|Year : 2019 | Volume
| Issue : 2 | Page : 159-165
A retrospective study on demographic and clinical characteristics of cutaneous leishmaniasis suspected cases in southern Israel, 2013-2016: Comparison between confirmed and negative cases
Hila Bufman1, Orli Sagi2, Yonat Shemer3, Amir Horev1, Naphtali Justman1, Elina Bazarsky2, Shalom Ben-Shimol1
1 Pediatric Infectious Disease Unit; Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer-Sheva, Israel
2 Faculty of Health Sciences, Ben-Gurion University of the Negev; Parasitology Laboratory, Soroka University Medical Center, Beer-Sheva, Israel
3 Faculty of Health Sciences, Ben-Gurion University of the Negev; Virology Laboratory, Soroka University Medical Center, Beer-Sheva, Israel
|Date of Submission||21-Nov-2017|
|Date of Acceptance||15-Nov-2018|
|Date of Web Publication||31-Jul-2019|
Pediatric Infectious Disease Unit, Virology Laboratory, Soroka University Medical Center, P.O. Box 151, Beer-Sheva–84101
Source of Support: None, Conflict of Interest: None
Background & objectives: Clinical diagnosis of cutaneous leishmaniasis (CL) may bear a high rate of false diagnosis. This study assessed CL-suspected episodes, in an attempt to differentiate confirmed CL and non-CL diagnoses.
Methods: In this retrospective, case-control study, medical files of CL-suspected episodes, tested by a biopsy for Leishmania-PCR, from 2013 to 2016, were collected and analysed statistically.
Results: Of 324 suspected CL episodes, 48.8% were PCR-confirmed CL (96.2% Leishmania major) and 51.2% were non-CL (57.1% bacterial infections). Overall, 59.3% episodes were in males. Mean (± SD) duration until diagnosis was 3.7 ± 7.2 months. Lesions (mean 2.9 ± 3.8 per episode) were mostly (60.8%) sampled from September through February. Ulcer, pain, itching, purulent discharge and fever were recorded in 55.2, 47, 42.9, 18.2 and 4.7% of episodes, respectively. Univariate analysis showed that male gender, multiple lesions, ulcer, >1-month duration until diagnosis, and seasonality were associated with CL. Empiric CL treatment was recorded in 63.4 and 16% of CL-confirmed and non-CL episodes, respectively (p <0.001); and was observed to be associated with Jewish ethnicity, seasonality, multiple lesions, ulcer, absence of fever and duration of >1-month until diagnosis. In multivariate analysis, seasonality (odds ratio, OR = 2.144), empiric CL treatment (OR = 5.144) and ulcer (OR = 2.459) were associated with CL. Empiric CL treatment was associated with Jewish ethnicity (OR = 2.446) and duration of >1-month until diagnosis (OR = 3.304).
Interpretation & conclusion: CL diagnosis should be laboratory confirmed, as clinical appearance is often misleading. Seasonality, ulcer appearance and gender may aid in correct identification and treatment of CL cases.
Keywords: Clinical manifestation; cutaneous leishmaniasis; Israel; Leishmania major; non-CL cases
|How to cite this article:|
Bufman H, Sagi O, Shemer Y, Horev A, Justman N, Bazarsky E, Ben-Shimol S. A retrospective study on demographic and clinical characteristics of cutaneous leishmaniasis suspected cases in southern Israel, 2013-2016: Comparison between confirmed and negative cases. J Vector Borne Dis 2019;56:159-65
|How to cite this URL:|
Bufman H, Sagi O, Shemer Y, Horev A, Justman N, Bazarsky E, Ben-Shimol S. A retrospective study on demographic and clinical characteristics of cutaneous leishmaniasis suspected cases in southern Israel, 2013-2016: Comparison between confirmed and negative cases. J Vector Borne Dis [serial online] 2019 [cited 2019 Aug 23];56:159-65. Available from: http://www.jvbd.org/text.asp?2019/56/2/159/263723
| Introduction|| |
Leishmaniasis, a vector-borne zoonosis, is endemic and notifiable in Israel. Leishmania parasites can cause different clinical manifestations, including cutaneous leishmaniasis (CL), mucocutaneous or visceral disease,,. In CL, the extra-cellular stages of the Leishmania parasites (promastigotes) are transmitted (during blood meal) by a sandfly vector into skin macrophages, where they transform into the cellular phase (amastigote). The disease manifests itself as a cutaneous lesion at the site of the sandfly bite,.
