NEW ORLEANS — A study doctors had hoped would provide some clear answers about which of the most common pain relievers was safest for people with or at high risk for cardiovascular disease has instead left experts puzzled about how to interpret the results.

The conclusion of the trial, called PRECISION, appears impressive at face value. Among people with arthritis taking ibuprofen, naproxen or celecoxib for pain relief, the COX-2 inhibitor celecoxib — known by the brand name Celebrex — posed no greater risk for heart attacks, strokes or cardiovascular-related deaths.

But the study, presented at the American Heart Association’s Scientific Sessions, doesn’t solidly back that up, experts said.

“The trial was so poorly designed that the conclusions are unsupportable,” said Garret FitzGerald, M.D., director of the Institute for Translational Medicine and Therapeutics at the University of Pennsylvania, who wrote about the study in an editorial published in Circulation.

“The trial is called PRECISION, and that’s the last thing it delivers,” he said. “It provides us no useful information to influence practice.”

The study’s lead investigator, Steven Nissen, M.D., had a different takeaway.

“The thing that everybody feared was that celecoxib would be like Vioxx. And we can say definitively that it is not,” said Nissen, chairman of cardiovascular medicine at the Cleveland Clinic, who presented the findings Sunday.

The three drugs studied are part of a class of painkillers called non-steroidal anti-inflammatory drugs, or NSAIDs. They are among the most commonly used drugs in the world, including more than 100 million U.S. prescriptions in 2013.

Well-known over-the-counter brands such as Advil, Motrin and Aleve are used to treat everything from fevers, headaches and menstrual cramps to back pain, arthritis and other chronic conditions. Some are available over-the-counter at low doses while higher-dose drugs require a prescription. A specific class of NSAIDs, the COX-2 inhibitors, were developed in an attempt to avoid gastrointestinal side effects.

It was Nissen’s own study in 2001 that raised a red flag about heart attack and stroke risk among one of the COX-2 inhibitors, Vioxx (rofecoxib). The findings, showing an increase in cardiovascular events, led Merck & Co. to pull Vioxx from the market in 2004. A year later, the FDA pulled another COX-2 inhibitor called Bextra (valdecoxib) and announced that all NSAIDs would carry a warning about cardiovascular risk on their labels.

Celebrex was the only COX-2 inhibitor left, and the FDA allowed Pfizer, the maker of Celebrex and Advil, to keep Celebrex on the market but required that they conduct a safety trial.

The study, simultaneously published in the New England Journal of Medicine, hoped to enroll people at risk for heart attacks and strokes and who needed to take medication for chronic pain. The European Medicines Agency refused to allow investigators to enroll patients in Europe, and the more than 24,000 who ultimately joined the study ended up being at low cardiovascular risk — only about 1 percent of patients each year had a heart attack or stroke or died from cardiovascular causes across all three groups.

“Doctors may have been nervous about giving people with high cardiovascular risk a COX-2 inhibitor, so they never enrolled them in the trial,” said Elliott Antman, M.D., a cardiologist at Brigham and Women’s Hospital in Boston and associate dean for clinical/translational research at Harvard Medical School, who was not involved in the study.

In addition, FitzGerald said, the study design included biases in favor of Celebrex, a drug with sales once topping $2 billion a year that is now available as a generic. Its patent expired in 2014, eight years after the PRECISION trial started.

The dose of celecoxib, which is only available as a prescription, was capped at the FDA-approved amount for patients with osteoarthritis arthritis, who made up most patients in the trial. But doctors could increase the prescription-strength dose of ibuprofen and naproxen to achieve better pain relief. The average daily dose of celecoxib was 209 milligrams.

“Less drug exposure means you’re going to have fewer side effects,” FitzGerald said.

The researchers also found no indication that naproxen was safer, as many had thought.

Aspirin use during the study, which wasn’t reported in Sunday’s results, might have given another advantage to celecoxib because ibuprofen and naproxen can undermine the blood thinner’s effectiveness in preventing heart attacks and strokes.

Making the results even harder to untangle, experts said, is that patients went on and off their assigned drug, and more than two-thirds stopped taking their assigned drug altogether by the 10-year study’s end.

Guidelines currently say NSAIDs should in general be avoided in people with or at high risk for heart disease and stroke. These latest findings won’t change that advice, Antman said.

“Use NSAIDs preferentially in the lowest-risk patient, give them the lowest dose required to manage their symptoms and for the shortest period of time needed,” said Antman, who helped write the AHA’s 2007 recommendations on NSAID use.

What the findings do highlight, FitzGerald said, is the need for a personalized medicine approach to predict which NSAIDs will work and which ones will cause problems based on a person’s unique genetic makeup and how they respond to environmental factors.

“That’s what counts for a patient. They don’t want to know the average effect. They want to know what’s going to be the effect for me,” said FitzGerald, who is part of a research consortium working to understand how different NSAIDs work differently in different people.

Nissen said an analysis is already underway to figure out how genes affected PRECISION participants’ responses to the three pain drugs.

“We have genetic material on the vast majority of these patients and we will be exploring it,” he said. “Stay tuned. We’re going to look at everything.”