Menu Close

Premier Safety Share® Newsletter – May 2010

May 24, 2010

Premier Safety Institute - Safety Share Newsletter

Check out our tools on readmissions, safety culture, safety data, CMS dashboard and more.

Gina Pugliese, editor


Follow us on

Follow us on Twitter 

Subscribe Now




Bloodstream infection rates at nearly zero for three years

Updated report Ninety Michigan
intensive care units (ICUs) reported the reduced
rates of catheter-related bloodstream infections
achieved in an initial 18-month study were sustained
for an additional 18 months. The mean and median
rates of catheter-associated bloodstream infections
(CLABSI) decreased from 7.7 and 2.7 percent at
baseline to 1.3 and 0 percent at 16-18 months, and
to 1.1 and 0 at the 34-36 month post-implementation
period, representing a greater than 60 percent
reduction in infection rates from baseline sustained
at the end of the 36-month period.    




Baseline 7.7 2.7
16-18 months 1.3 0
34-36 months    
Post implementation 1.1 0

Previously the New England
Journal of Medicine
, December 2006, reported that
the Keystone ICU project interventions, a group of
five evidence-based recommendations known as the
CLABSI reduction “bundle,” resulted in a 66 percent
decrease in the rates of CLABSI across 103 ICUs in
the state of Michigan. The current Keystone ICU
study reported on the18-month post-implementation
period, which included CLABSI rate measurement and
feedback to the ICU teams. (Pronovost
et al. 2010

Process During the
sustainability period, ICU teams were instructed to
integrate the interventions into staff orientation,
collect monthly data from hospital infection control
staff, and continue reporting infection rates to
appropriate stakeholders. The authors found that the
reduced rates of CLABSI achieved in the initial
18-month post-implementation period were sustained
for an additional 18 months as participating ICUs
integrated these quality improvement interventions
into standard practice. These reductions in CLABSI
represent a greater than 60 percent decrease in
infection rates when compared to the baseline
measurement reported prior to the combined 36-month

Although implementation of
CLABSI bundle is becoming common practice in ICUs,
few quality improvement projects have evaluated
their sustainability on a large scale. The authors
attribute their success to the continued utilization
of the quality improvement model from the
Keystone ICU project including maintaining a
team safety culture, orienting new staff, collecting
monthly data and reporting infection rates to
appropriate stakeholders after the initial
evaluation period. The authors also note that
integration of the CLABSI “bundle” into practice in
all ICUs across the country could result in
significant reductions in the estimated 82,000
infections and $2.3 billion costs of CLABSI
annually. See:

Safety Share: “New techniques dramatically decrease
ICU infections”

Safety Share: “Michigan’s success in preventing
infections expanded to 10 states”

More resources: See
Premier Safety Institute’s Bundling – Evidence-based



Baxter to recall COLLEAGUE infusion pumps

On May 3, 2010, Baxter
International announced that it will recall all
COLLEAGUE infusion pumps from the U.S. market as
part of an existing 2006 consent decree with the
Food and Drug Administration. Baxter noted in its
press release that it will work with the FDA to
ensure the recall process provides customers
appropriate alternatives for supporting patient
needs, including an offer to exchange COLLEAGUE
pumps for Baxter’s Sigma SPECTRUM.

Baxter’s action was prompted by
an FDA order for the company to “recall all of the
COLLEAGUE infusion pumps currently in use in the
United States” because of a “longstanding failure”
to correct problems with the pumps. The FDA
estimates that there are as many as 200,000 of these
pumps currently in use. FDA and Baxter are still
working out the specific details of the recall.

ECRI Institute prepares
ECRI Institute prepared a Special
Report, “FDA
Orders Recall of Baxter COLLEAGUE Infusion Pumps:
Hospitals May Continue to Use for Now,”
that is
publicly available. According to Eric Sacks,
director of Healthcare Product Alerts at ECRI
Institute, “Hospitals are in an awkward situation
having been notified that the pumps will be recalled
related to safety problems, but not having the
details or remedial options available yet.” The
recent recall order does not stem from any newly
identified design or safety risk, but rather, as FDA
explained in its press release, because of concerns
regarding the time frame in which Baxter could
adequately address previously identified safety
issues. “It is important that everyone understand
that FDA has stated that at this time users can
continue to use COLLEAGUE pumps,” added Sacks. “The
best way to respond for now is to organize your team
and be prepared to take action when the actual
recall is announced.” In preparation, hospitals
should identify internal stakeholders, notify them
of the latest announcement, and ask them to help
identify key institutional factors related to future
replacement of the pumps across the enterprise.

