The outbreak period was in September and October 2003, when a marked
increase in the number of cases was noted. The patterns indicate various strains
of multidrug-resistant Acinetobacter baumannii (MDR-Ab)
as defined by pulsed-field gel electrophoresis. After the outbreak was halted
in late October, the number of cases of MDR-Ab remained somewhat elevated
in November because of an increase in the number of patients admitted with
MDR-Ab. These cases were unrelated to pulsatile lavage therapy and the strains
did not match the outbreak strain.
This timeline depicts the 11 patients identified as infected or colonized
with multidrug-resistant Acinetobacter baumannii (MDR-Ab)
at Johns Hopkins Hospital during September and October 2003. Eight of the
11 patients had pulsatile lavage therapy, as indicated by the bars. Circles
denote cultures that grew MDR-Ab. During the outbreak investigation, it was
discovered retrospectively that patient 11 had a urine culture that grew MDR-Ab
in July 2003. This isolate was not available for pulsed-field gel electrophoresis
Strain A was isolated from blood, urine, sputum, or wound specimens
from 6 patients who were treated with pulsatile lavage therapy and from multiple
surfaces in the wound care treatment room. Strains B and C were isolated from
sputum or wound specimens from 2 patients who were not treated with pulsatile
lavage therapy. ETS indicates endotracheal suction. MDR-Ab, multidrug-resistant Acinetobacter baumannii; MW, molecular weight.
Maragakis LL, Cosgrove SE, Song X, Kim D, Rosenbaum P, Ciesla N, Srinivasan A, Ross T, Carroll K, Perl TM. An Outbreak of Multidrug-Resistant Acinetobacter baumannii Associated With Pulsatile Lavage Wound Treatment. JAMA. 2004;292(24):3006–3011. doi:10.1001/jama.292.24.3006
Author Affiliations: Division of Infectious
Diseases, Department of Medicine (Drs Maragakis, Cosgrove, Carroll, and Perl),
and Department of Pathology (Ms Ross and Dr Carroll), Johns Hopkins University
School of Medicine, Departments of Hospital Epidemiology and Infection Control
(Drs Maragakis, Cosgrove, Song, and Perl and Ms Rosenbaum) and Physical Medicine
and Rehabilitation (Ms Ciesla), Johns Hopkins Hospital, and Department of
Preventive Medicine, Johns Hopkins Bloomberg School of Public Health (Dr Kim),
Baltimore, Md; Division of Healthcare Quality and Promotion, Centers for Disease
Control and Prevention, Atlanta, Ga (Dr Srinivasan).
Context Pulsatile lavage is a high-pressure irrigation treatment used increasingly
in a variety of health care settings to debride wounds. Infection control
precautions are not routinely used during the procedure and are not included
in pulsatile lavage equipment package labeling.
Objectives To investigate an outbreak of multidrug-resistant Acinetobacter baumannii and to test the hypothesis that pulsatile lavage
wound treatment was the mode of transmission for the organism.
Design Outbreak case-control investigation including case identification, review
of medical records, environmental cultures, and pulsed-field gel electrophoresis.
Setting A 1000-bed tertiary care hospital in Baltimore, Md, during September
and October 2003.
Patients The investigation included 11 patients infected or colonized with multidrug-resistant A baumannii. Seven of these patients met the case definition
for the case-control study and were compared with 28 controls randomly selected
from a list of inpatients without multidrug-resistant A
baumannii who had a wound care consultation.
Main Outcome Measure Infection or colonization with multidrug-resistant A baumannii.
Results Eleven patients had cultures that grew multidrug-resistant A baumannii during the outbreak period. Of the 10 health care–associated
cases, 8 had received pulsatile lavage treatment. One strain of multidrug-resistant A baumannii was recovered from all 6 pulsatile lavage patients
who had isolates available for pulsed-field gel electrophoresis analysis and
from multiple surfaces in the wound care area. Six of 7 cases (86%) were treated
with pulsatile lavage vs 4 of 28 controls (14%) (odds ratio, 36; 95% confidence
interval, 2.8-1721; P<.001). These results confirm
that pulsatile lavage was a significant risk factor for acquisition of multidrug-resistant A baumannii.
