Sunday, 16 June 2013

Effective Data Transparency: Cloudier Than You Might Think
Data transparency, the development of a safe and open culture of data sharing between clinicians and patients, is a noble aspiration. I have no doubt that it will improve the quality of data collection and lead to improvements in clinical care. Tim Kelsey, NHS National Director for Patients and Information, says that data transparency is the future of the people's NHS (1). It is one of the core strategic principles of NHS England. As part of this drive to improve data transparency, outcome data for nine surgical specialities AND interventional cardiology are being published this summer. An important issue that has not been discussed is how this data will be presented and interpreted by the public.

Jeremy Hunt says that patients will see which surgeons apparently have the best outcomes (2). Patients will be able to choose the best surgeons. Really? Cardiac surgery is cited as an example of how this model can drive improvement. Cardiac surgeons have published their outcome data since 2004 and mortality rates have apparently dropped to half those in Germany and a fifth of those in Portugal. What is often not discussed is how this may have led to risk avoidance amongst some surgeons.

MACCE (major adverse cardiovascular and cerebrovascular events) data for all interventional cardiologists in England will be published shortly. This is not a bad thing. However, we should exercise caution when looking at the data and bear in mind the effect this may have on many cardiologists. BCIS (British Cardiovascular Interventional Society) explicity states that you CANNOT compare the performance of cardiologists on the basis of the data. This is unlikely to be the view of government and the public.

For one thing, the data does not discriminate between cardiologists working in high volume heart attack centres and those only doing planned (elective) cases. Cardiologist A in a high volume heart attack centre will have a high volume of complex emergency work (usually 1/3rd of his/her cases) of which a proportion will be of extremely high risk. Patients following cardiac arrest who have been intubated and ventilated (and possibly on a Lucas chest compression device) have a high risk of death even if their angioplasty is successful, but are not considered separately in the data. Cardiologist B may only perform low risk elective cases or avoid taking high risk patients to the operating theatre (known as a catheterisation laboratory or 'cath lab'). While Cardiologist A has an elective mortality of 0%, his/her overall mortality may be 5%. Cardiologist B may have an elective mortality of 1%, and yet appear better overall.

Imagine the following clinical scenario (imagined but based on cases we frequently treat). A 50 year old man has a cardiac arrest at home and receives CPR from his wife. An ambulance crew arrives within 5 minutes and shocks him out of ventricular fibrillation. He is intubated on the scene, airlifted to a Heart Attack Centre by an air ambulance where a specialist interventional cardiology team reopens his blocked coronary artery within 15 minutes of him landing on top of the hospital.

The interventional cardiologist who takes full responsibility for treating this man considers many factors. One, that he may already be brain dead and that saving his heart may not save his life. Two, although the situation appears futile, there is a real chance that this man’s life may be saved. To adopt a non-interventional treatment strategy is to guarantee death (3). Three, the best chance of saving this man’s life depends on an immediate angioplasty and cooling. This is evidence based. Four, the man’s wife, an ambulance crew and an air ambulance crew have fought hard to this point to save his life. Is it fair that the cardiologist may deny potentially life-saving interventional treatment when the patient arrives at the door of his/her 'cath lab'?
Now there is a further consideration, namely the effect that taking on this extremely high risk case may have on the cardiologist's published outcome data. Many of us will swiftly put this consideration to one side and get on with the urgent job in hand.

Data transparency and publication of outcomes are here to stay in a patient centred NHS. But as patients use complex data to make comparisons between cardiologists, will some cardiologists begin to avoid high risk emergency work? How do we help patients to navigate the jungle of data they will be faced with?

Vinod Achan is a Consultant Cardiologist and Lead Consultant for Primary Angioplasty at the Surrey Heart Attack Centre, Frimley Park Hospital