Researchers at MUSC Hollings Cancer Center have developed a machine learning tool to identify cancer patients who may be at high risk for financial toxicity – the financial stress and hardship that can accompany a cancer diagnosis and treatment. The study brought together several investigators from the Hollings Cancer Prevention and Control Research Program , reflecting the project's focus on cancer outcomes, survivorship and care delivery.
The study, published in JNCI Cancer Spectrum , describes a personalized risk prediction model that uses patient information to estimate the likelihood that someone will struggle with cancer-related financial burdens, such as medical debt, unpaid bills or anxiety about treatment costs.
"Cancer treatment is unfortunately expensive, and financial toxicity is a complex problem," said lead author Haluk Damgacioglu, Ph.D . "There are transportation costs, lodging costs, lost income, medical bills and out-of-pocket expenses. In some cases, financial stress may even lead patients to delay or discontinue treatment. We aimed to identify people at risk earlier, before these challenges escalate."
Connecting patients with support sooner
Nearly a quarter of people with cancer in the U.S. experience financial toxicity, which includes both financial hardship and psychological stress.
"There's the material side, like debt or unpaid bills, but there's also the psychological side," Damgacioglu explained. "Even worrying about how you'll pay for treatment can become a major source of stress."
Many studies have examined who is most likely to experience financial hardship during cancer care, but ways to predict a patient's risk have remained limited. Earlier identification could help to connect at-risk patients with counseling and other services before financial strain affects care decisions, treatment adherence or quality of life.
Hollings offers a wide range of patient services for patients and their families. That includes financial counseling staffed by counselors who specialize in cancer care.
"At Hollings, we have financial navigation and counseling resources," Damgacioglu said. "The first step is identifying the patients who may need additional support so we can connect them with those resources sooner."
Building a tool to predict financial risk
To address that gap, the research team analyzed national survey data from almost 800 cancer patients who were undergoing or had completed cancer treatment within the past year. Patients were classified as experiencing financial toxicity if they responded "yes" to at least one of several material hardship or psychological stress questions, such as borrowing money, being unable to pay medical bills, going into debt, filing for bankruptcy or worrying about future medical costs related to cancer care.
The researchers tested six machine learning models, using patients' demographic, clinical and financial information to predict who would experience financial toxicity. They then fine-tuned the models to maximize sensitivity, prioritizing the identification of as many at-risk patients as possible.
"We didn't want to miss anyone who may experience financial toxicity," Damgacioglu said. "That was one of the most important goals of the study."
The best-performing model identified patients at risk for financial toxicity with 84% sensitivity and 75% specificity, balancing the ability to detect patients who need support with minimizing false alarms. The model identified most patients likely to experience financial toxicity without unnecessarily flagging many false positives.
The study also used interpretable machine learning methods to identify the factors most strongly linked to financial risk. Among the strongest predictors were:
- Younger age.
- Lower income.
- Poorer overall health.
- Active cancer treatment.
- Higher out-of-pocket medical expenses.
To translate the research into clinical care, the team developed a publicly available web-based risk calculator that estimates how likely a patient is to experience financial toxicity and categorizes it as low, moderate or high. The researchers envision the tool as eventually helping to connect patients with financial and other support services earlier in care. They will now focus on improving the model, testing it in real-world clinical settings and exploring how financial stress affects patients' long-term health outcomes.
"Financial toxicity is another side effect of cancer," Damgacioglu said. "If we can identify risk early, there may be opportunities to help patients before that stress becomes overwhelming."
