Government will still publish hospital standardised mortality ratios despite criticism

BMJ 2009; 338 doi: (Published 21 April 2009) Cite this as: BMJ 2009;338:b1641
  1. Zosia Kmietowicz
  1. 1London

    The Department of Health has said it will publish mortality ratios of patients in hospitals in England in the next few weeks despite research indicating that the measures do not help to distinguish between the quality of care provided at different hospitals.

    The department has said that the publication of hospital standardised mortality ratios on the NHS Choices website will allow doctors and the public to more easily compare the performance of local services. The site already carries mortality rates for heart surgery and four other procedures, including hip and knee replacements, for every hospital in England.

    But a study published online by the BMJ last month (2009;338:b7808, doi:10.1136/bmj.b780) criticised the use of the ratios for rating hospitals in England and other countries because their use distorted results.

    Although the figures were standardised, some of the variables, such as primary diagnosis and previous admissions in the past year, were of dubious value, said the researchers, led by Mohammed Mohammed, senior lecturer in the public health unit at the University of Birmingham. Other variables, such as whether patients were admitted as emergencies and whether they had other illnesses, were not safe to use because of different logging and admissions criteria around the country.

    Dr Mohammed said, “Any claims that variations in hospital SMRs [standardised mortality ratios] reflect differences in quality of care are less than credible. They should not be used without a massive warning on them.”

    Michael Summers, vice chairman of the Patients’ Association, said the move to publish the ratios was very welcome. But he added that a “more professional and more standardised publication would be of greater benefit to patients.”

    However, Bruce Keogh, the NHS’s medical director, defended the use of such ratios. A high hospital standardised mortality ratio (HSMR) was one of the indicators that had prompted the NHS watchdog, the Healthcare Commission, to investigate deaths at Mid-Staffordshire NHS Foundation Trust (BMJ 2009;338:b1141, 18 Mar, doi:10.1136/bmj.b1141).

    Professor Keogh said: “It is a moral and social duty for all healthcare organisations to know what they’re doing and how well they’re doing it. The HSMR is one of many measures that will help them do this, but it is not enough on its own. The HSMR is an aggregate measure of mortality for the organisation and hence a rather blunt, but useful, indicator of trouble.”

    He added that because of the controversy over the measure, an explanation about its strengths and weaknesses would appear on NHS Choices. An extra 250 “more focused measures of quality” that have been approved by the new National Quality Board for Health and Social Care would appear on the site in due course, said Professor Keogh.

    Last month, in its report into the deaths in the emergency department at the Mid-Staffordshire trust, the Healthcare Commission called for all NHS hospitals to be able to access reliable information on comparative mortality and other outcomes. The report concluded that 400 more people died than would have been expected at the hospital between 2005 and 2008 and that there were deficiencies at “virtually every stage” of emergency care.

    The health minister Ara Darzi promised more detailed information about how hospitals are performing in his review last year of the NHS, Quality Care for All, as a way to drive up standards (BMJ 2008;337:a642, doi:10.1136/bmj.a642).


    Cite this as: BMJ 2009;338:b1641

    View Abstract

    Sign in

    Log in through your institution

    Free trial

    Register for a free trial to to receive unlimited access to all content on for 14 days.
    Sign up for a free trial