Health Care associated Infections

Common but - there are now many

interventions we can implement that will reduce them

Prof Peter CollignonThe Canberra Hospital

Australian National University

What are health-care associated infections?

• Any infection that occurs following a health care procedure

– All “hospital onset” infections

– But many now also have a “community onset” but related to medical care

– wound infection– Many blood stream infections

Examples

• Blood stream infections• IV catheters

• Wound infections• After surgery• May be deep seated

• Urinary tract• Catheters

• Respiratory tract• Ventilators• drugs

Why do these infections occur?

• Breach normal defense barriers– Skin– Respiratory tract– Acid in stomach

• Lowered immune defenses– Chemotherapy– Part of disease

• Increased exposure– Resistant bacteria

Health care infections are common

• Very common; – various studies in many countries– Likely between 5 -10% of all admissions

develop a new infection

• Most are relatively minor– UTI, superficial wound

• But many Serious and Life threatening– Blood stream– Prosthetic joints etc

Patient safety is important• Hospitalisation is inherently hazardous

– Drug errors most common misadventure

– But infections are 2nd biggest problem

– Occur in at least 10% of acute admissions

• 50-80% potentially preventable

• Misadventures primarily result from system failures

not incompetence

• We need national and comparative data

Clinical Excellence Commission, 2005; Leape 2000; Wilson et al 1995

Serious infections are common

• Blood Stream infections– Most from IV catheters– In Australia likely about 4,000 per year– In USA more than 200,000 per year

• High mortality and morbidity attached– With MRSA blood stream infections - 35%– Central nervous system - lower but still >5%

– In Australia - about 400 deaths per year and USA 20,000 from JUST IV catheters!

How hazardous is healthcare?Dr. Lucien Leape Harvard Medical School. USA

Dangerous Regulated Ultrasafe(>1/1000) (< 1/100,000

Total liveslost per

year

1

10

100

1000

10,000

100,000

1 10 100 1000 10,000 100,000 1M 10M

Bungeejumping

Mountainclimbing

Healthcare

Driving

Chemicalmanufacturing

Charteredflights

Scheduled airlines European

railroadsNuclearpower

Number of encounters for each fatality

How hazardous is healthcare?Dr. Lucien Leape Harvard Medical School. USA

Dangerous Regulated Ultrasafe(>1/1000) (< 1/100,000

Total liveslost per

year

1

10

100

1000

10,000

100,000

1 10 100 1000 10,000 100,000 1M 10M

Bungeejumping

Mountainclimbing

Healthcare

Driving

Chemicalmanufacturing

Charteredflights

Scheduled airlines European

railroadsNuclearpower

Number of encounters for each fatality

Hospital-Acquired Blood stream infections;

8th leading cause of death in USA

Emerging Infectious Diseases April 2001

http://www.cdc.gov/ncidod/eid/vol7no2/wenzel.htm

Staphylococcus aureus

• Common– Many sites esp blood, wounds

• Bacteraemia likely 7,000 per year in Australia– 50% hospital onset– 1/3 of community onset are health care related

• High mortality in bacteraemia– Pre-antibiotics 82%– MSSA median 25%– MRSA median 35%

Antibiotic Resistance is common

• Penicillin

• Beta-lactams– MRSA

• Other common agents– macrolides etc

• Vancomycin– New forms of resistance

• New agents– linezolid

Serious Morbidity also common

Prosthetic joint infection (eg hip)– To cure need 2 major operations, 8- 10

weeks incapacitated.– > $100,000 per episode– 1 to 2% of all joint replacements

– when things go well!

Blood stream infections; serious morbidity

• Blood stream infections– Renal failure, osteomyelitis, prolonged

antibiotic therapy etc

Blood stream infections are common;and more than 60% of these are health care

associated

The Canberra Hospital  1998 1999 2000 2001 2002 2003 2004

Significant 337 307 320 288 271 316 354

Indeterminate 37 37 37 36 30 32 25

Contaminant 245 200 195 197 217 210 235

Total positive Blood cultures 619 544 552 521 518 558 614

This means that at the Canberra Hospital each year over 200 BSI episodes are Health-care associated

Many primary sites for BSI; but IV catheters main site at all major hospitals

Body system (TCH data) 1998 1999 2000 2001 2002 2003 2004 Total

IV Device 109 72 81 54 39 45 42 442

Respiratory 50 36 54 31 41 49 47 308

GIT 47 38 46 43 40 41 59 314

Genito-urinary 43 38 38 43 45 54 70 331

Skin 24 22 22 19 18 27 35 167

Unknown 19 39 32 37 32 28 27 214

Cardiovascular 13 9 10 12 8 19 14 85

Musculo-skeletal 10 14 5 13 12 20 19 93

Haematology 9 17 10 15 16 15 20 102

Maternal 9 4 5 5 6 3 2 34

Neurology 4 13 8 7 6 5 5 48

Other 0 0 2 1 1 1 0 5

Prim Bacteraemia 0 5 7 8 7 9 14 50

Infections can be reducedBSI from IV catheter sepsis (The Canberra

Hospital)

0

0.5

1

1.5

2

2.5

3

Epi

sode

s/1,

000

Sep

s (in

clud

ing

sam

e da

y)

Interventions that decreased IV sepsis

INTERVENTIONS HOSPITAL UNITS

Prospective surveillance of BSI's Hospital wide

Tunnelling of vascaths Renal medicine

CVC retention by exception ICU

Prevention of septic flush by correct alcohol usage Oncology / Haematology

Monitoring of peripheral IV policy compliance Aged care / General surgery

Blood culture collection poster Hospital wide

Patient information pamphlet for CVC care Hospital wide / community

Introduction of Alcoholic chlorhexidine skin prep Hospital wide

Reduction in the use of TPN Hospital wide

Notification of IVD BSI as a critical incident Hospital wide / Medical officer

Dissemination of BSI project information Hospital wide / GP's / Media

IV catheter infections can be reduced

• Too many used• In for too long• Poor selection of most

appropriate catheters• Poor selection of sites• Almost every doctor inserts

them• including CVC’s - even if little training

• CVC’s used instead of peripheral catheters

• for convenience BUT much higher per day risk

IV’s; what can be done?

