the patients died with an undetectable viral load, and most were dying from preventable trauma-related conditions.

So, we set out to design a study that compared a novel and very rigorous method of determining causes of death and comparing that to what was written on death certificates.

Our method was to convene a structured group of all the clinical providers who had known patients well, often well over a decade or two. Those included their primary care physician, nurse practitioner, pharmacist, case manager, social worker and often a therapist. We called that "The Clinic Team," and we had a structured meeting where we all determined through consensus and conversation what the multiple intersecting causes of death were for that patient. Then we compared those with what was written on death certificates, and those results could not have been more different. The clinic team determined that the two most common causes of death in patients with women living with HIV were mental illness and substance abuse, each occurring in 58% of the cases.

Yet, when we compared those results to death certificates, death certificates had substance use written as a cause of death only 13% of the time and mental illness only 5% of the time, and that meant that death certificates missed mental illness 78% of the time and substance use 91% of the time.

We also found several causes of death that were common amongst our patients that weren't mentioned on a single death certificate — 30% of our patients had treatment non-adherence who are key cause of their death, yet none of the death certificates mentioned it. HIV stigma, which is just profound shame from the discrimination and perceived discrimination that you have for being an HIV-positive person, contributed to 13% of the deaths but wasn’t on a single death certificate, and similarly, intimate partner violence was attributed to 13% of the deaths, yet not a single death certificate mentioned that cause.

MHE: Why do you think that these causes were largely missing from the official death records?

Machtinger: Death certificates are very well known to be flawed. This has been known for decades. Death certificates are increasingly filled out by physicians, mostly in hospital settings, who don't know the patient well over time.

For instance, many of our patients die from lung disease or heart disease or something called pulmonary hypertension, which is a stiffening of the pulmonary artery. They will die from that, and the cause of death will be pulmonary hypertension. We happen to know that most of our patients have pulmonary hypertension because of either crack or crystal meth use, and so that pulmonary hypertension didn't happen in isolation.

Yet, that provider typically doesn't have access to the primary care team, and doesn't have time to call the primary care team, and the primary care teams don't have time to engage with them, and what ends up happening is a death certificate that gets filled out inadequately.

Conversely, a lot of advocacy happened early in the epidemic to include HIV on death certificates. Early in the epidemic, they wrote that someone died from pneumonia because they wouldn't want to stigmatize the family of a dead relative. So, we were very successful getting death certificates to include HIV.

But now clinicians, anytime someone dies with HIV, typically just put HIV as another contributing cause of death on the death certificate because they figure maybe it contributed.

We know that in the vast majority of cases it doesn't. So, 68% of the deaths on death certificates versus 15% in our review, most of whom had undetectable viral loads, mostly represents a misunderstanding of HIV and how well it's controlled now and how little, if any, it contributes to deaths currently of people living with HIV.

MHE: What would allow the primary care team and hospital physicians to communicate more effectively?

Machtinger: I'm a primary care clinician. I've also worked in the hospital for almost two decades at UCSF. Given the pressures that you're under to fill out a death certificate quickly, given the pressures that both teams, the hospital team and the primary care teams, are under — it'd be magical thinking to think that there'd be a simple prescription to better communication between those two teams. Frankly, they need to max out that communication when that patient is alive so that they can help save that person's life. I would focus my efforts there and acknowledge that the death certificates are flawed and that we need a system that is not too expensive to create for health conditions related to stigma and trauma that may provide you with more accurate guidance about what is causing deaths and what interventions are necessary to prevent deaths.

We have recommended in the paper that death certificate data be supplemented with focused studies like the one we did in different health conditions in different regions of the country that can more accurately determine the causes of death for health conditions that are often related to stigmatized conditions.

So, based upon our methodology, you could do the same thing with HIV-positive men, you could do something with transgender women, you could do something in the South, and then just from a series of studies, maybe four or five of them, you'd have a far more accurate picture of the causes of death of women and men living with HIV, and you could supplement that or combine that with the death certificate data.

MHE: Why are death certificates so important?

Machtinger: Knowing the reason for a death is critical. If all our work is guided by an understanding of what that patient is most at risk of getting hurt from, or dying from, and we have a misunderstanding of the risk faced by our patients, we will misallocate resources. We will focus principally on antiretrovirals in our case, when a focus really would be far more appropriate to be invested in mental health and substance use and stigma and treatment adherence. Interventions can't be adequately designed. Public health campaigns can’t be designed without an accurate understanding of death certificates. Advocates in the community, people living with HIV or these various health conditions, don't know how to advocate for themselves or their communities unless the threats to themselves in those communities and these patients are clear.

As clinicians, physicians, and leaders, all of our work is designed to help people survive and help people stay healthy and live, and it is really helpful for me as a clinic leader to know the primary threats to my patients' lives so that I can allocate resources to the things that will intervene.

MHE: Would you say that women with HIV face unique challenges when getting care? How would that stigma differ from being a gay man, for example?

Machtinger: HIV stigma in the women is, in my experience, much more profound.

In the gay male community, there's more community around HIV. It has just been much more normalized. Not completely normalized, but there is more community to be found.

When women get HIV, they feel like, “How could this have happened? This is a gay disease. How could I, as a Black woman, or Latina woman, or a White woman, have this happen to me?”

Women also often feel very isolated. Some of my patients have not told anybody they're HIV positive, and so when you have this stigma and you can't be out about your diagnosis, it’s hard to take medicines in front of people; it's hard to come to doctors because they're terrified of people finding out they're HIV positive.

Some of the biggest growth in people's health has come from overcoming stigma. My patient once said, "When I first got HIV, I felt nasty.” I think when she realized it was a treatable health condition, her health improved dramatically. She was hopeless before.

MHE: What is trauma-informed healthcare?

Machtinger: Trauma-informed care basically is a model of care that understands that much of the behavior and much of the medical and psychological conditions are rooted in experiences of trauma and sets a clinic up to be trauma-responsive and healing and less traumatizing.

Interestingly, many clinics themselves currently are traumatized. You have clinics that are understaffed or overstressed with providers who haven't had really adequate supervision and are kind of overwhelmed themselves in their lives and in their work lives. So that clinic becomes a traumatized clinic. It is reactive. It is hierarchical. It is fragmented in how it works, and when patients go into that type of clinic, it's traumatizing to them, and it's traumatizing to the people who work there.

What we're trying to do in trauma-informed healthcare is help clinics themselves be calmer and more collaborative.

The two principal attributes towards trauma in trauma-informed care are safety and connection. Connection, having a patient walk into a clinic and feel safe emotionally there, and then having the focus of the care be on their safety, their immediate safety, and what is contributing to risk in their lives, and then establishing a trusting, transparent relationship with them where they can connect with you and then reveal what's going on in their lives to you, which is often the things that are much more threatening to their lives than what you came into that appointment looking to address.

MHE: How do these results benefit women who are alive now?

Machtinger: Understanding what is hurting and killing women with HIV is critical for everybody who cares about women with HIV. It helps patients and advocates who themselves have HIV or are advocating for people with HIV know what to advocate for.

It helps clinicians and clinician leaders and hospital executives understand what the primary threats to their patients' health are and design care systems and resource care systems effectively to interrupt those leading causes of death. Without that, I think with the current death certificate data, most leaders would focus almost singularly on antiretrovirals to make the virus undetectable, and while that's critical and changed the epidemic, it does little to address the current causes of death of people living with HIV. Those leaders could then resource programs that are team-based, that are focused on addiction, that are focused on mental illness in environments of care that are more aware of trauma and the consequences of trauma, and the outcomes in those clinics we've seen are profoundly different when those conditions are understood and addressed appropriately.