Using real-time PCR to decrease time to appropriate management with staphylococcal blood cultures

Abstract number: R2202

Objectives: Blood cultures (BCs) are key in directing antibiotic therapy, often affecting the sickest patients in hospital. Staphylococci are the most common organism seen in Gram stain. Data from University Hospital Coventry and Warwick, June-August 2008, showed roughly 80% of these are contaminants. At the time of phoning the clinician to discuss an unwell patient, it can be difficult to differentiate a contaminant, a possible methicillin sensitive Staphylococcus aureus(MSSA) or methicillin resistant Staphylococcus aureus(MRSA). A more rapid test could:

1Reduce time spent by laboratory staff, microbiologists and clinicians working on and reviewing these results

2Prevent unnecessary anti-staphylococcal antibiotics commencing

3Allow earlier institution of optimal antibiotic therapy for MSSA/MRSA

All of this must be in conjunction with infection control teaching about correct technique for taking a BC to reduce the rates of contamination in the first instance.

Methods: A pre-study period for 1 month was observed before 100 sequential BCs were tested by GeneXpert Cepheid. This machine differentiates coagulase negative Staphylococci (CNS), MSSA and MRSA in 1 hour.

For each relevant Gram stain during these periods:

1Standard questionnaire was filled in: last MRSA screen, current antibiotics

2Advice given to clinicians based on clinical picture was documented, including changes made prior to PCR or culture results

3Consultant on duty was blinded from the PCR result

4PCR results reviewed and changes made to management as appropriate

5Time to review a single patient, contact clinicians and input data was assessed

Results: See tables.

There were 10 failures with the PCR kit (10%). The cause is not known. 1 mismatch of PCR result to culture results occurred. Repeat PCR concurred with the culture results. Dealing with initial Gram stains took 20minutes/culture on average.

Conclusion: Throughout the study a substantial number of BC were misinterpreted by microbiologists and clinicians. Up to 55% of MSSAs were not identified until further testing was obtained, a 24hour delay to best treatment. Time spent phoning contaminated BCs equates to 5.15PAs/month=£884 (average consultant salary). If all 71 BCs a month were tested at £29/ test=£2059. A saving of 1.7 bed nights/month would make this test cost efficient (Bed stay=£691/night) Importantly reducing the rate of contamination is key for best practice, which would in turn reduce the number of BCs that would require testing.