Friday 27 November 2009

Can we trust the data on hospitals?

By Emma Wilkinson
Health reporter, BBC News

Cleaning in a hospital
Inspections focus heavily on cleaning

The details of what was uncovered during an inspection in October at Basildon NHS Trust in Essex are shocking.

Floors, curtains and equipment stained with blood; soiled and stained mattresses; and re-use of tubes and other items designed to be thrown away after each patient.

It is the second time in less than a year that standards at a foundation trust - a status awarded to strongly performing NHS institutions - have been found wanting.

"Appalling" standards at Mid-Staffordshire NHS Foundation Trust were exposed in March when it was estimated that 400 patients had died needlessly over a three-year period.

The pattern is remarkably similar - an investigation initially prompted by data showing higher than expected death rates - and will further shake public confidence in how hospitals are monitored.

Especially as league tables published by the Care Quality Commission just last month gave Basildon an overall "Good" rating for quality of services.

It immediately poses the question, can the ratings be trusted and what does that mean for patient choice?

Regulation

The Care Quality Commission is a fairly new body having taking over the responsibility for the monitoring of hospitals from its predecessor the Healthcare Commission eight months ago.

"We need a much more sophisticated system that doesn't just rely on data but takes information from real inspections on the ground
Baroness Young
Care Quality Commission

CQC chair Baroness Young says that they inherited a rating system that is not fit for purpose and needs to be scrapped.

"We did make it very clear that this is not the way we want to regulate for the future.

"We need a much more sophisticated system that doesn't just rely on data but takes information from real inspections on the ground."

The "Good" rating given to Basildon hospital relates to an earlier time period - the year up until April 2009 - and there will be variation within each hospital, which makes the "single-word summary" misleading, she adds.

The CQC, she says, will have greater power and resources to carry out spot checks and inspections, rather than just relying on tick-box exercises which some argue enable trusts to fudge the data.

"By this time next year there will be a completely different system which will be much more able to spot issues, nip them in the bud, and take prompt action, rather than reporting months and months later."

Accurate data

The jury seems to be out on whether figures such as death rates are reliable indicators of poor performance - although in these two cases they started the alarm bells ringing.

Data from Basildon NHS trust showed a death rate 36% above the national average for the past three years.

When initially questioned, the trust had said the data was misleading - the same response given by Mid-Staffordshire when asked to account for higher than expected deaths.

Professor Sir Brian Jarman, who has developed systems for monitoring death rates - known as hospital standardised mortality ratios - and other data on a monthly basis, says the whole idea is to adjust for factors, which may affect the figures such as an unusually high elderly population.

He believes the data can be relied upon and that Basildon, along with some other hospitals, had higher than average rates since they started collecting the figures a decade ago but initially people had not taken any notice of the information.

"We now have a very good relationship with the CQC and they do listen closely."

"What we have found in virtually every hospital we have drawn attention to is there are usually dramatic reductions in the figures.

"Even in Basildon - from April this year they are down to normal."

Complaints

But Kieran Mullan, director of policy and public affairs at the Patients Association says death rates are a very "blunt tool" which only highlights problems once they get severe.

He is also critical of CQC plans to change the rating system.

"The bulk of it will still be data driven self-assessment, that's not going to change.

"It will just be a form where the trust can tick, yes we do this and then it's up to chance whether the CQC will choose your hospital to look at more closely.

"They say they will do more going into wards and talking to patients but the framework is basically the same."

A far simpler way to monitor whether hospitals are up to scratch is to pay closer attention to and follow up patient complaints, he says.

Almost every complaint they receive has already been logged with the trust and therefore accessible to the regulator.

"If you treat a complaint seriously we wouldn't need anything else because patients tell you when something has gone wrong."

Featured post

More patients in Scotland given antidepressants

More patients in Scotland given antidepressants 13 October 2015   From the section Scotland Image copyright Thinkstock Image ca...