Vancomycin-Resistant Staphylococcus aureus:
Perspectives on Measures Needed for Control
Michael B. Edmond, MD, MPH; Richard P. Wenzel, MD, MSc; and A. William Pasculle,
Given the dramatic increase in the incidence of vancomycin resistance
among the enterococci and experimental evidence for the transfer of vancomycin
resistance from enterococci to Staphylococcus aureus, there is concern that
strains of S. aureus wiIl emerge that are resistant to vancomycin. The result
would be a highly virulent pathogen for which effective antimicrobial therapy
would not be avaitable. To prevent the nosocomial transmission of such an
organism, stringent infection control policies need to be developed and
implemented. We offer proposals that are based on the limited data available
on the transmission and control of S. aureus and that may be used as starting
points for the development of formaI guidelines for the isolation of colonized
and infected patients and for microbiology laboratory precautions.
Ann .Intern .Med. l996; I24:329-334.
From Medical College of Virginia and Virginia
Commonwealth University, Richmond, Virgînia; and University of Pittsburgh
School of Medicine, Pittsburgh, Pennsylvania.
For current author addresses, see end of text.
Throughout recorded history, humans have been prey to infections caused
by virulent strains of staphylococci and streptococci. The advent of the
antibiotic era -50 years ago- brought with it great optimism about the control
of Staphylococcus aureus and Streptococcus pyogenes. Fortunately, despite
five decades of penicillin use, all strains of Streptococcus pyogenes remain
very susceptible to penicillin.
The story of S. aureus is more complex. By the late 1950s, almost 50% of
all strains were resistant to penicillin. The organism had developed the
ability to break the beta-lactam ring by producing a beta-lactamase referred
to as penicillinase. In 1960, however, methicillin-a penicillinase-resistant
beta-lactam effective for treating penicillin-resistant S. aureus-was
The subsequent availability of oral cephalosporins added greatly to outpatient
management of infections caused by this organism.
Methicillin-resistant strains cf staphylococci emerged by the late 1970s
and have become increasingly more prevalent as nosocomial pathogens. The
medical community was comforted by the fact that vancomycin-available since
1958-provided effective therapy for aIl strains of methicillin-resistant
Nevertheless, the emergence of vancomycin-resistant strains of
staphylococci (I, 2) caused concern that such observations might presage
similar developments in S aureus. Adding to these concerns were observations
1) vancomycin-resistant enterococci caused isolated infections or epidemics
in some U.S. hospitals and were becoming increasingly prevalent in critical
care units (3, 4);
and 2) high-level vancomycin resistance was experimentally transferred from
Enterococcus faecalis to S. aureus in both in vitro and in vivo models (5).
It seems likely that vancomycin-resistant S. aureus will emerge as a nosocomial
pathogen with disastrous consequences if widespread nosocomial transmission
occurs. Thus, we believe that hospitals should adopt a proactive approach.
To that end, we provide perspectives on isolation guidelines for care of
the patient with vancomycin-resistant S aureus colonization or infection
and for the handling of the organism in the clinical microbiology laboratory.
Developers of formal guidelines should take into account the background
issues and control measures described below. It is important to note that
our comments are based on the limited data available on the transmission
and control of S aureus.
Transmission and control of Staphylococcus aureus
The reservoir for S. aureus is the anterior nares (6). The prevalence of
nasal colonization is approximately 40% among healthy adults (7). Three
patterns of nasal colonization have been observed: Some persons are never
colonized, some are persistently colonized, and others are intermittently
colonized. Half of those with nasal colonization also carry the organism
on their hands (8); transmission by the hands is probably the major mode
of transmission (9-12).
Certain patient populations (patients with diabetes (13), patients receiving
hemodialysis (14), patients receiving continuous ambulatory peritoneal dialysis
(15), injecting drug users (16), and patients with human immunodeficiency
virus infection (17,18), for example) have higher rates of staphylococcal
colonization and infection.
