Literature Update - August, 2005
Compiled by Dr Deb Friedman, VICNISS ID Physician
Ventilator-Associated Pneumonia
A comprehensive review of ventilator-associated pneumonia (VAP) was published recently (Shaw MJ.
Curr Opin Pulm Med 2005;11(3):236-241) and below is a summary of important points about VAP.
VAP complicates the hospital course of up to 47% of intubated patients, it increases mortality, and length of
stay in both ICU (26 vs 4 days) and the hospital (38 vs 13 days)
are extended compared with uninfected patients. In addition, the costs of the hospital stay are more than 3 times higher
than for uninfected patients (US $70 568 vs US $21 620).
Several risk factors for VAP have been identified. Some factors are present prior to ICU admission, and include: male
sex, age, COAD, emergency surgery, and admission after trauma. There are also numerous factors present after ICU admission
including: elevated illness scores (eg: APACHE II score), coma, and Glasgow coma score of 6-13. In addition, other measures
of illness severity play a role, such as: uraemia and haemodialysis, hypoalbuminaemia, blood transfusion, duration of
ventilation and reintubation. Invasive devices and interventions like bronchoscopy, tracheostomy, nasogastric tubes/enteral
nutrition, TPN, CVCs, balloon pumps and medication (such as prior antibiotics, H2-receptor blockers, sucralfate, steroids,
continuous sedation) also increase the risk of VAP.
The diagnosis of VAP is very difficult because no gold standard exists. Numerous different methods (such as BAL,
quantitative ETA, blind telescopic or brush sampling) have been used with varying success, and are probably infinitely
more reliable than clinical judgement. In fact, 60% of BAL cultures rule out VAP, and allow for antibiotics to be discontinued.
Infection Control Consultants may well be frustrated with the surveillance definitions utilised to diagnose VAP- but they
are simply an indication of the complexity of the condition, and the difficulty in making clinical diagnoses.
There are numerous measures which may reduce the risk of VAP developing. These include: minimising the duration of
ventilation, good oral hygiene (via use of mouthwashes and oral decontamination, semirecumbent positioning and small bowel
feeding (to reduce aspiration), closed endotracheal suction systems, sputum clearance/suctioning below the endotracheal
cuff, humidification, and kinetic beds. Ultimately, these preventive measures can only be effective if staff are adequately
educated in their routine use.
The Clinical Significance of Coagulase-Negative Staphylococci Isolated from Blood Cultures
Researchers from a hospital in Iowa sought to develop an algorithm to assist in determining the clinical significance
of coagulase-negative staphylococci (CNS), as this organism is such a frequent contaminant of blood cultures (Beekman SE,
et al. Infect Control Hosp Epidemiol 2005;26(6):559-566).
The medical records of 960 patients with positive blood cultures were reviewed. 405 of the 960 contained CNS, and 316
of the 405 (78%) were considered to be contaminants. The algorithm which performed best at determining the clinical significance
of these CNS blood cultures was: at least 2 positive blood cultures for CNS within 5 days, or 1 positive blood culture plus
clinical evidence of infection (abnormalities in temperature, white blood cell count, and blood pressure). This algorithm had a
sensitivity of 62% and a specificity of 91%.
The use of such an algorithm could well reduce inappropriate antibiotic use, prescribed to treat positive blood cultures
which represent contaminants. Unfortunately, for clinicians, algorithms are not a gold standard, and do not replace careful
assessment of all relevant factors and good clinical judgement.
Public Reporting of Healthcare-Associated Infections
The issue of public reporting of healthcare-associated infections has become increasingly topical recently. HICPAC guidelines
have recently been published (McKibben L, et al. Infect Control Hosp Epidemiol 2005;26(6):580-587), as has an interesting article
about infection control report cards (Weinstein RA, et al. New Engl J Med 2005;353(3):225-227).
In response to consumer demand for healthcare information, hospitals, especially in the USA are publicly disclosing hospital
and physician performance. Typically they disclose both process and outcome measures of medical care. Process measures such as hand hygiene compliance and influenza vaccination are recommended, as are outcome measures, such as mortality, and selected complications or medical errors.
The complications of admission to hospital which are of direct relevance (to us) are nosocomial infections, and those
infection rates which are being reported are: CVC-associated BSI, VAP and SSI. At least 2 of these outcome measures are well-chosen
as they also provide a measure of adherence to the related processes of care, ie, CVC insertion practices and surgical antibiotic
prophylaxis).
Obviously, such reporting systems would need to ensure that they use standardised case-finding methods, and data validity
checks, and that they risk-adjust for infection risk. Regular feedback would also need to be provided to healthcare providers
to both increase the acceptability of the public reporting system, and to encourage poor performers to improve.
Although in certain circumstances it has been shown that public reporting reduced infection rates (for example; mortality
due to CABG surgery), overall the effectiveness of this practice is not yet proven, and will need to be proven in order to be
recommended in the future.