Epidemiological and clinical factors associated with lethality from Human Visceral Leishmaniasis in Northeastern Brazil, 2007 to 2018.

Human Visceral Leishmaniasis (HVL) presents a subacute clinical evolution with systemic involvement, which can result in high case fatality, especially among untreated individuals or those with low socioeconomic status. This study aimed to identify epidemiological and clinical factors associated with HVL case fatality in the Ceara State, from 2007 to 2018. This is an analytical cross-sectional study. The bivariate analysis was performed by Stata 15.1 using Pearson's Chi-square or Fisher's exact test; and Poisson regression for age-controlled multivariate analysis. From 2007 to 2018, there were 4,863 new confirmed cases and 343 deaths from HVL (case fatality rate=7.05%). The risk factors associated with case fatalities were: age group (RR=8.69; 95%CI:3.56-21.20); black population (RR=2.21; 95%CI:1.45-3.35); jaundice symptoms (RR=1.72; 95%CI:1.38-2.14); edema (RR=2.62; 95%CI:2.10-3.26) and hemorrhagic phenomena (RR=1.63; 95%CI:1.26-2.10); and no prescription drug intake (RR=4.03; 95%CI:2.98-5.46). Treatment with pentavalent antimonial was a protective factor (RR=0.35; 95%CI:0.27-0.45). The number of deaths increased among the elderly, illiterate, urban residents, and black skin color individuals. The drugs pentavalent antimonial and amphotericin B showed an association with death, but were not considered causal factors. Treatment failure led to a high risk of death. In multivariate analysis, the risk factors for fatal cases were age group, black skin, symptoms of jaundice, edema and hemorrhagic phenomena; and failure to take the prescription drugs. Treatment with pentavalent antimonial was shown to be a protective factor. Knowing the factors associated with the fatality of VL-HIV cases may help to improve public policies, in order to refine the epidemiological surveillance program and, consequently, prevent deaths related to the disease in Ceara.

INTRODUCTION

Human Visceral Leishmaniasis (HVL) is a zoonotic disease caused by an intracellular protozoan, Leishmania infantum. In Brazil, the disease is transmitted by the vector Lutzomyia longipalpis .

HVL is part of the group of neglected tropical diseases and its importance in public health is directly related to a further expansion of its area of influence. It evolves clinically with systemic involvement, which can result in high mortality, especially among untreated individuals or those with low socioeconomic status .

In Brazil, HVL is endemic and distributed in its five regions. The number of municipalities that have initiated the notification of cases has been increasing, with expansion to areas historically free of this disease, such as the municipalities in Southern Brazil . From 2010 to 2018, Brazil registered 42,067 cases, which led to an average incidence of 1.92 cases per 100,000 inhabitants and 2,704 deaths .

Northeastern Brazil stands out in the number of cases and deaths. Between 2008 and 2017, a total of 19,895 cases and 1,270 deaths were reported in this Brazilian region, with a case fatality rate of 6.4%. In this period, there was a gradual increase in the number of deaths from HVL, which increased from 517, in the period from 2008 to 2012, to 753 deaths from 2013 to 2017. From the total number of deaths in this period, 10.2% had been coinfected with HVL and human immunodeficiency virus (VL-HIV), with a case fatality rate of 8.5% in this specific group .

In the Ceara State, located in the north part of Northeastern Brazil, HVL has been detected since 1930, but only since 1986 has it been continuously reported by the Brazilian Ministry of Health (MH) . This fact has led Ceara to be one of the Brazilian states with the highest number of cases. According to the risk stratification defined for HVL, the composite index for the triennium from 2016 to 2018 showed that Ceara had 20.3% (29/184) of its municipalities classified as priorities. The municipality of Fortaleza, the state’s capital, was classified as having very intense transmission. Even though it is fully urbanized, this municipality is located in an area considered endemic and has high incidence coefficients . Another four municipalities: Barbalha, Caucaia, Itapipoca and Juazeiro do Norte were classified with high transmission, and another 24 had medium transmission .

The Brazilian Ministry of Health developed the Visceral Leishmaniasis Control Program (VLCP) to reduce the incidence, case fatality and degree of morbidity through early diagnosis and treatment as well as control of reservoirs and vectors. It is noteworthy that the early identification of patients is of fundamental importance to reduce case fatality through the implementation of timely prophylactic and therapeutic measures .

However, despite the surveillance and control actions established by the MH, the case fatality of the disease has been attributed to factors such as late diagnosis, comorbidities and clinical complications . The emergence of the disease in new geographical areas and the presence of hosts and intrinsic characteristics of the affected individual, – such as immunosuppression, malnutrition and coinfection with HIV, – are also factors associated with the case fatality due to HVL .

The drugs used for the treatment of HVL are pentavalent antimonial and amphotericin B, considering that the choice for each one depends on factors such as age, pregnancy and comorbidities . It is recommended that patients over 50 years of age, kidney, heart and liver transplant recipients, and individuals coinfected with HIV be treated with liposomal amphotericin B . The patient can only be considered cured if he or she remains clinically stable after a follow-up six months later .

One realizes that mortality data and related factors are quite scarce and generally represent only the deaths in healthcare facilities . Due to the current context of HVL transmission in the Ceara State, and because it is a serious disease that can have high lethality, depending on the susceptible population , there is a need to analyze its epidemiological aspects for effective planning of control strategies. Therefore, this study aimed to identify epidemiological and clinical factors associated with HVL case fatality in the Ceara State, from 2007 to 2018.

