SciELO - Scientific Electronic Library Online

vol.79 issue3Pediatric dermatoses at the Clinicas Hospital, Federal University of PernambucoStudy of nine observed cases of Tinea Nigra in Greater Vitória (Espírito Santo state, Brazil) over a period of five years author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Anais Brasileiros de Dermatologia

Print version ISSN 0365-0596On-line version ISSN 1806-4841

An. Bras. Dermatol. vol.79 no.3 Rio de Janeiro May/June 2004 



Erysipelas: a clinical study of 35 pacients hospitalized at the São Paulo Central Hospital of Irmandade da Santa Casa de Misericórdia*



Renata Mie Oyama OkajimaI; Thaís Helena Proença de FreitasII; Clarisse ZaitzIII

IStudent in the Medical Clinic Specialization program, Irmandade da Santa Casa de Misericórdia de São Paulo
IIAssistant Professor, Faculty of Medical Sciences, Santa Casa de São Paulo. Assistant Professor, Dermatology Clinic, São Paulo Hospital Central da Santa Casa de Misericórdia.
IIIAdjunct Professor, Faculty of Medical Sciences, Santa Casa de São Paulo. Assistant-Director, Dermatology Clinic, São Paulo Hospital Central da Santa Casa de Misericórdia





BACKGROUND: Erysipelas and cellulitis are common skin infections.
OBJECTIVES: The aim of this paper is to investigate the frequency, risk factors, clinical features, complications, principal drugs used for treatment and the disease course.
METHODS: The authors studied 35 ward patients who had been diagnosed with erysipelas, and were admitted to the Central Hospital of Irmandade da Santa Casa de Misericórdia de São Paulo between April and August 2002.
RESULTS: Among the patients in the ward during the study 0.87% had been diagnosed with erysipelas. The most common local risk factor was lymphedema, followed by previous episodes of erysipelas. Among the general risk factors, diabetes mellitus, alcohol abuse and cancer were most frequently observed. Local inflammatory signs were found in 97.8% of the patients. Four cases were observed to have complications, which were: necrosis, abscess, deep thrombophlebitis and septicemia. The course was satisfactory in more than 97% of patients.
CONCLUSIONS: Therapy with penicillin was associated with a decrease of complications (p<0.05) and at a lower cost compared to other antibiotic therapies (p<0.05). When anticoagulants were combined to the therapy, there was a lower incidence of complications (p<0.05).

Key words: cellulitis; erysipelas; heparin; therapeutic.




Erysipelas and cellulitis are frequent cutaneous infections, with an estimated incidence in France of 10-to-100 cases per 100,000 inhabitants per year.1 These are clinical entities that are different by definition. However, the clinical and bacteriological criteria used to differentiate them are not significant. It is recommended to consider the two diseases as a single one. 1

These diseases are infections whose main etiological agent is Lancefield Group A beta-hemolytic Streptococcus,2,3,4 qwhich affect both sexes especially during the fifth and sixth decades of life. The lower leg and face are the most affected sites. 1,2,5,6,7

Risk factors are considered to be any alterations facilitating cutaneous infection. Among the local risk factors are: preexisting skin diseases, traumas, operative injuries and vascular alterations, with venous insufficiency and lymphedema.1,2,8,9,10 General risk factors, like diabetes mellitus, alcoholism, corticotherapy, chemotherapy and cancer, produce leukopenia and involve cellular immunity. This damages chemotaxis and the phagocytosis of polymorphonuclears, thus facilitating the occurrence of skin infections. 1,11,12,13

The classical clinical picture is characterized by erythema, edema, heat and pain, accompanied by fever, shivering, malaise and, quite often, nausea or vomiting.1,14,15 The treatment of choice is crystalline penicillin G, which may give rise to cephalosporin or erythromycin and clindamycin in the case of patients who are allergic to penicillin.1,2,16 The anticoagulant therapy is indicated for confirmed or suspected cases of associated thrombophlebitis.1,5

