Risk Factors for Community-associated Methicillin-resistant Staphylococcus Aureus Cellulitis - and the Value of Recognition

Abstract

Objectives

To identify the risk factors for community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA) cellulitis.

Methods

A review of risk factors for CA-MRSA skin and soft tissue infection in previously published literature was first performed. A retrospective cohort study was then conducted in a teaching ambulatory-care clinic of a tertiary medical center in Honolulu, Hawai‘i.

Results

Of 137 cases with cellulitis diagnosed from January 2005 to December 2007, MRSA was recovered from 85 (62%) of patients who presented with either abscesses or skin ulcers. The recovery of MRSA was significantly associated with obesity (p=0.01), presence of abscesses (p=0.01), and lesions involving the head and neck (p=0.04). Independent risk factors by multivariate logistic regression analysis included the presence of abscesses [adjusted odds ratio (aOR) 2.72; 95% confidence interval (CI) 1.27–5.83; p=0.01] and obesity (aOR 2.33; 95% CI 1.10–4.97; p =0.03). Patients with CA-MRSA were less likely to receive an appropriate antibiotic (p=0.04) and were more likely to require antibiotic change at evaluation in one week (p=0.04) compared with patients infected with non-MRSA bacteria.

Conclusions

The presence of abscesses and obesity were signifi cantly associated with CA-MRSA cellulitis. Empiric therapy with antibiotics active against MRSA should be guided by these risk factors.

Introduction

Methicillin-resistant Staphylococcus aureus (MRSA) has become a significant healthcare problem, especially in hospital settings over the last 40 years. In the past decade, this formidable agent has emerged with a new strain in the community known as community-associated MRSA (CA-MRSA). Outbreaks of CA-MRSA infections have been increasingly reported worldwide with skin and soft tissue as the most common manifestation.^1,2 In the United States, the prevalence of CA-MRSA skin and soft tissue infection (SSTI) have been reported to range from 15 to 74% of all SSTIs depending on the location.³ Hawaii is recognized as a state with one of the highest rates of MRSA infections.^4,5 A study of 1,389 individuals hospitalized for MRSA infections in Hawai‘i demonstrated that 28% were community-acquired and the incidence of CA-MRSA infections had significantly increased during a recent 2-year study period.⁵

Risk factors for CA-MRSA infections reported in previous studies include African-American ethnicity, men who have sex with men, military personnel, athletes participating in contact sports, close contact with persons who have MRSA, injection drug use (IDU), recent antibiotic use and prior hospitalization.^3,6–11 The relative importance of these risk factors and potentially other factors may depend on geographic and climate differences. Identification of risk factors is important in initial decisions about antibiotic selection as clinical outcomes are dependent on their activity with SSTIs.^12,13

Methods

Literature review.

A comprehensive search was performed for clinical studies published in the English literature using the Pubmed databases. Search terms included “risk factors,” “methicillin-resistant Staphylococcus aureus” and “skin and soft tissue infection.” Only the studies that included patients with SSTIs were reviewed.

Study population and setting.

This retrospective cohort study included adult patients (age ≥ 18 years) with cellulitis who presented to the Queen Emma Clinic, a teaching clinic, which provides primary and specialty care to the underserved in Honolulu, Hawai‘i. The study was conducted from January 2005 through December 2007. The study was approved by the Queen's Medical Center Institutional Review Board.

Study design and definitions.

The study patients were identified from the clinic's electronic medical record system using the International Classification of Diseases (ICD-9) codes of 681 and 682. Only patients with a positive culture were included and those were divided into MRSA and non-MRSA categories. Demographic characteristics, clinical presentations, underlying conditions, general and microbiology laboratory data, treatment, outcomes data were collected. Obesity was defined as having a body mass index (BMI) ≥ 30. Patients were categorized as cigarette smokers, alcohol drinkers or IDUs if they were documented to be using these substances at evaluation. Patients identified their own ethnic group. Patients were documented as having chronic kidney disease (CKD) based on the K/DOQI clinical practice guidelines for chronic kidney disease (stage I = glomerular filtration rate (GFR) ≥ 90 mL/min/1.73m²; stage II = GFR 60–89 mL/min/1.73m²; stage III = GFR 30–59 mL/min/1.73m²; stage IV = GFR 15–29 mL/min/1.73m² and stage V = GFR < 15 mL/min/1.73m²).¹⁴ Recent hospitalization was defined as being hospitalized within the past month. A cellulitis severity scale was created specifically for this study. A score was given for clinical presentations and laboratory data (Table 1). Patients were then categorized as having mild, moderate and severe disease severity if their calculated scores were 0–1, 2–3 and 4–6 respectively.

