Download Clinical and epidemiological characteristics of adult - LADERM-Ba

Eur J Clin Microbiol Infect Dis
DOI 10.1007/s10096-012-1549-2
ARTICLE
Clinical and epidemiological characteristics of adult patients
hospitalized for erysipelas and cellulitis
M.-R. Perelló-Alzamora & J.-C. Santos-Duran &
M. Sánchez-Barba & J. Cañueto & M. Marcos &
P. Unamuno
Received: 10 October 2011 / Accepted: 6 January 2012
# Springer-Verlag 2012
Abstract The purpose of this investigation was to analyze
the clinical and epidemiological aspects of all cases of
erysipelas and infectious cellulitis admitted to a tertiary
hospital during a period of five years. All patients admitted
with the main diagnosis of erysipelas or cellulitis to the
Department of Dermatology of the author’s institution from
January 2005 to May 2010 were included. Seventy patients
were identified and their medical records were retrospectively reviewed so as to record the epidemiological and
clinical data. Univariate and multivariable analyses were
performed to analyze variables that predicted longer length
of stay. The frequency of cellulitis in the lower limbs was
higher in men and patients older than 65 years. Moderate/
severe cellulitis in patients with basal comorbidity followed
by a poor response to oral antibiotic therapy for 48 h were
the most common reasons for admission. At arrival, four
patients had abscessed areas. Fourteen patients developed
local complications and 18 cases developed general inM.-R. Perelló-Alzamora (*) : J.-C. Santos-Duran : P. Unamuno
Department of Dermatology, University Hospital of Salamanca,
Paseo San Vicente, 58-182,
37007, Salamanca, Spain
e-mail: [email protected]
M. Sánchez-Barba
CAIBER Research Unit, University Hospital of Salamanca,
Salamanca, Spain
M. Marcos
Department of Internal Medicine,
University Hospital of Salamanca,
Salamanca, Spain
J. Cañueto
Department of Dermatology, Hospital Nuestra Señora de Sonsoles,
Ávila, Spain
hospital complications. Most patients improved or were
healed with intravenous amoxicillin–clavulanate 1 g–
200 mg/8 h. Intravenous amoxicillin–clavulanate 1 g–
200 mg/8 h may be a good choice for empiric treatment in
our setting. The development of in-hospital complications
and the need for changing empiric antibiotic therapy were
significant and independent variables associated with longer
length of stay.
Introduction
Cellulitis is an inflammatory skin condition with an
infectious origin which affects the deep dermis and the
subcutaneous tissue [1]. Erysipelas affects the upper
dermis, and it is usually accompanied by lymphatic
involvement [1]. Patients with these diseases may usually receive outpatient treatment [2–4], but elderly
patients or those with important comorbidities or moderate–severe infections may require admission [5, 6],
and the length of hospitalization is usually determined
by other factors rather than cellulitis.
To date, there are scarce data regarding clinical aspects,
microbiological spectrum (probably due to the difficulty in
isolating a pathogen from the infected area), and pathogenesis of cellulitis [6]. Accordingly, there is no accepted consensus regarding admission criteria and empirical antibiotic
therapy in moderate-severe cellulitis patients admitted to
hospital [3, 6].
In view of this, the aim of this study was to analyze the
epidemiological aspects of all cellulitis and erysipelas cases
admitted to our department and to study the factors associated with the length of hospitalization.
Eur J Clin Microbiol Infect Dis
Patients and methods
Patients
This retrospective study was conducted in a 1,000-bed tertiary referral hospital in Salamanca, Spain. The hospital
serves an urban and rural area with a population of
350,000 inhabitants. All patients admitted to the Department
of Dermatology with a main diagnosis of cellulitis or erysipelas between January 2005 and May 2010 were included
and episodes were identified through the database maintained by the Admission Department of the University Hospital of Salamanca.
