16th Society for Healthcare Epidemiology of America (SHEA) Scientific Meeting
Chicago, April 2006
Mike Richards from the VICNISS Coordinating Centre recently attended the 16th Society for Healthcare Epidemiology
of America (SHEA) Annual Scientific Meeting 2006, held in Chicago, between the 18th and 21st March.
This meeting (intended for physicians, nurses and others involved in hospital epidemiology and hospital
infection prevention and control) was to present and discuss recent advances in these fields, and is one of
the two major hospital infection meetings in the US annually. This Conference also allowed us to receive
information on the recent changes from the NNIS Program to the National Health and Safety Network, and some
details of this is described elsewhere in the VICNISS April eBulletin.
One VICNISS poster was presented, titled "Surgical Antibiotic Prophylaxis in Smaller Hospitals, Victoria, Australia".
Below is an update of selected interesting topics and information presented at SHEA:
Methicillin-resistant
Staphylococcus aureus in the Community (CA-MRSA)
As in Australia, MRSA is an emerging community pathogen in the United States. The case definitions for CA-MRSA are
problematic. Epidemiological definitions based on where MRSA is first cultured, and lack of established MRSA risk
factors, were initially used, but hospital transmission of so-called CA-MRSA is now increasing, with reports of
infections in post-partum women, neonatal ICUs, prosthetic joint infections, and newborn infections in multiple
US states. CA-MRSA are at times identified by their antimicrobial resistance patterns. However, antibiotic
resistance is further developing, particularly clindamycin resistance, limiting the usefulness of this
definition. Many of the new MRSA strains have contained Panton-Valentine Leukocidin (PPL) toxin genes,
that have been rare in hospital-acquired MRSA strains. These PPL genes have also been found in some
Methicillin-sensitive S. aureus (MSSA) strains.
CA-MRSA seems less likely to colonise and more likely to cause disease than MSSA and hospital-acquired MRSA.
Thirty-eight percent of CA-MRSA colonised patientswent on to develop disease in one study, whereas previous
studies of MSSA colonised soldiers suggested only a 3% progression to disease. Seventy-seven percent of reported
infections with this pathogen were the skin and soft tissues, but there were also wound infections, urinary tract
infections, sinusitis, bacteraemia, and 10% of infections were pneumonia. A report from adult Emergency Department
patients attended a network of emergency ID hospitals in many States of the US, CA-MRSA was now responsible for 59%
of skin and soft tissue infections. Of 17 patients with CA-MRSA pneumonia, 71% had laboratory-confirmed influenza as
a pre-disposing factor, 94% were hospitalised, and 81% required intensive care support. The dramatic increase in this
pathogen in the US, is such that it is recommended that perulent skin infeciton in the community be routinely
cultured. Infection control recommendations for CA-MRSA in patients whilst in hospital are the same for hospital
strains. Strains that test erythromycin-resistant, but clindamycin-susceptible should be further tested for
inducible clindamycin-resistance, as this may impact on treatment.
Clostridium difficile-associated Diarrhoea (CDAD)
In US hospitals since 2000, there has been a sharp increase in the incidence of
Clostridium difficile-associated diarrhoea (CDAD). In contrast to a low attributable
mortality described in papers in the 1990's, in a Pittsburgh report there was an increase from 1.6 to 3.2%
of life-threatening disease. Forty-four patients required colectomies, and 20 patients died. In a Quebec
hospital in 2003, there was a 16.7% attributable mortality. Current isolates show increased resistance to
the quinolines gatifloxacin and moxifloxacin, which are infrequently used in Australia. In some cities,
hospital outbreaks were related to increased use of these drugs. A new epidemic strain, Toxinotype III B1
has emerged, that produces increased amounts of toxin A and toxin B as a result of genetic changes. These
strains also produce a toxin termed a binary toxin, a possible further virulence factor.
There are now reports of CDAD in groups of people previously considered at low-risk, including patients
without significant underlying illness, causing peri-partum infections in 4 US States and community-acquired
infections. Transmissions to close contacts has been seen, and also infections in patients that have not had
previous antimicrobial exposure.
Infection control measures include:
- Handwashing with soap and water, not alcohol-based disinfection;
- Bleach cleansing in rooms of C. difficile patients;
- Monitoring of isolation compliance for patients with C. difficile isolates; and
- Expanded duration of contact isolation for the entire length of stay for CDAD patients.
Research Methods and Infection Control
Randomised control trials are not feasible to evaluate many infection control interventions. They may be
unethical or impractical to implement, and the interventions may need changes in institution-wide policies
for institution-wide programs. Quasi experimental (QC) methods, involving no random assignment, no comparison
group involve measurements of points before and after interventions. Acceptable study designs include pre and
post measures where an appropriate control group or time series with three or more measures per segment.
A review was done of 73 quasi experimental studies in four of the most major infectious disease journals.
Sixty-eight percent of these had invalid designs, and only 4% justified their choice of designs. Only 22%
described the design correctly. In infection control there is a need for well-designed quasi experimental
studies, and if these are well performed the quality of data that can be obtained may be acceptable for
Cochrane reviews. This important review emphasises the need for strong epidemiological input into study
design in infection control.
Hand Hygeine
Two hand hygeines used by staff were compared for their impact on neonatal healthcare-associated infection
rates and skin conditions and microbial count of the nurses hands (Laos et al). They used a cross-over
design in two neonatal ICUs involving almost 52,000 patient days and 119 nurses. Traditional handwashing
with antiseptic detergent containing 2% Clorhexidene gluconate was compared with a handrub with 60% alcohol.
No significant differences in infection rates were observed, but the nurses skin condition improved with alcohol,
and hospital costs were less, predominately through differences in consumption of nursing time.
In one institution where educational efforts were ineffective, and outbreaks of antibiotic resistant
organisms occurred (Shadow et al), direct observation of unit staff was undertaken by the unit Director,
ICPs, nursing administration, and a hospital Epidemiologist, for 15 minute intervals daily. Non-compliant
employees would be counselled, given further education, and repeat offenders would be at first reprimanded,
then referred to Human Resources, then asked to take corrective actions, and if continuing to be non-compliant,
discharged. Handwashing compliance dramatically increased, reaching 100% in the Coronary Care Unit. A process
of physician accountability was included in the program, which could lead to restricted practice priveliges,
and referral to the hospital executive. This improvement in compliance was sustained, including over 90%
compliance by physicians during the 12 month period.
Reduction of Central line-associated Bloodstream Infections (CLABSIs) in a Large Healthcare System (Mato et al)
In a consortium of 20 tertiary care hospitals in Pittsburgh, a CLABSI "bundle" of preventative
measures was implemented, including:
- A standard procedure of documenting barriers and antisepsis used;
- Mandatory CLABSI education;
- Clorhexidine gluconate for skin antisepsis and Clorhexidene gluconate/silver Sulphadiazine coated central venous catheters;
- Use of five maximal barrier precautions; and
- Goals, infection rates and process compliance were reported monthly to comitted health care leaders.
With this approach, a dramatic reduction in CLABSIs was achieved and sustained for three years in many
types of ICUs. CLABSI rates per 1,000 central line days fell to 0.5 in Medical-Surgical ICUs, monitored
for 4,000 days, to 0 in Trauma ICUs, monitored for 1,800 central line days, and to 0.6 in teaching hospital
Medical-Surgical ICUs.
This dramatically effective intervention in a large number of ICUs in a large number of hospitals achieved
a sustained fall in BSI rates to almost zero, levels that had previously been thought to be unattainable.