Complications occurred in ratios varying from eight to 30% of cases.15,16 They may be manifested as areas of necrosis, abscesses, gangrene, necrotizing fasciitis, thrombophlebitis, acute glomerulonephritis, septicemia, septic arthritis, endocarditis up to and including death.3,15,16,17 Progression is favorable in 80-to-90.6% of cases.1,11 Mortality may vary from 0.5-to-20% of cases1 depending on the antibiotics used and associated comorbitities. Recurrence within six months may occur in 12% of cases. 9,16,18

In light of the scarcity of epidemiological studies on erysipelas, our present study aims to investigate the frequency of this disease, its local and general risk factors, the clinical picture, main complications, drugs being used, cost of treatment as well as the immediate course in patients hospitalized at the wards of the São Paulo Hospital Central da Irmandade da Santa Casa de Misericórdia.



In a five-month interval (April to August 2002), the investigation took account of prospective patients of both sexes, over 18 years of age and hospitalized at the Medical Clinic or Surgery Departments of the São Paulo Irmandade da Santa Casa de Misericórdia, who were diagnosed with erysipelas. The study was based on a protocol to observe epidemiological, clinical, laboratory, and therapeutic data and disease course during hospitalization.

Various local risk factors were considered such as preexisting skin diseases, traumas, operative injuries, venous and arterial insufficiency, lymphedemas, sequelae of poliomyelitis, osteomyelitis or previous surgeries at the affection site. As for general risk factors, conditions leading to immunodepression were analyzed, such as breast and liver cancer, acute myeloid leukemia, systemic corticotherapy, alcoholism and diabetes, obesity and comorbidities in association with the cutaneous infection picture, with renal insufficiency, cardiac insufficiency and arterial hypertension.

The clinical picture was divided into general and local signs and symptoms. Regarding general forms, the investigation was focused on fevers, prostration, nauseas, vomiting and local lymph node hyperplasia, while the local signs analyzed were: pain, edema, erythema, heat, presence of blisters, pustules, crusts, desquamations, secretion, ulcer, necrosis and abscesses.

The additional tests analyzed were the hemogram, electrolyte dose, urea, creatine, blood glucose, as well as specific examinations like hemoculture, secretion cultures or fragments of cutaneous lesions.

As for treatment and course, the authors analyzed the antimicrobial therapy used and the combination of anticoagulants with the therapy, hospitalization time, presence of complications and genotype of the patient's course. Other complications considered were abscesses, necrosis, deep venous thrombosis (DVT) and spread of infection vector. The cost of antimicrobial treatment was calculated based on the price of an entire day's treatment multiplied by the number of days of medication use. Medication prices were obtained from the Purchasing Sector of the São Paulo Hospital Central da Irmandade da Santa Casa de Misericórdia. Regarding its course, patients were split into three groups: those who had lengthy hospitalization and were cured of cutaneous infections (a), those who showed improvement of the infection, (b) and those who died (c).

The statistical analysis was carried out by SPSS 10.0 software. The Student's t-test was used to compare variables between two groups of patients, who were either using crystalline penicillin and anticoagulants in the treatment, or not. Correlations between the variables studied were analyzed by the Pearson and Spearman coefficient. The data are expressed in an average deviated ± pattern. The statistical significance obtained was p < 0.05.



In the period analyzed, there were 3,981 hospitalizations, with 35 (0.87%) diagnosed with erysipelas; 80% of cases were hospitalized at the Medical Clinic Department. Nineteen patients (54.3%) were female, while 45.7% were male; the average age was 51.2 ± 18.5 years, varying from 18-to-86 years, with 68% of patients over 40 years of age.

Regarding their point-of-entry, clinical lesions suggestive of superficial mycosis, such as onychomycosis and tinea of the foot, were present in 19 patients (54.3%), while other fistulae, such as feet fissures, lower leg ulcers, inframammary and inguinal folds, were present in 34.3% of cases, as demonstrated in table 1. Within local risk factors, lymphedema stood out as it was observed in 15 patients (42.8%), followed by a previous episode of erysipelas, identified in 34.3% of cases, and expressed in table 2. As for general risk factors, diabetes mellitus as well as chronic alcohol abuse were observed in seven patients each (20%); 17.1% were obese, 11.42% had breast or liver cancer, and acute myeloid leukemia, as shown in table 3. An association of local and general risk factors was present in 62.8% of cases, while the presence of isolated local risk factors was identified in 31.4% of the patients analyzed.

