Preventing Infections in Hemodialysis

Priti R. Patel, MD, MPHDivision of Healthcare Quality Promotion

Centers for Disease Control and PreventionNothing to disclose

July 28, 2010

The findings and conclusions in this presentation are those of the author and do not necessarily represent the views of the Centers for Disease Control and Prevention.

Have you..

Been involved in central line associated bloodstream infection (CLABSI) prevention efforts in your facility?

Had any involvement in your facility’s dialysis center?

Important Trends

• Growing dialysis population; ~350,000

• Mortality, increasing morbidity from infections

• Antimicrobial resistant infections, emerging patterns of resistance

United States Renal Data System (USRDS) 2008 Annual Data Report

Invasive Methicillin-Resistant S. aureus (MRSA) Infections, 2005• Incidence of invasive MRSA infections

45.2 cases per 1,000 dialysis population

= 100 X rate in general population (0.2 – 0.4 per 1000)• Dialysis patients

– ~0.1% of the U.S. population– 15% of all invasive MRSA infections

• Invasive MRSA in dialysis – 86% were bloodstream infections (BSIs) – 90% required hospitalization, mortality = 17%

CDC. MMWR 2007; 56(09):197-9

Delivery of Dialysis Care• 5,240 dialysis facilities nationwide

– ~850 are hospital-based• Increasingly consolidated ownership• 2 large, for-profit chains treat ~60% of all

patients• Medicare primary payor (ESRD program)• Economic incentives – major driver • Facilities frequently lack infection control

expertiseUnited States Renal Data System (USRDS) 2008 Annual Data Report

Nonhospital Healthcare Settings: The Next Frontier

What can be done about infections in outpatient populations?

• Improve infection control practices in outpatient settings – Regulatory efforts– Engage hospital infection control expertise

• Prevention research / initiatives– Demonstrate preventability– Could there be a dialysis “bundle”?

• Efforts in inpatient settings

Regulatory Changes

• April 2008 – The Centers for Medicare and Medicaid Services (CMS) released new conditions for coverage for End Stage Renal Disease (ESRD) facilities– First comprehensive revision since 1976– Incorporates CDC / HICPAC infection control

recommendations– First time infection control is a separate condition

New CMS Conditions for Coverage

• Includes by reference:– Recommendations for Preventing Transmission

of Infections Among Chronic Hemodialysis Patients, 2001

– Guidelines for the Prevention of Intravascular Catheter-Related Infections, 2002

Links: http://www.cdc.gov/mmwr/preview/mmwrhtml/rr5005a1.htm

http://www.cdc.gov/mmwr/preview/mmwrhtml/rr5110a1.htm

Early Impact & Perspective

• New Conditions went into effect October 2008• Has helped to highlight the importance of infection

control in dialysis settings• Infection control has been the most common

category of citation during the new survey process• Demonstrated gaps:

– In adherence to recommendations prior to the new conditions

– In the recommendations and conditions

CMS Conditions for Coverage: New Opportunities

• Improving infection control– Reduce infections and improve patient outcomes

• Infection prevention & the Conditions – Conditions are a floor, not a ceiling– Need to go beyond requirements to truly prevent

infections

BSIs in Hemodialysis: Capturing our Attention

Epidemiology of Infections among Hemodialysis Patients

• Infections are the 2nd leading cause of death (15% of deaths)

• Site of infection– 57% vascular access– 23% wound– 15% lung– 5% urinary tract

USRDS 2005 Annual Data ReportTokars, Miller, Stein. AJIC 2002;30:288-295

How Common are Vascular Access Infections, Including BSIs?

