Evaluation of PCR in the molecular diagnosis of endocarditis

Objective. Infective endocarditis (IE) is diagnosed by the Duke criteria, which can be inconclusive particularly when blood cultures are negative. This study investigated the application of polymerase chain reaction (PCR) to identify bacterial DNA in excised valvular tissue, and its role in establishing the diagnosis of IE. Methods. Ninety-eight patients undergoing valve replacement surgery were studied. Twenty-eight patients were confirmed as definite for endocarditis by the Duke criteria; nine were considered as possible and 61 had no known or previous microbial infection of the endocardium. A broad-range PCR technique was used to amplify prokaryotic 16S rRNA genes present within homogenised heart valve tissue. Subsequent DNA sequencing of the PCR amplicon allowed identification of the infecting microorganism. Results. PCR results demonstrated the presence of bacterial DNA in the heart valves obtained from 14 out of 20 (70%) definite IE patients with positive blood cultures preoperatively. The causative microorganism for one patient with definite culture negative endocarditis was identified by PCR. Two out of nine (22%) of the valves from possible endocarditis patients also had bacterial DNA present converting them into the definite criteria whereas in the valves of seven out of nine (78%) of these patients no bacterial DNA was detected. Conclusion. The application of PCR to the explanted valves in patients with possible or confirmed diagnosis can augment the Duke criteria thereby improving post-surgical antimicrobial therapeutic options. © 2003 The British Infection Society. Published by Elsevier Ltd. All rights reserved.

The prevention and diagnosis of central venous catheter-related infection

The potential source of CVC colonisation was assessed. Isolates of coagulase-negative staphylococci (CoNS) recovered from the skin and CVC components of 3 cardiothoracic surgery patients were characterised by pulsed-field gel electrophoresis (PFGE). The genetic heterogeneity of CoNS isolated from the skin was demonstrated and specific genotypes implicated in catheter colonisation. In addition, phenotypic and genotypic typing techniques were assessed for their ability to characterise strains of CoNS recovered from 33 patients who developed catheter-related bloodstream infection (CR-BSI) on a bone marrow transplant (BMT) unit and Siaphylococcus aureus recovered from 6 cardiothoracic surgery patients with surgical site infection (SSI) following median sternotomy. This epidemiological investigation revealed that common strains of CoNS and 51 aureus where not associated with infection in patients with CR-BSI or sternal SSI during the study period. Furthermore, there was no correlation between phenotypic and genotypic characterisation results. The variable expression of phenotypic traits within strains of staphylococci was evident whilst PFGE and randomly amplified polymorphic DNA (RAPD) were highly discriminatory for the molecular characterisation of S. aureus and CoNS. This was highlighted in 8 stem cell transplant (SCT) patients whereby it was demonstrated that routine identification and characterisation of CoNS by phenotypic techniques may not be adequate for the diagnosis of CR-BSI by current guidelines. The potential of the lipid S ELISA to facilitate the diagnosis of CR-BSI in 38 haematology/SCT patients and sternal SSI in 57 cardiothoracic surgery patients was also assessed. The ELISA proved to be a sensitive test for the rapid serodiagnosis of infection due to staphylococci in immunocompetent patients. The acridine orange leucocyte cytospin test (AOLC) was also evaluated for the rapid diagnosis of CR-BSI in 16 haematology/SCT patients with Hickman CVC in situ. Although the sensitivity of the test was low, it may provide a useful adjunct to conventional methods for the in situ sampling of catheters to predict and diagnose CR-BSI, preventing the unnecessary removal of CVC.

Chlorhexidine and the prevention of surgical site infection

Surgical site infections (SSI) are a prevalent health care-associated infection (HAl). Prior to the mid-19th century, surgical sites commonly developed postoperative wound complications. It was in the 1860's, after Joseph Lister introduced carbolic acid and the principles of antisepsis that postoperative wound infection significantly decreased. Today, patient preoperative skin preparation with an antiseptic agent prior to surgery is a standard of practice. Povidone-iodine and chlorhexidine gluconate are currently the most commonly used antimicrobial agents used to prep the patient's skin. In this current study, the epidemiology, diagnosis, surveillance and prevention of SSI with chlorhexidine were investigated. The antimicrobial activity of chlorhexidine was assessed. In in-vitro and in-vivo studies the antimicrobial efficacy of 2% (w/v) chlorhexidine gluconate (CHG) in 70% isopropyl alcohol (IPA) and 10% povidoneiodine (PVP-I) in the presence of 0.9% normal saline or blood were examined. The 2% CHG in 70% IPA solutions antimicrobial activity was not diminished in the presence of 0.9% normal saline or blood. In comparison, the traditional patient preoperative skin preparation, 10% PVP-I antimicrobial activity was not diminished in the presence of 0.9% normal saline, but was diminished in the presence of blood. In an in-vivo human volunteer study the potential for reduction of the antimicrobial efficacy of aqueous patient preoperative skin preparations compromised by mechanical removal of wet product from the application site (blot) was assessed. In this evaluation, 2% CHG and 10% povidone-iodine (PVP-I) were blotted from the patient's skin after application to the test site. The blotting, or mechanical removal, of the wet antiseptic from the application site did not produce a significant difference in product efficacy. In a clinical trial to compare 2% CHG in 70% IPA and PVP-! scrub and paint patient preoperative skin preparation for the prevention of SSI, there were 849 patients randomly assigned to the study groups (409 in the chlorhexidine-alcohol and 440 in the povidone-iodine group) in the intention-to-treat analysis. The overall surgical site infection was significantly lower in the 2% CHG in 70% IPA group than in the PVP-I group (9.5% versus 16.1 %, p=0.004; relative risk, 0.59 with 95% confidence interval of 0.41 to 0.85). Preoperative cleansing of the patient's skin with chlorhexidine-alcohol is superior to povidone-iodine in preventing surgical site infection after clean-contaminated surgery.

Rapid serodiagnosis of Staphylococcus aureus surgical site infection following median sternotomy

Objectives: To determine the sensitivity and specificity of a novel ELISA for the serodiagnosis of surgical site infection (SSI) due to staphylococci following median sternotomy. Methods: Twelve patients with a superficial sternal SSI and 19 with a deep sternal SSI due to Staphylococcus aureus were compared with 37 control patients who also underwent median sternotomy for cardiac surgery but exhibited no microbiological or clinical symptoms of infection. A further five patients with sternal SSI due to coagulase-negative (CoNS) staphylococci were studied. An ELISA incorporating a recently recognised exocellular short chain form of lipoteichoic acid (lipid S) recovered from CoNS, was used to determine serum levels of anti-lipid S IgG in all patient groups. Results: Serum anti-lipid S IgG titres of patients with sternal SSI due to S. aureus were significantly higher than the control patients (P<0.0001). In addition, patients with deep sternal SSI had significantly higher serum anti-lipid S IgG titres than patients with superficial sternal SSI (P=0.03). Serum anti-lipid S IgG titres of patients with sternal SSI due to CoNS were significantly higher than the control patients (P=0.001). Conclusion: The lipid S ELISA may facilitate the diagnosis of sternal SSI due to S. aureus and could also be of value with infection due to CoNS. © 2005 Published by Elsevier Ltd. on behalf of The Bristish Infection Society.