Conclusions Transmission was apparently caused by dissemination of multidrug-resistant A baumannii during the pulsatile lavage procedure, resulting
in environmental contamination. Appropriate infection control precautions
should be used during pulsatile lavage therapy and should be included in pulsatile
lavage equipment labeling.
Multidrug-resistant Acinetobacter baumannii (MDR-Ab)
has emerged worldwide as an important health care–associated pathogen,
causing infections such as ventilator-associated pneumonia, bloodstream infections,
and wound infections.1- 4 The
organism can survive on environmental surfaces for months, making nosocomial
transmission extremely difficult to prevent and control.5- 7 Many Acinetobacter isolates demonstrate extensive resistance
to antimicrobial agents, including carbapenems, which greatly complicates
treatment of these infections.2,3,7,8 Previously
reported outbreaks of MDR-Ab have frequently been associated with respiratory
care equipment and water sources such as humidifiers.1,9- 12
In October 2003, the Department of Hospital Epidemiology and Infection
Control at Johns Hopkins Hospital became aware of a cluster of 5 patients
infected or colonized with MDR-Ab. The number of cases of MDR-Ab was approximately
5 times the usual rate (Figure 1) and
the isolates were particularly resistant, susceptible only to colistin. The
patients had different primary diagnoses and were located in various units
throughout the hospital. Four of the 5 original cluster patients had wounds
and 3 had received pulsatile lavage therapy, a water-based, high-pressure
irrigation treatment with concurrent suction used for cleansing and debridement
of wounds.13 The technique originated in the
1960s for use in the operating room and is now performed widely in hospitals,
rehabilitation centers, outpatient clinics, and long-term care facilities.
This is the first report of a nosocomial outbreak related to pulsatile lavage
The Johns Hopkins Hospital is a 1000-bed tertiary care hospital in Baltimore,
Md. Prior to this outbreak, physical therapists performed wound care, including
pulsatile lavage, for inpatients and outpatients in a single procedure room.
The room was a large, open area with 3 whirlpools, 1 sink, and 2 stretchers.
It contained curtains for patient privacy, wound care carts, and open supply
shelves. There was minimal ventilation and high humidity.
Pulsatile lavage treatment is used to debride wounds of devitalized
tissue and debris. Pulsatile lavage devices are class 2 devices, which are
exempt from premarket notification or approval by the US Food and Drug Administration
(FDA). A battery-powered device that resembles a water gun is used to deliver
pressurized sterile saline to the wound. A small shield at the tip of the
device is placed in contact with the wound bed so that suction is created
and splash is minimized. Tubing connects the device to irrigation fluid and
to a suction pump. The water gun, splash shield, and tubing are for single
use and are disposable. In 2003, 2947 pulsatile lavage procedures were performed
at Johns Hopkins Hospital.
Microbiology records were reviewed to identify cases and to define the
baseline rate of MDR-Ab prior to the outbreak. A case was defined as any patient
infected or colonized with MDR-Ab susceptible to no more than 1 class of antimicrobial
agents, excluding colistin. The baseline rate of MDR-Ab for the 4 years prior
to the outbreak was a mean of 2 cases per quarter (range, 0-4). There were
only 2 cases of MDR-Ab between January 2003 and the beginning of the outbreak
in September 2003 (Figure 1). Medical
records were reviewed, including paper and electronic charts, the microbiology
database, and physical therapy wound care records. The pulsatile lavage procedure
was observed and multiple surfaces in the wound care room were cultured. Given
prior reports of MDR-Ab outbreaks associated with mechanical ventilation,
respiratory care records were also reviewed and respiratory care equipment
After identifying pulsatile lavage therapy as a potential risk factor
for transmission of MDR-Ab, certified letters were sent to all 58 patients
who received pulsatile lavage or whirlpool therapy in the wound care room
from September 1 through October 31, 2003. Patients were asked to return for
a surveillance wound culture. For patients who did not reply to the certified
letter, at least 2 attempts were made to contact them by telephone.
