Observational studies have previously reported a higher risk for venous thromboembolism (VTE) in people who use nonsteroidal anti-inflammatory drugs (NSAIDs) than in those who don’t use NSAIDs.

However, authors of a new study of more than 7 million patients wrote that the previous studies may have been vulnerable to bias because of potential confounders related to the condition of the patients who used NSAIDs, such as acute inflammatory states, frailty, and immobility.

In the study, published in the Journal of Internal Medicine, Yuichiro Matsuo, MD, MPH, with the Department of Clinical Epidemiology and Health Economics at The University of Tokyo in Bunkyō, Japan, and co-authors employed an active comparator, new-user cohort design, with acetaminophen as the active comparator.

Avoiding NSAIDs May Not Be a Reasonable Choice

The researchers concluded that, “Although new NSAID users had a lower incidence of VTE than new acetaminophen users, they had a higher incidence compared with nonusers. Assuming that acetaminophen does not increase VTE risk, NSAIDs themselves may not increase VTE risk.” They added that avoiding NSAIDs and choosing acetaminophen because of concerns about VTE risk may not be a reasonable choice.

The study included new users of either NSAIDs or acetaminophen, aged 18-65 years, from January 2005 to May 2022, taken from a Japanese health insurance claims database. The adjusted hazard ratios (aHRs) of VTE among new NSAID users (n = 4,282,421) and new acetaminophen users (n = 2,728,202) within 60 days of prescription were calculated using propensity score overlap modeling and Cox regression.

VTE Incidence Rate Low in Both Groups

The authors note that the incidence of VTE among new users of both NSAIDs and acetaminophen was low — about 0.02% within 60 days of prescription in both groups. That is still much higher than in the general population, where incidence ranges from “2 to 20 per 10,000 person-years” depending on factors including age, race, and underlying conditions, they note.

“The relatively high incidence rate of 13.2 per 10,000 person-years observed in our study appears reasonable,” the authors wrote, “given that new users of NSAIDs or acetaminophen include individuals with acute medical conditions that elevate the risk of VTE.”

New NSAID users in the current study had a significantly lower incidence of VTE than new acetaminophen users (aHR, 0.70; 95% CI, 0.62-0.80). NSAID users had a significantly higher incidence of VTE than nonusers (aHR, 3.18; 95% CI, 2.85-3.55).

The authors note that their conclusion that NSAIDs alone may not increase VTE risk “strongly relies on the assumption that acetaminophen does not increase the risk of VTE. Although no biologically plausible mechanism has been established between acetaminophen use and VTE,” they wrote, “and no studies have conclusively established its effect on the development of VTE, it cannot be definitely stated that acetaminophen has no impact on VTE risk.” Studies evaluating the risk for VTE with the use of NSAIDs in different populations with different indications for NSAIDs are needed, they note.

Trial Design Sets This Work Apart, Expert Says

Robert J. Glynn, PhD, ScD, professor at Harvard Medical School in Boston and biostatistician in the Divisions of Preventive Medicine and of Pharmacoepidemiology and Pharmacoeconomics at Brigham and Women’s Hospital, Boston, has studied VTE risk for decades, and he told Medscape Medical News that the design of this trial and its statistical power are important.

“The active comparator study design is really critical,” he said. Comparing NSAID use with NSAID nonuse, as many previous studies have done, “is a real problem for exposures like this because nonusers are inherently different in ways that are very hard to capture.”

Glynn cautioned that the study “is not a breakthrough” but “useful” and also novel in using a window of 60-day risk after the prescription because “these events don’t happen very often. The closer you are to the exposure initiation, the more likely it is that it’s related to the drug. When you’re doing administrative claims data, you can never be sure whether the patients continue to take the drugs. Data surveillance is different for the people who have gone to the doctor and sat down with the doctor and gotten a prescription than for the people who haven’t gone to the doctor in months.”

Glynn said he agreed with the conclusion that choosing acetaminophen over NSAIDs based on concern for VTE may not be reasonable and said he has never been comfortable with rejecting NSAIDs for that reason.

He said it’s very difficult to control for the reasons people need to fill a prescription. Additionally, “the full risk factor set for VTE is just incompletely known.”

“The authors don’t conclude, to their credit, from my perspective, that NSAIDs are safer than acetaminophen; they say they’re just not any worse,” Glynn said.

The study was funded by the Ministry of Health, Labour and Welfare of Japan. The authors reported having no relevant financial relationships. Glynn reported receiving grant funding to his institution from Pfizer, Amarin, and Novartis.

Marcia Frellick is an independent, Chicago-based healthcare journalist and a regular contributor to Medscape Medical News.