The eight researchers reported in The Lancet that in the year after the 2,587-patient study was halted, 34 people who had taken Vioxx and 18 who had taken placebo suffered a heart attack, for a 94 percent higher risk with Vioxx; strokes occurred in 19 Vioxx users and nine people on placebo, for a risk slightly more than double. Altogether, 76 Vioxx users and 46 placebo takers had a heart attack, stroke, blood clot or died during that follow-up year.

The increase in those complications was roughly the same as what was found during the three-year trial and the first two weeks after the study ended — the period included when Merck first reported results of APPROVe in the New England Journal of Medicine in February 2005.

Merck critics have said it was inappropriate to end patient follow-up 14 days after the study.

"This study is raising an important red flag about that" cutoff, showing the risk persisted for at least a year, although too few patients were followed longer than that to see if the risk subsided, said Dr. Harlan Krumholz, a Yale University cardiologist who has assisted Vioxx plaintiffs suing Merck.

"It adds another important chapter to the Vioxx story, but also is an important warning to us about how we assess the safety of medication" long-term, including after use stops, Krumholz said.

After Merck published its first analysis of APPROVe, numerous experts criticized it. Merck later admitted it used the wrong statistical method in the analysis, and editors of the New England Journal 2006 published a correction stating the risk of heart problems was elevated throughout the time people took Vioxx.

Merck posted all the data from APPROVe — the data used in the new study — on its Web site that same month.

In an editorial, doctors from the University of Oxford in England and Catholic University of Rome wrote that differences in the approach of the new analysis make it clear there's no "latent period" before Vioxx increases heart risks, and shows the danger appears to have been worse in people who already had risk factors — something they said is shown in a combined analysis of six studies of Celebrex.

Made by Pfizer Inc., Celebrex is the only drug from the same class as Vioxx still on the market in the United States.

The editorial writers also note that older nonsteroidal anti-inflammatory drugs — such as naproxen and ibuprofen — also can bring cardiovascular risk, plus risks of stomach bleeding and other complications.