Cellulitis and erysipelas were defined by the presence of
cutaneous tenderness, pain, and erythema, and both entities
were considered to be non-complicated soft tissue infections. Although there are clinical features differentiating
cellulitis from erysipelas, for the purposes of this study,
the term “cellulitis” was used invariably to denote infection
involving the skin and/or the subcutaneous tissue. Patients
who were admitted with deep tissue infections (myositis or
necrotizing fasciitis) were not included in this study. Other
exclusion criteria were as follows: patients younger than
14 years of age, patients with a main diagnosis or a reason
for hospitalization that was not cellulitis or erysipelas, and
patients who were admitted to other departments. Patients
with cellulitis and severe comorbidities, severe sepsis, or
those with cellulitis but other main reason for admission
were normally admitted to the General Internal Medicine
Ward.
The following data were retrospectively recorded for
each episode: age, sex, location of infection, risk factors,
origin of infection, length of hospitalization, reason for
admission, prior episodes of cellulitis/erysipelas, general
symptoms on admission and at discharge (fever, malaise),
local symptoms on admission and at discharge (including
classic manifestations such as erythema, edema, tenderness,
and pain, and non-classic manifestations: blisters, purpuric
or ecchymotic areas, pustule areas, and abscessed areas),
microbiological results (results of blood cultures and swabs
collected from infected areas, results of skin biopsies), the
use of imaging studies, treatment before admission (antibiotic therapy and topical treatment), empirical and definitive
antibiotic therapy, local and general complications, and 30day outcome.
with Bonferroni correction was used. Two-tailed tests of
significance were performed and p-values < 0.05 were
regarded as being significant. Variables associated with the
length of stay with p<0.20 in the univariate analysis were
selected for inclusion in a logistic regression model, using a
forward stepwise selection algorithm with entry and removal criteria of p00.05 and 0.10, respectively.
Results
From January 2005 to May 2010, 1,356 patients were admitted to the Department of Dermatology of our institution
and 70 patients presented with a main diagnosis of cellulitis
or erysipelas. A total of 69 patients had one episode and one
patient with chronic lymphedema had three episodes of
erysipelas in her left arm. For the purposes of this study,
we only considered one episode. Therefore, 70 episodes in
70 patients were included. The number of patients per year
with these diagnoses ranged from a minimum of 7 patients
(year 2008) to a maximum of 16 patients (year 2007), and
the number of hospital stays due to cellulitis in the Department of Dermatology divided by the total number of hospital stays over the year is shown in Fig. 1.
Clinical and epidemiological characteristics
Among the 70 patients, 43 (61.4 %) patients were men and
27 (38.6%) were women. The characteristics of the 70 cases
included in the study are detailed in Table 1. The most
common site of infection was the lower limbs (67.1%). Risk
factors for cellulitis were observed in 35 patients (50%), and
the most common was the presence of chronic venous
insufficiency or stasis dermatitis (21.4%). In our series, the
frequency of cellulitis on the lower limbs was higher in men
(odds ratio [OR] 1.913) and in patients over 65 years old
(OR 1.44).
Statistical analysis
All data were collected from the patients’ clinical records
and they were processed with SPSS (Statistical Package for
Social Sciences) version 18.0. For categorical variables,
Pearson’s χ2 or Fisher’s exact test was used; for continuous
variables, Student’s t-test or one-way analysis of variance
Fig. 1 Number of hospital stays due to cellulitis admitted to the
Department of Dermatology per 10,000 patient-days in our hospital
Eur J Clin Microbiol Infect Dis
Table 1 Characteristics of the 70 episodes of cellulitis/erysipelas
requiring hospital admission
Variable
n (%)
Age (years)
63.8
(18.6)
Gender
Female
23 (38.3)
Male
43 (61.4)
Location
Lower limbs
47 (67.1)
Upper limbs
12 (17.1)
Head and neck
Risk factors for cellulitis
11 (15.7)
Chronic venous insufficiency or stasis dermatitis
35 (50)
Diabetes mellitus
Mammary tumor and chronic lymphedema
8 (11.4)
4 (5.7)
Immunosuppression
Joint presence of diabetes mellitus and chronic venous
insufficiency
Previous surgery
3 (4.3)
2 (2.9)
Alcoholism
Source of infection
Wounds or ulcers
Web intertrigo
Arthropod bites
2 (2.9)
hospitalization were: face involvement (8), suspected complications such as necrotizing fasciitis or osteomyelitis (2),
immunosuppression (2), and inability of the patient to carry
out the outpatient treatment prescribed (14).
