Wednesday, 27 November 2013

Comparing hospital data worldwide will drive up quality

We learned recently that it costs about £3,700 to deliver a child in the UK.

In a new report, the National Audit Office revealed that the NHS has spent nearly a fifth of its maternity services budget covering compensation costs paid out when errors or negligence result in the death or injury of an infant. Margaret Hodge, chair of the Public Accounts Committee called it "scandalous". The Department of Health replied that the UK was one of the safest places in the world to have a baby; but the National Audit Office said that, although the situation is complicated, this "does not represent good value for money".

So who is right? To judge whether or not our maternity services are as safe as they should be and as cost-efficient as possible, we would have to compare them with the way in which other countries manage pregnancy and birth. Until recently, there were few direct comparisons of clinical outcomes at hospital level between various healthcare systems across international borders.

However, a new project means that staff across a wide range of disciplines will now be able to directly compare what happens inhospitals abroad with what is happening in the UK. Eleven hospitals in the UK have joined Dr Foster Intelligence's global comparators project, an international collaboration which so far has allowed 40 hospitals across Europe, Australia and the US to share data.

It is the first time that chief executive officers, chief medical officers and lead clinicians in world-class hospitals have been able to share cost and quality metric data on a global scale. The project empowers healthcare professionals to analyse caseloads similar to their own, drilling all the way down to patient-level data. Members can compare mortality, length of stay, re-admission and complication rates across 259 clinical diagnoses and 38 clinical procedure groups, all across national borders.

University hospitals Coventry and Warwickshire (UHCW) is one of the project participants. It was able to compare its performance on heart attack patients with Yale New-Haven hospital in the US and saw that there could be room for improvement.

UHCW found its ambulance crews were only alerting the cardiology team as they were approaching the hospital, giving staff little time for preparation. Now, the ambulance crews trigger an alert as soon as they make the diagnosis of heart attack on the ECG, so that the cardiology team can be waiting for the patient to arrive in the emergency department. By studying how patients were managed in Yale New-Haven and adopting many of their methods, UHCW has achieved a 25% reduction rate in emergency patient mortality.

The first tentative findings from the global comparators project are now beginning to appear in academic journals. US hospitals are found to have had the lowest in-hospital mortality rates; but that is partly due to shorter lengths of stay. When mortality rates for stroke patients were compared including discharged patients, previous differences disappeared. US hospitals also had a higher 30-day re-admission rate than other participating hospitals – 9.4% compared with 6.6% in England and 4.9% in the other hospitals.

The Australian state of Victoria recently decided it would use these tools to understand healthcare performance across all of its hospitals, following the Australian government's announcement of a big data strategy. It is increasingly being realised that data itself means nothing if it's not trusted and acted upon – it's of no use being gathered but then sitting in a filing cabinet or on the cloud.

The use of data to drive improvements in performance is still in its relative infancy in the healthcare sector, but projects such as global comparators are setting the standard for what is achievable.

Thanks to the availability of data, patients around the world will continue to experience ever-greater improvements in the quality of care they receive.

This article is courtesy from The Guardian.

No comments:

Post a Comment