The sharing of data sources is a big topic in the news due to the revelation that Facebook shared the data of individual users with Cambridge Analytica, which then targeted political ads towards people for the benefit of Donald J. Trump's 2016 election campaign. Closer to home, the gay dating/hookup site Grindr had to apologize to all of its users when it was revealed that it had sold their data, including their HIV status, to other business entities.
But social media isn't the only place where the ease of data collection is being utilized. Increasingly, public health and other medical providers -- health insurers, pharmacists, etc. -- either have data systems that speak to each other or have access across different institutions. The revolution in data collection has given public health a new set of tools to address HIV and other infectious diseases, but questions about how these data are used and whether they have the potential to cause harm will likely lead to burgeoning debate.
The Adherence 2018 Conference, organized by the International Association of Providers in AIDS Care (IAPAC), convened in Miami from June 8 to 10 to address the latest research on improving treatment adherence outcomes for people living with HIV, with the goal of increasing viral suppression. One pre-conference panel featured providers and researchers discussing the role of data and surveillance through several approaches and how they are being used to understand local and global epidemics, marshal resources, and mobilize stakeholders for interventions.
However, with all of the relatively new methods of capturing, tracking, analyzing, and visualizing data, questions remain about how much data we need and whether efforts to "intervene" in epidemics can help people with HIV access care without fueling stigma, discrimination, or criminalization.
Thinking Globally but Responding Locally
The revolution in data collection has helped produce many of the surveillance tools we now take for granted. The HIV care continuum is possible only due to the ability to better measure HIV diagnoses, engagement in care, antiretroviral therapy use, and viral suppression. And, once the framework for the HIV care continuum was established in 2011, seeing the visual representation of the major drop-offs in people engaged in care after receiving an HIV diagnosis made many providers and advocates realize how much work needed to be done to address low levels of viral suppression globally. But that work couldn't happen at the "global" level: Local, state, and national efforts needed to be developed and implemented to deal with the specifics of local epidemics.
"Two hundred cities account for 25% of people living with HIV worldwide, and there are several cities in sub-Saharan Africa that account for almost 50% of their national [HIV] burden," said Sindhu Ravishankar, M.Phil, director of the Fast-Track Cities Initiative with IAPAC. "The idea behind the cities initiative is the optimization of the HIV care continuum from diagnosis through suppression of viral load, but through a human rights lens."
In 2013, UNAIDS developed the 90-90-90 campaign as a way to mobilize global AIDS resources and provide some targets for individual countries to meet -- to increase the number of people who know their status, the number of people who have started antiretroviral therapy, and the number of people who are virally suppressed. One year later, UNAIDS, in partnership with IAPAC, the city of Paris, and UN-Habitat launched the Fast-Track Cities Initiative to mobilize cities around the world to adopt plans to meet these targets. Now, over 250 cities have signed on to the initiative globally, and because of the Initiative's capacity to look at the data from each city and convene leadership, they're able to help describe what's working in each locale; analyze trends; and disseminate best practices to public health researchers, providers, and advocates alike.
However, as much as HIV care continuum data has provided a framework to view an epidemic, it provides answers to the question "what" but doesn't tell us much about the "who" or "why."
"Currently a lot of what the cities have been doing is working to get their baseline care continuum data together, and that's not enough," said Ravishankar. "We need to be looking at key populations and stratifying by age and gender to start."
She also discussed the need for cities to look at social, cultural, and structural barriers that impact access to care. Some cities (she named Bangkok specifically) have been using geospatial data to identify the locations of highly concentrated new infections and overlaying maps of city resources (testing sites, medical facilities, pre-exposure prophylaxis [PrEP] providers, etc.) to make decisions about what kinds of resources need to be allocated. Using the data submitted by its member municipalities, Fast Track Cities has also developed a global dashboard -- using the surveillance information from cities to create data visualization products as both an information and an advocacy tool.
Finding People Who Are Out of Care Using "Data to Care" and Tracking Pharmacy Refills
Creating tools for the global surveillance of progress or "knowing your epidemic" is only one way data is being used to address the care continuum. Some jurisdictions are using data to intervene directly in the care of individuals who have fallen out of care.
