AN ELECTRONIC nose that sniffs out infections could help hospitals tackle outbreaks of the antibiotic-resistant superbug MRSA.

Culture tests routinely used to identify MRSA (methicillin-resistant Staphylococcus aureus ) take two or three days to complete. This hampers attempts to manage outbreaks as infected patients remain untreated and at risk of infecting others.

DNA-based tests are being trialled that promise to reduce test times to 2 hours, but now UK-based researchers have come up with a test using an electronic sniffer that could cut the time further, to just 15 minutes. Writing in the journal Sensors and Actuators B (vol 109, p 355), engineers at the University of Warwick and doctors at the Heart of England Hospital, Birmingham, say the electronic nose can recognise the unique cocktail of volatile organic compounds that S. aureus strains excrete.

E-noses analyse gas samples by passing the gas over an array of electrodes coated with different conducting polymers. Each electrode reacts to particular substances by changing its electrical resistance in a characteristic way. Combining the signals from all the electrodes gives a "smell-print" of the chemicals in the mixture that neural network software built into the e-nose can learn to recognise.

Each e-nose is about the size of a pair of desktop PCs and costs about £60,000. The food industry uses similar machines to root out rotten ingredients.

David Morgan, a surgeon at the hospital, says the idea of sniffing out superbugs came to him one day in the operating theatre. "I was operating on neck abscesses on two different patients and noticed their infections had slightly different smells, so I wondered if a machine could work out what the bacterial infections were from the smell alone."

Morgan approached Warwick engineer Ritaban Dutta and his colleagues to develop the idea. They first trained an e-nose to recognise the smell-prints of MRSA and the related but more easily treated MSSA (methicillin susceptible S. aureus ) by exposing them to nasal swabs from people carrying the infection. They then put their e-nose to the test using swabs from 150 patients whose infection status was already known from culture tests. The system correctly detected 96 per cent of those who had an S. aureus infection.

The e-nose system cannot yet distinguish the methicillin-resistant superbug from the methicillin-susceptible strain. If it cannot be trained to do so, Morgan suggests the e-nose be used as a quick screening system to prioritise which patients or healthcare workers are given the 2-hour DNA-based test, which can tell the difference.

The UK government's Department of Health says that while the panel which investigates new infection controls will study the e-nose, hopes for tackling MRSA remain focused on the emerging DNA-based tests.

Morgan is also using an e-nose to look for early signs of pneumonia in intensive care patients.

Journal reference: Sensors and Actuators B (vol 109, p 355)