The heart is the matter - feeding premature babies.
New guidelines on the intravenous feeding of sick premature babies were issued yesterday after a spate of deaths.
The Government's Chief Medical Officer, Professor Liam Donaldson, said doctors should stop the widespread practice of inserting feeding tubes into babies' hearts.
In future they will be expected to adopt an alternative, but also risky, method of placing the tube in a large vein attached to the heart.
Prof Donaldson ordered a review of the procedure after four gravely ill babies died in paediatric intensive care units in the Greater Manchester area.
All had been kept alive by having nutrients injected straight into their bloodstream via a catheter pushed into the heart.
But in each case there was a fatal complication. The babies died from a condition called cardiac tamponade - a build up of fluid in the heart which eventually stops it beating.
The new guidance says that in the absence of more supportive evidence, catheter tips should no longer be placed inside the heart chambers.
Medical opinion is divided over whether it is safer to insert catheters into a baby's heart or into the vena cava, the large vein that returns blood to the heart.
But experts agree that sometimes intravenous feeding is the only way premature babies can be kept alive.
This was spelled out by a health service chief last November at the Stockport inquest on one of the dead babies, ten-day-old Daniel Carpenter.
Dr Chris Burke, chief executive of Stockport NHS Trust, said: 'Inserting a catheter into a baby's heart or the vein that leads into the heart is the only way to feed nutrients to such premature babies who cannot feed normally.'
The review was carried out by consultant paediatrician Dr Unni Wariya, from the Royal Victoria Infirmary, Newcastle, and consultant anaesthetist Dr David Hallworth, from the Royal Hospital for Sick Children in Glasgow.
Among 14 recommendations it called for a wide-ranging audit to look at complications that can result from all methods of feeding babies in hospital.
Dr Wariya explained the difficulties inherent in the procedure. 'The size of the heart of these babies is not much bigger than a walnut.'
Many experts believed it was safer to insert the catheter into the heart, he added.
The high rate of blood flow quickly diluted the feeding solution and reduced the chances of it causing corrosive tissue damage.
But targeting the heart carried with it the risk of fluid leaking into the cavity beneath the heart membrane and causing cardiac tamponade.
The complication was extremely rare, occurring in between 0.5 per cent and two per cent of the 1,600 procedures of this type carried out each year.
Dr Hallworth said none of the clinicians involved had ever experienced it before.
The Manchester cases, which all occurred at specialist centres, had been spread over a period of two years.