Leptospirosis: Epidemiological and Clinical Characteristics at the National Reference Hospital in Panama

ABSTRACT.

Leptospirosis represents a public health problem in Panama, with an incidence rate of 1 in 100,000 inhabitants in 2014. Despite active surveillance and reports of outbreaks in the news, publications about human leptospirosis in Panama are scarce. The objective of this study was to describe the epidemiological and clinical features of leptospirosis in a cohort of patients admitted to the national reference hospital from January 2013 to December 2018. A total of 188 patients with suspected leptospirosis were identified, but only 56.9% (107 of 188) of the medical records could be retrieved. Microagglutination assays were completed in 45% (48 of 107) of the patients, confirming leptospirosis in 29.2% (14 of 48) of the patients. The most prevalent serogroup identified was Leptospira interrogans icterohemorrhagiae (4 of 14, 28.6%). The majority of patients with confirmed disease were middle-aged (36.4 ± 15.7 years), male (11 of 14, 78.6%), and symptomatic for 6.8 ± 0.7 days before admission. The predominant clinical presentation was fever (13 of 14, 92.9%), abdominal pain (7 of 14, 50%), and jaundice (8 of 14, 57.1%). Respiratory failure (8 of 14, 57.1%), elevated creatinine levels on admission (8 of 14, 57.1%), transfusion of blood-derived products (6 of 14, 42.9%), and required use of vasopressors (4 of 14, 28.6%) were common complications. Mortality was 28.6% (4 of 14). Empiric antibiotic therapy was initiated in almost all patients (10 of 12, 83.3%), and was appropriate in 90% (9 of 10) of them. Our study highlights the high prevalence of severe disease and reveals the diagnostic challenges concealing the true burden of leptospirosis in Panama. However, the small number of confirmed patients limits the generalization of these findings.

INTRODUCTION

Leptospirosis is a globally distributed zoonotic disease caused by pathogenic spirochetes from the genus Leptospira.¹ Leptospirosis is seen in approximately 1.03 million cases and causes 58,900 deaths globally each year, approaching or even exceeding the impact of other hemorrhagic fevers.² The disease is acquired through direct contact with urine or tissues of infected animals via skin abrasions or mucous membranes. Changes in the epidemiology of leptospirosis are expected as a consequence of climate change and its impact on rodent populations and the incidence of extreme weather events.³

Panama is a Central American country with a population of 4.2 million,⁴ located just north of the Equator. It has a tropical climate influenced by the El Niño–Southern Oscillation/El Niño–La Niña events, tropical cyclones, and sea-surface temperatures. It has a long rainy season that lasts from May to November,⁵ and has achieved historical records of maximum accumulated precipitation in the past decade.⁶ It is calculated that 40% of Panama’s population lives in the countryside, earning a modest living through small-scale farming,⁷ whereas a quarter of its urban population inhabits slums.⁸ An estimated 77% of the national population had access to at least basic sanitation services in 2015.⁹

The first reports of leptospirosis in Panama date back to the 1950s, after researchers from the Walter Reed Military Institute began serological testing in humans and animals, as well as an investigation of an outbreak among American soldiers who had been training in proximity to the Panama Canal.^10–13 From 1999 to 2016, the GeoSentinel Network diagnosed 167 travelers with leptospirosis, of whom five had acquired the infection in Panama.¹⁴ The small number of patients may be explained by the lack of freshwater activities for international travelers in the country. On the other hand, a study in rural areas of Panama found a 90.6% (78 of 86) positivity rate for leptospirosis among serum samples of asymptomatic participants with epidemiological risk for the disease, with Leptospira borgpetersenii Tarassovi being the most frequent serogroup (53 of 86, 62%).¹⁵

Despite an active surveillance system and reports of outbreaks in the news, publications today about human leptospirosis in Panama are scarce. The objective of this study was to describe the epidemiological and clinical features of leptospirosis in a cohort of patients admitted to the national reference hospital in Panama.