Clinical cultures were performed using standard practices on routine
media. Susceptibility testing was performed by the agar dilution method. Environmental
samples were cultured on brain-heart infusion plates supplemented with 5%
sheep blood and gentamicin at a concentration of 10 μg/mL. Pulsed-field
gel electrophoresis (PFGE) (Bio-Rad Gen Path, Hercules, Calif) was performed
on all available isolates. DNA was digested with SmaI
and gels were analyzed with Molecular Analyst Fingerprinting Plus software
(Bio-Rad, Hercules). Isolates were considered genetically related if their
PFGE patterns differed by 3 or fewer bands.14
We conducted a case-control study to assess risk factors for acquisition
of the outbreak strain of MDR-Ab in our institution. The case definition was
any patient who had the outbreak strain or unidentified strain of MDR-Ab,
in a clinical culture taken at least 48 hours after admission, during an inpatient
hospitalization between September 1 and October 31, 2003. This study definition
did not include patients with strains other than the outbreak strain by PFGE
or patients admitted to our institution already harboring MDR-Ab. Patients
whose MDR-Ab was detected by surveillance culture during the epidemiologic
investigation were not included to avoid biasing the study toward pulsatile
lavage as a risk factor. Controls were randomly selected from the list of
inpatients without MDR-Ab who were seen for wound care consultation during
the outbreak period. The ratio of controls to cases was 4 to 1. All controls
had wounds and at least 1 clinical culture of urine, sputum, wound, or blood
during the study period. Controls were not matched with cases on any other
Data were collected from medical records to obtain the following information
for each study patient: age; sex; race; hospital days; admission to an intensive
care unit; mechanical ventilation; surgery; invasive procedures; hemodialysis;
presence of a central line, urinary catheter, or arterial line; receipt of
parenteral nutrition or antibiotics; recent residence in a nursing home; McCabe
disease severity score15; tobacco use; history
of diabetes, hypertension, human immunodeficiency virus infection, malignancy,
or pulmonary, cardiac, renal, or neurologic disease; immunosuppression; recent
trauma; site and etiology of wound; and exposure to whirlpool therapy or pulsatile
lavage treatment. Data were collected for the period September 1 through October
31, 2003, for controls and September 1, 2003, through the date of first MDR-Ab
culture for cases.
The institutional review board at Johns Hopkins University granted the
study exemption status and a Health Insurance Portability and Accountability
For the case-control study, we performed a 2-tailed Fisher exact test
to calculate odds ratios (ORs), 95% confidence intervals (CIs), and P values. We used the Wilcoxon rank-sum test to compare
group medians. With α = .05, the study had 70% or higher power
to detect an absolute difference of 70% or higher between controls and cases,
assuming a control proportion of 0.15. Multivariate logistic regression was
not performed because of small sample size. All statistical analyses were
performed with Stata software, version 7 (Stata Corp, College Station, Tex).
A total of 11 case patients infected or colonized with MDR-Ab were identified
between September 1 and October 31, 2003 (Table
1 and Figure 2). This includes
the 5 original cluster patients and 6 other patients who were identified during
the outbreak investigation. The patients had different primary diagnoses and
were located in various units throughout the hospital. Isolates of MDR-Ab
were found in blood, sputum, urine, and wounds. All MDR-Ab isolates were susceptible
to colistin but were otherwise either completely resistant or showed only
intermediate susceptibility to amikacin. The mean age of case patients was
59 years (range, 24-83 years). One patient was culture-positive for MDR-Ab
on admission, but this strain of MDR-Ab was unrelated to the outbreak strain
by PFGE. Of the 10 patients who acquired the organism in the hospital, all
had wounds and 8 had received pulsatile lavage treatment. Thirty-one (53%)
of 58 wound care patients returned for surveillance cultures and 1 patient
(2%) had MDR-Ab grown from the surveillance culture.
