A new "checklist" system for treatment has cut patient deaths by 15% in pilots at three hospitals, according to research published in the British Medical Journal.

The lists were used in a pilot study at three hospitals operated by the North West London Hospitals Trust. Termed "care bundles,'' the lists provided people in nurse manager jobs and frontline staff clear details of best practice treatment such as when to administer antibiotics, what kind of blood samples are needed or the optimum level of oxygen saturation in the blood.

Effectively, the checklists operated like any other list of this type; providing clinical staff with a clear outline and reminder of correct procedure when treating a patient. Rather than add extra bureaucracy and oversight to the frontline of healthcare treatment, they were designed to simplify the steps that professionals could take to cut down on avoidable deaths.

The care bundles covered the 13 different diagnostic areas identified as having the highest number of deaths at North West London Hospitals Trust between 2006 and 2007, including stroke, heart failure and chronic obstructive pulmonary disease. The researchers looked at the effect of the care bundles on the hospitals standardised mortality ratio - otherwise known as HSMR.

In the HSMR system, a rating of 100 equates to the same risk of death to the national average for English hospitals. A rating of 120 would represent a mortality rate 20% higher than the average, and a rating of 80 would be 20% lower.

Following the introduction of the care bundles, researchers reported that the Trust's HSMR fell from the 2006-07 rate of 89.6 to 71.1 for 2007-08, making it the lowest among acute trusts in England. In total they calculated that 255 fewer deaths occurred, with 174 of these in the targeted diagnostic areas.

Cutting down on the number of avoidable deaths is a key responsibility for doctors and nurse managers. Jobs in healthcare often come under fire for the rate of patient admissions which result in some form of mishap or adverse effect; national statistics suggest that some 7% of all admissions result in something going wrong. Previous research indicates that around half of these are avoidable.