The number of deaths involving the superbug clostridium difficule (C.diff) at the John Radcliffe Hospital in Oxford and Banbury's Horton Hospital have soared.

But the NHS trust that runs both facilities said patients should not be worried as rates of infection were actually dropping.

New figures show that in the five-year period from 2002-06, 111 out of the 7,608 people who died at the JR had contracted C.diff - 1.46 per cent.

Between 2001-05, 57 of the 7,351 people who died had the infection - 0.78 per cent.

At the Horton Hospital, cases rose from seven to 24 for the same periods.

Both Oxford Radcliffe Hospitals NHS Trust and the Oxfordshire joint health overview and scrutiny committee, however, have said people should not panic because incidents of patients contracting superbugs at both hospitals were on the way down.

They said the latest figures - issued by the Office of National Statistics - reflected a greater awareness of C.diff and a drive to include the infection on death certificates, rather than anything more worrying.

Scrutiny committee chairman Dr Peter Skolar said: "I would take the statistics with a very large pinch of salt. I'm very wary.

"If you actually look at the figures for the John Radcliffe, they are extremely good. You can't have bad infection control and good figures."

A trust spokesman said rates of clostridium difficule had fallen by 40 per cent since 2005.

He added: "At the same time we know the reporting of C.diff on death certificates has gone up because of increasing awareness of the issue.

"It is also important to remember only a small fraction of patients who contract C.diff will die as a direct result of it.

"Death certificates included in these statistics include all those involving' C.diff, even if it isn't the main cause of death.

"The ORH sees nearly 200,000 inpatients every year, and treats some of the sickest patients, including the terminally ill.

"It is these seriously ill patients who have damaged immune systems which increase the risk of contracting an infection.

"Separating out the individual causes of death in these patients can be difficult, and research is ongoing to more accurately determine the level of mortality caused directly by infections."

Nationally, from 2002-06, 15,683 people died either as a direct result of C.diff or had contracted the disease prior to death - up from 0.40 to 0.60 per cent of the total number of deaths recorded during the same time.