• Estimate in the literatureCatheter-related BSI:

2.5 – 5.5 per 1000 patient-days

0.9 – 2.0 episodes per patient-year• Surveillance data

CDC’s National Healthcare Safety Network

(NHSN) dialysis event module

Allon. AJKD 2004; 44:779-91

Outcomes of S. aureus BSI

Among hemodialysis patients admitted with S. aureus bacteremia1:

– Avg. length of stay: 13 days– Cost of hospital admission = $20,685– 31% had complications– 21% had to be readmitted – Within 12 weeks,

• 19% died from any cause • 11% died due to S. aureus

1. Engemann. ICHE 2005(26):534-9 2. Nissenson. AJKD 2005(46):301-8

Hospitalization Rates

• Cause-specific hospitalization rates among hemodialysis patients, 2006:– Vascular access infection = ~125 admissions / 1000 pt-yrs– Bloodstream infection = 103 admissions / 1000 pt-yrs– Pneumonia = 76 admissions / 1000 pt-yrs

• Since 1993, rates* have increased for:

All infections (+34%)Bloodstream infection (+31%)Cellulitis (+20%)Pneumonia (+7%)

USRDS 2008 Annual Data Report%

cha

nge

sinc

e 19

93

Change in hospitalization rate

Year

(* adjusted for age, race, sex, and cause of ESRD)

Trends in Incidence of Central Line-Associated Bloodstream Infections by ICU Type—United

States, 1997-2007C

LA

BS

Is p

er 1

,000

C

entr

al L

ine

Day

s

Year

†

Slide courtesy: Deron Burton, CDC Source: NNIS (< 2005) and NHSN (> 2005). Data represents 1,681 units, 16,225,498 patient days, and 33,587 CLABSIs

0

1

2

3

4

5

6

7

8

9

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Cardiothoracic CoronaryMedical Medical/Surgical--Major TeachingMedical/Surgical--Non-Major Teaching PediatricSurgical

Distribution of Facility BSI Rates by Vascular Access Type,

NHSN 2007-2008 (N=49)BSIs per 100 patient-months

PercentileFistula Graft Tunneled CVC

10th 0.00 0.00 1.50

25th 0.27 0.00 2.54

50th (median) 0.66 1.02 4.76

75th 1.13 1.88 8.89

90th 3.17 3.81 14.39

Pooled mean 0.68 1.14 3.93

Distribution of Facility BSI Rates by Vascular Access Type, NHSN 2007-2008 (N=49)

BSIs per 100 patient-months

PercentileFistula Graft Tunneled CVC

10th 0.00 0.00 1.50

25th 0.27 0.00 2.54

50th (median) 0.66 1.02 4.76

75th 1.13 1.88 8.89

90th 3.17 3.81 14.39

Pooled mean 0.68 1.14 3.93

~1.6 per 1,000 catheter-days

~0.5 per 1,000 catheter-days

~4.8 per 1,000 catheter-days

BSI Rate in Patients with Tunneled CVC by Facility Type,

NHSN 2007-2008

# BSIs in tunneled CVC patients

# Tunneled CVC patient-months

Pooled mean rate (per 100 CVC patient-months)

RR (95% CI)

Hospital-based (n=26)

490 13,018 3.76 Ref.

Other outpatient (n=23)

433 8,208 5.28 1.40 (1.23,1.60)

Vascular Access

Vascular Access InfectionsRisk Factors

• Type of access– catheter >> – graft >– fistula

• Lower extremity access• Recent access surgery• Trauma, hematoma,

dermatitis, scratching

• Poor hygiene• Poor needle insertion

technique• Older age• Diabetes• Iron overload• Others

Differences in Event Rates: Fistula vs. Catheter

Fistula CatheterInfections at access site Lowest Highest

Infections at other sites Lowest Highest

Hospitalizations Lowest Highest

Deaths from Infection Lowest Highest

Deaths from all causes Lowest Highest

Tokars, Miller, Stein. AJIC 2002;30:288-295Pastan, Soucie, McClellan. Kidney Int 2002;62:620-626

Rate of Access-Related Bloodstream Infection by Vascular

Access Type

0

1

2

3

4

5

6

7

8

Ac

ce

ss

-re

late

d b

ac

tere

mia

ra

te

(pe

r 1

00

pa

tie

nt-

mo

nth

s)

Fistula Graft CuffedCatheter

Non-cuffed

catheter

Dialysis Surveillance Network 1999-2005

Types of Vascular Access, U.S. Hemodialysis Patients, by Year

0

20

40

60

80

95 96 97 99 00 01 02

Year

% o

f P

atie

nts

Dia

lyze

d

Th

rou

gh Graft

Fistula

Catheter

Finelli, Miller, Tokars. Semin Dial 2005;18:52-61

Prevalent Hemodialysis Patients with AV Fistula

USRDS 2008 Annual Data Report

USRDS 2008 Annual Data Report

Prevalent Hemodialysis Patients with AV Graft

USRDS 2008 Annual Data Report

Prevalent Hemodialysis Patients with Catheter

Fistula First Initiative

Spergel LM. Seminars in Dial.

