[Skip Global Navigation]

Customers

Customers Home

St. George’s Hospital

Situation

Almost 39% of the deaths of patients who are moved out of intensive care units to a bed elsewhere in the hospital could be prevented if those most at risk remained in the unit for 48 hours longer, according to new research. The shortage of intensive care beds in Britain became a hot political issue in the NHS winter crisis of two years ago when a flu epidemic killed scores of elderly and vulnerable people.

These results were the conclusion of a Phd. research paper written by Kathleen Daly, Research Nurse at St Thomas’s Hospital, London. Dr Rene Chang, Consultant Transplant Surgeon, St George’s Hospital, London, was also part of the research team. The paper looked at the fates of 14,000 patients admitted to intensive care between 1989 and 1998, just before the NHS crisis began, and was recently published in the British Medical Journal May 2001.

Challenge

The hypothesis was that there would be a marked difference in death rates of intensive care patients they categorised as "at risk" - a third of the total - compared with those who were not. The final analysis found that a quarter of the at-risk patients died, compared with just 4% of those who were not. Dr Chang explains, “Mortality after discharge from intensive care could be reduced by nearly 39% if these patients stayed another two days before discharge".

It was also concluded that 16% more intensive care beds are needed to prevent those at risk being sent to other wards too early.

Solution

With the help of SPSS software, the researchers created a triage model which took into account the patient's age, disease, length of stay in the unit, whether they had heart/lung surgery and other factors in order to predict who was most in danger. This model could help doctors assess which patients are ready to be moved from intensive care to the wards. But unless the shortfall in intensive care beds is corrected, they say, neither this model nor Department of Health guidelines on discharge times will have much impact.

“SPSS software tools were used for data pre-processing, data checking and re-coding of variables before we went on to some descriptive statistical analyses”, said Dr Chang. SPSS was then used for the following:

This is the first study to pin-point a major cause of post ICU deaths in the UK and it would have been difficult, time consuming and expensive without the use of SPSS software.

— Dr. Rene Chang, Consultant
Transplant Surgeon, St.
George’s Hospital NHS Trust

Results

The model was developed from 5,475 patients discharged from Guy's intensive care unit. The logistic regression analysis identified the best predictive variables being age, presence of chronic end stage disease of major organs, whether had undergone cardiac surgery, length of ICU (intensive care unit stay) and the APACHE II acute physiology points on last day of ICU stay.

A probability >=0.6 was used to define patients at risk.

The model was then applied to 8,449 patients from 20 intensive care units. The post ICU discharge mortality of the cohort was 11.3%. 34% of the cohort was identified as at risk. Among these 2875 at risk patients the mortality was 25% (718). Among those not at risk the mortality was only 4%. If the at risk patients were to stay another 48 hours, their mortality could be reduced by 39% (280). To do this would require an increase in the number of fully funded ICU beds by 16%.

Dr Chang concludes on some alarming findings by saying, “This is the first study to pin-point a major cause of post ICU deaths in the UK and it would have been difficult, time consuming and expensive without the use of SPSS software”. The study was published in the British Medical Journal on the 25th May, 2001 and received saturation media coverage in all the TV, broad sheets and radio bulletins.