Each year the number of deaths in every hospital in England is recorded and compared with national averages for the range of patients and conditions treated. The results are published by a company called Dr Foster in The Hospital Guide. The Guide has a solid reputation. Its findings are studied and used by leaders of the NHS. Dr Foster's statistical expert says that high mortality statistics should act as a 'smoke alarm' raising investigation of standards at a hospital. The Care Quality Commission praises Dr Foster's "powerful analysis of hospital trusts" and the Health Secretary says: "We expect all hospitals to examine this data carefully and take action wherever services need to improve". But some leading statisticians question the reliability of mortality statistics as an indication of clinical quality. And they believe that many pockets of poor practice go undetected in hospitals with good mortality scores. Critics also see the publication of such data as an invitation to the press to distort the available evidence by calculating numbers of 'needless deaths' within the NHS. Such calculations have in fact been produced and then given widespread publicity. The NHS Medical Director calls them "clinically meaningless and academically reckless". But they continue to make the front pages. Gerry Northam reports from hospitals which have "worryingly high" mortality statistics according to Dr Foster and asks how much this really shows about their quality of care. Producer : Ian Muir-Cochrane Editor : David Ross.