Table 1.

Clinical Presentations Used in Cellulitis Severity Scoring

Clinical Presentations	Score
Largest lesion >5cm. in a longest diameter	1
More than 3 affected areas	1
Concurrent ulcers and/or abscesses	1
Fever (temperature >38.0 °C or >100.4 °F	1
Leukocytosis (white blood cell >12,000 cells/mm3 in male, >11,000 cells/mm3 in female) or leukopenia (white blood cells <4,000 cells/mm3 in both sexes)	1
Hypotension (systolic blood pressure <90 mmHg or 20 mmHg dropped from baseline)	1

Treatment success was defined as clinical improvement or resolution of signs and symptoms of cellulitis when they returned for evaluation a week later. Cases were considered failures if there was a change in antibiotics, surgical intervention or hospitalization. Identification and susceptibility testing of isolates from infected sites were performed in accordance with the Clinical and Laboratory Standards Institute guidelines.^15,16

Statistical analysis.

Categorical and continuous variables were compared using the Pearson's χ² test, Fisher's exact test and student's t-test respectively. All tests were two-tailed with p value < 0.05 considered significant. Logistic regression analysis was performed to identify independent risk factors for MRSA cellulitis. All statistical analyses were conducted using SPSS for Window software, version 15.0.

Results

A total of 462 episodes of cellulitis were identified during the study period. Three hundred and twenty-five episodes were excluded because no culture was done or no organism was recovered. The final cohort consisted of 137 patients with 137 independent episodes of cellulitis. Eighty-eight patients presented with abscesses while 49 patients had concurrent ulcers.

Patient characteristics.

The mean age of the cohort patients was 48 years (range, 18–73). Eighty four (61%) were male. Self-identified race or ethnic groups included: 54% Pacific Islanders (PIs) (including Hawaiian, Chuukese, Samoan, American Samoan, Marshallese, and Tongan), 33% Caucasians, 9% Asians and 4% other groups. Fifteen percent were homeless. Comorbidities included obesity (45%), diabetes mellitus (36%), cigarette smoking (27%), alcohol excess (12%), CKD (stage III–V) (9%) and IDU (4%). The mean duration of symptoms before the presentations was 7 days (range, 1–30 days). Infections were located on the lower extremities in 47%, torso in 34%, upper extremities in 21% and head and neck in 8%. Forty-nine percent had mild-severity cellulitis and 51% had moderate cellulitis. There were no cases of severe cellultis.

Therapy and microbiological data.

Fifteen patients were hospitalized. Of the remaining 122 patients, 118 (97%) patients were empirically treated with oral antibiotics. Trimethoprim-sulfamethoxazole (TMP-SMX), cephalexin, and clindamycin were the three most commonly prescribed oral antibiotics (48%, 33% and 13% respectively). Other prescribed antibiotics included amoxicillin-clavulanate and cloxacillin. Cultures of the 137 patients revealed MRSA in 62%, methicillin-sensitive Staphylococcus aureus (MSSA) in 23%, β-hemolytic streptococci in 4%, Pseudomonas aeruginosa in 3%, Klebsiella pneumoniae in 2%, Proteus mirabilis in 2%, Escherichia coli in 1%, Serratia marcescens in 1%, Acinetobacter baumanii in 1% and Propionibacterium acnes in 1%.

Risk factors for MRSA cellulitis, treatment and outcomes.

Patient characteristics, treatment and outcomes were compared between MRSA and non-MRSA-infected patients and are displayed in tables 2 and 3. Antimicrobial susceptibility data of MRSA was shown in table 4. Patients with MRSA cellulitis were significantly more obese (p=0.01), present with abscesses (p=0.01) and have lesions that involved the head and neck compared with those who had other bacteria (p=0.04). The non-MRSA group had more diabetic patients (p=0.03) and presented with concurrent ulcers (p=0.01) and lower extremity involvement (p=0.01) more often than MRSA patients. By multivariate logistic regression analysis (Table 5), independent risk factors for MRSA cellulitis were presence of abscesses [adjusted odds ratio (aOR) 2.72; 95% confidence interval (CI) 1.27–5.83; p=0.01] and obesity (aOR 2.33; 95% CI 1.10–4.97; p=0.03).

Table 2.

