Crowdsourced data on overdoses pinpoints where to help

Men lined up outside the Corner of Hope, a homeless resource center, eyeing free supplies on plastic shelves inside a white van.

Some wanted bags with toiletries or condoms, but others took kits that help them safely use drugs or naloxone, an opioid overdose reversal medicine.

Gilbert Shepherd, an outreach worker for Punto de Partida, a nonprofit that serves people who misuse drugs, gently questioned those who took the drug safety kits. A man wearing sunglasses and a black T-shirt explained that not long ago he took a pill he bought for $1.50 and, within minutes, he passed out for hours. A man in a plaid shirt and khaki pants described seeing someone overdose after taking a blue pill a month before.

Those two overdoses would be added to a new Texas database called Texans Connecting Overdose Prevention Efforts, which aims to improve drug overdose tracking across the nation's second-largest state.

The University of Texas project, known as TxCOPE, is one attempt to solve a problem exasperating officials nationwide who are trying to lower the record number of drug deaths: getting an instant, accurate picture of both nonfatal and fatal drug overdoses. Community groups are now using TxCOPE's data dashboards and heat maps to see where overdoses are spiking and then target those hot spots with prevention efforts such as naloxone training and supplies, said Christopher Bailey, project coordinator at Project Vida, a health center in El Paso.

It is one of the few projects in the U.S. pooling crowdsourced overdose data from harm reduction groups in a systematic way, according to Leo Beletsky, a public health law expert at Northeastern University. Such projects compensate for the lack of an accurate picture of the decades-long overdose crisis. "It's scandalous," Beletsky said.

More than 107,000 Americans died of drug overdoses in 2021, according to the Centers for Disease Control and Prevention. But there is no national count of how many people survive drug overdoses. The CDC doesn't even have a standard method that states can use to count nonfatal overdoses. It aggregates overdose data from emergency room visits based on clinical and billing codes from participating states, but that excludes people who don't interact with the medical system, said Bradley Stein, director of the Rand Opioid Policy Center. It is a "huge blind spot," Stein said.

Plus, fatal overdose data is often published weeks or months later, once an official medical examiner's report or toxicology results show what substances caused the deaths. "We're looking in the rearview mirror with opioids," Stein said.

Other projects that count nonfatal overdoses, such as ODMAP, rely on reports from law enforcement or first responders. But many drug users won't call emergency services or report overdoses for fear of arrest, deportation, or other consequences such as the loss of their children or housing due to drug use, said Traci Green, a professor and the director of the Opioid Policy Research Collaborative at Brandeis University.

"Simply put, current national data systems have not kept up with the scale of the overdose epidemic," wrote Dr. Rahul Gupta, director of the Office of National Drug Control Policy, in a call for action published June 30 in JAMA. He added that building a better data system is essential and that his agency is convening with other federal agencies to improve the tracking of nonfatal overdose data.

Nationally, Green estimated, about 50% of nonfatal overdoses go unreported, with a higher undercount in places with stricter law enforcement and among communities of color. The missing part of the picture "is a very diverse one," she said.

In Texas, up to 70% of overdoses, mostly nonfatal, go unreported, estimated Kasey Claborn, lead researcher on the TxCOPE project and an assistant professor at UT's Dell Medical School and Steve Hicks School of Social Work.

Officially, about 5,000 Texans died of a drug overdose in 2021. Claborn believes that is an undercount, too, because the state has medical examiners' offices in only 15 of its 254 counties. Most counties have justices of the peace who don't always request pricey toxicology tests to determine the cause of death. The state recorded nearly 4,000 opioid-related calls to the Texas Poison Control Network last year and nearly 8,000 opioid-related emergency room visits in 2020. Claborn is analyzing how the data TxCOPE collects compares with those official statistics.

Drug experts are frustrated the U.S. doesn't treat the overdose epidemic with the urgency it does for covid-19. Drug deaths surged during the pandemic as illegal fentanyl, which is 50 to 100 times more potent than morphine, flooded the nation's street drug supply and people were cut off from substance abuse support. But while public health authorities based pandemic restrictions on local covid caseloads and death counts, experts and outreach workers have lacked real-time data that would allow them to react with interventions that could save drug users' lives.

"How is that helping in a public health emergency?" said Daniel Sledge, a paramedic testing TxCOPE in Williamson County, just north of Austin.

That information could help health workers identify which areas to blanket with naloxone or whether they need to educate people about a batch of drugs laced with lethal fentanyl.

TxCOPE, funded by the state's opioid grant and the federal Substance Abuse and Mental Health Services Administration, started in El Paso in June 2021 and then expanded to Austin, San Antonio, and later Williamson County. An official launch is planned for Sept. 1, with a rollout to the rest of the state in stages.

Before the group started, outreach was more haphazard. "It was catch as catch can," said Bailey, with Project Vida. Like many harm reduction groups, they would informally track overdoses, which often occurred among an itinerant population. But they didn't have a way to pool that information with other city groups or generate maps to drive outreach. TxCOPE has helped the group find pockets of at-risk people so they can provide them with overdose prevention, peer support, or treatment referrals. Now "you are able to really home in on those areas with laserlike focus," he said.

Punto de Partida outreach worker Paulina Hijar, for example, said she routinely meets people who injected their friends with dangerous and ineffective homemade overdose remedies — either milk or a mixture of water and salt — or got naloxone and never called authorities. Because outreach workers have built trust in their communities, they say, they can gather information about overdoses, including when and where they occurred, that would normally be omitted from official statistics.

Privacy is a key feature of the TxCOPE project — people need to be able to share overdose information without fear of consequences, Claborn said. Texas passed a law in 2021 intended to shield from arrest people who call emergency services during an overdose, but it's narrowly tailored. People who have a felony drug conviction, for example, don't qualify. And someone is protected from arrest only once. TxCOPE has a federal certificate of confidentiality that protects it from court orders, and Claborn doesn't share raw data with the state.

Claborn wants to use the project, which is being revamped this summer before the launch, to bring more federal dollars to the state. "We've had difficulty proving there is an actual problem in Texas, because it's been hidden," she said.

She has been working on a feature that would allow anyone in the community to report overdoses, an effort to improve counts in parts of the state without harm reduction groups. Eventually, Claborn wants to check the crowdsourced data against toxicology reports.

For now, though, the project relies on anecdotal evidence from drug users and others in the community who report reviving someone with naloxone or seeing someone lose consciousness or suffer other effects of taking too much of a drug.

On a recent afternoon, Shepherd and Hijar recognized a man in El Paso's Houston Park. He told them about a woman who had died alone in her apartment about a week and a half earlier after taking a mixture of drugs. He also mentioned a guy he was able to revive with naloxone about two months earlier. Details were minimal.

Researchers and harm reduction groups say this nebulous data is better than what they've had in the past. Even though the project's anecdotal data isn't thoroughly vetted, the step toward timeliness is great, Stein said. "We've got nothing else right now," he said.