Goals:

66% AV fistula use

<10% long term catheter use

Vascular Access at First Outpatient Dialysis, 2006

USRDS 2008 Annual Data Report

Catheter – 82%

17% maturing fistula

3% maturing graft

AV graft – 4%

AV fistula – 12%

Vascular Access Distribution Among Patient Census, NHSN 2007-2008 (N=49)

Fistula Graft Tunneled CVC

Nontunneled CVC

Median 48.1% 15.5% 34.8% 0.0%

Range 26.1 - 66.8% 2.3 - 38.5% 14.8 - 69.8% 0.0 - 15.3%

Hospital-Affiliated vs. Other Outpatient Facilities,

NHSN 2007-2008

Median by Facility Type

Fistula Graft Tunneled CVC

Nontunneled CVC

Hospital-based (n=26) 45.9% 14.2% 37.7% 0.0%

Other outpatient (n=23) 51.7% 17.8% 27.7% 0.0%

p=0.10 p=0.02

Prevention

Fact: Indwelling catheters are the single most important factor contributing to bloodstream infection in hemodialysis patients.

Actions: Hemodialysis: Use catheters only when essential Maximize use of fistulas Remove catheters when they are no longer

essential

Prevent Infection:Get the catheters out

Fact: Careful infection control can prevent dialysis-related infections.

Actions: Follow established guidelines for access

careUse proper insertion and catheter-care

protocolsRemove access device when infected

Prevent InfectionOptimize access care

BSIs in Hemodialysis: Achieving Success

Surveillance & Feedback

Intervention Bundle

Prevention Collaboratives

Bloodstream Infection Interventions: Pittsburgh Regional Healthcare Initiative,

32 hospitals, 66 ICUs

Intervention:• Promotion of evidence-based catheter insertion practices• Development / promotion of educational module• Promotion of standardized tools for recording catheter insertion

practices• Promotion of standardized catheter insertion supply kits• Regular feedback of BSI rates

• Standardized definitions and case finding methods• Process to share information and experience

MMWR 2005;54:1013-16

Overall rate reduction of 68%

MMWR 2005;54:1013-6

Semi-Annual Central Line-associated Bloodstream Infection Rates in Medical-Surgical Intensive Care Units Participating in the Southwest

Pennsylvania Collaborative and NNIS, 2001-2005

0

1

2

3

4

5

Apr 2001-Sept 2001

Oct 2001-Mar 2002

Apr 2002-Sept 2002

Oct 2002-Mar 2003

Apr 2003-Sept 2003

Oct 2003-Mar 2004

Apr 2004-Sept 2004

Oct 2004-Mar 2005

Semi-annual period

Ra

te p

er

10

00

ce

ntr

al lin

e-d

ay

s

*NNIS data only available for Oct-Dec 2004

*

*

p<0.001

p=NS

Michigan Keystone Initiative

• Intervention:• Training of team leaders in science of safety• Standardized central-line cart with necessary supplies• Checklist was used to ensure adherence to catheter-

insertion practices • Providers were stopped (in nonemergency situations) if

these practices were not being followed• Removal of catheters was discussed at daily rounds• Regular feedback of BSI rates

Pronovost et al. NEJM 2006;355:2725-2732

Michigan Keystone ICU Project(103 ICUs, 67 hospitals)

Pronovost et al. NEJM 2006;355:2725-2732

Michigan Keystone ICU Project(103 ICUs, 67 hospitals)

Overall rate reduction of 66%

Conclusions from Pittsburgh and Michigan Experiences

• Decreases in central line-associated BSI rates >60% achieved in hospital ICUs of varying types