• Protocols already exist• CDC, Australia, WHO• Guidelines for the Prevention of Intravascular Catheter-Related Infections,

2002http://www.cdc.gov/ncidod/hip/iv/iv.htm

• They need to be followed

• Will be discussion and disagreements on these protocols

• eg peripheral IV catheters – remove after 2-3 days • but these are relatively minor issues

Australian Guidelines

http://www.safetyandquality.org/intravascdevicejun05.pdf

We can have an impact on all types of infections

• Surgical site• Infection rates can be decreased• Hobart, Victoria, TCH, internationally

• Blood stream infections– Especially IV catheter

• Urinary tract• Pneumonia

• All types– If you recognize there is a problem

Alcohol-chlorhexidine hand-rub solution

+ culture change program• A new standard of healthcare

– CDC, WHO, AICA

• Does it work?

• Does it increase hand hygiene compliance?

• Does it reduce nosocomial MRSA infections?

250

0

25

50

75

100

50 75 100Opportunities for hand hygiene per hour of care

Co

mp

lian

ce w

ith h

and

hyg

ien

e (%

)

Pittet et al, Ann Intern Med 1999, 130:126

MRSA colonisation rates and hospital

contamination

Johnson et al. Med J Aust 2005 – 21st November issue

or www.mja.com.au

Health care worker hand-hygiene compliance

Johnson et al. Med J Aust 2005 – 21st November issue

or www.mja.com.au

Use of alcohol/chlorhexidine solution

Johnson et al. Med J Aust 2005 – 21st November issue

or www.mja.com.au

MRSA isolates and patient-episodes of bacteraemia

After 36 months:

Total MRSA isolates:

• 40% reduction (95% CI, 23%–58%)

• 1008 fewer clinical isolates

Patients with MRSA bacteraemia:

• 57% reduction in monthly rate (95% CI, 38%–74%)

• 53 fewer bacteraemias than expected (95% CI, 36–68 episodes)

Johnson et al. Med J Aust 2005 – 21st November issue

or www.mja.com.au

Program costs & financial impact

• $180,000 per year to maintain• Saved $325,000 per year on BSI*• 72,000 separations per year (inc.

day cases)

– $2.50 per patient – BigMac in Australia = $3.20

* Estimated cost: $20,000 AUD per case of MRSA BSI

What can we do?

• Recognize/admit there is a problem

• No self justification• Do we really need to hide the data?

• Measure what is happening

• Meaningful and easy• Research

• Change things• Education• Interventions “but –ins”

• Measure again

Epidemiologists; are they a hindrance?

• Too much time and effort to get the perfect denominator

• This is Not research but quality improvement

Need to collect and have readily available some easy to measure

but important RATES

• Will not be popular with hospitals

– Always reasons why my rates are worse than someone else's

BUT

• We need to do it

US; report cards

What do we need to measure in all hospitals:

Infections

• S. aureus blood stream infection rates

– All episodes- community and hospital onset

– Separate MRSA and MSSA– Per 1,000 hospital separations– Should be on the web for each hospital– Based on pathology systems

AGAR: Rates at different hospitals (total)

Staph. aureus Bacteraemia

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

rate

/ 10

00 a

dm

issi

on

s

Collignon P, Nimmo GR, Gottlieb T, Gosbell IB; Australian Group on Antimicrobial Resistance.Staphylococcus aureus bacteremia, Australia.Emerging Infect Dis. 2005 Apr;11(4):554-61.

Hospital onset

Hospital Acquired Staph. aureus Bacteraemia

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

rate

/100

0 ad

mis

sion

s

MRSA Bacteraemia 1998 - 2004 By separations at Canberra Hospital

0.27

0.21

0.06

0.64

0.50

0.14

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

MRSA hosp. acq MRSA comm. acq MRSA

Ra

te /

1,0

00

se

ps

/1000 separations

/1000 seps >24 hrs

We can improve things

• Need to be motivated• Both internal and external

pressure for better QA is needed• We need to aim for major

improvements

• This can be achieved

Conclusions• Hospital safety is important

• Data can be measured reliably using existing practical, commonsense definitions

• “Simple” interventions can make a huge difference

• But changing human behaviour is not simple and commonsense is not common

• Open transparent reporting is the best form of “risk management”

Conclusions• Hospital safety is important

• Data can be measured reliably using existing practical, commonsense definitions

• “Simple” interventions can make a huge difference

• But changing human behaviour is not simple and commonsense is not common

• Open transparent reporting is the best form of “risk management”

Conclusions• Hospital safety is important

• Data can be measured reliably using existing practical, commonsense definitions

• “Simple” interventions can make a huge difference

• But changing human behaviour is not simple and commonsense is not common

• Open transparent reporting is the best form of “risk management”