Eight of the case patients had wound infections and 3 had both bloodstream
infections and pneumonia. Three patients required admission to the intensive
care unit for sepsis and respiratory distress, and 2 patient deaths were possibly
related to their infections.
Isolates were available from 6 of the case patients who had undergone
pulsatile lavage treatment and from 3 case patients who did not have pulsatile
lavage. Pulsed-field gel electrophoresis of the isolates from all 6 patients
who received pulsatile lavage demonstrated 1 identical strain of MDR-Ab that
was distinct from isolates obtained from 2 patients who did not have pulsatile
lavage (Figure 3). One patient who did
not receive pulsatile lavage had the outbreak strain of MDR-Ab in her urine.
No risk factor for transmission was identified for this patient. The disposable
pulsatile lavage gun and suction canister insert grew the outbreak strain
of MDR-Ab after being used to treat 1 of the case patients at the time of
the outbreak investigation. Environmental cultures from a cleaned stretcher,
the sink, and a supply shelf above the treatment area all grew the outbreak
strain of MDR-Ab, as confirmed by PFGE (Figure
Investigation revealed that a change in pulsatile lavage procedure had
occurred approximately 2 months prior to the cluster of cases that brought
the outbreak to our attention. As a cost-saving measure, the disposable suction
canister inserts that were previously discarded after each patient was treated
were changed once a day or when full.
Seven patients met the case definition of the case-control study. Of
the 4 patients who did not meet the definition, 1 grew MDR-Ab within 48 hours
of admission, 1 was detected by surveillance culture, and 2 had MDR-Ab strains
different from the outbreak strain. The cases and controls were similar with
respect to age, race, sex, smoking status, McCabe disease severity score,
and mean number of cultures performed (Table 2). The mean number of hospital days was greater for the patients
with MDR-Ab than for the controls (50 days vs 33 days); the difference was
not statistically significant. Univariate analysis confirmed that pulsatile
lavage was a significant risk factor for acquisition of MDR-Ab. Six (86%)
of 7 cases vs 4 (14%) of 28 controls were treated with pulsatile lavage (OR,
36; 95% CI, 2.8-1721; P<.001). Prior residence
in a nursing home was also a significant risk factor. Four (57%) of 7 cases
vs 1 (4%) of 28 controls resided in a long-term care facility in the 6 months
prior to admission (OR, 36; 95% CI, 2.2-1833;P = .003).
Three (43%) of 7 cases vs 3 (11%) of 28 controls received hemodialysis, which
approached but did not achieve statistical significance (OR, 6.3; 95% CI,
0.9-42.5; P = .08). All other variables
tested, including mechanical ventilation and administration of antimicrobial
agents, were not significantly different between case and control patients
All pulsatile lavage treatment was stopped on October 24, 2003. The
outbreak was halted by aggressive infection control measures, including temporary
closure of the wound care treatment room, thorough cleaning and disinfection
of environmental surfaces, strict isolation of infected and colonized patients,
and termination of pulsatile lavage wound care until renovation of the wound
care area and procedural changes occurred. When pulsatile lavage resumed on
November 18, treatments occurred in newly constructed private rooms and patients
known to be colonized or infected with MDR-Ab were scheduled at the end of
the day whenever possible. After these interventions, MDR-Ab cases remained
slightly elevated in the final months of 2003; however, no further cases of
the outbreak strain were detected. Follow-up environmental cultures of surfaces
in the new private rooms constructed for pulsatile lavage treatment have been
performed monthly for 11 months and none has grown MDR-Ab.