Microbiological results
With regard to the microbiological diagnosis, cutaneous
swabs were taken from patients with open skin lesions in
19 cases (35.2%), and a pathogen was identified in 13 of
them (Table 2). Blood culture was performed in 26 patients
(37.1%), and only one patient had bacteremia secondary to
Pseudomonas aeruginosa infection. Skin biopsy was performed in two patients with erysipeloid symptoms in the
breast, with a personal record of mammary tumor and under
the diagnostic suspicion of neoplastic infiltration. In both
cases, the skin biopsies revealed perivascular dermatitis with
no criteria for specific causation.
Antibiotic therapy
1 (1.4)
38 (54.3)
8 (11.4)
4 (5.7)
Categorical variables are presented as absolute (relative) frequencies
and numerical variables as mean (standard deviation)
With regard to systemic clinical symptoms that were
observed on admission, malaise was found in 37% of
patients (26 cases). Local signs and symptoms of inflammation, including soft tissue swelling, erythema, and pain,
were present in all patients. With regard to unusual manifestations of cellulitis, blisters were observed in 7 cases
(10%), purpuric or ecchymotic areas were seen in 18 cases
(25.71%), pustule areas were observed in 5 cases (7.14%),
abscessed areas were seen in 4 cases (5.71%), traumainduced hematoma was observed in 3 cases (4.28%), and
there was one case of hematoma due to coagulation disorder
or plateletpenia (1.43%).
Before admission, 45 patients (64.3%) had previously
consulted with their primary care physician and 24
(53.3%) were correctly diagnosed with cellulitis. The rest
of the patients were diagnosed with microcrystalline arthritis
(1), vasculitis (1), thrombophlebitis (10), septic arthritis (1),
allergic reaction (1), and herpes simplex infection (1). In 6
cases, the primary care report does not include a diagnosis.
The most common reason for admission was moderate–
severe cellulitis in elderly pluripathologic patients or in
patients with several comorbidities (27 patients, 38.6%),
followed by a lack of improvement despite oral antibiotic
therapy for 48 h (17 patients, 24.3%). Other reasons for
Concerning treatment prior to admission, eight patients were
given amoxicillin–clavulanate 875/125 mg/8 h, and five
patients were given cloxacillin 500 mg/6 h. Topical treatment would have to be prescribed in 23 cases (79%), but
they were only administered in 4 cases (13%). The prescribed topical treatments were acyclovir and mupirocin,
one case each. Drainage of a sero-hematic fluid/abscess
was performed in two patients.
With regard to empirical therapy after admission, 50
patients (71.4%) received systemic antibiotic therapy plus
topical treatment; in 12 cases (17.1%), only systemic antibiotic therapy was used; in 5 cases (7.1%), systemic antibiotic therapy was prescribed together with topical treatment
and drainage of a purulent or serohematic collection; and in
3 cases (4.3%), systemic antibiotic therapy and topical treatment were combined with ulcer debridement. The empirical
antibiotic treatment on admission is detailed in Table 3.
Amoxicillin–clavulanate was not administered in two cases
Table 2 Microorganisms isolated in the 70 episodes of cellulitis
Microorganism
n
MSSA (methicillin-sensitive Staphylococcus aureus)
Pseudomonas aeruginosa
Morganella morganii
Serratia marcescens
MRSA (methicillin-resistant S. aureus)
Polymicrobial infection
MSSA and Streptococcus pyogenes
P. aeruginosa and S. pyogenes
MSSA and Proteus vulgaris
5
1
2
1
1
3
1
1
1
Eur J Clin Microbiol Infect Dis
due to allergy to penicillin and its derivatives, and macrolide
(erythromycin) plus clindamycin was administered instead.
Antibiotic doses were adjusted to the creatinine clearance in
three patients.