The lower limbs were the most affected site, corresponding to 74% of cases, while the face was afflicted in 11%. Among the general symptoms, fever was present in 82.8%, while phlogistic signs were detected in more than 97.8% of the patients studied. Laboratory tests were not relevant regarding the course of the disease. Cultures were carried out in 10 patients (28.57%) (five hemocultures, four cultures of secretion from the lesions, and one culture of a lesion fragment). However, only the culture from the lesion fragment was positive, thereby isolating Enterobacter sp, which might be correlated to the pathogen involved in cutaneous infection.

Thirteen patients (37%) were using antimicrobials prior to hospitalization, during which crystalline penicillin (CP) was used in 24 patients (68.4%). On the other hand, only five patients (14.3%) made use of this antibiotic as monotherapy. Regarding any associations, the most common one was aminoglycosides, observed in 89% of patients.

As for complications, four patients (11.4%) presented with them as precipitated by cutaneous infection, like necrosis, abscess, DVT and spread of the infection vector. Diabetic, obese and immunodepressive patients showed a tendency to develop a higher incidence of complications (14%, 17% and 17%, respectively), compared to the whole sample. On the other hand, the incidence of complications was lower in the group of patients who made use of CP (18.2% vs. 8.3%; p < 0.05) just as with those who made use of the anticoagulant therapy (20.0 vs. 6.9%; p < 0.05), as illustrated in graphic 1.



Average hospitalization time was 9.9 ± 5.9 days. Average cost of antibiotic therapy was R$ 745.90 ± 173.50. The group treated with CP showed hospitalization time equal to those who did not received CP (11.9 ± 2.6 vs. 9.0 ± 0.8 days; p > 0.05). However, the group treated with CP had lower costs for antimicrobial treatment (R$ 114,099.00 ± 53,654.90 vs. R$ 56,480.00 ± 5,358.00; p < 0.05), as shown in graphic 2.



Hospitalization time was directly proportional to age and presence of complications (r = 0.35; p < 0.05 and r = 0.49; p < 0.01; respectively), i.e. the greater the number of local or general risk factors, the higher the presence of complications (r= 0.43; p < 0.01). Twenty nine (83%) patients discharged after the infection had improved, but there was one death.



Erysipelas, also known during antiquity as "Saint-Anthony's Fire", was a bacterial infection with a high mortality rate existing in the pre-antibiotic era usually affecting the periorbital region.15 Nowadays, the definition of erysipelas cannot exclude other dermohypodermitis, like cellulitis. Although the latter is based on clinical or bacteriologic criteria, these two diseases are considered to be a single one.1

In this study, erysipelas was responsible for 0.87% of hospitalizations in the period analyzed. This is a much lower incidence to that reported in the literature.1 On the other hand, it is worth emphasizing that only the patients requiring the most prolonged hospitalization were included. Diagnosed cases treated at first aid centers or hospital ambulatory clinics were not considered. Regarding sex and age range, the study material agreed with the literature, and showed a similar prevalence in both sexes and a higher number of cases in patients as of the fifth decade of life.1,6,15,19,20

As opposed to studies from the previous century, showing the face as the main site of affiction, the lower limbs are currently the most affected areas by these infections. 4,6 In agreement with other authors, this study shows that the lower limbs were affected in more than 70% of patients.2,3,5

It was possible to identify the point-of-entry in 91.4% of cases analyzed. Clinical lesions suggestive of superficial mycosis are the most common points (54.3%). These data back up the findings in the literature, in which most cases of erysipelas show some fistulae on the cutaneous surface. 5,15