MATERIALS AND METHODS

Population and study area.

Panama’s national reference hospital is Hospital Santo Tomas, located in Panama City. It is in close proximity to the Instituto Conmemorativo Gorgas de Estudios de la Salud, a medical research institution dedicated to investigating tropical diseases and promoting preventive medicine. The investigation of leptospirosis cases in Panama is centralized in the laboratory of this institution.

A cohort of patients admitted to Hospital Santo Tomas with suspected leptospirosis from January 2013 to December 2018 was reviewed retrospectively. The national guidelines on epidemiology published by the Ministry of Health of Panama defines suspected leptospirosis based on symptoms, signs, and exposure.¹⁶ However, for the purposes of our study, we defined cases of suspected leptospirosis as those with requested serological testing for leptospirosis upon admission, regardless of whether the sample was obtained. Demographic data, clinical signs and symptoms, laboratory results, and treatment regimens were collected from medical charts. Possible exposures to leptospirosis were ascertained from the leptospirosis laboratory request form, where physicians were asked this information in an open-ended question. Only those patients whose medical records were available for review in the Hospital Santo Tomas archives were included in the analysis.

Laboratory methods.

The microscopic agglutination test (MAT), the current gold standard for the diagnosis of leptospirosis, was used to confirm or rule out the disease, and to identify the presumptive Leptospira serogroup. This test requires obtaining paired serum or blood samples—the first one taken 10 to 12 days after the onset of symptoms and the second one taken at least 10 days after the first. A minimum 4-fold increase in MAT titers was used to confirm leptospirosis (the panel of Leptospira strains used in the MAT is presented in Table 1). In addition, the reference laboratory runs confirmatory molecular tests (polymerase chain reaction) in cultures plates with suspicious growth, but this information was not available in the medical records.

Table 1.

Panel of Leptospira strains used in the microscopic agglutination test

Code	Species	Serogroup	Serovar	Strain
CLEP 0001	L. interrogans	Icterohaemorrhagiae	Icterohaemorrhagiae	RGA
CLEP 0002	L. interrogans	Icterohaemorrhagiae	Copenhageni	M 20
CLEP 0003	L. interrogans	Canicola	Canicola	Hond Utrecht IV
CLEP 0004	L. kirshneri	Gripothyphosa	Gripothyphosa	Moska V
CLEP 0005	L. interrogans	Pomona	Pomona	Pomona
CLEP 0006	L. interrogans	Australis	Australis	Ballico
CLEP 0007	L. interrogans	Bataviae	Bataviae	Swart
CLEP 0008	L. borgpetersenii	Ballum	Castellonis	Castellon 3
CLEP 0009	L. kirshneri	Cynopteri	Cynopteri	3522 C
CLEP 0010	L. borgpetersenii	Javanica	Javanica	Veldrat Batavia 46
CLEP 0011	L. noguchii	Panama	Panama	CZ 214K
CLEP 0012	L. interrogans	Pyrogenes	Pyrogenes	Salinem
CLEP 0013	L. interrogans	Sejroe	Hardjo	Hardjoprajitno
CLEP 0014	L. borgpetersenii	Sejroe	Sejroe	M 84
CLEP 0015	L. biflexa	Semaranga	Patoc	Patoc I
CLEP 0016	L. borgpetersenii	Tarassovi	Tarassovi	Perepeletsin
CLEP 0017	L. interrogans	Autumnalis	Autumnalis	Akiyami A
CLEP 0018	L. interrogans	Hebdomadis	Hebdomadis	Hebdomadis
CLEP 0019	L. interrogans	Sjeroe	Wolffi	3705
CLEP 0021	L. biflexa	Andamana	Andamana	CH11

Data analysis.