• The prevention practices utilized during these interventions were not novel– Improving adherence to existing evidence-based

practices can prevent BSIs– Collaboration may be helpful in identifying and

overcoming commonly shared barriers to adherence

Conclusions from Pittsburgh and Michigan Experiences

• Results from successful collaborative demonstration projects may be an important strategy for influencing global changes in practice in ways that improve quality– Disarms uncertainties about preventability that

can hamper improvement efforts– Helps identify practical strategies that can be

successful across many facilities

A Successful, Multi-Center BSI Prevention Collaborative For Hemodialysis Patients

Will Have National Impact

• Motivated hemodialysis centers who are interested in working in partnership with others to:– Identify setting-specific barriers and challenges (because dialysis

centers are very different from ICUs)– Identify workable and practical solutions to those barriers– Be open to innovation– Collect and share data in a uniform fashion – Contribute to an effort that will likely have major and enduring

impact on the health of hemodialysis patients not only in your center, but across the Nation

CDC-Sponsored Dialysis Collaborative

Share information related to best practices

Work to develop and implement practical solutions

Prevent BSI & improve patient outcomes

Establish collaboration of outpatient dialysis facilities all reporting to NHSN

BSI Prevention, Bundles & Collaboratives

• Good evidence-base for interventions• Logical extension of efforts to reduce BSIs in

inpatient settings– Recognizing challenges unique to dialysis

• Early evidence supports the preventability of these infections

Proposed Core Interventions for CDC Dialysis Collaborative

1. Surveillance for positive blood cultures, antimicrobial starts, and hospitalizations using NHSN – Conduct monthly surveillance for dialysis events and enter events into NHSN. CDC will provide facility rates and comparisons to other facilities. Facilities should actively share results with front-line providers.

 2. Chlorhexidine for skin antisepsis – Use chlorhexidine (2% or greater) as the

first line agent for skin antisepsis. Povidone-iodine, preferably with alcohol, is an alternative.

 3. Antimicrobial ointments – Apply bacitracin/gramicidin/polymixin B or povidone-

iodine ointment to catheter exit sites at each dressing change. 4. Hand hygiene surveillance – Perform monthly hand hygiene audits with

feedback of results. 

Proposed Core Interventions for Dialysis Collaborative

 

5. Catheter care/ access observations – Perform monthly audits of catheter care and accessing practices to ensure adherence to facility guidelines. This may include use of a mask while connecting and disconnecting catheters and during dressing changes.

 6. Patient education/engagement – Provide standardized, basic education to all

patients including (but not limited to) care of vascular access, hand hygiene, cleansing vascular access, and instructions for access management when away from the dialysis unit.

 7. Staff education and competency – Provide regular training for staff on infection

control topics, including care of access and aseptic technique. Perform evaluation of competency for skills such as catheter care and accessing at least yearly and upon hire.

 8. Catheter reduction – Incorporate efforts within the facility (e.g., patient

education) to reduce catheters by identifying barriers to permanent vascular access placement and catheter removal.

The Value of Surveillance • Busy London dialysis unit: 112 patients• Implemented CDC dialysis surveillance; described

their experience over 18 months• After initial set up, required 2 hours per month• Outcomes: Reductions in

– Access-related bacteremia – Antibiotic usage – Hospital admissions

George A, Tokars JI, Clutterbuck EJ, et al. BMJ 2006; 332:1435-1439

George A, Tokars JI, Clutterbuck EJ, et al. BMJ 2006; 332:1435-1439

Antimicrobial Starts

George A, Tokars JI, Clutterbuck EJ, et al. BMJ 2006; 332:1435-1439

Access-Related Bacteremia

Observations• “Surveillance raised awareness and provided a

cornerstone for improved infection control and line care involving all staff of the dialysis unit.”

• “The data feedback generated unit led programmes of risk reduction and infection control.”

