Achiam CC, Fernandes CMB, McLeod SL, Salvadori M, John M, Seabrook JA, Theakston KD, Milburn S, Huss

1
What Is the Prevalence of Methicillin-Resistant
Staphylococcus aureus in Skin and Soft Tissue
Infections presenting to the Emergency
Departments of a Canadian Academic Health Care
Center?

• Achiam CC, Fernandes CMB, McLeod SL, Salvadori M,
  John M, Seabrook JA, Theakston KD, Milburn S,
  Hussain Z

June 7, 2009
2
Disclosure
X I do not have an affiliation (financial or
otherwise) with any commercial organization that
may have a direct or indirect connection to the
content of my presentation.
3
Background

• 1960s Methicillin-resistant Staphylococcus
  aureus (MRSA) identified as a nosocomial pathogen
• 1981 First Canadian case of MRSA reported
• More recently, infections caused by MRSA have
  been identified in persons who have no known
  traditional risk factors, such as exposure to a
  health care facility
• Community-acquired (CA) MRSA

4
CA- MRSA in the Emergency Department

• Recent studies have shown
• In United States, MRSA has emerged in many cities
  as the most common pathogen in skin and soft
  tissue infections (SSTIs)
• 2003 Canadian Nosocomial Infection Surveillance
  Program (CNISP)
• Data suggested that, in Canada, MRSA remains
  primarily a HA pathogen
• Only 8 of cases community-acquired

5
CA- MRSA Canadian Emergency Departments

• Borgundvaag, et al (2008) CJEM
• 7 hospitals in the Greater Toronto Area
• Of purulent SSTI caused by S. aureus 18 MRSA
• No published study that prospectively examines
  the importance of CA-MRSA in managing purulent
  and non-purulent skin and soft tissue infections
  of all etiologies in the Canadian ED setting

6
Study Objectives

• Primary Objective
• Prospective observational study
• Estimate the city-wide prevalence of MRSA and
  CA-MRSA
• Adult patients (gt18yrs old)
• Presenting with skin or soft tissue infections
  (SSTIs)
• Emergency departments (EDs) of a Canadian
  academic tertiary care center

7
Study Objectives

• Secondary Objectives
• Identification of demographic and clinical
  variables associated with MRSA
• Characterization of MRSA antimicrobial
  susceptibilities and genotypes

8
Methods

• Prospective prevalence study
• Approved by the Ethics Review Board at The
  University of Western Ontario
• 3 Emergency departments in London, Ontario
• 150,000 visits annually
• Estimated sample size of 138 patients

9
Methods

• Inclusion Criteria
• gt 18 years of age
• Diagnosed with any of the following skin or soft
  tissue infections
• Cellulitis, necrotizing soft tissue infection,
  wound infection, ulcer, septic bursitis, acute
  lymphadenitis, pilonidal cyst without abscess,
  impetigo, or abscess (including furuncle,
  carbuncle, superficial skin abscess, paronychia,
  hordeolum, or pilonidal abscess)
• Written, informed consent

10
Methods

• Exclusion Criteria
• Previously enrolled
• Diagnosed with
• Bartholin gland abscess, odontogenic infection,
  or perianal abscess

11
Methods

• At enrolment, the following were obtained from
  each participant
• Standardized health history and lifestyle
  questionnaire
• Anterior nares and throat cultures
• Infection site cultures

12
Statistical Analysis

• Descriptive statistics were used to summarize
  patient characteristics and prevalence of MRSA
• Patient characteristics associated with MRSA
  colonization or infection were determined with a
  backwards stepwise multivariate logistic
  regression model

13
Results

• During 6 weeks from July to August 2008
• 205 subjects enrolled
• Mean (standard deviation) age was 45.3 (17.9)
  years
• 52.2 Male

14
Results

• Primary Outcome
• 35 patients (17.1) were found to be infected or
  colonized with MRSA
• 27 (13.2) patients were found to be infected /-
  colonized with MRSA
• 8 (3.9) patients were found to be colonized with
  MRSA

15
Results

• Primary Outcome
• MRSA isolates were present in
• 27 (13.2) infection site cultures
• 21 (10.2) of the nares cultures
• 12 (5.8) of the throat cultures
• S aureus grew in 38.0 of infection site cultures
• 34.6 MRSA

