Nursing shortages are linked to an increase in patient death rates, a study of English hospitals has found.
Patients in hospitals with more nurses per patient fared better
Scientists discovered mortality was 26% higher for the hospitals with the worst staffing levels compared with those with more nurses per patient.
The Royal College of Nursing said the study, published in the International Journal of Nursing Studies, showed it was essential to retain nursing posts.
But the Department of Health said nurse numbers had risen in recent years.
To establish the link between nursing levels and patient outcomes, UK researchers analysed just under 120,000 patient records and compared them with information from 4,000 nurses from 30 hospital trusts in England.
The data was taken between 1998-99.
They found the patient-to-nurse ratios varied from 6.9 to 14.3 across the trusts. And patients in the hospitals where nurses had the highest workloads were more likely to suffer complications and/or die than those in hospitals with better staffing ratios.
The researchers also discovered the nurses caring for the most patients were 71% more likely to suffer "burn out", and 91% more likely to be unsatisfied with their jobs compared with the nurses with a lighter workload.
Professor Anne Marie Rafferty, lead researcher on the study and a health services researcher from Kings College London, said: "We calculate that some 246 fewer deaths would have occurred in these 30 trusts had all the patients been treated in hospitals with the most favourable staffing levels.
"The number of lives that could potentially be saved through investments in nursing throughout NHS hospitals could be thousands every year."
The research forms part of a collaboration between five countries: US, Scotland, Germany, Canada and England, and is called the International Hospital Outcomes Study.
The scientists said the findings in England closely mirror those of the US and Canadian components, which were published recently.
The delay between the time the data was collected and the date of publication of this latest paper was because of the length of time taken to design how the study should be carried out, Professor Rafferty said.
Dr Beverly Malone, general secretary for the Royal College of Nursing, said: "This new independent research backs up what nurses have always known: that nurse numbers really do matter and that nurses make a life and death difference to how well their patients recover.
"That is why we must not allow nursing posts to be sacrificed to ease financial deficits. This is short-termism in the extreme and will end up costing the health service more in the long run as patients with complications that should have been picked up in the first place are re-admitted."
A Department of Health spokeswoman said: "Patient safety is always a top priority for the NHS and we now have around an extra 89,000 nurses working in the NHS delivering high quality care for patients than we did in 1997.
"Nursing ratios are a complex subject which has yet to gain a definitive consensus within the nursing profession and we welcome this addition to this debate."
Liberal Democrat health spokesman Steve Webb said: "This is further evidence of the damage that will be done to patient care if the NHS financial crisis results in even more frontline staff cuts."