The painkiller Vioxx was withdrawn from the market in 2004 following a three-year study which showed it doubled the risk of heart attack and strokes in patients taking it for at least 18 months.

Vioxx belongs to a class of drugs known as COX-2 inhibitors which were originally designed to be a safer option to aspirin, ibuprofen and other non-steroidal anti-inflammatory drugs or NSAIDs, which can cause sometimes fatal stomach bleeding.

Now a correction to a key study on Vioxx has revealed the risk of heart problems was elevated throughout the time people were on the drug and did not just develop following 18 months of use of the drug as manufacturer Merck has stated.

The correction backs the claims of many doctors who believe the risks showed up after as little as three months of use.

Dr. Jeffrey Drazen, editor-in-chief of the New England Journal of Medicine who published both the study and the correction, worked with the authors to correct their original findings of the APPROVe trial which was published in March 2005.

He says heart attacks and strokes occurred more frequently after people had been on the drug for at least 18 months, but the actual harm occurred much earlier, and it is a subtle but critical point.

Cleveland Clinic cardiologist Dr. Steven Nissen, who has repeatedly challenged the study’s conclusions in the past, says a key legal defence in the liability cases has been the suggestion that there was no risk until patients had taken the drug for 18 months, now, with the correction the authors have removed any claim that there was a delay in risk.

The main authors, Dr. Robert Bresalier of the University of Texas’ M.D. Anderson Cancer Center and Dr. John A. Baron of Dartmouth Medical School, have apparently acknowledged that their statistical analysis had been flawed.

Drug company Merck is facing more than 11,500 lawsuits from people claiming to have been harmed by the arthritis drug and an ongoing trial in Atlantic City, N.J., is the seventh case. Merck has won three to date.

Vioxx earned the company $2.5 billion a year before it was withdrawn.

The correction and comments on it appear in the New England Journal of Medicine.