NHS data reveals ‘huge variation’ in Covid-19 death rates across England

“The five trusts with the highest death rates are in the south-west (80%), north-west (68%), south-west (62%), east of England (60%) and London (54%).”

Analysis such as this may make uncomfortable reading, and it may not be perfect, but Owl has consistently argued for more analysis of emerging data. Readers also need to be reminded that the “south-west” referred to above is the NUTS definition extending eastwards to Gloucestershire and Wiltshire.

Niamh McIntyre www.theguardian.com 

A wide disparity in coronavirus mortality rates has emerged in English hospitals, with data seen by the Guardian showing that one hospital trust in south-west England had a death rate from the disease of 80% while in one London trust it was just 12.5%.

The figures, which NHS England has compiled but never published, show the age-standardised mortality rates that all of the country’s 135 acute hospital trusts have recorded during the pandemic. Doctors regard age as the single biggest predictor or risk factor for dying from Covid-19.

They cover the period from the start of the coronavirus crisis in March, through its peak in late March and April, up until 15 May, by which time 42,850 – or 85% – of the 50,219 deaths so far in all settings had occurred in England and Wales.

It is the first such data to emerge about how many people have lived or died in each trust after being treated there because they had been left critically ill by the disease. They are based on patients who were treated in an intensive care or high-dependency unit or on a ward.

Senior doctors said the dramatic gap in death rates of 67.5 percentage points between the trusts with the highest and lowest rates was notable and may mean that some hospitals needed to learn lessons from others.

“That is a huge variation, a huge range. I’m surprised at the degree of variation. A spread between 12.5% and 80% is quite stark,” said Dr Alison Pittard, the dean of the Faculty of Intensive Care Medicine. It represents the intensive care specialists who have played the lead role in treating what is now more than 100,000 people hospitalised in England with Covid-19.

Dr Nick Scriven, a former president of the Society for Acute Medicine (Sam), said: “The range does look larger than you would expect and should prompt further analysis and thought as to why this may appear as it does, which for the general population will be concerning.”

However, both Pittard and Scriven cautioned that the data did not give a full picture of differential death rates between hospitals because it did not take account of four other key factors for risk of death from Covid-19, namely gender, ethnicity, deprivation and underlying health problems. All four have been found to significantly increase a patient’s chances of dying.

NHS England has plotted each trust’s death rate, and the number of patients with Covid-19 each of them admitted, on a graph which it has shared with some senior doctors. Crucially, though, it has not identified the trusts on it. It has only disclosed which of the NHS’s seven regions the trust is in.

The five trusts with the highest death rates are in the south-west (80%), north-west (68%), south-west (62%), east of England (60%) and London (54%).

The five trusts with the lowest death rates are in: London (12.5%), Midlands (13%), London (14%), London (15%) and the south-east (15%).

Doctors pointed out that some trusts’ apparently high mortality rates could be skewed because they were based on them having treated fewer than 100 patients by 15 May, which makes their rates less reliable. But other trusts with notably high or low mortality had treated up to 2,350 patients over the same period, so their rates are more likely to be reliable.

Research published by the Intensive Care National Audit and Research Centre shows that of 9,995 patients treated in intensive care units with known outcomes, 5,985 (59.9%) have been discharged and 4,010 (40.1%) have died, while 426 others are still receiving critical care.

Mortality among such patients in intensive care has improved from 50% early in the pandemic to 41% now, reflecting in part medical teams’ better understanding and treatment of Covid-19.

There is no suggestion that a high or low death rate indicates that patients have received a worse or better standard of care at any particular hospital. The makeup of the local population that a trust serves is the single biggest factor underlying a high or low death rate, Pittard said.

“We know that poorer communities and BAME communities have a higher risk of mortality, so if a trust is in an area of higher-risk individuals you woud expect that trust’s Covid-related mortality rate to be higher,” added Pittard.

NHS England’s data shows that 26 of the 135 trusts had a death rate between 12.5% and 25%. “That’s reassuring, absolutely. It’s positive that 26 trusts had that low mortality,” said Pittard.

However, at least half of patients treated at 11 trusts died. “Eighty percent does seem an extremely high number,” she added.

Dr Sue Crossland, the president of the Sam, said hospitals’ use of non-invasive ventilation, the early involvement of critical care teams and lying ventilated patients on their front – “proning” – have saved patients’ lives as doctors have better understood how to deal with Covid-19.

An NHS spokesperson said: “We do not recognise these figures, which appear to be experimental analysis of unverified data. But there is now a wide range of published data on the role that health inequalities, including pre-existing conditions and other health factors, have played when it comes to the impact of Covid, including from the ONS and Public Health England.

“The NHS is accelerating work to tackle health inequalities, and will shortly be providing local services with a range of actions they should build in to their plans for the coming months.”