AI and Aspirin

Nearly a quarter of Americans use three or more medications simultaneously. Because the risk of adverse reactions to drug combinations increases as a patient takes more medications, it is widely known that mixing medications can be unsafe. However, with 40 percent of American adults suffering from two or more chronic conditions, cutting back on medications is not so easy.

Drug interactions can occur through two broad mechanisms: synergistic reactions, where the effects of a drug are exacerbated, or antagonistic reactions, where the effects of a drug are minimized. Pharmacodynamic reactions are reactions caused by the effects of drugs on the body, while pharmacokinetic reactions occur through the body’s uptake of drugs.

Mixing aspirin and ibuprofen, two common household drugs, actually diminishes, rather than amplifies, the blood-thinning abilities of aspirin. This is because aspirin and ibuprofen compete for the binding sites on the cyclooxygenase (COX) enzyme. Binding COX catalyzes hormones that help modulate inflammatory responses. However, while aspirin irreversibly inhibits COX, ibuprofen does not. This means that the competitive inhibition between aspirin and ibuprofen prevents the full effects of aspirin from occurring.

More dangerous, however, is the way that non-steroidal anti-inflammatory drugs (NSAIDs), including aspirin and ibuprofen, can interact with selective serotonin reuptake inhibitor (SSRI) antidepressants. SSRIs are the most common category of antidepressants and can cause an increased risk of internal bleeding when taken with NSAIDs.

Combining some medications used to lower blood pressure, such as ACE inhibitors and spironolactone, also produces a synergistic effect. ACE inhibitors and spironolactone can react to cause hyperkalemia, or high potassium levels in the blood. This is because ACE inhibitors dilate the urinary tract blood vessels, thus decreasing the level of urine filtration, causing more potassium to remain in the blood. Spironolactone prevents sodium reabsorption from happening and thus increases potassium levels. The interactions between these hypertension medications are especially important with nearly half of American adults suffering from hypertension.

The nation’s chronic disease epidemic and subsequent overmedication are problems that require longer-term solutions. In the present, artificial intelligence may be used to mitigate adverse drug combination reactions. Computer programs that detect these drug interactions function in a number of ways. Some programs gather all of the data on the diseases that affect a particular organ in order to find what drugs could interact to affect that organ. Penn State researchers, for example, developed an algorithm using this model in 2019. Other algorithms rely on analyses of the target proteins of drugs. While some programs aim to work with data about existing drugs, others are created to aid the development of new pharmaceuticals.

For programs that directly work with doctors and patients, well-designed alert systems are crucial. Some researchers have even used autoencoder models that mimic how the human brain processes information. For these alert systems, researchers have developed ways to limit alerts to severe adverse reactions only. This can help mitigate alert fatigue, which could result in notifications about serious side effects going unaddressed.

Detecting new dangerous drug interactions will become more necessary as our dependence on medication continues to grow. The early stages of the COVID-19 pandemic saw a 21 percent increase in depression, anxiety, and insomnia medication prescriptions, for example. With dozens of new drugs being approved by the FDA each year, these algorithms could help speed up drug testing and increase the efficacy of drugs.