Dr. Tony Holohan:

We have provisional perinatal mortality data for 2014, which we got from the national perinatal reporting system. Deputy Creighton mentioned the work we did last year on perinatal statistics. We have discovered that perinatal outcomes and deaths are reported through four mechanisms, namely, the national perinatal reporting system, which was with the Central Statistics Office and has moved to the HSE; the HSE's National Perinatal Epidemiology Centre, which was recently established following the Lourdes inquiry and which is based in Cork; the CSO, which reports on statistics; and the General Register Office, GRO. Not surprisingly, with four different systems, we get four different answers, for reasons that are complex but explicable. They relate to the variations between the systems, how they get their data and the mandatory nature of the reporting.

Among the recommendations we made as a consequence of some of the analysis we did last year was to make mandatory the notification of early neonatal deaths, which was not a requirement under the Civil Registration Act. The Department of Social Protection amended the Act in 2014 as a consequence of this piece of work, and a set of regulations is to be commenced to bring the provision into force. There is a recommendation on standardising a single, common definition of perinatal death. If we end up with different reporting systems each producing what appears to be a different answer to the same question, it undermines public confidence in our collective ability to describe. The one outstanding one remains for the HSE regarding the bringing together of the national perinatal reporting system and the NPEC which use different processes and methods and have different answers on perinatal mortality levels.

I would like to strike a note of caution regarding the interpretation of perinatal mortality statistics. The statistics have a value, which is reassuring, especially at national level, where they show that we stand up well in comparison with other countries, notwithstanding what I said about the variations between the reporting methods. When one examines smaller units and hospital level data, the numbers on which it is based are much smaller and the variation between centres is not picked up in a way that will show statistical significance. This can lead to an over-reliance on clinical staff and undue reassurance being taken from a rate when there were many other issues that might have suggested there were problems that were not being taken into account. I encountered this in Portlaoise.

The perinatal mortality rate should be seen as an alarm, not a description of a standard of care. It tells us we should look further. A perinatal mortality rate which is reassuring in the context of a series of so-called never events and a rise in adverse events, etc. is like saying the fire alarm is not ringing and ignoring the evidence that there is a fire. Undue reliance on perinatal mortality statistics can lead to a situation in which people are not as open as they might be to the possibility that things are not as reassuring as the perinatal mortality rates might suggest. This is why we made a recommendation around so-called patient safety statements in order that on a monthly basis some of the kinds of information about which I have spoken, including increases in adverse incidents, staff numbers, increased numbers of births, which happened in Portlaoise, and a range of other information, can be examined as means of quickly identifying safety problems rather than relying, particularly at the small unit level, on data that are produced only once a year in arrears.