Data were analyzed using Stata IC (version 14.2; StataCorp LLC, College Station, TX).¹⁷ χ2 tests and Fisher’s exact tests were used to analyze the relationship between categorical values, and one-way analysis of variance was used to detect differences in the means. Missing data were deleted from the analysis. P < 0.05 was used to indicate statistical significance. The protocol was reviewed and approved by the local ethics committee from Hospital Santo Tomas (1263-2016/CIDI/HST) and the institutional review board of the University of Alabama at Birmingham (IRB-300004245). Informed consent from participants was not obtained, because it was not required by the review boards for this type of research.

RESULTS

A total of 188 patients had Leptospira testing requested during the study period. Medical records of 107 patients (107 of 188, 56.9%) were obtained, of which only 98 (98 of 107, 91.6%) had samples collected. The analysis of paired samples ruled out leptospirosis in 34 of 48 patients (70.8%) and confirmed it in 14 of 48 patients (29.2%). Fifty-nine of 107 patients (55.1%) did not have complete diagnostic testing (Figure 1). The time between the onset of symptoms and the first sample among patients with at least one sample collected (82 of 107) was an average of 12.7 ± 6.9 days. Other causes of infectious fever were investigated in some of the 107 patients, including dengue (5 of 26, 19.2%), syphilis (3 of 57, 5.3%), hantavirus (0 of 4, 0%), and HIV (4 of 102, 3.9%). Our description of the findings focuses on the features of the patients with confirmed leptospirosis.

Figure 1.

Flow chart legend. ¹Suspected case: clinical suspicion and diagnostic test for Leptospira requested. ²Ruled-out case: clinical suspicion and negative microscopic agglutination test result. ³Confirmed case: clinical suspicion and positive microscopic agglutination test result.

The mean age of patients was 36.4 ± 15.7 years. The majority were male (11 of 14, 78.6%), and worked in construction (3 of 14, 21.4%) or were unemployed (3 of 14, 21.4%). The most common risk factors were reported use of latrines (3 of 14, 21.4%) and contact with animals (4 of 14, 28.6%) (Table 2). Half the patients came from Panama City (8 of 14, 57%), and almost all cases (12 of 14, 85.7%) occurred during the rainy season (Figure 2). The presumptive Leptospira serogroup most commonly identified was Leptospira interrogans icterohemorrhagiae (4 of 14, 28.6%).

Table 2.

Demographic data among confirmed patients, ruled-out patients, and patients with incomplete testing

Demographic data	Confirmed patients (n = 14)	Ruled-out patients (n = 34)	P value
Age, mean ± SD	36.4 ± 15.7	42.8 ± 15.9	0.2
Male	11 (78.6)	26 (76.5)	0.8
Profession,* n (%)
Unemployed	3 (21.4)	9 (26.5)	1
Construction	3 (21.4)	3 (8.8)	0.3
Farmer	2 (14.3)	3 (8.8)	0.6
Risk factors,† n (%)
Latrines	3 (21.4)	3 (8.8)	0.3
Contact with animals	4 (28.6)	6 (17.6)	0.4

^*Includes the three most common professions.

^†Includes the two most common risk factors.

Figure 2.

Number of patients per month of admission.

On average, confirmed patients were symptomatic for 6.8 ± 0.7 days before hospital admission, significantly less than the average duration of illness of ruled-out patients (15.6 ± 2.6, P = 0.02). The most prevalent signs and symptoms were fever (13 of 14, 92.9%), chills (6 of 14, 42.9%), nausea and vomiting (6 of 14, 42.9%), abdominal pain (7 of 14, 50.0%), jaundice (8 of 14, 57.1%), and myalgias (4 of 14, 28.6%). In addition, three patients presented with altered mental state (3 of 14, 21.4%), two presented with gastrointestinal bleeding (2 of 14, 14.3%), one with painful hepatomegaly (1 of 14, 7.1%) that was not investigated through ultrasound, and one with a conjunctival injection (1 of 14, 7.1%) (Table 3). None of the patients presented with hemoptysis. Common laboratory abnormalities included anemia (10 of 14, 71.4%), leukocytosis (10 of 14, 71.4%), thrombocytopenia (9 of 14, 64.3%), and elevation of creatinine (8 of 14, 57.1%), alanine aminotransferase (12 of 14, 85.7%), aspartate aminotransferase (11 of 14, 78.5%), and total bilirubin (11 of 14, 78.5%) levels.