Cutaneous leishmaniasis is common in the Old World (including the Middle East) and the New World regions,. Outbreaks of CL have been reported in the Old World, mostly in the context of human invasion to the zoonotic cycle,,,,. In Israel, the major endemic species are Leishmania major (prevalent in southern Israel) and L. tropica (prevalent around Jerusalem and the northern Jordan valley),,. These species cause disease in both the humans and reservoir host animals. For L. major, the reservoir hosts are small rodents,.
The cutaneous disease is characterized by one or more skin lesions that may appear as a papule or ulcer, or characteristically as a round “wet ulcer” lesions (especially in case of L. major infections),. In general, CL cannot be diagnosed solely on the basis of lesion’s appearance,. Commonly, CL lesions are painless, non-purulent, and non-itching,. However, it is not uncommon for CL lesions to be infected with bacteria (secondary infection), usually Staphylococci, resulting in painful, purulent and itching lesions,,. Similarly, systemic features, such as fever or organomegaly, are not characteristics of CL, unless a secondary bacterial infection is present,,. Therefore, findings of purulent discharge from lesions, which are usually attributed to bacterial infection, do not exclude CL diagnosis. Cutaneous leishmaniasis lesions do not respond to local or systemic antimicrobial and steroidal treatment.
Typically, CL diagnosis is delayed for several months following the sandfly’s bite,. As sandflies are mostly active during the summer months in the Old World, the lesions appear typically during the winter months,,. The sandfly’s bite does not hurt, thus when the lesion appears most of the patients do not remember having been bitten,.
In the past few years there has been a continuous outbreak of CL by L. major in southern Israel, especially in the northwest region, possibly caused by human invasion into the zoonotic endemic cycle. The Parasitology Laboratory of the Soroka University Medical Center (SUMC), Israel, where lesion biopsies from suspected-CL cases are evaluated, serves all southern Israel residents, including military bases in the area. Following the outbreak in recent years, the parasitological laboratory at the SUMC switched from microscopy-based diagnosis to PCR diagnosis.
This study aimed to assess the demographic and clinical characteristics of CL-suspected episodes/cases, in an attempt to differentiate confirmed CL and non-CL episodes.
| Material & Methods|| |
Study setting and design
This is a retrospective, case-control study, conducted by the Pediatric Infectious Disease Unit (PIDU) at the SUMC, Israel. This report is based on the CL data obtained from the SUMC’s computerized database, spanning over a three-year period, from 2013 through 2016. All the cases of suspected-CL with skin lesions (irrespective of age), who were treated at the SUMC during the study period and whose lesions were sampled and tested for Leishmania by PCR at the SUMC’s Parasitology Laboratory were included in the study. Only episodes with at least one recorded visit at the emergency room, outpatient clinics or hospitalization wards of the SUMC were included, to allow comprehensive evaluation of clinical data.
Cutaneous leishmaniasis was defined as a skin lesion with a positive PCR diagnosis of any Leishmania species (L. major, L. tropica, L. donovani/infantum), with or without any microscopic diagnosis (amastigotes or pro-mastigotes in culture). Empiric CL treatment was defined as either systemic treatment with intravenous liposomal amphotericin B (AmBisome®—Gilead Sciences Ltd., Ida Business & Technology Park, Carrigtohill, Co., Cork, Ireland) or sodium stibogluconate [Pentostam-Glaxo Operations (UK) Limited, Barnard Castle, UK] or local/topical treatment with intra-lesional pentostam, paromomycin (as sulfate) 15% and methylbenzethonium chloride 12% ointment (Leshcutan©—Teva Pharmaceutical Industries Ltd, Israel) or liposomal amphotericin B gel (Super Pharm Professional, Israel).
Demographic and clinical data were retrospectively obtained. Demographic data included seasonality according to diagnosis date (September through February as the cold months and March through August as the warm months), age, gender and ethnicity. Clinical features of the lesions included ulceration, itching, pain and secretion based on the information available in the medical files. Additionally, treatment data were also recorded.
The smears were taken under sterile condition using a disposable scalpel blade (No. 11). A small incision was made in the margins of the lesion with the blade. The blade was scraped along cut edge to pick up skin tissue, which was then smeared on a clean glass microscope slide for staining (later on) with Giemsa stain.
