Performance of the PREVI colour Gram automated staining system

Abstract number: P889

de Montclos H., Verdier I.

Gram stain is a critical stage of the microbiological diagnosis. Even though it is usually carried out manually, automation has become an alternative. Bacteriologists used to the manual method may, however, consider that automation allows to subtly adapt the decolorisation step to the thickness of the smear.

Objectives: A study was performed to evaluate the performance of the PREVI™ Color Gram system on pure strains and biological samples in comparison with the manual staining method. This new system ensures an automatic Gram staining of slides disposed on a carrousel. Dyes are sprayed on slides during rotation.

Methods: Manual and automatic stainings were carried out in parallel. Manual Gram stain was performed according to the conventional procedure. The Gram staining on PREVI Color Gram system was performed using the Decolorizer 2 and 3 programs. Slides were then mixed and read by a trained bacteriologist not knowing the bacterial identity and the staining type carried out on each slide.

The first part of the study consisted of the staining of 40 reference strains. Results were considered to be concordant when morphology, arrangement and Gram reaction of bacteria were identical. The second part of the study included 107 clinical specimens from various origins. Smears were randomly attributed to the manual or the 2 automated protocols. Preparations were considered to be concordant when the main bacterial populations were found in equivalent proportions.

Results: 321 preparations were compared, 8 of them were declared non concordant, but not significant as real discrepancies. These discrepancies involved 4 samples and were due to:

ainsufficient decolorisation by the manual technique,

bwashing out of the smear during manual staining,

cdifference in thickness of the smears in the case of a very mucoid expectoration, and

dstaining problem which could not be allocated to one or the other of the techniques.

The 2 automated protocols gave the same results.

The PREVI Color GRAM using the Decolorizer 3 provided 100% and 98% agreement with the manual method for reference strains and clinical samples respectively. Discrepancies analysis was never in discredit of the automatic technique.

Conclusion: Even using a manual standardised protocol, significant staining differences may persist between operators. The PREVI Color GRAM system provides distinct and reproducible results, thus contributing to stain standardisation, besides saving dyes and technician time.