Empirical antibiotic was changed in 16 cases because of
poor clinical and/or analytical evolution (7 cases, 43%), due
to resistance to the initial antibiotic (6, 37.5%), and due to
adverse side effects or intolerance (3 cases, 18.8%). Antibiotics used in these cases were: ceftriaxone (1), ciprofloxacin plus clindamycin (2), erythromycin (2), clarithromycin
(1), cloxacillin (2), imipenem (2), linezolid (1), vancomycin
(1), meropenem plus vancomycin (2), and piperacillin–tazobactam (2). This second antibiotic choice had to be replaced
in two patients, due to poor clinical evolution, and the third
option was cefotaxime plus cloxacillin and imipenem in one
case each. A fourth antibiotic alternative was not necessary.
The topical treatments are also detailed in Table 3 and
other topical treatments for concomitant processes were
applied in 16 cases: treatment of interdigital tinea pedis with
topical antifungal (ketoconazole or terbinafine) in eight
cases, treatment of candidal intertrigo on the inguinal or
inframammary folds with topical antifungal (nystatin or
clotrimazole) in seven cases, treatment of myiasis with
topical metronidazole and occlusion of the lesion in one
case.
Outcome and complications
With regard to local complications, they appeared in 14
patients. The most common was the presence of vesicles,
blisters, denudation or erosion (9 cases), followed by dissecting hematoma (2 cases), abscess that required draining
Table 3 Empirical treatment on
admission in our series
(2 cases), and diabetic foot with involvement of deep tissue
that required debridement (1 case). With regard to general
complications, they appeared in 18 patients (25.7%), and the
most common was the decompensation of comorbid conditions (16 cases), followed by sepsis (2 cases). The mean
length of hospitalization was 6.99 days ± 4.27 days, with a
minimum stay of 1 day and a maximum stay of 23 days
(range: 21 days). Variables associated with hospital stay
longer than 7 days are shown in Table 4. After multivariable
analysis, the development of in-hospital complications and
the need for changing the empiric antibiotic therapy were
independently and significantly associated with longer
length of stay (OR: 4.27, 95% confidence interval [CI]:
1.27–14.49; p 00.019 for in-hospital complications and
OR: 3.66, 95% CI: 1.06–12.66; p00.019 for substitution
of the initial antibiotic).
The evolution was very good (with rapid and complete
healing without after-effects) in 34 cases (48.6%); it was
good (with slower healing, but complete and without aftereffects) in 31 cases (44.3%); it was poor (with an acceptable
improvement but with residual after-effects) in 3 cases
(4.3%), and it was bad (death) in 2 cases (2.9%). Overall,
68 patients (97.1%) had total or partial recovery, and two
patients died due to uncontrolled sepsis (2.9%).
With regard to clinical symptoms, at discharge, 97% of
patients did not have fever or malaise. Local erythema still
persisted in 75% of the patients, and 21% of them reported
pain at discharge. Topical treatment, analgesics, and oral
antibiotic therapy were prescribed in 36 cases at discharge
(51.4%); oral antibiotic therapy and analgesics in 16 cases
(22.9%); topical treatment and analgesics in 5 cases (7.1%);
and analgesics alone in 11 cases (15.7%).
