News

Company denies that its coding advice is aimed at improving mortality data

BMJ 2013; 346 doi: http://dx.doi.org/10.1136/bmj.f1488 (Published 05 March 2013) Cite this as: BMJ 2013;346:f1488
  1. Nigel Hawkes
  1. 1London

The health analytics company CHKS has denied that it sought to help hospitals improve their mortality figures by advising changes in coding practices.

The charge was implied in evidence given to the public inquiry into Mid Staffordshire NHS Foundation Trust by Brian Jarman, director of the Dr Foster Unit at Imperial College London, who presented as part of his evidence a press release from CHKS dated 11 December 2008.

In it CHKS recorded that advice it had given to Medway NHS Foundation Trust about coding hospital care at the end of life would have the effect of reducing the trust’s hospital standardised mortality ratio (HSMR) by just over a third, adding that “most hospitals would consider a reduction of 5 per cent a good achievement.”

Under the headline “Hospitals ‘pay private company to hide death rates’” the Sunday Times on 3 March accused hospital trusts of using taxpayers’ money to cover up high death rates by hiring CHKS to reclassify the data.1 The story quoted Jarman, the originator of HSMRs,2 as agreeing that the changes to the way deaths were classified would have covered up problems for a while. “It would have done,” he is quoted as saying. “Correct, it would have had that effect.”

CHKS said that its advice was not designed for that purpose but aimed simply to put right errors in Medway’s coding for palliative care. “CHKS auditors follow Connecting for Health [the NHS’s information technology agency] coding rules,” said Jason Harries, the company’s managing director.

“If, for example, CHKS is recommending coding for palliative care, it is because specialist or generalist palliative care is recorded in the clinical notes and the coders have missed it, not because it improves mortality rates.”

So CHKS’s position is that it did give coding advice to Medway, but not with the intention of affecting its mortality rate data. What the hospital did with that advice was not the company’s responsibility.

In the wake of the Francis report into Mid Staffordshire the coding issue has come to the fore. Bolton Clinical Commissioning Group has asked Dr Foster, a rival to CHKS that markets analytical data based on Jarman’s HSMRs, to investigate changes in coding at Bolton NHS Foundation Trust that led to a rise in deaths there being coded as caused by septicaemia at the same time as the hospital’s mortality ratio showed improvements. Potentially, by coding more deaths as being caused by septicaemia, a hospital could increase its number of “expected deaths.” Because HSMRs are the ratio of actual to expected deaths, this could improve the figure, assuming that actual deaths remained unchanged.

CHKS said that it was not working with the Bolton trust over the period in which its mortality ratio improved.

Notes

Cite this as: BMJ 2013;346:f1488

References