Once nasal colonization has been established, inections occur through
of the hands and subsequent inoculation of any traumatized area of skin.
Studies show that staphylococcal infections in colonized persons are often
due to the staphylococcal strain responsible for colonization (19-22); thus,
the infections have an endogenous origin.
Calcium mupirocin ointment has been shown to be effective in eliminating
staphylococcal nasal coloniration (8, 23), and it significantly decreases
hand carriage as well (8). In patients receiving hemodialysis, it has been
shown that elimination of nasal colonization is associated with a decrease
in S.aureus infection (19, 24).
Staphylococci can survive desiccation for days to weeks and can travel
distances through the air (25). Studies done in the 1950s and later established
that nasal carriers of S.aureus can shed the organism into the air (25-28)
and that aerial dissemination is greatest in those with the heaviest burdens
of organisms on their nasal mucosa (28). It is unclear whether aerial
from nasal carriers is associated with transmission of the organism to others.
Patients with large burns have been shown to shed large numbers of staphylococci
into the air (29, 30).
Washing with medicated soaps has been shown to remove S. aureus from the
hands (31-35). For vancomycin.-resistant enterococci, 60% isopropyl aIcohol
has a direct effect, but chlorhexidine has been shown to eliminate the organism
from the hands both directly and residually (36, 37). Although alcohol-based
hand.washing agents are used extensively in Europe, limited data suggest
that health care workers in the United States wash more frequently with
chlorhexidine than with alcohol (38).
Anecdotal data from the University of Iowa Hospitals and Clinics suggest
that concentrating on infection control activities decreases nosocomial
transmission. Asking the primary nurse of the infected or colonized patient
to monitor visitors' compliance with barrier precautions has proved useful.
Using intensive microbiologic and epidemiologic surveillance (admission
and weekly stool surveillance cultures and concurrent unit-based surveillance
for nosocomial infections, for example), we discovered only three infections;
furthermore, using molecular typing, we documented only one case of nosocomial
transmission, in which the patient became colonized only (unpublished data).
Haley and colleagues (39), in an effort to eradicate endemic
S.aureus infections from a neonatal intensive care unit, had an infection
controI nurse dedicated to the unit. One of her duties was to observe compliance
with infection controI practices. Although the authors could not determine
the independent effect of the nurse due to the concomitant implementation
of several other control measures. they deemed that her presence had led
to successful eradication of the pathogen and increased compliance with
aseptic practices (39).
The duration of colonization with methicillin-resistant S. aureus is long;
the estimated half-life of nasal colonization is 40 months (40). However,
twice-daily intranasal application of mupirocin ointment for 5 days has
been shown to have a long-term effect with a downward trend in the rate
of nasal colonization and a statistically significant decrease in the rate
of hand carriage at 6 months (41). Although resistance to mupirocin among
S. aureus isolates has been reported, it remains uncommon (42-48). High-level
resistance bas been observed primarily when mupirocin therapy has been prolonged
(44) or extensive (as in dermatology wards)(42, 43, 45,48).
Identification of Vancomycin-Reslstant Staphylococcus aureus
The Centers for Disease Control and Prevention (CDC) recommend that all
clinical isolates of S. aureus be tested for susceptibility to vancomycin
(49). Laboratoiy personnel should notify the laboratory director if
S. aureus is discovered.
Many isolates of S. aureus presumed to have been vancomycin-resistant have
been found to be mixed with other organisms in cultures; therefore, vancomycin
resistance should be confirmed by re-streaking the colony to certify that
the culture is pure (49). The hospital epidemiology program should be notified
so that it can institute appropriate isolation procedures. The public health
department, other hospitals in the vicinity and the CDC should also be notified.
Precautions for Vancomycin-Resistant Staphylococcus aureus
The proposals we outline below are designed to help standardize the approach
to infected or colonized patients, and are based on the limited data cited
above. Indîvidual institutions may adopt some or aIl of our proposals,
depending on local circumstances and resources.
