Interest in hospital mortality indicators will “wane” within three years as clinicians produce dozens of service-specific quality measures, the NHS medical director has predicted.

Sir Bruce Keogh’s comments follow the first results from the government-backed whole hospital death rate measure which had originally been designed to bring consensus to this contentious area. The summary hospital-level mortality indicator singled out 14 trusts as outliers with high numbers of unexpected deaths and another 14 trusts with low rates. Whittington Hospital Trust in north London had the lowest (see list, bottom of page)

The indicator was developed by a group of experts and interested groups commissioned by the Department of Health-hosted national quality board. Results will now be published quarterly and it is likely its methodology will change in response to feedback.

Sir Bruce told HSJ there would soon be several measures of hospital-wide mortality, all with different results.

The NHS Information Centre has published two ways of interpreting the SHMI; Dr Foster is expected to publish two different hospital-wide measures next month; data analyst CHKS already has a measure; and University Hospitals Birmingham Foundation Trust is developing two further indicators.

Sir Bruce told HSJ the variety would “help stop the knee-jerk response to one single figure and encourage people to think about the component parts that make up the figures”.

He predicted that within two or three years the development of service-specific quality indicators would have advanced enough for little attention to be paid to hospital-wide measures.

Sir Bruce said: “When everybody is measuring quality in each and every service line, the utility of single mortality measures will wane.”

He said he supported the new revalidation process, designed to ensure doctors remain competent, considering how well medics were using quality measures to analyse the care they provided. This could speed up the development of measures. Sir Bruce said: “The essence of professionalism is being able to describe what you do and how well you do it.”

Media coverage of the new indicator was less dramatic than the outcry over Dr Foster Hospital Guide mortality rates in recent years, but national and regional press raised concern about outlier trusts.

Nine outliers said their high figure was due to clinical coding rather than care quality. Four trusts said they were rated high because the indicator does not take into account whether patients are admitted to receive palliative care.

York Teaching Hospital Foundation Trust medical director Alastair Turnbull believed his trust’s rating had been distorted by the accidental inclusion of data from a hospice. The hospice is based on the same site and uses the same software, but is not run by the trust. “The way the SHMI was published made it extraordinarily difficult to validate the data in the short time available. To be honest, I think we have been a victim of that,” he said.

“I feel very strongly about that because our organisation has been portrayed in a way that doesn’t reflect [the reality].”

However, George Eliot Hospital Trust in Warwickshire – which had the largest gap between expected and observed deaths under the SHMI – said it was taking the rating “extremely seriously” and had commissioned an external review.

British Medical Association consultants committee chair Mark Porter said the measures should not be used in isolation. But he added: “It’s not an appropriate option to say there must be something wrong with statistics. As doctors working for the NHS we have an absolute obligation to look at these statistics.”

King’s Fund senior fellow Veena Raleigh, a member of the group that developed the new indicator, said the impact of palliative coding was a “valid issue”.

She said: “The SHMI is a useful global measure… looking ahead, we should be moving towards using richer clinical audit data and more precise measures.”

Higher than expected deaths

  • George Eliot Hospital Trust (highest)
  • Isle of Wight PCT
  • East and North Hertfordshire Trust
  • Blackpool Teaching Hospitals Foundation Trust
  • Tameside Hospital Foundation Trust
  • Medway Foundation Trust
  • York Teaching Hospital Foundation Trust
  • Northern Lincolnshire and Goole Hospitals Foundation Trust
  • Basildon and Thurrock University Hospitals Foundation Trust
  • Hull and East Yorkshire Hospitals Trust
  • Northampton General Hospital Trust
  • East Lancashire Hospitals Trust
  • University Hospitals of Morecambe Bay Foundation Trust
  • Western Sussex Hospitals Trust

Lower than expected deaths

  • Whittington Hospital Trust (lowest)
  • Barts and the London Trust
  • University College London Hospitals Foundation Trust
  • Imperial College Healthcare Trust
  • Royal Free Hampstead Trust
  • St George’s Healthcare Trust
  • Chelsea and Westminster Hospital Foundation Trust
  • Cambridge University Hospitals Foundation Trust
  • Newham University Hospital Trust
  • North West London Hospitals Trust
  • Sheffield Teaching Hospitals Foundation Trust
  • Kingston Hospital Trust
  • James Paget University Hospitals Foundation Trust
  • Ealing Hospital Trust

Source: NHS Information Centre