Baxter to offer replacement
of COLLEAGUE pumps
What is known at this time is
that FDA’s action will require Baxter to compensate
users of the COLLEAGUE in some way. In its May 3,
2010, press release, Baxter said it anticipates
that, among alternatives to be provided to
customers, the company will offer to exchange
Baxter’s Sigma SPECTRUM infusion pumps for COLLEAGUE
infusion pumps without charge to customers. For
those who choose not to adopt the SPECTRUM, Baxter
will be required to offer some form of monetary
compensation. The company is currently negotiating a
value for COLLEAGUE pumps with the FDA. Baxter also
has indicated that facilities that lease the pumps
may be entitled to terminate leases without penalty.

Meanwhile, ECRI Institute has
reminded hospitals that “transitioning to new
infusion pumps is no easy task” and should not be
done in haste. For example, regardless of which
replacement pump is selected, facilities using
COLLEAGUE will need to develop a new drug library –
a collaboration between pharmacy, nursing, and
medical staff that typically takes three to six

Resources for users More
details, including ECRI Institute’s Special Report,
the FDA press release, and Baxter’s press release,
are available on the

ECRI Institute
website. The FDA also has issued
a “Questions and Answer” and

feedback form
on its website. Current users are
urged to use the feedback form to communicate any
concerns to the FDA regarding this action, including
suggesting an adequate timeline to permit planning
and a safe transition to an alternative infusion

FDA announces a new
initiative to address infusion pump safety

Infusion pumps have been the source of persistent
safety problems and not confined to one manufacturer
or type of device, according to the FDA. In the past
five years, the FDA has received more than 56,000
reports of adverse events associated with the use of
infusion pumps. Those events include serious
injuries and more than 500 deaths. Between 2005 and
2009, 87 infusion pump recalls were conducted to
address safety concerns, according to FDA data. In
April, FDA announced a new initiative to address
safety problems and will be establishing additional
premarket requirements that will include static
testing in FDA’s facilities before device
submissions. More on FDA’s initiative:

More on FDA’s initiative:



Methods to detect adverse events are unreliable; hospital incident reports miss 93 percent: OIG study

Shortcomings in current
screening methods for adverse events have
implications for Medicare payments and federal
initiatives to monitor them, according to a study by
the Office of Inspector General (OIG); the study
found that present on admission codes and incident
reporting systems are missing the majority of

Methodology Five methods
of identifying adverse events (AE) in hospitalized
Medicare patients were examined by the OIG in its
report to Congress titled, “Adverse Events in

Methods for Identifying Events

A sample of 278 Medicare
beneficiary hospitalizations from acute care
hospitals were screened using five methods:

  • Nurse review of medical
    records (IHI global trigger tool);
  • Interviews of Medicare
  • Agency for Healthcare
    Research and Quality (AHRQ) Patient Safety
    Indicators (PSI);
  • Hospital billing data –
    Present on admission (POA) diagnosis codes; and
  • Review of internal hospital
    incident reports.

Overall, 63 percent of the
possible events identified by these five screening
methods were not associated with actual events.

POA codes and hospital
incident reports found to be unreliable
diagnosis codes for POA were inaccurate or absent
for seven of the 11 Medicare hospital-acquired
conditions (HACs) identified by the physician
reviewers. This inaccuracy would prevent Medicare’s
automated payment software from identifying the HACs,
resulting in both overpayments and limiting use of
billing data to monitor quality of care in

Hospital incident reports were
not generated for 93 percent of the events
identified, including some of the most serious
events. This lack of complete reporting prevents
hospitals from tracking and reporting AEs as
required by regulation. It also suggests that
hospital reporting systems may be unreliable as a
source of information for patient safety
organizations (PSO), entities that aggregate and
analyze information about AE voluntarily reported by

Recommendations The OIG
recommended that CMS and AHRQ explore ways to
identify AEs when conducting medical record reviews
for other purposes, e.g., during hospital CMS
compliance surveys, possibly requiring new
interpretative guidelines for tracking AEs, which
currently do not exist. The OIG stated that CMS
needs to verify that claims are coded accurately and
completely to ensure identification of Medicare HACs,
and AHRQ should inform PSOs that internal hospital
incident-reporting systems may provide insufficient
information about AE.  