Comparison of Demographic and Clinical Characteristics of Outpatients with Cellulitis Caused by Methicillin-resistant Staphylococcus aureus (MRSA) and Non-MRSA Bacteria

Characteristics	MRSA (n = 85)	Non-MRSA bacteriaa (n = 52)	OR (95% CI)	p-valueb
Age, mean years (range)	47 (18–70)	50 (18–73)	- - -	0.19c
Male sex	54 (64)	31 (60)	1.18 (0.58–2.39)	0.72
Ethnicity
Pacific Islander	43 (51)	31 (60)	0.69 (0.35–1.39)	0.38
Caucasian	33 (39)	12 (23)	2.12 (0.98–4.56)	0.06
Asian	6 (7)	6 (12)	0.58 (0.19–1.82)	0.37
Homelessness	13 (15)	7 (13)	1.16 (0.44–3.04)	0.81
Underlying conditions
Obesity (BMI ≥ 30)	45 (53)	16 (31)	2.53 (1.23–5.20)	0.01
Diabetes mellitus	25 (29)	25 (48)	0.45 (0.22–0.92)	0.03
Cigarette smoker	22 (26)	15 (29)	0.86 (0.40–1.85)	0.70
Alcoholic drinker	11 (13)	5 (10)	1.40 (0.47–4.09)	0.79
CKD (stage III–V)	6 (7)	7 (13)	0.49 (0.16–1.48)	0.24
Injecting drug use	5 (6)	1 (2)	3.19 (0.47–21.04)	0.41
Recent skin infections	10 (12)	5 (10)	1.25 (0.42–3.72)	0.78
Hospitalization within a month	6 (7)	5 (10)	0.71 (0.22–2.33)	0.75
Cellulitis presentations
Head/neck involvement	10 (12)	1 (2)	6.80 (1.08–42.16)	0.04
Upper extremity involvement	21 (25)	8 (15)	1.81 (0.75–4.35)	0.28
Trunk involvement	32 (38)	14 (27)	1.64 (0.78–3.45)	0.26
Lower extremity involvement	33 (39)	32 (62)	0.40 (0.20–0.80)	0.01
Cellulitis with ulcers	23 (27)	26 (50)	0.37 (0.18–0.76)	0.01
Cellulitis with abscesses	62 (73)	26 (50)	2.70 (1.31–5.54)	0.01
Severity of cellulitis
Mild	43 (51)	24 (46)	1.14 (0.57–2.27)	0.73
Moderate	42 (49)	28 (54)	0.84 (0.42–1.67)	0.73
Severe	0 (0)	0 (0)	- - -	- - -

Data are number (%) of episodes, unless otherwise indicated. ^aInclude methicillin-sensitive Staphylococcus aureus (n=31), β-hemolytic streptococci (n=6), Pseudomonas aeruginosa (n=4), Klebsiella pneumoniae (n=3), Proteus mirabilis (n=3), Escherichia coli (n=2), Serratia marcescens (n=1) Acinetobacter baumanii (n=1) and Propionibacterium acnes (n=1). ^bDetermined by the χ² test, unless otherwise indicated. ^cDetermined by the student's t-test. Abbreviations: CI = confidence interval; BMI = body mass index; CKD = chronic kidney disease; I&D = incision and drainage; MRSA = methicillin-resistant Staphylococcus aureus; OR = odd ratios.

Table 3.

Comparison of Treatment and Outcomes of Outpatients with Cellulitis Caused by Methicillin-resistant Staphylococcus aureus (MRSA) and Non-MRSA Bacteria

Variables	MRSA (n = 80)	Non-MRSA bacteriaa (n = 42)	OR (95% CI)	p-valueb
Initial treatment
Antibiotics	77 (96)	41 (98)	0.63 (0.09–4.56)	1.00
TMP-SMXc	40 (52)	17 (41)	1.53 (0.71–3.26)	0.34
Cephalexinc	25 (32)	14 (34)	0.93 (0.42–2.05)	1.00
Clindamycinc	6 (8)	9 (22)	0.30 (0.10–0.89)	0.04
Cloxacillin or amoxicillin-clavulinic acidc	6 (8)	1 (2)	3.38 (0.51–21.96)	0.42
Active antibioticsc	45 (58)	32 (78)	0.40 (0.17–0.93)	0.04
Outcomes
Clinical success	51 (64)	33 (79)	0.48 (0.21–1.13)	0.09
Treatment change	30 (38)	10 (24)	1.92 (0.84–4.40)	0.16
Antibiotic change	20 (25)	4 (10)	3.17 (1.05–9.50)	0.04
Antibiotic change and additional I&D	3 (4)	1 (2)	1.60 (0.22–11.44)	1.00
Hospitalization	5 (6)	4 (10)	0.63 (0.17–2.31)	0.49

