New statistics released for the first time today show that death rates are above expected levels in 11 out of 17 of Wales' District General Hospitals.
The Risk Adjusted Mortality Index (RAMI) shows the ratio of the actual number of deaths in a hospital, compared to the expected number of deaths. The baseline is set at 100 - so anything above that means a higher than expected death rate.
The figures show the Royal Glamorgan Hospital has the second highest level of unexpected deaths in Wales, behind the University Hospital of Wales in Cardiff.
The Welsh Labour Government says that the statistics are part of a drive to make the NHS more transparent in the wake of the Stafford Hospital scandal. However, they also say that the statistics may not be reliable.
The BBC reports that
"Officials warned that higher scores may be due to problems gathering data, rather than lapses in the quality of patient care. The Rami takes into account other factors about patients, including their underlying health, social factors and their lifestyles.
The way the data is collected is known as coding. Welsh government officials say that coding in Welsh hospitals has not been as good as in the English NHS.
The Welsh government uses Rami scores as one way of measuring performance in the NHS. However, it says Rami cannot be looked at in isolation and that reviewing case notes after a patient dies is a better way to find out whether improvements to care are needed."
What then is the point in publishing the data? Surely if such statistics are to have any meaning at all and play any part in driving up standards then they need to be reliable? There needs to be a uniform method of collecting data.
Yet again people in Wales are being let down by the Labour Welsh Government - "coding in Welsh hospitals has not been as good as in the English NHS." Coding is not the only thing where there are failures. Let's hope the new Health Minister has more success in dealing with some of the problems, and indeed shows more willingness to do so, than his predecessor.