Expanding MRSA Surveillance: The Options

1
Expanding MRSA Surveillance The Options
?

• Ari Robicsek, MD
• Hospital Epidemiologist
• Evanston Northwestern Healthcare

2
Disclosures

• BD
• Speaking honoraria
• Research Funding
• 6 month old baby

3
Why Expand Surveillance?

• Culture-based surveillance often misses colonized
  patients
• 437 patients found to be colonized by
  surveillance
• Only 66 (15) were found to be MRSA-positive by
  clinical cultures
• Average missed isolation days 7.4 per patient

Salgado and Farr. Infect Control Hosp Epidemiol.
2006 27(2)116-21.
4
Why Expand Surveillance?

• Patients in contact isolation are less likely to
  transmit MRSA to other patients
• Neonatal ICU MRSA outbreak
• 15.6-fold decrease in MRSA transmission to other
  patients by neonates being cared for using
  contact precautions (0.009 transmissions per day
  vs 0.14 transmissions per day for the standard
  precautions group)
• Multicenter study in the Netherlands
• 95 nonisolated patients 19 patients (20)
  transmitted pathogens at least once
• 73 isolated patients 4 patients (5) transmitted
  pathogens at least once

Jernigan et al. Am J Epidemiol. 1996143496-504.
Esveld et al. Nederlands Tijdschrift voor
Geneeskunde 1999 143205-208.
5
Why Expand Surveillance?

• Patients colonized with healthcare-associated
  MRSA frequently develop disease
• 60 of 209 (29) patients identified as
  MRSA-colonized or infected developed MRSA disease
  within the next 18 months 28 were bacteremic
• 5 of 26 (19) patients colonized with MRSA on
  admission developed disease in 12 months of
  follow up

Huang and Platt. Clin Infect Dis. 2003
36(3)281-5.
Davis et al. Clin Infect Dis. 2004 39776.
6
Why Expand Surveillance?

• CA-MRSA may pose a particular risk for disease

Ellis et al, Clin Infect Dis 2004 39971-9.
7
Experience of Other Groups

• 850 bed hospital
• Nasal surveillance in high risk patients (e.g.
  ICU, BMT, long term care) admission and periodic
• isolation and decolonization
• Quasi-experimental 18 months pre and 24 months
  post
• Reduction from 0.64 to 0.3 HA-MRSA BSI/1000
  admissions

Pan et al. Infect Control Hosp Epidemiol. 2005
Feb26(2)127-33
8
Experience of Other Groups

• 800 bed hospital, 80 ICU beds
• Nasal surveillance in ICU admission and weekly
• isolation
• Quasi-experimental 14 months pre and 16 months
  post segmented regression analyses
• Chose to look at all MRSA BSI not just 48h
• 67 hospital-wide reduction in any MRSA BSI
  (compared to projection)

Huang et al. Clin Infect Dis. 2006 Oct
1543(8)971-8
9
The ENH Experience

• 3 hospitals on North Shore
• 850 inpatient beds ? 40,000 annual admissions
• 2 hospitals primarily private rooms
• Single central microbiology laboratory
• In-house molecular diagnostics expertise

10
Local MRSA Burden

• MRSA HA-BSI 0.42 infections/1000 admissions in
  two years prior to universal intervention
• MRSA HA pneumonia 1.1 infections/1000 admissions
• Overall, for the two years prior to our
  intervention, 100 hospital-acquired MRSA
  infections/year ? 2.5 infections/1000 admissions

11
ENH timeline

• 2001 through 2003
• Several MRSA outbreaks (USA 300) in neonatal ICU,
  orthopedic ward
• 2004
• ICU-based surveillance implemented
• Point prevalence survey
• MRSA prevalence 8.5!
• 2/3 not previously known
• Conclusion to detect all MRSA need to screen all
  admissions

12
Local Cost Modeling

• FY 2004 and 2005
• Located all intra-admission MRSA HAI blood,
  wound/abscess, respiratory, urine cultures
  obtained 2 days into a hospitalization
• Cohort matching by DRG

13
Local Cost of MRSA HAI
14
Local Cost of MRSA HAI

• But
• Length Of Stay is a major driver of cost
• MRSA infections, in addition to causing increased
  LOS, occur in patients with already-long LOS
• E.g.
• ENH median LOS 3 days
• ENH median LOS at which MRSA HAI occurs 8 days
• Therefore, control for DRG and restrict
  comparators to patients with at least median LOS
  of MRSA HAI patients

15
Local Cost of MRSA HAI
16
Local Cost of MRSA HAI
LOS matching Bloodstream
LOS matching Respiratory
17
Local Cost of MRSA HAI

• 20,000 excess cost per hospital-acquired
  infection
• 600,000 to 1,000,000 excess cost of MRSA
  surveillance
• Need to prevent 40-50 hospital-acquired MRSA
  infections per year

18
Preparation

• Measures taken to ensure high compliance
• recruitment, education and involvement of nursing
  leadership at each hospital in the program
• educational documents and video for staff and
  patients
• Grand Rounds for physicians
• documented education with annual competency
  evaluations for the patient care technicians
  primarily performing the swabs
• streamlined computerized test order entry
• development of a test kit that included a swab
  and instructions that was placed in each patient
  room prior to admission
• real-time compliance surveillance with feedback

19
The program

• Surveillance of all patients on admission
• Generally done by Patient Care Technician
• Electronic Record forcing function