16
Results

• Secondary Outcomes
• Genotype
• All 35 patients infected or colonized with MRSA
  had at least one isolate genotyped
• Twenty-seven (77) had isolates identified as
  CMRSA 10 (USA 300), a recognized
  community-acquired strain of MRSA

17
Results- Antimicrobial Susceptibilities
18
Results Univariate Analysis of Predictor
Variables

• Variables associated with colonization
• Currently homeless

• Variables associated infection
• Currently homeless
• Age 18-44 years
• Incarcerated in past yr
• Physical contact with incarcerated person in last
  year
• Know anyone else with skin infection
• Competitive sports
• Known exposure to MRSA
• Previous abscess within past year
• IV drug user
• Purulent skin infection currently

19
Results Table 2
20
Results Table 3
21
Limitations

• Prevalence and clinical predictors of MRSA may
  lack of external validity to other times of year,
  practice settings and communities
• Estimated prevalence of MRSA can not be
  generalized to our overall ED population
• Definitive conclusions about the pathogenic
  etiology of participants with pure cellulitis can
  not be drawn from our data

22
Implications

• In appropriate patients with SSTIs
• MRSA coverage should be considered when choosing
  empiric antimicrobial therapy

23
Implications

• Cultures of purulent SSTI are not part of routine
  care in many EDs
• Recommended by The Centers for Disease Control
  and Prevention
• We suggest that all SSTI with purulent drainage
  should be sent for culture to
• Definitively identify the pathogenic organism
• Track the changing epidemiology and antimicrobial
  susceptibility of MRSA in the community
• Reduce nosocomial infections

24
Thank you
25
(No Transcript)
26
Future Areas of Research

• External validation of patient characteristics
  identified as risk factors for MRSA
  carriage/infection
• May lead to a clinical algorithm to guide
  empirical antibiotic therapy
• Best practice for empiric antimicrobial therapy
  in patients suspected of being infected with MRSA

27
Sample Size Estimate

• Sample Size Estimate
• n Z2 P (1-P)
• d2
• Assuming Power of 80
• Estimated MRSA prevalence of 10 based on a local
  retrospective review
• Desired precision level of 5
• To be 95 confident that the true proportion of
  MRSA positive patients was between 5 and 15
• Estimated 138 patients needed to be enrolled in
  the study
• Decided to increase our sample size by 10 to
  account for potential missing data from the
  questionnaire

28
Clinical Definition of CA-MRSA

• Community-dwelling persons without risk factors
  for HA-MRSA
• Hospitalization or surgery within the previous
  year, residence in a nursing facility, dialysis,
  or presence of an indwelling bladder catheter
• 68.6 fit the clinical definition for CA-MRSA
• 11 patients did not meet the definition
• 10 hospitalized within last 12 months
• 1 indwelling bladder catheter

29
Estimated Enrollment at Each Site

• University Hospital ED
• 96 patients diagnosed with SSTI
• 11 enrolled (11.5)
• Victoria Hospital ED
• 126 patients diagnosed with SSTI
• 81 enrolled (64.3)
• St. Josephs Health Care Urgent Care Center
• 244 Patients diagnosed
• 113 enrolled (46.3)

30
Statistical Analysis

• Statistical Analysis of Patient Characteristics
• Descriptive statistics were used to summarize
  patient characteristics and MRSA prevalence with
  95 confidence intervals (CIs)
• Using contingency tables, univariate analysis was
  conducted for each independent variable collected
  in the health history and lifestyle questionnaire
• The likelihood ratio chi-square test with k-1
  degrees of freedom was used to measure the level
  of association between the dichotomous outcome
  variable (whether or not a patient was MRSA) and
  primary predictor variables.
• In the univariate analysis, variables with a p
  value lt 0.1 were found to be associated with
  being colonized or infected with MRSA were
  considered in the multivariate logistic
  regression model
• Likelihood ratio tests were then used to
  determine the appropriate inclusion of variables
  in the model.
• Backwards stepwise logistic regression (using a
  likelihood ratio removal criterion of 0.1) was
  used to obtain odds ratios with 95 confidence
  intervals for the final model

31
Types of SSTI with MRSA Isolates on Infection
Site Culture
32
Genotypic Concordance

• 27 subjects with MRSA at infection site
• 15 (55.6) colonized with MRSA
• 13 (86.7) had genetic typing of a colonized site
  infection site
• 5 (38.5) had genetically identical clones in all
  3 sites
• 8 (61.5) had genetically identical clones in at
  least 2 sites