For culturing the parasites, the skin lesion aspirates were inoculated into a culture tube containing Rosewell Park Memorial Institute (RPMI) medium, supplemented with 30% fetal calf serum and then incubated at 25–28 °C for up to three weeks and was checked microscopically for promastigotes.
Multiplex real-time PCR (multiplex qPCR) analyses were performed, as previously described. A swab applicator was taken from the edge of the wound, placed into a tube containing transfer media (FLOQswabs, Copan, Murrieta, CA, US), and nucleic acid extraction was performed using Nimbus IVD system with Seegene Universal Cartridge kit, according to the manufacturer’s instructions. Following extraction, the samples were tested in parallel, in a test tube for Leishmania species, L. major and L. tropica. Amplification was carried out using the DNA qRT-PCR enzyme (Light Cycler 480 Probes Master; Roche Diagnostics, Mannheim, Germany). Primers, common to all Leishmania species, for the internal transcribed spacer [ITS-1] (5.8S ribosomal RNA gene) conserved region were used. Samples that were positive for Leishmania species but negative for L. major and L. tropica were considered positive and were tested further by additional qPCR for L. braziliensis and L. infantum/ donovani.
Descriptive statistical analyses, univariate analyses and multivariate analyses were calculated using SPSS 24. Univariate analyses were performed using either two-tailed chi-square test or Student’s t-test or Mann-Whitney U test to assess the association between potential risk factors and confirmed or negative Leishmania PCR results (CL and non-CL, respectively). The threshold for statistical significance was p <0.05.
Logistic regression models were used to evaluate potential risk factors, co-variate and confounders for CL. Several variables that were statistically significant at the level of p < 0.05 in the univariate analyses were included in multivariate logistic regression models. Additionally, logistic regression models were used to evaluate factors associated with administration of CL empiric treatment.
The study was approved by the SUMC’s Ethics Committee (Approval No. 0054-14-SOR; dated February 2, 2014).
| Results|| |
During the study period, 352 suspected CL episodes were identified. After exclusion of missing data, 324 episodes were eligible cases [Figure 1]. Out of the 324 suspected CL cases, 158 (48.8%) were confirmed as CL and 166 (51.2%) were negative for CL (non-CL). Of the confirmed CL episodes, 152 (96.2%) were caused by L. major, 3 (1.9%) by L. tropica, 2 (1.3%) by L. braziliensis, and 1 (0.6%) by L. infantum. In the non-CL group, bacterial infections, allergy, insect bites and unknown etiology comprised 43.4, 9.6, 9 and 24.1% of all the non-CL episodes, respectively.
|Figure 1: Study population of cutaneous leishmaniasis suspected cases in southern Israel from 2013-2016. The percentages indicated refer to the overall number of suspected cases (n = 324); *Others included insect bite and unknown etiology.|
Click here to view
Demographic and clinical characteristics of suspected- CL episodes
Overall, mean age ± SD was 33.4 ± 23.8 yr, with most episodes recorded in adults aging >18 yr (73.1%), males (59.3%) and patients of Jewish ethnicity (79.6%). The majority (60.8%) of episodes were diagnosed in the autumn and the winter months (September through February) [Table 1]. Analysis of residence distribution showed that 35.2% were from the Northwest Negev; 49.7% from other parts of the Negev region; and 12.6% from non- Negev regions. The lesions were primarily observed in the limbs (73.1%), followed by face (28.8%) and trunk (10.8%). Number of lesions per episode ranged from 1 to 26, with a mean of 2.9 ± 3.8 lesions. Overall, 52.5% of episodes were with a single lesion. The size of the lesions ranged from 0.2 to 15 cm (mean size 2.7 ± 2.4 cm). Ulcer appearance, purulent discharge, painful and itchy lesions were recorded in 55.2, 18.2, 47, and 42.9% of all the episodes, respectively, while systemic fever was recorded in only 4.7% of all the episodes. The majority (62.5%) of CL-suspected episodes were diagnosed within <1-month of duration.
|Table 1: Demographic and clinical characteristics of cutaneous leishmaniasis (CL) suspected cases in southern Israel, 2013-2016|
Click here to view
Initial treatment regimens
Initial empiric treatment for CL was recorded in 39.8% (115 of 289 episodes for which treatment data were available) of episodes. Cutaneous leishmaniasis treatment regimens included use of Leshcutan© ointment (n = 87, 75.7%), intra-lesional pentostam (n=23, 20%), liposomal amphotericin gel (n = 4, 3.5%) and intra-venous pentostam (n = 1, 0.9%). Other, non-CL (n = 174) targeted, treatment regimens included systemic (oral or intra-ve- nous) antibiotics (n = 67, 38.5%) and topical antibiotics (n = 56, 32.1%). Topical steroids were given in 60 (34.5%) episodes (20 CL and 40 non-CL episodes).
