When Penn LDI Fellows looked at nearly 7,000 Penn Medicine patients who used stimulants, they found something disturbing: Their cardiovascular risk was exceptionally high. Nearly 30% had suffered a heart attack, stroke, or heart failure within just three years.

Statistics like these led researchers Rebecca Arden Harris, Sean Hennessy, and colleagues to design the first heart disease risk calculator for patients who use cocaine or methamphetamine. Until now, researchers had fashioned calculators only for the general population.

Stimulants—specifically cocaine and amphetamines—are no minor threat to the heart. They can cause sudden spikes in blood pressure, induce spasms of the coronary arteries, trigger dangerous heart rhythms, and damage the heart muscle, often in people in their 30s and 40s.

Harris and Hennessy built their model using health records from 6,940 adults with documented stimulant use from 2016 to 2024. The tool is intentionally simple, so its elements—age, smoking status, blood pressure, and stimulant use, among others—can be gathered during office visits. Their calculator showed excellent calibration, meaning its predicted risks closely tracked the number of heart events that occurred over three years.

More work remains. The calculator was developed using data from one hospital system, so it needs more validation. But the team hopes that the three-year risk estimates will spur physicians to provide more preventive care, especially to patients who use cocaine, which causes more heart problems than methamphetamine.

Harris and Hennessy discuss more about their heart risk calculator below:

Why does a new calculator matter for people who use stimulants?

Harris: Because standard risk calculators may give clinicians the wrong picture for these patients. Most estimate risk over 10 or 30 years and were developed using populations that included very few people who use stimulants. For a 40-year-old who uses cocaine, a reassuring “low 10-year risk” score can be misleading because the greatest danger may come much sooner. A three-year window may better reflect the time frame in which serious cardiovascular events occur in this population. It also fits the settings where these patients are often seen, such as emergency departments and addiction treatment programs, where clinicians need information they can act on right away.

How will the heart calculator work in practice?

Hennessy: It’s deliberately simple. The calculator uses information that is typically available during a clinical visit: the patient’s age, whether they currently smoke, their blood pressure, whether they use cocaine only or also methamphetamine, their race, and whether they are already taking medication for heart disease or high blood pressure. It returns a personalized three-year risk estimate, with no blood tests required.

That last point matters more than it sounds. People who use stimulants often receive fragmented care, so laboratory results like cholesterol levels are frequently missing from their records. We designed the tool to work with information available at every clinical encounter, without waiting for laboratory test results. Ultimately, we envision it being integrated into the electronic health record so clinicians can use it where patients receive care.

What are the implications for heart care?

Harris: Our hope is that it supports more informed conversations between patients and clinicians. We have also proposed a preliminary clinical framework, but it is an early idea rather than a consensus guideline. For example, a higher estimated risk could prompt clinicians to consider more intensive preventive care, such as closer attention to blood pressure, statin therapy when appropriate, referral for a cardiac evaluation, and treatment for stimulant use. A lower estimated risk might support a less intensive approach. We see this as the beginning of a conversation with colleagues in addiction medicine and cardiology about how best to care for this underserved population. Whether using the calculator this way improves patient outcomes remains to be tested, but that is the next step.

What are the heart risk calculator’s main limitations?

Hennessy: This is a first-generation, proof-of-concept model, and we’re clear about that. It was developed using data from a single health system in the Mid-Atlantic, so it needs to be tested in other populations and regions, especially in the western U.S., where methamphetamine use is much more common. Electronic health records also don’t capture everything that matters, such as how long or how often a person has used stimulants. The model performed well in our study, but it is not a substitute for clinical judgment.

Where is your research headed next?

Harris: Three directions. First, external validation: testing and refining the model in other populations and regions to confirm that it performs well elsewhere. Second, implementation studies to determine whether providing clinicians with individualized risk estimates changes clinical decisions and, ultimately, improves patient outcomes. Third, revising the model with richer clinical information as new data become available. Ultimately, we hope to develop a tool that clinicians can trust and use routinely so this population receives the preventive cardiovascular care it has often lacked.

---

The article, “Three-Year Cardiovascular Risk Prediction Among People Who Use Cocaine or Methamphetamine,” was published in Drug and Alcohol Dependence Reports on June 22, 2026. Authors include Rebecca Arden Harris, Fengge Wang, Warren B. Bilker, Renae Judy, Michael G. Levin, Scott M. Damrauer, and Sean Hennessy.

---

Author

---