Prof Peter Collignon The Canberra Hospital Australian National University

1. Health Care associated InfectionsCommon but -there are now many interventions we can implement that will reduce them Prof Peter Collignon The Canberra Hospital Australian National University

2. 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

3. Examples • Blood stream infections • IV catheters • Wound infections • After surgery • May be deep seated • Urinary tract • Catheters • Respiratory tract • Ventilators • drugs

4. 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

5. 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

6. Patient safety is important 0 • 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

7. 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!

8. How hazardous is healthcare?Dr. Lucien Leape Harvard Medical School. USA 0 Dangerous Regulated Ultrasafe (>1/1000) (< 1/100,000 Total lives lost per year Healthcare 100,000 Driving 10,000 1000 Scheduled airlines 100 Chemical manufacturing Mountain climbing European railroads 10 Chartered flights Bungee jumping Nuclear power 1 10 100 1000 10,000 100,000 1M 10M 1 Number of encounters for each fatality

9. How hazardous is healthcare?Dr. Lucien Leape Harvard Medical School. USA 0 Dangerous Regulated Ultrasafe (>1/1000) (< 1/100,000 Total lives lost per year Healthcare 100,000 Driving 10,000 1000 Scheduled airlines 100 Chemical manufacturing Mountain climbing European railroads 10 Chartered flights Bungee jumping Nuclear power 1 10 100 1000 10,000 100,000 1M 10M 1 Number of encounters for each fatality

10. 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

11. 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%

12. Antibiotic Resistance is common • Penicillin • Beta-lactams • MRSA • Other common agents • macrolides etc • Vancomycin • New forms of resistance • New agents • linezolid

13. 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!

14. Blood stream infections; serious morbidity • Blood stream infections • Renal failure, osteomyelitis, prolonged antibiotic therapy etc

15. Blood stream infections are common;andmore than 60% of these are health care associated This means that at the Canberra Hospital each year over 200 BSI episodes are Health-care associated

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

17. Infections can be reducedBSI from IV catheter sepsis(The Canberra Hospital)

18. Interventions that decreased IV sepsis

19. 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

20. 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

21. Australian Guidelines http://www.safetyandquality.org/intravascdevicejun05.pdf

22. 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

23. 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?

24. Pittet et al, Ann Intern Med 1999,130:126 100 75 Compliance with hand hygiene (%) 50 25 0 0 25 50 75 100 Opportunities for hand hygiene per hour of care

32. MRSA colonisation rates and hospital contamination Johnson et al. Med J Aust 2005 – 21st November issue or www.mja.com.au

33. Health care worker hand-hygiene compliance Johnson et al. Med J Aust 2005 – 21st November issue or www.mja.com.au

34. Use of alcohol/chlorhexidine solution Johnson et al. Med J Aust 2005 – 21st November issue or www.mja.com.au

35. 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

36. 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

37. 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

38. Epidemiologists; are they a hindrance? • Too much time and effort to get the perfect denominator • This is Not research but quality improvement

39. 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

40. US; report cards

41. 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

42. AGAR: Rates at different hospitals (total) 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.

43. Hospital onset

44. MRSA Bacteraemia 1998 - 2004 By separations at Canberra Hospital

45. 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

46. 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”

47. 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”

48. 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”