In the authors' material, local risk factors were present in 94% of the patients studied, and 63% of cases were associated with general risk factors. Among local risk factors, lymphedema was the most frequent, found in 43% of cases; and previous episodes of erysipelas were present in 34% of patients. Prior affliction by a dermohypodermitis is important for the recurrence of these infections due to the fact of occasioning local anatomic and functional alterations that in turn give rise to the lymphedema.4,9 Among general risk factors, in agreement with the literature, those compromising the patient's immunity stand out in the study, like diabetes mellitus, alcoholism, cancer and systemic corticotherapy, which were present in 54.3% of cases.1,2,11,15,20

The diagnosis was made clinically, as advised in the literature,1,3,19 given that in the physical examination, phlogistic signs, like edema, erythema, heat and pain were present in 97.8% of the patients analyzed. Fever, in spite of being described by some authors as an obligatory condition for the diagnosis of erysipelas,2,5,6 was observed during hospitalization in only 82.8% of the patients studied. By constrast, other authors assert the presence of fever to not be indispensable for a clinical diagnosis,15 mainly owing to the possibility of previous antibiotics use, as observed in 37% of the patients in this study.

Cultures were carried out in 28.7% of patients. But in only one case, whose material was obtained by a lesion biopsy, could the isolated germ be identified as a possible causative agent. Taking cultures of these infections, mainly hemocultures, has been questioned as to its applicability, given that it is difficult to isolate the causative agent and by the fact that, when positive, it does not change the initial therapeutic approach.2,6

Crystalline penicillin is the drug of choice for treating erysipelas or cellulitis when requiring hospitalization.1 Meanwhile, the authors observed in this study that there was no standardization in choosing antibiotic therapy because 68.6% of patients made use of CP as a part of treatment and 14.3% used CP alone, as advised by the literature. The use of a wide range of antibiotics did not influence patients' course. However, it did represent a lower cost when compared to CP. Still, it must be emphasized that this study only included patients hospitalized in the wards. Perhaps CP is suitable for cases presenting with greater involvement of a patient's general state. More serious cases, though, require a broad range of antibiotics.

Patients who made use of CP showed lower hospitalization time and lower cost related to antibiotic therapy, as well as a tendency for higher incidence of complications. By contrast, and in accordance with this study, there was no standardization of antibiotic therapy in cases of erysipelas or cellulitis in the hospitalized ward patients investigated.

The use of anticoagulants is indicated by the majority of authors when facing or suspecting deep venous thrombosis or thrombophlebitis, but not in its prophylaxis.1,5 On the other hand, patients who made use of anticoagulants showed a lower incidence of complications when compared to those not using this therapy. This finding might be due only to the prophylaxis of thrombotic events, but also to the anti-inflammatory effect recently attributed to heparin. 21



The authors concluded that the incidence of erysipelas and cellulitis in the period studied was 0.87%, that females (54.3%) and patients over the fifth decade of life were the most affected (68%), and that the lower limbs were the sites to be most frequently involved (74%), followed by the face (11%).

General and local risk factors that stood out were diabetes, cancer, lymphedema and prior episodes of erysipelas or cellulitis. As for erysipelas' point-of-entry, the lesions suggestive of superficial mycosis were the main alterations that could be responsible for the bacteria's penetration.

The clinical picture observed in the patients studied matched with those described in the specialized literature. The complications observed were one case each of abscesses, necrosis and DVT.

Treatment with CP proved to be effective, and showed a good cost/benefit ratio due to its low price. Regarding the use of heparin as a back-up therapy, new studies are necessary to confirm what the benefits associated with this drug might be.



