They said patients treated in hospitals that ranked highest in use of health information technology to manage patient records and physician notes were 15 per cent less likely to die compared with patients in hospitals that ranked lower.

"If these results were to hold for all hospitals in the United States, computerizing notes and records might have the potential to save 100,000 lives annually," Dr Neil Poe of Johns Hopkins University School of Medicine in Baltimore, who worked on the study, said in a statement.

The study, published in the Archives of Internal Medicine, lends fresh evidence that information technology can improve health quality and cut costs by reducing errors.

President Barack Obama has pledged to encourage hospitals to shift to such systems as part of healthcare reform. But a number of recent studies, including one last May by the Congressional Budget Office, suggest the promise of health information technology was overstated.

The current study led by Dr Ruben Amarasingham of the University of Texas Southwestern Medical Center in Dallas took a different approach. Rather than simply looking at the presence of information technology, the researchers looked at whether doctors were actually using it.

URBAN HOSPITALS

They surveyed doctors at 41 urban hospitals in Texas and divided the hospitals into three groups according to their use of technology.

Then, they checked the records of more then 160,000 patients over age 50 to see if there was a link between information technology and the care given for one of four conditions: heart attack, heart failure, heart bypass and pneumonia.

The researchers looked at four types of information systems - those that automate notes and records, manage tests results, manage doctor's orders for patient care and those that help doctors make medical decisions.

Overall, patients treated at hospitals where information technology was used the most by doctors had 16 per cent lower odds of having a complication than patients where information technology was used less.

"They found that increased use of information technology was associated with both lower costs and better outcomes," said Dr David Bates of Brigham and Women's Hospital in Boston, whose editorial on the study appears in the same journal.

Different software programs were associated with different benefits. Hospitals that scored high in the use of software to automate patient records and notes had 15 per cent lower odds that patients would die during their hospital stay.

At hospitals with high scores in use of software to track a doctor's orders, heart attack patients had 9 per cent lower risk of dying and bypass patients had 55 per cent lower odds of dying compared with hospitals used the technology less.