Related study The
discrepancies found are consistent with another
study conducted at the Mayo Clinic (Premier
Safety Share Sept 09 Mayo clinic
), which also
found major variations in identifying AEs, depending
on the methods used. This study similarly used
external (AHRQ PSI) and internal (incident-reporting
system) indicators, plus a subset sample using the
IHI global trigger tool. It found inconsistencies in
the definitions and detection methods. Based on their
findings, these authors suggested that measuring AE
requires a multifaceted approach, cautioning against
use of any one indicator for public reporting and
performance comparisons.



Lab errors – Call for more automation to prevent mislabeling, misidentification and related harm to patients

Mislabeling of clinical
laboratory and pathology specimens results in repeat
diagnostic procedures and delayed or unnecessary
surgical procedure. Use of electronic patient
identifiers is one key prevention strategy.

Analysis More than 200
root cause analyses (RCA) over an eight-year period
in the
VA system identified system vulnerabilities
in specimen collection, processing, analysis, and
reporting associated with patient misidentification
involving the clinical laboratory, anatomic
pathology, and blood transfusion services. Data were
categorized by three stages of the laboratory test
cycle. Patient misidentification accounted for
nearly three quarters of 253 adverse events, and
occurred in all three stages of the test cycle:
pre-, analysis, and post-analysis. Examples follow:

  • Pre-analysis The
    largest percentage (73 percent of the 182
    misidentification errors) occurred during
    collection at admission such as wrist bands
    mislabeled or orders for the wrong patient due
    to similar names and Social Security numbers (SSN).
    This included two-source patient identification
    for clinical laboratory specimens and failure of
    two-person verification of patient identity for
    blood bank specimens. 
  • Analysis 20 percent
    of errors included relabeling, as well as
    misidentified microscopic slides due to a
    failure of two-pathologist verification for
    cancer diagnosis as well as wrong patient
    transfusion due to mislabeled blood products or
    failure of two-person verification for blood
    products before release by the blood bank.
  • Post-analytic phase
    The smallest number (7 percent) included results
    being reported into the wrong patient medical
    record or incompatible blood transfusions
    associated with failed two-person verification
    of blood products.

Findings Patient
misidentification figured greatly in patient errors,
whether in the clinical laboratory, anatomic
pathology, or blood transfusion processes. Errors
were analyzed with data mining software, and
determined that the pace of reporting increased in
2007. Authors speculated that beginning in 2006, the
VA expansion of the use of wireless bar-coding
alerted the system to the problems of mislabeling
and may have led to increased awareness and
reporting due to mislabeling during specimen
collection. Another important issue emerged causing
misidentification – that is, patients with similar
names, birthdays, and the last four digits of their
Social Security number.

Solutions As a result of
meticulous analysis and identification of the root
causes, the authors provide a valuable table which
identifies an intervention for each type of error,
along with the positive effect in reducing errors.
For example, one change includes: 1) the use of the
full SSN and 2) use of electronic means for
at least two of three patient identifiers (name,
birth date, or SSN). The overall theme identified to
reduce error is increased automation for as many
processes as possible, with an expectation that such
a focus will lead to systemic mitigation and
prevention strategies.

More resources: Premier
Safety Institute

UDI and Bar-coding



CDC updates H1N1 infection control guidance and reports flu vaccination rate the highest this season

The Centers for Disease Control
and Prevention (CDC) is updating the

“Interim Guidance on Infection Control Measures for
2009 H1N1 Influenza in Healthcare Settings,
Including Protection of Healthcare Personnel,”

noting the availability of an effective vaccine and
experience with the virus. In a separate report, the
CDC reported higher flu vaccination rates this
season among healthcare workers than ever before

In a May 3, 2010, posting on
the “Question and Answer” section of CDC’s H1N1 Flu
website, the question of how infection control
measures may be revised for the upcoming season was
addressed. The posting explains the revision will
take into consideration the availability of a safe
and effective vaccine and new information about the
health impact of the novel H1N1 influenza strain –
both of which will significantly change the
circumstances and offer guidance on how best to
control this infection. The response describes the
process of how this guidance is being updated to
reflect these changes.

Highest reported influenza
vaccination rates among healthcare workers
On a
related note, the

CDC’s April 2, 2010, “Morbidity and Mortality Weekly
notes that since 1989, overall
influenza vaccination coverage among healthcare
personnel has never exceeded 49 percent in any
season. However, the results of a population-based
survey during January 2010 estimated vaccination
coverage among healthcare workers was 37.1 percent
for 2009 pandemic influenza A (H1N1) and 61.9
percent for seasonal influenza. Overall, 64.3
percent received either of these influenza vaccines,
higher coverage than any previous season.