1. Chistmann D et al. Erysipèle et fasciite nécrosante:prise en charge. Ann Dermatol Venereol 2000;127(12):1118-37.         [ Links ]

2. Bishara J, Golan - Cohen, Robenshtok E, Leibovici L, Pitlik S. Antibiotic use in patients with erysipelas: a retrospective study. Isr Med Assoc J 2001;3(10):722-4.         [ Links ]

3. Chartier C, Grosshans E. Erysipelas: an update. Int J Dermatol 1996;35(11):779-81.         [ Links ]

4. Bonnetblanc JM. Infections cutanés bactériennes: impétigo, furoncle, érysipéle. Etiologie, diagnostic, évolution, traitement. Rev Prat 2001; 51(2):223-8.         [ Links ]

5. Crickx B, Chevron F, Signal-Nahum M et al. Érysipèle: donnés èpidémiologiques, cliniques et thérapeutiques. Ann Dermatol Venereol 1991;118:11-6.         [ Links ]

6. Bernardes CHA, Cardoso KT, Augusto JCA, Santos JR, Lopes LT. Experiência clínica na avaliação de 284 casos de erisipela. An bras Dermatol 2002;77(5):605-9.         [ Links ]

7. Duvanel T, Harms M. Erysipèle et cellulites infectieuses:classification, approche diagnostique, traitement. Schweiz Rundschau Med 1987;76:216-9.         [ Links ]

8. Carpentier PM, Colomb M, Poensin D, Satger B. Incidence de l´érysipèle des membres inférieurs en milieu thermal phlébologique. J Mal Vasc 2001;26(2):97-9.         [ Links ]

9. Dupuy A. Épidémiologie discriptive et connaissance des facteurs de risque de l érysipèle. Ann Dermatol Venerol 2001;128:312-6.         [ Links ]

10. Cestari SCP, Petri V, Castiglioni MLV, Lederman H. Linfedemas dos membros inferiores: estudo linfocintilográfico. Rev Assoc Med Bras 1994;40(2):93-100.         [ Links ]

11. Lanoux P, Penalba C, Legin C, Kivade M, Reveil JC. L´érysipèle. A propos de 118 observations. Med Mal Infect 1993;23:908-12.         [ Links ]

12. Harrison TR, Fauci AS, Braunwald E et al. Medicina Interna. 14ª ed. Rio de Janeiro: Mc Graw Hill, 1998:2967p.         [ Links ]

13. Wilson JD, Foster DW, Kronenberg HM et al. Textbook of endocrinology. 9ª ed. Philadelphia: Saunders, 1998:1819p.         [ Links ]

14. Guberman D, Gilead LT, Zlotogorski A, Schamroth J. Bullous erysipelas: a retrospective study of 26 patients. J Am Acad Dermatol 1999;41(5):733-7.         [ Links ]

15. Jégo P, Resche S, Karacatsanis C et al. L´érysipéle: une sèrie rétrospective de 92 patients dans un service de Médicine interne. Ann Med Interne 2000;151(1):3-9.         [ Links ]

16. Chartier C, Grosshans E. Erysipelas. I J Dermatol 1990; 29(7):459-67.         [ Links ]

17. Eriksson B, Jorup-Ronstrom C, Karkkonen K, Sjoblom AC, Holm SE. Erysipelas: clinical and bacteriologic spectrum and serological aspects. Clin Infect Dis 1996;23:1091-8.         [ Links ]

18. Jorup-Ronstrom C, Britton S. Recurrent erysipelas: predisposing factors and costs of prophylaxis. Infection 1987;15:25-6.         [ Links ]

19. Petit A. Érysipèle. Données récentes et questions d´actualité. Ann Dermatol Venereol 1996;123:585-93.         [ Links ]

20. Dupuy A, Benchikhi H, Roujeau JC et al. Risk factors for erysipelas of the leg (cellulitis): case control study. Br Med J 1999;118:1591-4.         [ Links ]

21. Tyrrell DJ, Horne AP, Holme KR, Preuss JMH, Page CP. Heparin in inflammation: potencial therapeutic applications beyond anticoagulation. Adv pharmacology 1999;46:151-89.         [ Links ]

Correspondence to
Renata Okajima
Rua Borges Lagoa, 908 / 22 - Vila Clementino
04038-002 São Paulo SP
Telefone: (11) 5082-2152
Fax: (11) 5084-2751

Received in September, 22nd of 2003.
Approved by the Consultive Council and accepted for publication in March, 12th of 2004.



* Work done at "Irmandade da Santa Casa de Misericórdia de São Paulo. Departamento de Clínica Médica e Dermatologia".

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License