Isolation of Infected or Colonized Patients
A patient who is infected or colonîzed with vancomycin-resistant S.
aureus should be placed in a private room, and all persons entering the
room should wear clean, nonsterîle gloves and a disposable gown. Gloves
and gowns should be removed before leaving the room. After the gloves are
removed, hand washing with 4 % chlorhexidine or 60 % isopropyl alcohol is
required (37). A monitor could be placed at the door to prevent unauthorized
access and to enforce hand washing and barrier precautions (39). The names
of all persons entering the room should be recorded for future use should
obtaining nasal surveillance cultures become necessary
A standard surgical mask and safety glasses must be worn by persons doing
procedures that might generate an aerosol (suction, bronchoscopy, sputum
induction, or aerosol treatment, for example). Patients with
S.aureus pneumonia requiring mechanical ventilation should have a filter
or condensate trap placed on the expiratory phase tubing of the mechanical
ventilator circuit. If oxygen therapy by nasal cannula is required, a standard
surgical mask should be worn by all persons entering the room. Although
few data support the idea that airborne transmission of staphylococci is
possible (25>, we prefer a conservative approach until the epidemiology
of vancomycin-resistant S. aureus is delineated.
If the patient is colonized in the nares, decolonization with mupirocin
should be attempted (8, 50-52). However, because clinical isolates have
not been available for the performance of susceptibility testing, the activity
of mupirocin against vancomycin-resistant S. aureus is unknown. We do not
recommend adding other drugs, such as rifampin or trimethoprim-sulfamethoxazole,
because these drugs have not been necessary with mupirocin and -unlike
may cause serious adverse effects.
Sharing of noncritical equipment (such as electronic thermometers, blood
pressure cuffs, stethoscopes, intravenous poles, bedside commodes, and
is not permitted.
-Infectious disease consultants should review the patient's antimicrobial
therapy, making every effort to reduce the selection of vancomycin-resistant
S. aureus by eliminating or substituting antibiotics. Prudent use of
agents should be stressed in both inpatint and outpatient settings, even
before the emergence of vancomycin-resistant S.aureus. Vancomycin use should
be reduced throughout the hospital. Oral metronidazole rather than oral
vancomycin should he used to treat antibiotic-associated colitis when possible
The number of health care workers who have contact with the patient infected
or colonized with vancomycin-resistant S. aureus should be limited. Care
of the patient should be done by no more than one nurse and one physician
per shift when possible. Phebotomy and other ancillary services should be
done by the primaiy nurse or primary physician. Until more is learned about
the epidemiology of vancomycin-resistant S. aureus aIl health care workers
caring for the patient should have nasal surveillance cultures done every
2 weeks. Health care workers known to be at higher risk for staphylococcal
colonization (those with exfoliative dermatitides or diabetes mellitus requiring
treatment with insulin) should not care for patients with vancomycin-resistant
S aureus colonization or infection. The recommended hand-washing agents
(alcohol and chlorhexidine) may themselves cause dermatitis; health care
workers who develop dermatitis should be reassigned.
Housekeeping personnel should be instructed to clean all horizontal surfaces
in the patient's immediate vicinity daily with a quaternary ammonium compound.
Cleaning cloths used in the room should not be used to clean other patients'
rooms and equipment, but should be carefully discarded.
Isolation must continue for the duration of the hospital stay. After the
infected or colonized patient is discharged and housekeeping personnel have
completed terminal disinfection of the room, environmental cultures should
be obtained. The room should remain closed to new admissions until negative
cultures have been reported. All equipment used in the room must be disinfected.