Data are number (%) of episodes. ^aInclude Methicillin-sensitive Staphylococcus aureus (MSSA), β-hemolytic streptococci, Pseudomonas aeruginosa, Klebsiella pneumoniae, Serratia marcescens, Acinetobacter baumanii, Escherichia coli and Proteus mirabilis. ^bDetermined by the χ² test. ^cTotal patients with MRSA infection = 77; total patients with non-MRSA infection = 41 Abbreviations: CI = confidence interval; I&D = incision and drainage; MRSA = methicillin-resistant Staphylococcus aureus; OR = odd ratios; TMP-SMX = trimethoprim-sulfamethoxazole.

Table 5.

Independent Risk Factors for CA-MRSA Cellulitis by Multivariate Logistic Regression Analysis

Characteristics	Adjusted Odd Ratio (95% confidence interval)	p-value
Presence of abscesses	2.72 (1.27–5.83)	0.01
Obesity (body mass index 30)	2.33 (1.10–4.97)	0.03

Choice of an antibiotic determined to be inactive by standard laboratory methodology was more likely in the MRSA group (p=0.04) and was associated with a change in prescription and continuation of antibiotics for an additional week (p=0.04) (Table 3).

Discussion

Cellulitis is one of the most common infections encountered by physicians in outpatient settings and is usually treated with oral antibiotics. The increase in methicillin resistance among these organisms worldwide has made the treatment more complicated and challenging. Whether empiric antibiotic therapy needs to be active against MRSA is difficult to determine but evaluation of risk factors can help.^11,17 We found a high incidence rate of MRSA infection of 54% in our clinic, which increased significantly from 32% at the beginning of the study 4 years prior.⁵ We believe the majority of our isolates should be considered CA-MRSA as they were recovered from outpatients and their susceptibility patterns were similar to those of isolates in other studies of CA-MRSA infections.^3,5,6 While obesity is a known risk factor for SSTIs,¹⁸ the association between obesity and CA-MRSA-related SSTIs has not been established.^11,19 Another significant association identified in our study is the presence of abscesses. This may represent a manifestation of virulence factors such as Panton-Valentine Leukocidin or unique characteristics associated with the new CA-MRSA strains.⁹ Together, obesity with abscess formation should lead to strong consideration of empiric therapy to include MRSA in an outpatient setting in our community.

This study is limited in that it is a retrospective study with a relatively small sample size. These factors may have been responsible for our failure to find a significant difference between MRSA and non-MRSA patients associated with ethnicity or injection drug use. The study also lacks information on history of close contact with persons with skin infections which was shown to be a significant risk factor in previous studies.^3,8,10 Lastly, we did not perform molecular testing such as pulse-field gel electrophoresis, or typing of resistance and toxin genes on our isolates.

The clinical importance of active antibiotic treatment for CA-MRSA SSTIs has been recently established in the literature.¹³ This study demonstrated a significantly higher rate of clinical failure in patients treated with antibiotics judged inactive by standard laboratory methods against MRSA.¹³ In our study, patients with CA-MRSA were more likely than those with other bacteria to receive inactive empiric antibiotics (58% vs. 78%) and required antibiotic change at the follow-up visits (25% vs. 10%). These findings reflect that physicians might not be aware of the incidence and susceptibility of CA-MRSA as a cause of cellulitis in the community. Although statistical significance was not reached, there was a trend towards lower clinical success rate among CA-MRSA patients in comparison to other bacterial groups (64% vs. 79%) possibly as a result of inactive empiric antibiotic use.

In summary, our study suggests that presence of abscesses and obesity are strongly associated with CA-MRSA cellulitis. Physicians should be aware of the incidence and risk factors of CA-MRSA infections in their community and consider empiric antibiotics with MRSA coverage when indicated. Further studies are needed to determine the common risk factor of CA-MRSA infections and assess the outcomes of risk factors-defined empiric antibiotic therapy.

The choice of initial antibiotic therapy should depend on multiple factors including the individual, geographic area, local antibiotic susceptibility patterns, history of illness, prior antibiotic therapy, and physical examination. Our review of the literature highlights the most common risk factors for CA-MRSA SSTIs, which included being African-American and close contact with persons with skin infections (Table 6). However, most of the studies were conducted in North America, which has different ethnic factors and climate compared with Hawaii. We have identified obesity and abscess formation as important factors in Hawai‘i and encourage others to assess the value of these measures in evaluating their patients elsewhere.

