Sarah Boseley, health editor
Friday April 15, 2005

Using statistics to assess competence is controversial. Some doctors will have many patients who do not make a full recovery because they take the highest risk cases, probably because they are at the top of their profession.

However, the scandal over babies dying in heart surgery at Bristol made scrutiny of data inevitable. Last month, using the Freedom of Information Act, the Guardian published death rates of every cardiac surgeon in the UK.

Today a team from the University of Birmingham demonstrate in the British Medical Journal how episode statistics (routine data on what happens to every patient) could identify gynaecologists with higher rates of complications and long-stay patients than others. Such data might indicate they take sicker patients, the team say, but could also identify the likes of Mr Ledward.

Mike Harley and colleagues say it is not feasible to quantify gynaecologists by death rates, because deaths are rare; it is possible to consider outcomes from episode statistics: percentage of cases with complications, length of stay, and cases requiring more than one operation.

The team collected data for 143 consultants: Rodney Ledward, three colleagues at his Kent hospital, and all those from the West Midlands. The data, made anonymous, covered 1991-92 to 1995-96, when he was suspended.

Mr Ledward was indeed an "outlier" - his performance was not in line with that of the majority - in three out of five years. Eight other consultants were also outliers - two in all five years - but the researchers say they do not know whether they are poor performers or have a lot of difficult cases.