Premier – Safe injection practices survey

The Premier Safety Institute
invites you and your colleagues to take a short
survey to identify current practices of healthcare
professionals who prepare and administer parenteral
and injectable medications. All data are anonymous
and aggregated results will be summarized on the
Safety Institute website. The information will also
be shared with professional associations, the
Centers for Disease Control and Prevention, the Food
and Drug Administration and other groups, to guide
research and education related to reducing risks to
patients. You are also invited to forward this
invitation to staff who may prepare and administer
parenteral and injectable medications in your
organization and any affiliated locations, such as
clinics, surgicenters and urgent care. Staff may
include pharmacists, anesthetists, physicians,
nurses, respiratory therapists and techs, for
example. To invite others to participate, please
forward this link
where the
survey is located. To go directly to the survey go


HRET/AHA – Guide to address avoidable readmissions

A free guide to help hospital
leaders address avoidable hospital readmissions, the
“Health Care Leader Action Guide to Reduce Avoidable
Readmissions,” is a quick, simple resource that
outlines four steps to avoid readmission:

  1. Examine your hospital’s
    current state of readmissions;
  2. Assess and prioritize your
    improvement opportunities;
  3. Develop an action plan of
    strategies to implement; and
  4. Monitor your hospital’s

was funded and produced by the Commonwealth Fund,
the John A. Hartford Foundation, and the Health
Research & Educational Trust (HRET) of the American
Hospital Association.


AHRQ – 2010 patient safety culture survey data report

Survey on Patient Safety Culture 2010 Comparative
Database Report”

is an update of the
2009 report that provides results from 885 hospitals
and 338,607 hospital staff respondents using the
safety survey tool. The 2010 report also includes a
chapter on trending that presents results showing
change over time for 321 hospitals that administered
the survey and submitted data more than once. It can
be used for comparisons with other hospitals using
the survey, for internal assessment and learning, to
identify strengths and areas for potential
improvements, and to assess changes in patient
safety culture over time. Hospitals seeking to
submit data for inclusion in the comparative
database can do so each year between May 1 and June


FGI – 2010 design/construction guidelines

The 2010 Facility Guidelines
Institute (FGI)’s “Guidelines for Design and
Construction of Health Care Facilities” provides
guidance on minimum program, space, and design
guidelines for clinical and support areas of
hospitals, outpatient facilities, long-term care,
residential care facilities and mobile units. The
Joint Commission, many federal agencies, and
authorities in at least 42 states use the guidelines
either as a code or a reference standard when
reviewing, approving, and financing plans;
surveying, licensing, certifying, or accrediting
newly constructed facilities; or developing their
own codes. Prominently featured are terms,
requirements and best practices with infection
prevention implications. The guidelines may be
purchased at the

American Society for Healthcare Engineering (ASHE)
or at the

FGI website
. For more information, see Premier’s
Green, Safer Building



HHS – Data transparency initiative “CMS Dashboard”

The Department of Health and Human Services (HHS) has launched a tool for viewing national, state, hospital and DRG claims payment and volume data from Medicare’s inpatient prospective payment system. The
CMS Dashboard tool is part of the Obama administration’s new “open government” initiative. The CMS Dashboard BETA is a beta release and offers statistical views of the Inpatient Prospective Payment System (IPPS) data as it relates to claims payment and volume as collected by CMS. The data contained in this version is current as of April 2010 for inpatient discharges from January 2006 to March 2010. Future releases may contain additional CMS program data.


AHRQ – Nursing handbook on patient safety and quality

Nurses play a vital role in improving the
safety and quality of patient care, not only in the
hospital or ambulatory treatment facility, but also
with community-based care and the care performed by
family members. Nurses need to know what proven
techniques and interventions they can use to enhance
patient outcomes. To address this need, the Agency
for Healthcare Research and Quality (AHRQ) has
prepared a comprehensive handbook:
“Patient Safety
and Quality: An Evidence-Based Handbook for Nurses.”


Michigan’s MARR – Antimicrobial resistance prevention toolkit for LTC

The Michigan Antibiotic Resistance Reduction
(MARR) Coalition’s “Long
Term Care Toolkit
” is designed to help long-term
care facilities implement the “12 Steps to Prevent Antimicrobial Resistance Among
Long-Term Care Residents,” a set of
recommendations developed by the CDC as part of its
campaign to prevent antimicrobial resistance in
healthcare settings.

Safety Institute