This report describes the first recognized nosocomial outbreak associated
with pulsatile lavage wound care and identifies a novel mode of transmission
for A baumannii. Six (13%) of 46 patients treated
with pulsatile lavage during the outbreak period were infected or colonized
with a single strain of MDR-Ab that was also cultured from multiple environmental
surfaces in the wound care room. Only 1 patient acquired the outbreak strain
of MDR-Ab without known exposure to pulsatile lavage treatment. The attributable
risk of acquisition of MDR-Ab in patients receiving pulsatile lavage therapy
was approximately 12.4%. Prior to the outbreak, the manufacturer of our pulsatile
lavage equipment did not include any infection control information in its
product insert. Based on our findings in this outbreak, 1 equipment manufacturer
added information on infection control measures to the product insert. However,
other manufacturers of pulsatile lavage equipment are not currently required
to include infection control information with their product.
The extensive resistance of the organism and the severity of the patient
outcomes make this outbreak significant, though it involved a small number
of patients. Three case patients required admission to the intensive care
unit for sepsis, and MDR-Ab infections possibly contributed to 2 deaths. Our
outbreak investigation identified a common procedure as a previously unrecognized
means for transmission of a multidrug-resistant nosocomial pathogen. Although
MDR-Ab was responsible for this outbreak, it is reasonable to hypothesize
that other aquaphilic organisms such as Pseudomonas might
be spread by this means.
The case-control study confirmed that pulsatile lavage was a significant
risk factor for acquisition of MDR-Ab. The results must be interpreted with
caution given the small sample size and wide 95% CIs. Case-control studies
can demonstrate an association between factors such as pulsatile lavage therapy
and acquisition of MDR-Ab but cannot lead to conclusions regarding causality.
When taken with the results of the epidemiologic investigation, however, these
data suggest a strong association between pulsatile lavage wound care and
the acquisition of MDR-Ab in this outbreak.
The association of pulsatile lavage with this outbreak highlights the
need for appropriate infection control measures when using this method of
wound care. We hypothesize that MDR-Ab was disseminated during the pulsatile
lavage treatments leading to contamination of the surrounding environmental
surfaces. Two of the case patients in this outbreak presented with MDR-Ab
pneumonia and sepsis, suggesting an inhalational route of exposure. The change
in procedure allowing suction canister inserts to fill before replacing them
may have played a role in transmission of the organism. It is likely that
multiple factors contributed to the development of this outbreak, including
simultaneous treatment of patients on adjacent stretchers, open supply shelves
that make cleaning and disinfection difficult, and high room humidity from
the whirlpools. The outbreak strain of MDR-Ab may have been introduced into
this environment by one of the case patients or, possibly, by an unidentified
Manufacturers of pulsatile lavage equipment note that proper technique
requires close proximity of the device’s suction tip with the wound
bed at all times. Practically speaking, however, uneven wound contour, momentary
breaches in technique, or patients pulling away from the device will certainly
lead to the chance for splash and environmental contamination during these
procedures. Aerosolization during pulsatile lavage treatment was previously
reported in a study in which organisms from wounds were recovered by air samplers
at 3 ft and 8 ft from patients during the procedure.16 In
addition, 2 recent published reports recognize that there is often splash
during pulsatile lavage treatment and describe barrier precautions to reduce
Our findings emphasize that pulsatile lavage treatment must be performed
under controlled circumstances with appropriate infection control measures
in place. The Centers for Disease Control and Prevention (CDC) now recommend
that health care workers performing pulsatile lavage procedures use appropriate
infection control procedures to minimize aerosols and protect themselves and
their patients from potentially infectious materials. Because of the risks
of splash, health care workers performing pulsatile lavage should use personal
protective equipment, including fluid-resistant gowns, gloves, surgical masks,
eye protection, and shoe and hair covers.13 In
addition, patients receiving pulsatile lavage treatment should wear surgical
masks and all intravenous lines and other wounds must be covered during the
treatment. The procedure should be performed in a private room with easily
washable surfaces, only essential equipment in the room, and no open supply
shelves. Thorough cleaning and disinfection of the room must occur after each
procedure and at the end of the day. As with any medical procedure, all staff
performing pulsatile lavage treatment should be thoroughly trained in proper
technique for use of the device.