Comparison of demographic and clinical characteristics between CL and non-CL episodes
In univariate analysis, mean age, region of residence, ethnicity, distribution of lesions and fever rates were similar when comparing CL and non-CL episodes [Table 2]. Cutaneous leishmaniasis (compared with non- CL) was associated with male gender (66.5 in CL vs 52.4% in non-CL episodes), multiple lesions (60 vs 37.5%), ulcer (66.9 vs 43.5%), September through February seasonality (72.8 vs 49.4%), >1-month duration until diagnosis (74.2 vs 50.8%) and empiric CL treatment (63.4 vs 16.0%). In contrast, itching (25.7 vs 52.4%) was associated with non-CL.
|Table 2: Comparison of demographic and clinical characteristics between CL and non-CL episodes, and episodes with and without CL empiric treatment in southern Israel, 2013-2016|
Click here to view
Demographic and clinical characteristics associated with empiric CL treatment administration
In univariate analysis, mean age, gender, region of residence, distribution of lesions, itching, painful and purulent lesion rates were similar when comparing empiric CL treatment with other treatment episodes [Table 2]. Empiric CL treatment administration was associated with Jewish ethnicity (88.7% in empiric CL vs 72.4% in other treatment), September through February seasonality (68.7 vs 56.3%), multiple lesions (58.7 vs 40.8%), ulcer (65.9 vs 52.1%) and >1-month duration until diagnosis (77.9 vs 51.4 %). In contrast, fever (0.9 vs 8%) was associated with other, non-CL, treatment regimens.
CL diagnosis: The depending variables were—Duration of >1-month until diagnosis, empiric CL treatment, gender, seasonality and ulcer. Empiric CL treatment (odds ratio, OR = 5.144), ulcer (OR = 2.459) and seasonality (OR=2.144) were significantly associated with CL, while for male gender (OR = 1.908), and duration of >1-month until diagnosis (OR = 1.924), a trend (p = 0.06, both) towards association with CL was observed [Table 3].
|Table 3: Multivariate analyses [Odds ratio (OR)] for CL vs non-CL episodes; and CL empiric treatment vs other non-CL treatment regimens|
Click here to view
Empiric CL treatment: The depending variables were: Jewish ethnicity, duration of <1-month until diagnosis, seasonality, ulcer and fever. Jewish ethnicity (OR = 2.446) and duration of >1-month until diagnosis (OR = 3.304), were significantly associated with empiric CL treatment [Table 3].
| Discussion|| |
In the current study, CL diagnosis was confirmed only in approximately half of all the suspected CL episodes sent for PCR diagnosis. Factors associated with CL diagnosis included male gender, seasonality, ulcer appearance of the lesions and >1-month duration until diagnosis. The rate of empiric treatment for CL was higher in CL-confirmed episodes (63%) than in non-CL episodes (16%) and was associated with Jewish ethnicity and >1-month duration until diagnosis.
The relatively low (<50%) positivity rate of CL diagnosis emphasizes that CL diagnosis should be laboratory confirmed, preferably by molecular diagnosis, and not rely solely on the basis of clinical presentation,.
Indeed, the appearance of skin lesions following CL infection varies depending on the species as well as host genetic factors and extent of immune activation. Molecular diagnosis has several advantages, not only over clinical diagnosis, but also over microscopy-based diagnosis, as it is less observer dependent, highly sensitive, and strain-specific, with the latter being a crucial determinant in the choice of treatment. Cutaneous leishmaniasis differential diagnosis is broad and includes, among others, bacterial (mainly staphylococcal and streptococcal) and fungal infections, hypersensitivity reactions, and cutaneous malignancy. Differentiating CL from those diagnoses is difficult. In the present study, we have focused on previously described demographic and clinical features that may help differentiate CL from these entities, including its chronicity, painlessness, number of lesions, the presence of purulent discharge, ulcerations with well-defined borders, and fever.