Table 6.

Summary of Clinical Trials Demonstrating Risk factors for CA-MRSA Skin and Soft Tissue Infection

Study design, reference	Geographic area	Study setting	Study population (Total number)	Comparator	CA-MRSA Infection ratea	Risk factors for MRSA
Prospective, 3	11 cities, USAb	ED	SSTI (422)	Other bacteriac	59%	Antibiotic use in the past month, non-Hispanic African-American, reported spider bite, history of MRSA infection, close contact with persons with skin infection
Prospective, 6	Atlanta, GA, USA	OP, H	SA (389)	CA-MSSA	63%	Female gender, African-American, hospitalization within the past year
Prospective, 7	Chicago, IL, USA	ED, OP, H	CA-SA (1,222)	CA-MSSA	42%	Younger age, HIV infection and incarceration within the past year, African-American
Prospective, 8	Los Angeles, CA, USA	OP	MSM, HIV (111)	Control	NA	Sex partner with skin infection, methamphetamine use, routine hand-on contact with customers at work, frequent fingernail biting, routine use of a public hot tub or sauna
Prospective, 9	Vancouver, BC, Canada	ED	MRSA (190)	HA-MRSA	61%	Presence of abscesses or cellulitis, injection drug use
Prospective, 10	San Diego, CA, USA	OP	Military personnel (202)	Control	11%	Close contact with persons with skin infection, having family or friends working in healthcare settings
Prospective, 20	St. Louis, MO, USA	OP	Football players (53)	Non-SSTI	9%	Being the lineman or linebacker position, high BMI
Prospective, 21	New York, USA	OP	Residents in a religious community (175)	Non-SSTI	14%	Antibiotic use within 12 months, sauna use
Prospective, Our Study	Honolulu, Hawai‘i, USA	OP	SSTI (137)	Other bacteriad	62%	Presence of abscesses, obesity (BMI ≥ 30)

^aDenominators are total numbers of the study populations.

^bIncludes 1) Albuquerque, New Mexico 2) Atlanta, Georgia 3) Charlotte, North Carolina 4) Kansas City, Missouri 5) Los Angeles, California 6) Minneapolis, Minnesota 7) New Orleans, Louisiana 8) New York, New York 9) Philadelphia, Pennsylvania 10) Phoenix, Arizona and 11) Portland, Oregon.

^cIncludes methicillin-susceptible Staphylococcus aureus (n=71), Streptococcus spp. (n=30), coagulase-negative Staphylococci (n=12) and Proteus mirabilis (n=6).

^dIncludes methicillin-sensitive Staphylococcus aureus (n=31), β-hemolytic streptococci (n=6), Pseudomonas aeruginosa (n=4), Klebsiella pneumoniae (n=3), Proteus mirabilis (n=3), Escherichia coli (n=2), Serratia marcescens (n=1) Acinetobacter baumanii (n=1) and Propionibacterium acnes (n=1). Abbreviations: BC = British Columbia; BMI = body mass index; CA = California; CA-MRSA = community-associated methicillin-resistant Staphylococcus aureus; CA-MSSA = community-associated methicillin-susceptible Staphylococcus aureus; CA-SA = community associated Staphylococcus aureus; ED = emergency department; H = Hospitalization; HA-MRSA = healthcare-associated methicillin-resistant Staphylococcus aureus; HIV = human immunodeficiency virus-infected; MO = Missouri; MRSA = methicillin-resistant Staphylococcus aureus; MSM = men who have sex with men; NA = not available; Non-SSTI = non-skin and soft tissue infected; OP = outpatient; SA = Staphylococcus aureus; SSTI = skin and soft tissue infection; USA = the United States of America.

Table 4.

Antimicrobial Susceptibility Data of Methicillin-resistant Staphylococcus aureus Isolates^a

Antibiotics	Number of susceptible isolates/number of tested isolates (%)
Erythromycin	32/85 (38)
Clindamycinb	72/85 (85)
Levofloxacin	75/85 (88)
Tetracycline	82/85 (97)
Trimethoprim-sulfamethoxazole	85/85 (100)
Gentamicin	85/85 (100)
Rifampicin	85/85 (100)
Vancomycin	85/85 (100)

^aAll isolates were recovered from culture specimens obtained during outpatient visit or within 48 hours after hospitalization

^bThe double-disk diffusion test was performed routinely for all erythromycin-resistant Staphylococcus aureus isolates that were susceptible to clindamycin.