Before discharge, an epidemiology alert sticker should be affixed to the
cover of the patient's chart and a notation should be made in the hospital's
information system. Any patient with previons vancornycin-resistant S. aureus
infection or colonization who is readmitted should be placed in isolation
immediately. Isolation should continue until surveillance cultures at the
nares and of any previously infected, open sites have been obtained and
If a nosocomial transmission is documented on a hospitai unit, the unit
should be closed to new admissions. Any previousiy uninfected patient from
this unit who requires transfer to another hospital unit should be placed
in isolation in the receiving unit until, two nasal cultures- 48 hours apart-
Diagnostic and therapeutic procedures that require the patient to leave
the isolation room should be postponed. When testing is done at the bedside
(portable radiography, electrocardiography), equipment should be wiped down
with a disinfectant when the test is complete. Collection of microbiologic
and other specimens for clinical testing should be done in the patient's
room with health care workers wearing protective attire as described above.
Specimens should be kept in a leakproot container and placed in a sealable,
leakproof plastic bag for transport (53). Laboratory forms should not be
placed in the bag with the specimen. Care must be taken to prevent contamination
of the outside of the bag. The specimen should be taken to the laboratory
immediately; it should not be sent through a pneumatic tube system.
MicrobioIogy Laboratory Precautions
To minimize the possibility for colonization or infection of hospital staff,
as few staff members as possible should handle specimens from a patient
with vancomycin-resistant S aureus infection. All specimens from an infected
or colonized patient should be delivered directly to the laboratory without
routing through a centralized specimen receiving area.
When a specimen for culture from a patient with vancomycin-resistant S.
aureus infection is delivered, it should be immediately placed in a biological
safety cabinet until it can be processed. The laboratory director should
review aIl culture requests before plates are inoculated, eliminating the
unnecessary processing of cultures that contain vancomycin-resistant S aureus
Specimen processing requires two persons, one working only within the biological
safety cabinet (technologist 1) and the other (technologist 2) assisting
technologist 1. The biological safety cabinet should contain a squirt bottie
filled with a disinfectant, a beaker containing a disinfectant for the disposaI
of loops, a heavy, clear biohazard bag for the disposaI of specimens and
other material, and clear specimen bags for holding inoculated plates.
Specimens should be inoculated onto as few plates as possible. Each plate
should be placed directly into a clear specimen bag as soon as it is inoculated.
When all plates have been inoculated, the specimen bag should be closed
with a twist closure, sprayed on the outside with a disinfectant, and wiped
off with a paper towel. The used paper towel should be placed in the biohazard
bag for disposaI, along with the remainder of the specimen and any rejected
The bag containing the inoculated plates should be placed in a large anaerobic
jar with no catalyst and no gas-generating envelope. For this step, technologist
2 should hold the jar outside the biological safety cabinet while technologist
1 gently transfers the bag containing the plates to the jar. Once the plates
are inside, technologist 2 should close the jar and place it in the incubator.
All remains of the specimen and transport material should be placed in the
biohazard bag and closed with a twist tie. While still in the biological
safety cabinet, the biohazard bag should be placed into a second biohazard
bag and closed with another twist tie. Technologist 1 should spray the work
surface, sides, and windshield of the biologîcal safety cabinet with
a disinfectant. After a 10-minute exposure time, the disinfectant residue
should be wiped from aIl surfaces with 70 % ethanol. Technologist I should
hand the closed bag to technologist 2, who should immediately take the bag
to the autoclave. It should be autoclaved for 30 minutes.
All cultures should be worked up in the biological safety cabinet using
the same two-technologist system . Technologist 1 should stay in the biological
safety cabinet at all times while technologist 2 obtains supplies and monitors
the technique of technologist 1. All subcultures, susceptibility tests,
and biochemical tests will be placed in either zip-lock specimen bags or
anaerobic jars before being placed in the incubator.
All reagents used during the examination of any culture from a patient colonized
or infected with vancomycin-resistant S. aureus should be placed in a biohazard
bag contained in the hood and discarded. No nondisposable reagents or equipment
should be used to work on cultures from a patient colonized or infected
with vancomycin-resistant S. aureus Plates that need to be held until test
results are finalized should be placed in zip-lock specimen bags and left
in the safety cabinet until discarded. All discarded plates, slides, and
biochemicals should be placed in bags and autoclaved as described above.
