Q&A with Rochelle Walensky Cost-Effective HIV Care

Rochelle Walensky photo by © IAS/Deborah W. Campos
Rochelle Walensky photo by © IAS/Deborah W. Campos

After graduating from Johns Hopkins School of Medicine in 1995, Dr. Rochelle Walensky worked in an inner-city hospital in Baltimore. “There were floors of patients with AIDS for whom we could do very little,” said Walensky. “Many of them died.” Highly Active Antiretroviral Therapy (HAART), also known as the triple-drug cocktail, was introduced the next year. Suddenly, AIDS patients were getting better. Fewer were hospitalized. “In my formative training years there was hope,” said Walensky. “I thought, I’ve got to be around to see how this all plays out.”

Today, Walensky is Co-Director of the Medical Practice Evaluation Center at Massachusetts General Hospital. She recently talked to Spotlight Editor Martha Henry about strategies for cost-effective HIV care.

Your background is as an infectious disease doctor. How did you get involved with cost-effectiveness research?

I have a strong math background from college. When I was pursuing my MPH (HSPH ’01), I took Milt Weinstein’s course in Decision Science. It was mathematically oriented and I loved it. Infectious diseases are among the diseases frequently modeled, so it was a perfect fit.

What is cost-effectiveness?

Definition of Qaly and DalyBy definition, cost-effectiveness is a ratio. It is costs in the numerator divided by effectiveness in the denominator. Generally, we talk about effectiveness in terms of life expectancy, so the units of the ratio are dollars per year of life saved.

QALY is based on the number of years of life that would be added by an intervention. Each year in perfect health is assigned the value of 1.0, down to a value of 0.0 for being dead. If the years would not be lived in full health, for example if the patient has chronic dementia or paralysis, then the years are given a reduced value between 0 and 1 to account for the diminished quality of life.

A study in the early 1990s suggested that the incrementalcost-effectiveness ratio of hemodialysis, compared to no dialysis, was about $50,000 per QALY (quality-adjusted life year). This ratio quickly became a benchmark. Any intervention with a cost-effectiveness ratio less than $50,000/QALY should also be considered an economically attractive one.The history of this question dates back to hemodialysis for those in chronic renal failure. We, as a society, are willing to pay for long-term dialysis in those with chronic kidney disease. It follows, then, that we should be willing to pay for other medical interventions that are “cheaper” (on a dollars per year of life saved basis) than dialysis.
Over the years, because of inflation and new technologies, healthcare became more expensive. People started asking why everything was going up except our willingness to pay for health. Over time, our “willingness to pay” ratio in the United States has generally increased. Now that ratio is felt to be around $100,000 to $150,000 per QALY.

How does cost-effectiveness translate to value in international settings, especially resource-poor countries in Africa?

Policy makers needed to decide what a reasonable value was in international settings. The World Health Organization (WHO) convened a committee that provided guidance based on the relative wealth, or per capita Gross Domestic Product (GDP), of a country.

If the incremental cost-effectiveness ratio of an intervention in $/DALY (disability-adjusted life year) is less than the per capita GDP of a country, it may be considered very cost-effective. If it is less than three times the per capita GDP of a country, it may be considered cost-effective.

In the case of the U.S., our per capita GDP is about $48,000. Under WHO guidance, then, an intervention with an incremental cost-effectiveness ratio less than $48,000 per QALY might be considered very cost-effective; less than $140,000 per QALY might be considered cost-effective.

In international settings, this brings up large ethical issues. The per capita GDP in South Africa is about $8100, which sounds low when compared to the U.S., until you look at a country like Mozambique, where the per capita GDP is about $530.

And should one really be using an individual country’s own GDP as a benchmark for “willingness to pay” when there are outside partners helping to finance care? When the World Bank finances an intervention in South Africa and Mozambique, should they finance more in South Africa than Mozambique because South Africa’s GDP is higher? There are a lot of ethics around this, but those are the benchmarks often discussed.

Cost-Effectiveness Treshold

We – Ken Freedberg and I and the entire Cost-Effectiveness of Preventing AIDS Complications (CEPAC) team -have partnered with the HPTN052 team. We presented data at the International AIDS Society in 2012 examining the cost-effectiveness of treatment as prevention. We looked at two countries, South Africa and India. We believe the cost structure is different in those two countries and certainly the per capita GDPs and benchmarks are different. In our analyses, we showed that treatment as prevention is very cost-effective in both South Africa and India over lifetime horizons.The HPTN052 study in 2011 showed that treatment of HIV with antiretroviral therapy (ART) also prevents the spread of HIV. How does that change the cost-effectiveness equation of large-scale ART programs?

It’s important to highlight that the question of the cost-effectiveness of treatment as prevention and the question of the cost-effectiveness of test and treat are different. Treatment as prevention implies that you have somebody in front of you with HIV and a high CD4 count [indication of strength of immune system] and the question is, do you provide ART to that person now, or do you wait until s/he might otherwise meet WHO treatment criteria? In a test and treat analysis, you have to test many people to find that same person. There are a lot of resources and costs that must be considered when an additional prerequisite is to identify the pool of people who might merit early treatment.

Does providing ART for people with HIV save money?

Policy makers do not necessarily want to hear that an intervention is cost-effective; they want to hear that it is going to save money. Almost none of our analyses demonstrate that we save money. Why? Our work is about keeping people alive, and the longer people live, the more they cost. Sadly, dying is cheap and staying alive costs more.

In treatment as prevention, we will be treating people who have HIV. The wonderful news is that this treatment will allow them to live longer. When they live longer, they will need ART for longer because there is no cure for HIV. They will need ART for the duration of their lives. There are many millions of people for whom that is true.

We will also, fortunately, be averting infections down the line. For those people—yes—we will save money. Will the money saved offset the current cost of the ART that it will require to save that money? I suspect not.

XIX AIDS Conference 2012

You and your colleagues make a lot of predictions. How do you know if they’re true?

So as I sit in my office and create these complex models with my colleagues, how do I know that they really work? We at CEPAC spend a reasonable amount of time validating our models. We must do this to ensure that what we are saying truly has reasonable predictive power with the current knowledge and science available. We spend a lot of time calibrating and validating the model, both retrospectively to see where we might have missed something, and prospectively as we look at our current analyses. We will often do analyses that project clinical trials forward into the future, beyond the planned duration. And then when the trial results become available, we go back and see how close we were and where we might have done better.

With limited resources, how do we decide where to spend them? Does cost-effectiveness give us clear answers?

Perverse things will happen if we look at cost-effectiveness alone. It’s one of the tools, and I think a much-ignored tool, to help guide us in making decisions.

What it tells you is whether an intervention is worth paying for. Is this good value for money compared to other things possible or compared to things you’re already paying for? What it doesn’t tell you is if the intervention is affordable. Something may be good value for the money, but you still may not have the money to pay for it.

My work is focused in the U.S. and resource-limited settings to help people decide what to do when resources are really limited. There is a huge menu of things that we could do. All is not possible. My goal is to maximize human life with those limited resources that we have. This is one way of doing it.