Patients having an operation may do well or badly. Knowing the percentage that do badly or die can tell us if a particular hospital or doctor are any good at doing the operation. For example, if we hear that in St Mungo Hospital, 2% of patients having coronary bypass surgery die, whereas in St Hilda, the figure is 5%, we might (quite reasonably) conclude that St Mungo is better than St Hilda. But (and it is a big "but"), this assumes that the two hospitals treat similar patients. Say St Mungo operates only on fit young people with no other illnesses, and St Hilda takes all the sick, the old, the emergencies and all the patients turned down by St Mungo, then it is quite possible that St Hilda is in fact the better hospital. One way of sorting this out is to find a measure of risk for a particular patient having a particular operation. If we have a system to work out the expected death rate for a group of patients, we can easily tell who is doing better. If our risk system tells us St Mungo should have a death rate of only 1% and St Hilda of 7%, then we can conclude, with some confidence, that St Hilda is the better hospital despite apparently poorer results.

Yes, and there are many of them. Really simple, basic ones make adjustments for one or two risk factors only such as age and sex. More sophisticated ones study a number of risk factors and add up their contributions to the outcome of the operation. In heart surgery, two well-known systems are the Parsonnet system (from America) and the EuroSCORE (from Europe, as its name implies). EuroSCORE have a website which explains many of these issues and allows you to work out your own risk of dying from a particular heart operation, as long as you have some knowledge of your medical condition. Even more complex and highly accurate risk models have been developed both in Europe and America, so that the assessment of heart surgery risk is now an area of intensive and rapidly advancing research.

The most important feature is that risk stratification helps guide the surgeon and the patient towards deciding whether or not to go ahead with surgery. Decisions about an operation can only be made if you know the likely risk and the benefit of the operation so that you can "weigh them up" against each other. For example, if you have angina, and you know that the benefit of an operation will be to get rid of the angina and help protect you against a heart attack, you might well wish to proceed with surgery if your risk of dying is 3%, but you may have second thoughts if your risk is, say, 30%.  Understanding risk is also important in knowing which hospital or surgeon you choose. League tables may place hospitals (and, soon, surgeons) in the order of their outcome results (death after a procedure). To make sense of such tables, having an idea of risk is essential.