Pathogen detection is essential without time delay especially for severe diseases like sepsis. Here, the survival rate is
dependent on a prompt antibiosis. For sepsis three hours after the onset of shock the survival rate of the patient drops
below 60 %. Unfortunately, the results from standard diagnosis methods like PCR or microbiology can normally be
received after 12 or 36 h, respectively. Therefore diagnosis methods which require less cultivation or even no cultivation
at all have to be established for medical diagnosis. Here, Raman spectroscopy, as a vibrational spectroscopic method, is a
very sensitive and selective approach and monitors the biochemical composition of the investigated sample. Applying
micro-Raman spectroscopy allows for a spatial resolution below 1 μm and is therefore in the size range of bacteria.
Raman spectra of bacteria depend on the physiological status. Therefore, the databases require the inclusion of the
necessary environmental parameters such as temperature, pH, nutrition, etc. Such large databases therefore require a
specialized chemometric approach, since the variation between different strains is small. In this contribution we will
demonstrate the capability of Raman spectroscopy to identify pathogens without cultivation even from real
environmental or medical samples.