George A, Tokars JI, Clutterbuck EJ, et al. BMJ 2006; 332:1435-1439

Median Facility

Collaborative Feedback Report

Example of an Intervention Involving A Vascular Access “Bundle”

• Healthcare worker education (May 2006)– Hand hygiene, aseptic technique, access site care

• Feedback of VAA-BSI surveillance data to facility staff and physicians (May 2006)

• Use of 2% chlorhexidine-70% alcohol solution for catheter site care and prior to accessing A-V fistulas and grafts (July 2006)

• Patient education (January 2007)– Access site care– Benefits of an A-V fistula– Vascular Access Liaison (May 2007)

Data presented at SHEA Annual Conference, Mar. 2009Slide courtesy: David Calfee, MD, Mount Sinai School of Medicine

Results:Incidence of VAA-BSI Over Time

0

1

2

3

4

5

6

2005 Q1-Q2

2005 Q3-Q4

2006 Q1-Q2

2006 Q3-Q4

2007 Q1-Q2

2007 Q3-Q4

2008 Q1-Q2

2008 Q3-Q4

VA

A-B

SI

pe

r 1

00

pt-

moOverall Catheter AVG-AVF

Data sharingHCW education

Chlorhexidine HCW education

Patient educationVascular Access

Liaison

p=0.03

p=0.01

p=0.16

Data presented at SHEA Annual Conference, Mar. 2009Slide courtesy: David Calfee, MD, Mount Sinai School of Medicine

• Highlights from their “expanded” bundle:– Catheter hub disinfection with chlorhexidine gluconate 3.15% – Hand hygiene plus gloving prior to contacting patients or machines– Relocating supplies, from near the patient to a central area– Strengthening environmental cleaning practices – Chlorhexidine-impregnated sponge dressing for catheters deemed high risk– Strengthening of a comprehensive fistula placement program

• Results:– Reduction in central line BSI rate from 2.4 per 100 patient-months to 0

Getting to Zero: Outpatient Hemodialysis Catheter-Associated Bloodstream Infections

Virginia R. Bren, RN, MPH, Altru Health System, Grand Forks, ND

Friday, March 19, 2010 SHEA poster presentation

Where Do We Go From Here?

• Prevention Efforts– Collaborative approach

• Improving adherence to evidence-based practices– Expanding surveillance & enhancing it’s utility

• Studies – Target prevention efforts– Identify new strategies

• Creative Thinking & Strategic Partnerships– How to bridge the inpatient-outpatient gap – Overcoming challenges in resources and expertise– Role of infection preventionists

BSIs in Hemodialysis: Achieving Success

Surveillance & Feedback

Intervention Bundle

Prevention Collaboratives

Join Us!

http://www.delmarvafoundation.org/providers/ambulatory/dialysis/index.html

Contact: NHSN@cdc.gov or Ppatel@cdc.gov

What else can be done?• Check out practices in your inpatient unit• Interface with dialysis staff• Join our conference calls• Check out new APIC Guide

Thank you!

PREVENTION IS PRIMARY!

Methicillin-resistant Staphylococcus aureus bloodstream infections

Epidemiological category

2005 pooled mean incidence (per 10,000 person years)

2008 pooled mean incidence (per 10,000 person years)

Modeled yearly percent change (2005-2008)

P-value

Hospital-Onset 0.88 0.62 -11.2% 0.001

Healthcare-associated Community-onset

1.97 1.62 -6.6% <0.001

Dialysis in last year 501.5 404.48 -6.4% 0.02

No dialysis in last year 1.58 1.31 -7.2% 0.006

Median Facility

Summary of Vascular Access Infections

• Major cause of morbidity & mortality• Indicators moving in the wrong direction:

– Increasing morbidity, catheter use• New regulatory efforts

– CMS requirements probably not sufficient to solve the problem

• Prevention efforts are underway; more needed– Evaluate current initiatives– Strategies to improve adherence– New technologies

Landscape & New Requirements

• Gaps– Lack of reimbursement for HCV screening

• Regulations can’t solve every problem– Requiring components (e.g., surveillance & QI)

doesn’t necessarily equate to a functional IC program

– Overcoming challenges in resources and expertise

Landscape of Infection Control in Dialysis: New Requirements

• Early successes– Increased awareness of infection control issues– Improved adherence to HBV testing & isolation

requirements– Essentially eliminated some breaches: re-use of single

dose medication vials

• Innovative uses– Promote interaction with public health & reporting

• Promising – Focus on immunizations– Dialysis technician certification requirements

Bridging the Gap: What is the Role of Acute Care Hospitals?

• Why do 60% of all patients start dialysis with a catheter and no permanent access?

• Could this be addressed prior to discharge?

• Can hospitals improve pre-ESRD vaccinations?

• Communication of laboratory and other information during a hospitalization