Literature Update - June, 2007
Compiled by Dr Leon Worth, ID Physician
Australian Study of Interventions to Reduce MRSA Infection or Colonisation
Harrington GH et al (Infect Control Hosp Epidemiol. 2007; 28:837-844) reported the outcome of serial interventions
used at an Australian centre to reduce the incidence of MRSA acquisition and infection. In the ICU, an antimicrobial
hand-hygiene product (0.5% chlorhexidine and 70% v/v ethyl alcohol gel) was introduced, in addition to signage indicating
the isolation of ‘antibiotic-resistant organism’. After 10 months, an standard hospital alternative measures for hand hygiene
product was were used (4% chlorhexidine hand washing agent; 1% chlorhexidine and 70% v/v ethanol rub). After a further one month,
an alternative hand hygiene gel (61.5% v/v ethanol) replaced the standard hand rub. Throughout
the hospital, statistical process charts (SPCs) were used to provide feedback of MRSA colonisation and infection rates.
A 2-year pre-intervention period, followed by a 28-month intervention period were analysed.
The rate of new patients colonised or infected with MRSA was significantly reduced both in the ICU and hospital-wide.
The rate of central line-associated BSI due to MRSA in the ICU was also significantly reduced in the intervention period,
although overall rates were comparable in the pre- and intervention periods. Furthermore, hospital-wide episodes of MRSA
bacteraemia were significantly reduced in the intervention period (0.45 episodes per 100 patients admitted pre-intervention,
0.27 episodes per 100 patients admitted during the intervention period; p=0.02).
Of note, the studied outcome measures did not require active surveillance cultures, and the studied hospital did not employ
isolation measures for patients with known MRSA infection or colonisation. Although the impact of the individual interventions
is not able to be determined, findings suggest that multiple interventions using minimal resources are effective for reduction
of MRSA acquisition and infection, and that the employed strategies are sustainable.
VICNISS Validation Study of Victorian CABG Surgical Site Infection Surveillance Data
Friedman ND et al (Infect Control Hosp Epidemiol. 2007; 28:812-817) assessed the positive and negative predictive value of
surveillance data submitted to VICNISS for coronary artery bypass surgery (CABG) surgical site infections (SSIs) during a
6-month period. Using data captured by the Australasian Society of Cardiac and Thoracic Surgeons, the completeness of data
captured by VICNISS from January 2003 until March 2005 was also assessed.
The study found that the case ascertainment rate for CABG procedures performed in Victoria was 95%. During the study period,
Infection Control Consultants (ICCs) identified 34 patients with SSIs - 33 of these were identified by the study reviewers,
with agreement in 31 cases (91%). Of note, 15 patients were reported as having deep or organ space sternal SSI, with a
discrepancy in reports of SSI depth for 9 of 15. A higher rate of discrepancy was evident for donor SSIs. Both ICCs and
reviewers identified 18 patients with 1 or more donor site SSI. However, ICCs reported 3 patients with 4 donor SSIs not
identified on review, and study reviewers identified 5 patients with 10 donor SSIs not reported by ICCs. Using the study
review findings as gold standard, the PPV for ICC reported SSI was 96% (95% CI 86-99%), and NPV was 97% (95% CI 92-99%).
Sensitivity was 55% and specificity was 100%. When only sternal SSIs were analysed, PPV and NPV were 91% and 98% respectively,
and sensitivity was 62%.
Importantly, this study suggests that modifications to the NNIS methodology for CABG SSI surveillance should be considered with
regard to the use of SSI depth (for example, combining deep and organ space sternal SSI into one category), and the reporting
of donor SSI rates (for example, consider reporting only sternal SSI rates, or reporting sternal and donor SSI rates separately).