Although clinical diagnosis was not reliable, several factors were associated with CL diagnosis which may serve the treating physicians in the diagnostic process of CL suspected episodes. Male gender association with CL diagnosis is possibly related to the zoonotic nature of the disease. Indeed, male gender predominance has been reported in other zoonoses,. Seasonality, and specifically presentation of skin lesions in the colder months of the autumn and the winter, is an important factor in this regard. As CL is usually acquired in the summer months (the peak activity period of female sandflies), lesions typically appear only after 3–4 months. This is probably derived from both CL pathogenesis (formation of a marked lesion over a period of several months) and the delay in diagnosis that may be related to delay in seeking medical advice or to the wrong treatment regimens initially prescribed by physicians for assumed bacterial skin infections. Indeed, in the current study, prolonged episode durations (until the confirmation of diagnosis and treatment provided) were observed in CL episodes.
The appearance of ulcers is another important factor which assists in CL diagnosis, especially in case of L. major infections. Leishmania major lesions are typically described as “wet ulcer” lesions and as such, this specific appearance, combined with other clinical and demographic factors, should raise a high level of suspicion for CL in endemic regions. Nevertheless, it should be noted that ulcers are not typical in other CL cases caused by, non-L. major, strains. Furthermore, of all the clinical characteristics evaluated in the present study, pathogno- monic characteristics were not recognized, albeit some unique manifestations (e.g. “Volcano shape” appearance) and could not be evaluated due to missing data.
Several clinical risk factors could not be assessed by multivariate analysis in the present study, and were only analyzed in univariate analysis, due to relatively high rates of missing data. Nevertheless, it is of crucial importance for treating physicians to realize that CL episodes may be complicated with bacterial skin infections, and that the characteristics like single lesion, purulent discharge, fever and itching merely suggest other diagnosis and do not rule out CL diagnosis. The substantial disparity between rates of confirmed diagnosis (~50%) and rates of administration of specific empiric CL treatment (63% in CL- confirmed episodes vs 16% in non-CL episodes) is striking. We offer two explanations for this disparity. First, treating physicians may be willing to conduct diagnostic assay and reluctant to start empiric treatment, which may be associated with side-effects. Second, administration of treatment may be related to other, non-clinical factors, including drugs availability, financial considerations and treating physicians’ disease epidemiology perception. In this regard, the higher rate of empiric CL treatment recorded in the Jewish population, compared with the Bedouin population (treated more commonly with non-CL treatment regimens), may be related to these factors.
The main limitation of this study is its retrospective nature. As a consequence, some data regarding risk factors and clinical parameters are missing. Moreover, clinical signs and symptoms are prone to observer and interviewer bias. Additionally, we recognize the possibility of selection bias, as all episodes were collected from the in-hospital setting, and may not accurately reflect the rates of “accurate” diagnosis in the community setting. The strength of the study includes the relatively large number of cases and the fact that all CL diagnoses were made by PCR.
| Conclusion|| |
The results of the study suggest that, CL diagnosis should not rely on clinical appearance, which is often misleading. Laboratory confirmation of CL, preferably by PCR, should be sought in all the suspected cases to allow suitable treatment. Seasonality, ulcer appearance, number of lesions and gender may aid in correct identification and treatment of CL.
Conflict of interest
The authors declare that they do not have any conflict of interest.
| References|| |
Anis E, Leventhal A, Elkana Y, Wilamowski A, Pener H. Cutaneous leishmaniasis in Israel in the era of changing environment. Public Health Rev
Hepburn NC. Cutaneous leishmaniasis: An overview. J Postgrad Med
Murray HW, Berman JD, Davies CR, Saravia NG. Advances in leishmaniasis. Lancet
2005; 366(9496): 1561-77.
Pace D. Leishmaniasis. J Infect
(Suppl 1): S10-8.
Magill AJ. Leishmania
species: Visceral (Kala-azar), cutaneous, and mucosal leishmaniasis. In: Mandell, Douglas, Bennett’s, editors. Principles and practices of infectious diseases
. VII edn. Philadelphia: Churchill Linvingstone Elsevier 2010; 2:
Arce A, Estirado A, Ordobas M, Sevilla S, Garcia N, Moratilla L, et al
. Re-emergence of leishmaniasis in Spain: Community outbreak in Madrid 2009 to 2012. Euro Surveill
2013; 18(30): 20546.