Table 3.

Clinical variables among confirmed patients, and ruled-out patients

Clinical variables	Confirmed patients (n = 14)	Ruled-out patients (n = 34)	P value
Duration of illness, days; mean ± SD (range)	6.8 ± 0.7 (1–11)	15.6 ± 2.6 (0–61)	0.02
Hospital stay, days; mean ± SD (range)	26.7 ± 34.9 (0–142)	28.2 ± 33.43 (0–182)	0.89
Fever, n (%)	13 (92.9)	27 (79.4)	0.2
Nausea/vomiting, n (%)	6 (42.9)	20 (58.9)	0.3
Chills, n (%)	6 (42.9)	11 (32.4)	0.4
Abdominal pain, n (%)	7 (50.0)	18 (52.9)	0.8
Jaundice,* n (%)	8 (57.1)	17 (50.0)	0.6
Headache, n (%)	4 (28.6)	9 (26.5)	1
Myalgia, n (%)	4 (28.6)	6 (17.6)	0.4
Hemorrhage, n (%)	2 (14.3)	5 (14.7)	1
Photophobia, n (%)	2 (14.3)	1 (2.9)	0.2
Rash, n (%)	1 (7.1)	2 (5.9)	1
Conjunctival injection, n (%)	1 (7.1)	2 (5.9)	1
Altered mental state, n (%)	3 (21.4)	4 (11.8)	0.4
Hepatomegaly, n (%)	1 (7.1)	10 (29.4)	0.1
Pulmonary infiltrate on chest X-ray, n (%)	3 (21.4)	8 (23.5)	1
Cardiomegaly on chest X-ray, n (%)	2 (14.3)	3 (8.8)	0.6
Admission to intensive care unit, n (%)	7 (50.0)	11 (32.4)	0.2
Intubation, n (%)	8 (57.1)	12 (35.3)	0.2
Transfusion of blood products, n (%)	6 (42.9)	14 (41.2)	0.9
Use of vasopressors, n (%)	4 (28.6)	9 (26.5)	0.5
Death, n (%)	4 (28.6)	9 (26.5)	0.5

^*Mean total bilirubin value at admission was 7.1 ± 2.4 mg/dL among confirmed patients and 7.8 ± 1.8 mg/dL among ruled-out patients.

Common complications included respiratory failure requiring intubation (8 of 14, 57%). Transfusion of blood-derived products (6 of 14, 42.9%) and use of vasopressors (4 of 14, 28.6%) were common events among confirmed patients. Chest X-rays from patients requiring mechanical ventilation showed diffuse bilateral pulmonary infiltrates (3 of 8, 37.5%), unilateral pulmonary infiltrates (1 of 8, 12.5%), and cardiomegaly (1 of 8, 12.5%), whereas the rest had normal findings (4 of 8, 50%). Half the patients (7 of 14, 50.0%) required admission to the intensive care unit (ICU). Mortality was 28.6% (4 of 14), with only one patient having been admitted to the ICU. Although the percentage of patients admitted to the ICU and those requiring intubation was greater among confirmed cases than ruled-out cases, the difference lacked statistical significance. The average number of days in the hospital was 26.7 ± 34.9 days (Table 3).