Empirical antibiotic therapy
n (%)
Amoxicillin–clavulanate IV 1 g/200 mg q8h
Amoxicillin–clavulanate IV 1 g/200 mg q8h and IV clindamycin 1.2 g q8h
Amoxicillin–clavulanate IV 500 mg/50 mg q8h and IV clindamycin 1.2 g q8h
60 (89.5)
16/60 (26.7)
2/60 (3.3)
Amoxicillin–clavulanate IV 500 g/50 mg q8h and ciprofloxacin VO 500 mg q12h
Ceftriaxone IV 1 g q24h and cloxacillin IV 1 g q8h
Azithromycin VO 500 mg/24 h and clindamycin IV 1.2 g q8h
Cefoxitin IV 1 g q8h
Vancomycin IV 1 g q12h and metronidazole IV 500 mg q8h
Topical treatments
Zinc sulfate fomentations 1/1,000 and mupirocin
Zinc sulfate fomentation and bacitracin and neomycin sulfate and polymyxin B sulfate
Zinc sulfate fomentation 1/1,000 and silver sulfadiazine
Zinc sulfate fomentation 1/1,000
Mupirocin
Silver sulfadiazine
Bacitracin and neomycin sulfate and polymyxin B sulfate
1/60 (1.6)
1 (1.5)
1 (1.5)
1 (1.5)
1 (1.5)
n (%)
22 (41.4)
2 (3.8)
1 (1.9)
3 (5.7)
16 (30.2)
6 (11.3)
3 (5.7)
Eur J Clin Microbiol Infect Dis
Table 4 Factors associated with prolonged hospital stay in the 70 patients admitted to hospital with cellulitis
Variables
Number of patients with stay longer
than 7 days/number of patients (%)
Univariate
p-value
Multivariate OR
(95% CI), p-value
Age >60 years
23/44 (52.3)
0.08
NS
Male sex
15/43
0.046
NS
Unusual manifestations on admission (blisters,
pustule areas, abscessed areas, hematoma)
Risk factors for cellulitis
Local complications during admission
12/25 (48.0)
0.641
NS
20/35 (57.1)
9/14 (64.3)
0.03
0.092
NS
NS
Medical complications during admission
13/18 (72.2)
0.006
4.27 (1.27–14.49), p00.019
Prior cellulitis episodes
Isolated pathogen
3/12 (2.5)
9/13
0.140
0.392
NS
NS
Empirical antibiotherapy on admission
14/39
0.209
NS
Substitution of the initial antibiotic
5/17
0.012
3.66 (1.06–12.66), p00.019
NS non-significant
Discussion
Cellulitis and erysipelas are common among the general
population, although few series have been published on
the last 10 years and their exact prevalence is still unknown
[5, 6]. In our series of cases, as previously described [6],
more cases are described in patients older than 60 years,
with the most common location being the lower limbs. We
also report a similar prevalence of risk factors for cellulitis,
such as chronic venous insufficiency and diabetes mellitus,
as published in the literature [6]. With regard to the origin of
the infection, it has been registered more cases of cellulitis
in patients with ulcers than in patients with web intertrigo.
As expected, all patients presented with the classical local
symptoms of cellulitis (a painful erythematous plaque with
not well-defined edges). We also included patients with
abscess, since its presence on admission might be related
to a higher risk of complications and, therefore, to a potential increase in the hospitalization stay. We acknowledge,
however, that the presence of abscessed areas on the cellulitis plaque is a rare non-classical manifestation of cellulites
and that it may be difficult to differentiate abscess on admission from cellulitis. Being that cellulitis prognosis is
highly variable and microbiological diagnosis difficult to
carry out, clinical findings gain a special importance. For
this reason, we believe that the presence of this clinical sign
at the time of diagnosis should be taken into account in
classifying the risks and the approaches taken in the Emergency Room in order to decide on hospitalization versus
outpatient treatment.
The mean duration of hospitalization was 6.99 days ±
4.27, in accordance with previously published data [2, 5].
The mean duration of hospitalization for cellulitis in a
tertiary hospital ranges between 5 and 10 days [2]. With
regard to microbiological diagnosis, we confirm in our study
that the lack of a microbiological diagnosis is common in
patients with cellulitis [2, 5–7]. As for specific microorganisms, the leading pathogen was methicillin-sensitive S. aureus (MSSA), which is in accordance with the data
published by Carratalà et al. [5], although other authors have
found that the most common pathogen is S. pyogenes,
followed by MSSA [2, 3]. Besides, we have not found that
methicillin-resistant S. aureus (MRSA) is a pathogen of
concern in our setting [1, 3], although more studies are
needed in order to confirm this aspect . In our hospital, the
prevalence of MRSA in S. aureus bacteremia was 27.3%
last year and there are no other data available about the
prevalence of MRSA in our region of Spain (Castilla-Leon).
Some data has been published in other regions showing that
the cumulative incidence of MRSA in long-term care facilities appears to be approximately 20% per year [8].
In our series, the most common reason for hospitalization
was moderate or severe cellulitis in patients with basal
comorbidities, followed by a poor response to oral antibiotic
therapy after 48 h. With regard to empirical antibiotic therapy, most patients showed improvement or healing with
intravenous amoxicillin–clavulanate 1 g–200 mg/8 h, with
no serious adverse effects. In some cases, this drug could not
be administered due to allergy to penicillin, and in other
cases, it had to be replaced due to a lack of response or to the
appearance of adverse effects. In these circumstances, quinolones or clindamycin were prescribed with satisfactory
results.