In addition to pulsatile lavage, our case-control study found that residence
in a nursing home within the previous 6 months was a significant risk factor
for acquisition of MDR-Ab. Antimicrobial-resistant pathogens are known to
be present in long-term care facilities.19 The
4 case patients who had recent residence in a nursing home were in 4 different
facilities, and it is unclear what role nursing home residence played in this
outbreak. Because wound care, including pulsatile lavage, is frequently performed
in long-term care facilities, practitioners in these facilities must be cognizant
of appropriate infection control procedures to help prevent transmission of
Finally, this outbreak highlights the importance of careful evaluation
of the infection control implications of new therapies, existing therapies
used in a new way, and cost-saving procedural changes. Manufacturers must
ensure that product labeling for new medical devices includes infection control
information and recommendations on appropriate precautions, proper storage,
cleaning and disinfection, and potential infectious risks. Pulsatile lavage
is an emerging technology that is being used with increasing frequency throughout
the United States in a variety of inpatient, long-term care, and outpatient
settings. Infection control information and recommendations must be disseminated
to the health care workers who perform these procedures in diverse settings.
Corresponding Author: Lisa L. Maragakis,
MD, Department of Hospital Epidemiology and Infection Control, Johns Hopkins
Hospital, 600 N Wolfe St, Osler 425, Baltimore, MD 21287 (email@example.com).
Author Contributions: Dr Maragakis had full
access to all of the data in the study and takes responsibility for the integrity
of the data and the accuracy of the data analysis.
Study concept and design: Maragakis, Cosgrove,
Song, Kim, Rosenbaum, Srinivasan, Perl.
Acquisition of data: Maragakis, Kim, Rosenbaum,
Ciesla, Ross, Carroll.
Analysis and interpretation of data: Maragakis,
Cosgrove, Song, Kim, Srinivasan, Ross, Carroll, Perl.
Drafting of the manuscript: Maragakis.
Critical revision of the manuscript for important
intellectual content: Maragakis, Cosgrove, Song, Kim, Rosenbaum, Ciesla,
Srinivasan, Ross, Carroll, Perl.
Statistical analysis: Maragakis, Cosgrove,
Obtained funding: Maragakis, Perl.
Administrative, technical, or material support:
Song, Kim, Rosenbaum, Ciesla, Ross, Carroll.
Study supervision: Maragakis, Cosgrove, Perl.
Funding/Support: This study was supported in
part by a Prevention Epicenters cooperative agreement (grant UR8/CCU315092)
from the CDC, by grant 5-T32/AI07291 from the National Institutes of Health,
and by grant 1 K01 CI000300-01 from the CDC.
Role of the Sponsors: The sponsors had no role
in the design and conduct of the study, in the collection, analysis, and interpretation
of the data, or in the preparation, review, or approval of the manuscript.
Dr Srinivasan works in an investigative capacity at the Division of Healthcare
Quality and Promotion and has no involvement with the funding aspect of the
Acknowledgment: We are indebted to Beryl Rosenstein,
MD, for support and for facilitating the implementation of appropriate patient
safety measures; John G. Bartlett, MD, for advice and for reviewing an early
version of the manuscript; David Blythe, MD, MPH, Maryland Department of Health
and Mental Hygiene, for advice and assistance; Melissa Johnson, PT, BS, and
the members of the Department of Physical Medicine and Rehabilitation, Johns
Hopkins Hospital, for reviewing treatment logbooks and contacting and evaluating
patients; Henry Garbelman, for design and construction of new pulsatile lavage
treatment rooms; the members of epidemiology and infection control departments
of other institutions affected by MDR-Ab, for their advice on prevention and
control of the organism; and Harriett Loehne, PT, CWS, for assistance with
development of new infection control guidelines.