The NHS Information Centre has been forced to amend the new summary hospital mortality indicator following complaints from the members of the steering group set up agree a consensus on measuring mortality rates, HSJ has learned.

The information centre, which has been leading the indicator’s development, published the latest version of the SHMI methodology on its website at the end of July, claiming it had been “completed” and approved by the technical group.

However, the group had not seen the methodology, developed by Sheffield University’s School of Health and Related Research, prior to publication. Following the complaints the information centre launched a formal two week consultation.

The consultation resulted in criticism of the methodology for removing some maternity and gynaecological episodes from the calculation, in an attempt to remove still births, and removing all deaths within a day of admission in an attempt to discount patients dead on arrival. It has also been claimed the new methodology produces vastly different results to its predecessor.

The group, led by Ian Dalton, chief executive of NHS North East, included representatives from the Academy of Royal Colleges, the Care Quality Commission, Dr Foster Intelligence and CHKS.

It was set up reach a high level agreement on disputes about how far mortality rates can be used as an indicator of a hospital’s quality.

The group agreed to include all deaths, rather than the 80 per cent covered by the hospital standardised mortality ratio developed by Dr Foster; to measure deaths within 30 days of discharge and remove palliative care coding which has led to accusations of gaming.

University Hospitals Birmingham Foundation Trust, which was also involved in the group, sent out a SHMI score to all trusts in July based on the methodology as it existed at that time, a move branded “unhelpful” by other steering group members.

Trust executive medical director David Rosser told HSJ it was preparing to send out new figures based on the latest methodology to give trusts time to prepare for their publication. He said initial calculations showed nearly half of hospitals were ranked differently under the new methodology.

A third version of the methodology is due to be published before September. While the information centre stresses this can be developed further it will be used to calculate the SHMIs that will be made public in October.

An information centre spokeswoman said the body had “acted quickly” to “make it clear” feedback was expected and to avoid the original documents’ “impression that we were relying unduly on the Sheffield work at the expense of the work of the technical group”.