Currently, there is no consensus regarding empirical antibiotic therapy in erysipelas and cellulitis. Recent studies
[2] recommend an empirical antibiotic therapy of cloxacillin
or first-generation cephalosporins as the first therapeutic
choice, and levofloxacin, moxifloxacin, or clindamycin as
therapeutic alternatives, except for the case of MRSA suspicion or beta-lactam allergy. In these cases, linezolid is the
Eur J Clin Microbiol Infect Dis
first therapeutic choice, and vancomycin–teicoplanin, clindamycin, or cotrimoxazole are listed as alternatives [1].
With regard to the complications, most of the patients
that were admitted to our department did not present severe
complications, and only in two cases was the outcome fatal,
with uncontrolled sepsis followed by death. Although cellulitis is a disorder that has traditionally been considered a
benign entity, its global mortality rate after 30 days is 5%
[5], and mortality after one year is estimated at 20% [9]. In
our series, the global mortality rate was 2.85%. We must
admit, however, that a potential limitation of our work
regarding this point is that patients admitted to our Department of Dermatology usually have fewer comorbidities than
patients admitted to the General Internal Medicine Ward. As
a result, the rate of complications as well as the mortality
rate may be lower in our work when compared with other
series.
Local complications appeared in 14 patients, the most
common being the presence of vesicles, blisters, denudation,
or erosion in the plaque of cellulitis. Skin abscesses have been
traditionally considered as the most common local complication [2] . The development of general complications on admission and the need for changing the antibiotic therapy are
the only variables significantly associated with longer hospital
stay. We think that the small sample size of the study could
limit the results in the multivariable analysis.
We believe that more studies on empirical antibiotic
therapy and prophylaxis for cellulitis recurrence are needed
in order to confirm the findings observed in our case series.
Conflict of interests
interests.
We declare that we do not have any conflict of
References
1. Stevens DL (2009) Treatments for skin and soft-tissue and surgical
site infections due to MDR Gram-positive bacteria. J Infect 59
(Suppl 1):S32–S39
2. Figtree M, Konecny P, Jennings Z, Goh C, Krilis SA, Miyakis S
(2010) Risk stratification and outcome of cellulitis admitted to
hospital. J Infect 60(6):431–439
3. Stevens DL, Bisno AL, Chambers HF, Everett ED, Dellinger P,
Goldstein EJ et al (2005) Practice guidelines for the diagnosis and
management of skin and soft-tissue infections. Clin Infect Dis 41
(10):1373–1406
4. Weigelt J, Itani K, Stevens D, Lau W, Dryden M, Knirsch C (2005)
Linezolid versus vancomycin in treatment of complicated skin and soft
tissue infections. Antimicrob Agents Chemother 49(6):2260–2266
5. Carratalà J, Rosón B, Fernández-Sabé N, Shaw E, del Rio O, Rivera
A et al (2003) Factors associated with complications and mortality
in adult patients hospitalized for infectious cellulitis. Eur J Clin
Microbiol Infect Dis 22(3):151–157
6. Lazzarini L, Conti E, Tositti G, de Lalla F (2005) Erysipelas and
cellulitis: clinical and microbiological spectrum in an Italian tertiary
care hospital. J Infect 51(5):383–389
7. Lebre C, Girard-Pipau F, Roujeau JC, Revuz J, Saiag P, Chosidow
O (1996) Value of fine-needle aspiration in infectious cellulitis.
Arch Dermatol 132(7):842–843
8. Manzur A, Gudiol F (2009) Methicillin-resistant Staphylococcus aureus in long-term-care facilities. Clin Microbiol Infect 15(Suppl 7):26–30
9. Tan R, Newberry DJ, Arts GJ, Onwuamaegbu ME (2007) The
design, characteristics and predictors of mortality in the North of
England Cellulitis Treatment Assessment (NECTA). Int J Clin Pract
61(11):1889–1893