Cardiac Statistics

Performance Data

Although Mortality is a crude indicator and we expect to be using a range of more sensitive data in future editions on our site , we are placing this data here because it is already available and be compared nationally.

Early mortality after cardiac surgery is defined as those patients who die within 30 days of operation. All patients who are discharged home after successful surgery are reviewed and closely followed. Any patients dying outside hospital within this period are also included in our mortality figures.

The society of Cardiothoracic Surgeons of Great Britain & Ireland collects mortality data from all Paediatric Cardiac Units in the U.K. From this data, it is possible to compare our results with the national average in terms of overall mortality, and mortality for broad age ranges and diagnostic sub-groups.

It is essential when assessing an individual Unit's overall performance compared with the national average, to consider the complexity of procedures performed in a particular unit and also the age of patients undergoing surgery. Some procedures are of great complexity and innately of a higher risk than less complicated operations. This so-called "case mix" can give an incorrect impression of a units performance where a large number of high-risk patients are treated. This also applies to the number of neonates and infants operated on, where again mortality is higher because of small size and complex cardiac abnormalities requiring surgery in this population.

Table 1, Table 2 and Table 3 show the mortality data for our unit for the past 3 years broken down into broad age ranges which can then be directly compared to the data published by the Association of Cardiothoracic Surgeons of Great Britain (ACTS) & Ireland for that particular year. It is gratifying that for each year not only our overall mortality, but also the mortality in infants, has been consistently lower than the national average.

It is immediately apparent that by dividing the population into smaller groups according to surgical procedure can be very misleading if the number of patients are very small. In the year 1997/1998 only one patient underwent operation in the group of closed heart procedures over 16 years of age. This patient did not survive giving mortality of 100% for the group that year. In the following year there were no deaths in 5 patients undergoing operation in this age group giving 0% mortality for those years. We believe that it would be more appropriate to report this data as cumulative results over a long period with statistical confidence limits to assess the significance of results. Unfortunately there are no comparative National published data to be able to make meaningful comparisons in this way at present.

In analysing this data meaningfully it is also important to appreciate the concept of "case mix" as mentioned above. Not all operations have the same risk and indeed the risk of not surviving an operation can vary from less than 1% to over 50%. The balance between the number of high risk and low risk procedures operated on can therefore have a big effect on the overall mortality of a centre. This has to be taken into account when comparing data between centres and also comparing mortality results with the national average which obviously will be skewed by those centres not performing this procedure.

This point is well illustrated if we consider the effect of a single operation - the First Stage Norwood procedure for hypoplastic left heart syndrome - on the surgical results of a centre. Only 3-4 other centres in the U.K perform this inherently high risk procedure. The majority of centres performing paediatric cardiac surgery do not have a Norwood program and it is a valid argument that, between-centre comparison can only be realistic, if this procedure is not included in the calculation of the institutional mortality

To illustrate this point the following graph shows results both including and excluding the Norwood procedure. In 1997/1998 and 1999/2000 the mortality is much lower excluding this high risk group. In 1998/1999 there was no mortality in our institution for the Norwood procedure and therefore the overall annual mortality is increased slightly excluding these patients.

Table 4 shows a breakdown of our results for 11 major diagnostic groups where there is U.K national data from the registry to allow comparison for 1997/1998. The data is further sub-divided according to 3 broad age bands of under 1 year, between 1-15 years and over 16 years. In 1998-1999 the number of diagnostic sub-groups where U.K data was available increased to 13.

Table 5 & Table 6 show this data for 1998-1999 & 1999-2000 compared with the national figures for 1998-1999 (The most recent National figures available).

This data is illustrated graphically in the following graphs. As discussed above comparison of mortality where numbers of patients operated on each year are small, is of limited validity, for these groups we need to examine results over a longer period where total experience is accumulated and accurate statistical assessment can be made compared to the U.K data. The data collected by ACTS does not exist in a format where these comparisons can be made currently.