Here in the U.S., the Centers for Disease Control and Prevention (CDC) has been using two methods to find people who are out of care. First, it's empowering local health departments to use National HIV Surveillance System data to make sure that people with HIV are in care, with the hope of increasing viral suppression, a strategy called "Data to Care." The system is used to collect information on people as they are newly diagnosed, which is also a marker of when they entered care. People who are HIV positive have viral load and CD4 count data collected at the time they enter care and each time they make a visit and give blood to check their numbers. People who aren't virally suppressed or who have fallen out of care can be identified and ultimately, through their health care providers or health department staff, reached to be brought back into care.
"One of the headaches is figuring out what time do you decide to go looking for people," said Paul Weidle, Pharm.D., M.P.H., acting director of the CDC's Division of Viral Hepatitis. Weidle said that the CDC is working on a study looking at the best time to reach people who are out of care and analyzing the cost effectiveness of finding people who are out of care and linking them back to care. "[Research shows that] somewhere after nine months to 12 months seems to be the time in which people lose viral suppression after falling out of care," he said.
Prescription data provides a second method of finding people who are out of care. Weidle stated that the CDC, health departments, and providers are able to find out whether prescriptions have been picked up. According to the data Weidle presented, over 90% of all prescriptions filled in the U.S. come from just five prescription benefit managers. The fact that prescriptions aren't being picked up provides an indication that a person is not in care and not virally suppressed. One of the drawbacks, as Weidle noted, is that tracking pharmacy prescriptions alone won't tell us why people haven't filled prescriptions. The "why" may not have to do with a person's willingness to adhere but with a complicated health care system where a person's insurer may change from year to year, depending on employment, income, and whether insurers continue to sell plans in a state market. He stated that data need to be collected by health departments to determine why people aren't filling prescriptions.
When asked how a person living with HIV's privacy is protected by Health Insurance Portability and Accountability Act of 1996 (HIPAA) laws as more entities have access to people's HIV status, what antiretrovirals they take, and what their CD4 or viral load count is, Weidle answered: "[T]hat all has to be worked out, the clinics and the data-to-care strategies. The clinics and health departments have to work out the arrangements on how they'll share the data. Health departments can look at the names, but they can't give over the names of other people's patients."
Studying HIV Molecular "Networks" to Understand Transmission Trends
One of the latest trends in epidemiological surveillance is the use of HIV molecular data to understand related connections among people who recently contracted HIV. Molecular surveillance, according to the CDC, is "the collection, reporting, and analysis of HIV genetic sequences generated through HIV drug resistance testing." When people are diagnosed with HIV, they receive a test for drug resistance that helps determine the best antiretroviral regimen for them, and that data can now be used to look at relationships of viruses between individuals, although it cannot tell whether a person contracted the virus from any one person. A molecular cluster is defined as at least five people with similar HIV strains being diagnosed within the last 12 months of one another. Health department officials may use information about a cluster of new infections to do follow up linkage-to-care and partner services work, as well as to also identify people within that sexual network who may need testing and prevention services (PrEP for instance), or people who may be undiagnosed.
At this year's CROI conference in Boston, the CDC released two HIV studies in the U.S. using molecular surveillance to understand rapidly growing clusters around the country and clusters among transgender women in Los Angeles.
Nanette Benbow, M.A.S., a research assistant professor in the Department of Psychiatry and Behavioral Sciences at the Northwestern University, presented her work on understanding the role of molecular clusters in HIV surveillance, as well as some of the ethical concerns about collecting and using such data. A meeting in May 2017 organized by Project Inform and Benbow in Chicago resulted in a report on legal and ethical concerns with the use of molecular surveillance data. Much of Benbow's presentation on this topic at Adherence 2018 came from the results of that report.
"Whenever there is the use of surveillance data, there is a reason for pause," said Benbow.
While understanding the potential uses for molecular surveillance, Benbow also discussed some of the risks involved. She noted that, while the technology does not show the direction of transmission between people, in the hands of the wrong prosecutor, attempts could be made to indicate a relation between HIV infections, leading to more HIV criminalization. And even if this never occurred, the perception by people living with or at risk for HIV that this technology was being used to unfairly target them could keep many people from testing and care. Benbow also said that the manner in which disease intervention specialists in health departments use or talk about this technology with communities and people living with HIV could create increased stigma.
"Community needs to be a part of the entire process, said Benbow. "If people have concerns about how this data is being used, we need to hear from them. And the burden of the health departments is to explain to people exactly what this data is, what is not, and how these data are being used."