All patients had both admitting and discharge diagnoses reported in their medical charts. Eleven of 14 patients (78.6%,) received a diagnosis of leptospirosis upon both admission and discharge. One patient without a discharge diagnosis of leptospirosis had not received proper antibiotic treatment and died. An admission diagnosis of dengue was more common among confirmed cases than among ruled-out cases (P = 0.005). Other admission diagnoses included sepsis (2 of 14, 14.3%), pneumonia (3 of 14, 21.4%), digestive bleeding (3 of 14, 21.4%), malaria (1 of 14, 7.1%), and hantavirus infection (1 of 14, 7.1%) (Table 4).

Table 4.

Most common attributed diagnosis among confirmed patients, ruled-out patients, and patients with incomplete testing at the time of admission and discharge

Diagnosis	Confirmed patients (n = 14)	Ruled-out patients (n = 34)	P value
On admission, n (%)
Leptospirosis	11 (78.6)	18 (52.9)	0.09
Sepsis	2 (14.3)	10 (29.4)	0.4
Dengue	4 (28.6)	0 (0.0)	0.005
Pneumonia	3 (21.4)	5 (14.7)	0.7
Gastrointestinal bleeding	3 (21.4)	0 (0.0)	0.02
Malaria	1 (7.1)	0 (0.0)	0.3
Hantavirus	1 (7.1)	0 (0.0)	0.3
At discharge, n (%)
Leptospirosis	11 (78.6)	4 (11.8)	0.001
Sepsis	1 (7.1)	8 (23.5)	0.2
Multiorgan failure	1 (7.1)	5 (14.7)	0.6
Dengue	1 (7.1)	0 (0.0)	0.3
Pneumonia	2 (14.3)	8 (23.5)	0.7
Acute kidney failure	2 (14.3)	2 (5.9)	0.6

The majority of patients (10 of 12, 83.3%) received antibiotic therapy, which was initiated within the first day of admission and lasted between 5 and 10 days (Table 5). This therapy was considered appropriate in 9 of 10 patients (90%), and included antibiotics such as penicillin, cefotaxime, ceftriaxone, doxycycline, or azithromycin.^16,18–20 Only one patient received inappropriate antibiotics—a combination of ciprofloxacin and metronidazole. Unfortunately, data on antibiotic use were missing in 2 of 14 patients (14.3%).

Table 5.

Antibiotic therapy among confirmed patients, and ruled-out patients

Therapy	Confirmed patients (n = 14)	Ruled-out patients (n = 34)	P value
Received antibiotics	10/14 (71.4%)	29/34 (85.3%)	0.3
Appropriate antibiotics*	9/10 (90.0%)	26/29 (89.7%)	0.9
Days until treatment, mean ± SD (range)	0.07 ± 0.27 (0–1)	3.15 ± 10.89 (0–62)	0.2
Duration of treatment, mean ± SD (range)	9.1 ± 3.4 (5–15)	8.14 ± 5.06 (1–23)	0.8

^*Penicillin, ceftriaxone, cefotaxime, doxycycline, and azithromycin.

DISCUSSION

Leptospirosis is a disease of mandatory notification in Panama. Epidemiological surveillance between 2010 and 2014 revealed a greater prevalence of cases during the rainy season, in urban areas, (64%) and among men (74%). It also revealed that the districts most affected by the disease were Panama and Chiriquí.²¹

Following a similar trend, most of the cases in our cohort occurred in middle-aged men. This gender difference is seen in many other case series and has been attributed to exposure-related reporting bias, as affected athletic events have similar rates of illness among men and women.²² In our cohort, there was no predominant profession or risk factor associated with the disease, which could suggest a rather widespread presence of Leptospira in the environment. In addition, most of the cases came from the Panama District, where the hospital is located, and occurred during the rainy season. The latter finding is of particular importance because different prediction models estimate large increases in Panama’s mean annual precipitation—as much as 80% by 2080⁶—which could lead to changes in the epidemiological trends of leptospirosis.²³

The predominant clinical presentation was fever and gastrointestinal symptoms, including jaundice. Conjunctival suffusion (1 of 14, 7.1%) and calf tenderness (0 of 14, 0%)—two characteristic physical findings of leptospirosis—were uncommon. Hemoptysis was not reported. Laboratory and radiographic findings were in accordance with previously described profiles in patients with leptospirosis.¹ Overall, ruled-out cases had similar characteristics to confirmed cases, suggesting either appropriate screening or a lack of statistical power as a result of the small sample size.