A stock culture of the organism may be made for future studies (molecular
typing, study of resistance mechanisms, testing of new antimicrobial agents).
A suitable stock can be made by placing 5 to 10 colonies in 1 mL of nutrient
broth containing 15% (v/v,) glycerol. The stock culture should be frozen
at -70 °C in a freezer located in an area of the laboratory to which
there is limited access.
In the event that a cluster of patients with vancomycin-resistant S. aureus
infection or colonization is discovered (two or more patients on the same
ward), all bacterial isolates should be stored for further microbiological
studies. Isolates should be frozen at -70 °C in a secure area of the
Nares cultures should be obtained from all exposed health care workers;
those workers found to be colonized with vancomycin-resistant S. aureus
should be treated with intranasal mupirocin ointment twice daily for 5 days
(23). Nasal cultures should be done 72 hours after the completion of mupirocin
treatment . A follow-up culture 48 hours later should also be done. If both
of the cultures done after treatment yield vancomycin-resistant S. aureus,
the health care worker should be removed from patient care duties, and
with mupirocin should be reattempted. If either culture yields
S. aureus, a third culture should he obtained at least 48 liaurs later.
If the third culture is positive, the health care worker should be removed
from patient care duties, and decolonization with mupirocin should be
Documentation of two negative nasal cultures done at least 48 hours apart
should be obtained from health care workers who have been relieved of patient
care duties before they are allowed to return to work.
All patients on the same ward should be cultured (nares and open wounds),
and isolation precautions should be instituted as outlined above.
The purpose of any guideline is to narrow variation in approaches to clinical
problems for which data defining optimal practices are limited. Guidelines
ensure that thousands of institutions do not reinvent the wheel and commit
unnecessary hours to the design of separate approaches. Guidelines may become
standards of care, but in their inception they are meant to be clinical,
not legal, documents.
These are our starting points as we offer our perspectives for managing
vancomycin-resistant S. aureus. As more information becomes available on
the epidemiology of vancomycin-resistant S aureus, definitive guidelines
for control can be developed.
Infection control guidelines are usually develeped in response to a problem.
Formal guidelines for the control of vancomycin-resistant enterococci were
published several years after infections with the organism were first reported.
Our goal, however, is to have a thoughtful plan of action in place before
the organism emerges. In doing so, we can attack the problem as it arises,
limiting nosocomial transmission and the ability of the organism to establish
itself as an endemic pathogen.
We attempt to provide health care workers in the acute care setting with
an understanding of the epdemiology and control measures for S. aureus that
may be useful for the control of vancomycin-resistant strains. Our proposaIs
could be adapted for use in long-term care institutions. However, the
of this orginism into long-term care facilities would pose a problem, given
the relatively limited financial and human resources available in most
and the current lack of infection-control infrastructure.
The emergence of vancomycin-resistant S. aureus would represent the most
important issue in antibiotic resistance since the dawn of the antibiotic
era. A common, virulent, and transmissible bacterial agent with no known
effective therapy would set infectious diseases back 60 years. The ability
of S.aureus to acquire genes coding for the toxic shock syndrome only adds
more concern. For these reasons, we propose conservative measures aimed
at minimizing nosocomial transmission as much as possible.
References: see in the original article.
Requests for Reprints:: Michael B. Edmond MD, MPH, Medical College
of Virginia, PO Box 980019, Richmond, VA 23298-0019.
Current Author Addresses:
Dr. Edmond: Medical College of Virginia, Division of Quality Health Care,
PO Box 980019, Richmond, VA 23298-0019.
Dr Wenzel: Medical College of Virginia, Department of InternaI Medicine,
PO Box 980663, Richrnond, VA 23298-0663.
Dr. Pasculle: University of Pittsburgh Medical Center, NW-630, 200 Lothrop
Street, Pittsburgh; PA. 15213.