AI Discovers Optimal Drug Combination to Prevent COVID-19 Recurrence

News

AI Discovers Optimal Drug Combination to Prevent COVID-19 Recurrence

AI Discovers Optimal Drug Combination to Prevent COVID-19 Recurrence.

A groundbreaking machine learning study has unveiled the best drug combination to prevent the recurrence of COVID-19 after the initial infection. Interestingly, the ideal combination varies among different patients.

Led by the University of California, Riverside, this research utilized real-world data from a Chinese hospital to find that factors like age, weight, and other health conditions determine which drug combinations can most effectively reduce the recurrence rate. This finding has been published in the journal “Frontiers in Artificial Intelligence.”

AI Discovers Optimal Drug Combination to Prevent COVID-19 Recurrence

An AI-driven study conducted by the University of California, Riverside, based on data from China, has revealed that the optimal drug combination to prevent COVID-19 recurrence varies based on individual factors such as age and weight.

This unique dataset considered patients who received treatment with up to eight different drugs and were monitored after discharge, allowing for a more in-depth analysis of reinfection rates and treatment effectiveness.

The data from China was chosen for two significant reasons:

Firstly, in the United States, patients undergoing COVID-19 treatment usually receive one or two drugs. However, in the early stages of the pandemic, doctors in China could prescribe up to eight different drugs, allowing for a more diverse analysis of drug combinations.

Secondly, COVID-19 patients in China were required to undergo government-operated hotel quarantine upon discharge, enabling researchers to understand reinfection rates in a more systematic manner.

“This makes the study unique and intriguing. You can’t get this kind of data anywhere else in the world,” stated Xinpeng Tu, a professor of statistics at UCR and an author of the study.

The project commenced in April 2020, around a month after the start of the pandemic. During that time, most research was focused on mortality rates. However, doctors near Hong Kong, in Shenzhen, were more concerned about the recurrence rate due to lower death numbers there.

“Surprisingly, almost 30% of patients tested positive again within 28 days of discharge,” noted Jason Liao, a co-author of the study and associate professor of bioengineering.

The study included data from over 400 COVID-19 patients, with an average age of 45 years, mostly having a moderate level of infection severity and an even gender distribution.

Most individuals received various combinations of antiviral drugs, anti-inflammatory drugs, and immune modulators like interferon or hydroxychloroquine.

Different populations saw better treatment outcomes with different drug combinations, influenced by the virus’s mode of action.

“COVID-19 can suppress interferons, which are proteins cells produce to counter invading viruses. With reduced defense, the virus can replicate within the body until the immune system ramps up, damaging tissues,” explained Liao.

Individuals with weaker immune systems before COVID-19 infection require immune-boosting drugs to effectively fight the infection.

Younger individuals tend to have an overactive immune response after infection, leading to excessive inflammation and even death. To prevent this, young patients require immune-suppressing agents during treatment.

“When treating a disease, many doctors often provide a one-size-fits-all solution for those over 18. We now need to rethink age differences and other conditions such as diabetes and obesity,” Liao stated.

In most cases, when testing drug efficacy, scientists design clinical trials that randomly assign individuals with the same disease and baseline characteristics to either a treatment or control group. However, this approach doesn’t account for other medical conditions that might influence whether a drug works or not for specific subgroups.

As this study utilized real-world data, researchers had to adjust for factors that could influence observed outcomes.

For instance, if a particular drug combination was mainly used for older individuals and proved ineffective, it’s unclear whether the issue lies with the drug or the individuals’ age.

“In this study, we pioneered a technique to address confounding factors by virtually matching individuals with similar characteristics who received different treatment combinations,” said Tu. “This way, we could infer the efficacy of treatment combinations in different subgroups. While there’s now a deeper understanding of COVID-19, and vaccines have significantly lowered mortality rates, there’s still much to learn in terms of treatment and preventing reinfection. Now, recurrence is a more concerning issue, and I hope people can make use of these findings.”

Machine learning has been applied to various aspects related to COVID-19, such as disease diagnosis, vaccine development, and drug design, along with novel analyses of multiple drug combinations. Liao believes that the technology will play an even more significant role in the future.

“In the medical field, the impact of machine learning and artificial intelligence hasn’t been as substantial as I believe it will be. This project serves as a great example of how we can move towards true personalized medicine,” he concluded.