Jacobson RL, Eisenberger CL, Svobodova M, Baneth G, Sz- tern J, Carvalho J, et al
. Outbreak of cutaneous leishmaniasis in northern Israel. J Infect Dis
Nateghi RM, Saghafipour A, Vesali E. A newly emerged cutaneous leishmaniasis focus in central Iran. Int J Infect Dis
2013; 17(12): e1198-206.
Sanchez JL, Diniega BM, Small JW, Miller RN, Andujar JM, Weina PJ, et al
. Epidemiologic investigation of an outbreak of cutaneous leishmaniasis in a defined geographic focus of transmission. Am J Trop Med Hyg
Biton A, Namir B, Mosovich D, Vardi D. Cutaneous leishmaniasis: Incidental outbreak in a new endemic area? Harefuah
1997; 132(6): 385-7.
Gandacu D, Glazer Y, Anis E, Karakis I, Warshavsky B, Slater P, et al
. Resurgence of cutaneous leishmaniasis in Israel, 2001-2012. Emerg Infect Dis
2014; 20(10): 1605-11.
Klaus S, Axelrod O, Jonas F, Frankenburg S. Changing patterns of cutaneous leishmaniasis in Israel and neighbouring territories. Trans R Soc Trop Med Hyg
1994; 88(6): 649-50.
Gholamrezaei M, Mohebali M, Hanafi-Bojd AA, Sedaghat MM, Shirzadi MR. Ecological Niche modeling of main reservoir hosts of zoonotic cutaneous leishmaniasis in Iran. Acta Tropica
Doudi M, Setorki M, Narimani M. Bacterial superinfection in zoonotic cutaneous leishmaniasis. Med Sci Monit
2012; 18(9): BR356-61.
Edrissian GH, Mohammadi M, Kanani A, Afshar A, Hafezi R, Ghorbani M, et al
. Bacterial infections in suspected cutaneous leishmaniasis lesions. Bull World Health Organ
1990; 68(4): 473-7.
Ziaei H, Sadeghian G, Hejazi SH. Distribution frequency of pathogenic bacteria isolated from cutaneus leishmaniasis lesions. Korean J Parasitol
2008; 46(3): 191-3.
Monge-Maillo B, Lopez-Velez R. Therapeutic options for Old World cutaneous leishmaniasis and New World cutaneous and mucocutaneous leishmaniasis. Drugs
2013; 73(17): 1889-920.
Ben-Shimol S, Sagi O, Codish S, Novack V, Barrett C, Frucht- man Y, et al
. Dramatic increase in laboratory-diagnosed human cutaneous leishmaniasis cases in southern Israel, 2007-2013. Infect Dis (Lond)
2015; 47(3): 161-7.
Orshan L, Szekely D, Khalfa Z, Bitton S. Distribution and sea- sonality of Phlebotomus sandflies in cutaneous leishmaniasis foci, Judean Desert, Israel. J Med Entomol
2010; 47(3): 319-28.
Sagi O, Berkowitz A, Codish S, Novack V, Rashti A, Akad F, et al
. Sensitive molecular diagnostics for cutaneous leishmaniasis. Open Forum Infect Dis
2017; 4(2): ofx037.
Blonski KM, Blodorn-Schlicht N, Falk TM, Faye RS, Clausen OP. Increased detection of cutaneous leishmaniasis in Norway by use of polymerase chain reaction. APMIS
2012; 120(7): 591-6.
Lavergne RA, Iriart X, Martin-Blondel G, Chauvin P, Menard S, Fillaux J, et al
. Contribution of molecular diagnosis to the management of cutaneous leishmaniasis in travellers. Clin Microbiol Infect
2014; 20(8): O528-30.
Kubba R, Al-Gindan Y, el-Hassan AM, Omer AH. Clinical diagnosis of cutaneous leishmaniasis (oriental sore). J Am Acad Dermatol
1987; 16(6): 1183-9.
Theiler RN, Rasmussen SA, Treadwell TA, Jamieson DJ. Emerging and zoonotic infections in women. Infect Dis Clin North Am
2008; 22(4): 755-72.
[Table 1], [Table 2], [Table 3]