• Additional surveillance
• ICU admission
• In-house long-term care center

20
The program

• Nasal testing
• Double-swab, both nares
• One swab used for MRSA real-time PCR
• Other swab inoculated onto agar to allow mupA
  testing, strain banking

21
The program

• If MRSA positive
• Contact isolation
• Cohorting of MRSA-positive patients
• Attempted decolonization
• 5-day regimen
• Nasal mupirocin twice daily
• Chlorhexidine wash every second day
• Can be retested 7 days after decolonization to
  allow removal from isolation room

22
August 1, 2003
August 1, 2004
August 1, 2005
April 30, 2007
Period 2
Period 3
Period 1
Surveillance
None
ICU admission
Any admission retesting on ICU admission
Admissions tested ( of intended)
0
3334 (75.9)
62 035 (84.4)
Positive tests ( of total)
0
277 (8.3)
3926 (6.3)
Routine therapy for colonization
No
No
Yes
30 day phase-in
30 day phase-in
Follow-up for MRSA disease
180 days post discharge
180 days post discharge
180 days post discharge (less if discharged in
final 180 days of period)
Robicsek et al. Ann Int Med. 2008 In Press
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
MRSA infections occurring during hospitalization
6
Surgical Site
Urine
5
Respiratory
Bloodstream
4
MRSA infections per 10,000 patient days
3
2
1
0
Q1
Q2
Q3
Q4
Q5
Q6
Q7
Q8
Q9
Q10
Q11
Q12
Q13
Q14
27
1.1
1.0
0.5
p
28
p NS
29
Intra-admission MRSA infections
-51
30
Timing of MRSA disease relative to admission
25
20
15
No surveillance
ICU surveillance
MRSA Infections per 10,000 admissions
Universal surveillance
10
5
0
During Admission
1-30
31-60
61-90
91-120
120-150
151-180
Days since most recent admission
31
Decisions

• Who to test
• Which test method to use

32
Risk Factors for Colonization
33
Risk Factors for Colonization
34
Prediction rule

• Step 1
• B -4.02 0.38 (if Emergency admission) 0.69
  (if Urgent admission) 0.32 (if Unknown
  admission type) 0.65 (if LTCF residence in the
  past year) 0.30 (if had surgery in the past
  year) - 0.18 (if female) 0.72 (if American
  Indian) 0.27 (if Asian) 0.24 (if Black) -
  0.33 (if Hispanic) 0.01 (if Other, non-White,
  race/ethnicity) 0.42 (if has congestive heart
  failure) 0.24 (if has chronic lung disease)
  0.44 (if has diabetes mellitus) 2.31 (if had
  positive culture for MRSA in the past 2 years)
  0.01Age (in years)
• Step 2
• Predicted probability of MRSA colonization

35
(No Transcript)
36
Simple Strategies

• No surveillance
• Test if positive in the past 24 months (Past
  Positive)
• Test if LTCF resident or Past Positive
• Test if age 90 OR LTCF resident or DM or Past
  Positive
• Test if inpatient in the past 12 months or Past
  Positive
• Test if ICU admission or Past Positive
• Universal surveillance

37
Past positive or 90 or LTCF or DM
Past positive or recent admission
Past positive or LTCF
Past positive or ICU
Past positive
38
The similarity across three different hospitals
suggests wider generalizability of these models
39
Who to test?

• BOTTOM LINE
• Strategy of testing patients from LTCF or with
  Prior Positive clinical culture for MRSA
• requires testing 14 of patients
• yields 56 of all necessary isolation days
• as efficient as a complex prediction rule

40
Which test?

• Impact of test type
• What are people using right now?
• Recent Survey of clinical microbiologists on
  ClinMicroNet
• 78/88 responding labs perform some sort of MRSA
  surveillance test
• 13 (17) use mannitol salt agar
• 37 (47) use chromogenic agar
• 2 (3) use enrichment broth chromogenic agar
• 16 (21) use PCR-based test
• 10 (13) use other method

41
Which test?

• TIMING
• Traditional media
• 2-3 day turnaround time
• Chromogenic media
• 1-2 day turnaround time
• PCR-based tests
• 3 hour turnaround time

42
(No Transcript)
43
(No Transcript)
44
Which test?

• COST (materials handling)
• Traditional media
• 6 per patient
• Chromogenic media
• 7.50 per patient
• PCR-based tests
• 25-35 per patient

45
Which test?

• TEST SENSITIVITY
• Traditional media
• 80 (up to 90 with broth enrichment step)
• Chromogenic media
• 80 (up to 90 with broth enrichment step)
• PCR-based tests
• 98

46
Which test?

• TEST SPECIFICITY
• Traditional media
• 100
• Chromogenic media
• 99.4
• PCR-based tests
• 97.5

47
(No Transcript)
48
(No Transcript)
49
Ideal World
50
PCR
51
Chromogenic Agar
52
Which test?

• BOTTOM LINE
• Traditional media
• Slow, insensitive, cheap
• 50 of isolation-days captured
• Chromogenic media
• Faster, insensitive, cheap
• 60 of isolation-days captured
• PCR-based tests
• FAST, sensitive, EXPENSIVE, non-specific
• all possible isolation-days captured

53
(No Transcript)
54
18 of isolation-days
All possible isolation-days
33 of isolation-days
55
(No Transcript)