First of its kind algorithm may help predict treatment adherence in MS

Swoop, a consumer health data company, is launching a first of its kind algorithm that’s designed to predict treatment adherence in people with multiple sclerosis (MS) or other conditions.

The aim of the new algorithm — which is based on artificial intelligence (AI) and machine learning (ML) strategies that use real-world data from more than 300 million patients — is to identify those individuals who are more likely to fail to maintain treatment adherence.

Specifically, the algorithm will predict the likelihood of patients stopping their medication or therapeutic regimen within the next 30 days. That would enable pharmaceutical companies and healthcare providers to timely engage with patients with a particular condition to prevent them from stopping their medication.

“Until now, real world data targeting has focused on what has historically occurred in the healthcare ecosystem, such as a diagnosis or a prescription, rather than predict what is likely to happen in the future,” Scott Rines, president of Swoop, said in a company press release.

“Through advanced ML and AI, this breakthrough targeting allows brands to proactively intercept patients and their HCPs [healthcare providers] at one of the most critical moments in the treatment journey: just prior to a patient becoming non-adherent,” Rines said, noting that interventions could then be introduced to try to prevent non-adherence.

Recommended Reading

May 18, 2022 News by Marisa Wexler, MS

Marriage, Education, DMT Affect Patients’ Treatment Adherence

Swoop says its algorithm is accurate for over 90% of MS patients

The real-world data used by the new algorithm was collected over more than 10 years from more than 300 million people whose data was de-identified. It also includes 65 billion anonymized social determinants of health signals — nonmedical factors that influence health outcomes.

As many as half of patients with chronic conditions such as MS fail to adhere to their therapeutic regimens, according to researchers. This often leads to added healthcare costs to manage symptoms that would otherwise be controlled with medication.

According to Swoop, its AI-powered algorithm accurately predicted 92% of MS patients who became non-adherent to their treatment in the following 30 days.

By using Swoop’s algorithm, pharmaceuticals now would have the chance to reach patients in a timely manner before they stop their treatment. Then, companies can target those patients for interventions that better educate them about their disease and about therapies that could effectively treat their conditions, enabling them to be active players in their treatment journey.

“The end result is that patients are more likely to stay on life-improving treatments, benefiting their health and the healthcare system,” Rines said.

Advances in AI and machine learning have proven a game-changer in treating patients more proactively, according to Swoop.

The end result is that patients are more likely to stay on life-improving treatments, benefiting their health and the healthcare system.

“This represents just the beginning of what predictive modeling can bring to healthcare marketing, allowing brands to better understand their audience, accurately target them and optimize engagement before a real world event has even occurred,” Rines said.

Swoop predictive algorithm can be used with any condition, including rare diseases. The company says it is compliant with the U.S. Health Insurance Portability and Accountability Act, known as HIPAA, and is a member of the Network Advertising Initiative.