There was also a high prevalence of severe disease and mortality (28.6%) among patients with confirmed leptospirosis, despite the young age of the cohort. In general, severe cases of leptospirosis tend to be confirmed through diagnostic tests and reported more often than non-severe cases.²⁴ The small sample size and lack of standardized ascertainment disallowed the identification of risk factors, but the predominance of the serogroup Leptospira interrogans icterohemorrhagiae was highlighted in our study, and has been found to be associated to severe illness in other case series.^25,26 Nevertheless, Leptospira serogroups were obtained through MAT, which can only provide a presumptive identification.

Clinical suspicion of leptospirosis did not necessarily translate into initiation of appropriate antibiotic treatment. However, there is controversy regarding the use of antibiotics for this disease.²⁷ Anecdotally, from the two patients with confirmed leptospirosis who did not receive antibiotic therapy, one died and one survived. Similarly, from the two confirmed patients with missing data on antibiotic therapy, one died and one survived. Last, the two patients with leptospirosis who died despite antibiotic therapy received penicillin and meropenem, respectively.

Although confirmed cases represent a significant portion of the total cases with complete diagnostic tests (14 of 48, 29.2%), the small absolute number is the main limitation of our study. This may have been a result of the lack of access to 48% (81 of 188) of the medical charts and an incomplete diagnostic workup in 63% of patients (68 of 107). Nevertheless, the small number of cases matches with the previously reported annual incidence of leptospirosis in Panama,^21,28 which is less than the incidence in the neighboring countries of Colombia and Costa Rica.²⁴ The disease is underreported because of a lack of awareness and overlapping clinical presentations with other febrile illnesses of the tropics.²⁹ However, the rate of simultaneous testing for dengue, yellow fever, malaria, chikungunya, Zika, and hantavirus was low. HIV was the exception, with 95.3% of patients screened. Although ordering a large battery of tests is common in high-income countries, physicians in low- and middle-income countries must use their clinical judgment to adhere to a strict budget and limited diagnostic capabilities. To this end, the development of a cost-effective, syndromic broad pathogen screening assay,³⁰ adapted to different regions of the world, would revolutionize the practice of tropical medicine.

Another reason that leptospirosis is underreported includes noticeable diagnostic challenges. Collecting paired samples is a cumbersome process in which the second sample is often not obtained because it is perceived unnecessary if the patient has already completed antibiotic treatment or an alternative diagnosis has been reached. In our study, only 78.6% (11 of 14) of confirmed cases received a discharge diagnosis of leptospirosis. This may have been a result of the release of final MAT results after the patient was discharged or deceased. Unfortunately, there is no record of the dates on which the final MAT results were reported to the physicians to confirm this hypothesis. The implementation of a tool to offer timely feedback between laboratory staff and physicians could strengthen the diagnostic process.

To the best of our knowledge, this is the first article to be published on epidemiological and clinical features of leptospirosis in hospitalized patients in Panama in the past few decades.^31,32 It is largely limited by the small absolute number of confirmed cases, but does highlight the high prevalence of severe disease. Most important, it reaffirms the diagnostic difficulties that conceal the true burden of leptospirosis. Because most patients presented within the first week of illness, the cost-effectiveness of polymerase chain reaction testing for leptospirosis should be considered in this setting. In addition, further evaluation of simultaneous testing or sequential testing for dengue and leptospirosis, and its cost-effectiveness, should be considered in the right clinical scenario. This approach could potentially increase the detection rate of leptospirosis as well co-infection and help physicians implement timely rapid and effective management.