MATERIALS AND METHODS

The study was conducted with data from the Ceara State, located in Northeastern Brazil. It has a population of almost nine million inhabitants in an area of approximately149,000 km2 divided into 184 municipalities. It has irregular rainfall distribution, average temperatures ranging from 29.4 °C (March) to 30.7 °C (November) and caatinga as the predominant biome .

An analytical transversal study was conducted based on secondary data from new cases of HVL reported in the Information System for Notifiable Diseases (SINAN). We used all the new and confirmed resident and autochthonous cases with laboratory or clinical-epidemiological confirmation criteria and whose final classification was the evolution to cure or death from HVL.

In this study we defined as a case of HVL every person reported in SINAN whose diagnosis was made through rapid immunochromatographic tests and/or parasitological tests and/or immunological tests, like Indirect Immunofluorescence Reaction – IFTN; notifications in which “municipalities of residence” was outside the Ceara State and “diagnosis field” was inconsistent, that is, with errors or not filled in, were excluded.

The clinical evolution of the case to death by HVL or cure was considered as the dependent variable. The independent variables of this study included: sex, age group (distributed by 0 to 4, 5 to 9, 10 to 19, 20 to 39, 40 to 59, and 60 years or older); skin color (white, black, yellow, brown and indigenous); education (illiterate, elementary, high school, and higher education); municipality of residence; presence of VL-HIV coinfection; type of initial drug for treatment (pentavalent antimonial, amphotericin B, pentamidine, liposomal amphotericin B, others, and not used); and presence of clinical manifestations (fever, weakness, weight loss, cough and/or diarrhea, splenomegaly, hepatomegaly, infectious condition, edema, jaundice, hemorrhagic phenomena, and pallor).

The database was processed and analyzed in Stata software (version 15.1, StataCorp, College Station, Texas, USA). For the bivariate analysis, Pearson’s Chi-square test or Fisher’s exact test with 95% significance was used, and the relative risks (RR) and their respective 95% confidence intervals (CI) were calculated.

Thereafter, the variables that presented statistical significance with p-value ≤0.25 were selected for the multivariate analysis. This analysis was performed by the generalized linear model (GLM) with logarithmic linkage and the Poisson distribution family to estimate the relative risks and confidence intervals, using robust error variances that evaluated the differences between categories of explanatory variables for the outcome in the final model. Initially, the multivariate analysis was controlled by age group and those variables that remained significantly associated with the outcome (p<0.05) were taken into the final model.

The study was approved by the Research Ethics Committee of the Federal University of Ceara (CAAE Nº 22785819.6.0000.5054).

RESULTS

Between the years of 2007 and 2018, 6,066 confirmed cases of HVL were diagnosed in the Ceara State. Of these, 4,863 (80.17%) were classified as new cases and 343 evolved to death from HVL, with a case fatality of 7.05%. The male case fatality was 7.20%, 1.06 times higher than the female, but this difference was not statistically significant (95%CI:0.85 – 1.33; p=0.572) (Table 1).

Table 1

Sociodemographic characteristics of cases associated with deaths from Human Visceral Leishmaniasis in the Ceara State, 2007-2018.

Variables	Cases (n)	Deaths (n)	Case fatality (%)	Relative Risk (95%CI)	p-value
Sex (n=4,863)
Female	1,477	107	6.76	1.0
Male	3,043	236	7.20	1.06 (0.85 – 1.33)	0.572
Age Group (4,853)
< 1-year-old	825	52	6.30	4.58 (2.51 – 8.35)	0.012
1 to 10 Years Old	945	13	1.38	1.0
10 to 20 Years Old	522	14	2.68	1.95 (0.92 – 4.12)	0.395
20 to 50 Years Old	1,780	110	6.18	4.49 (2.54 – 7.94)	0.002
50 to 60 Years Old	387	54	13.95	10.14 (5.60 – 18.37)	<0.001
> 60 Years Old	394	100	25.38	18.45 (10.48 – 32.49)	<0.001
Race (n=4,278)
White	346	26	6.99	1.0
Black	87	14	13.86	1.98 (1.08 – 3.65)	0.027
Yellow	14	2	12.50	1.79 (0.46 – 6.89)	0.404*
Brown	3,808	269	6.60	0.94 (0.64 – 1.39)	0.772
Indigenous	23	1	4.17	0.60 (0.08 – 4.21)	0.595*
Education (n=1,631)
Illiterate	81	27	25.00	12.00 (1.68 -85.77)	<0.001
Elementary School	1,022	81	7.34	3.52 (0.50 – 24.79)	0.165
High School	358	14	3.76	1.81 (0.24 – 13.43)	0.555
Higher Education	47	1	2.08	1.0
Residence Zone (n=4,387)
Urban	3,398	266	7.26	1.0
Rural	946	65	6.43	0.88 (0.68 – 1.15)	0.362
Peri-urban	43	2	4.44	0.61 (0.16 – 2.38)	0.468*

*Fisher’s exact Chi-square test. The difference between the number of observations for the variables was due to the ignored and blank records excluded in all variables.

It is noteworthy that the age group over 60 years old has an 18.45 times higher risk of death (95%CI: 10.48 – 32.49) compared to the 1 to 10 years old age group; and the risk of death for those with black skin color was 1.98 times higher (95%CI: 1.08 – 3.65) than for those with white skin color (Table 1).

Among the illiterates, the case fatality rate was 12 times higher (95%CI: 1.68 – 85.77) when compared to cases with higher education; and 1.14 times higher (95%CI: 0.87 – 1.47) in urban compared to rural areas, but not statistically significant (p=0.617) (Table 1).

A case fatality rate of 8.51% was identified in people coinfected with HIV, but it was not statistically significant (p=0.162). On the other hand, the groups of people treated with amphotericin B, liposomal amphotericin B, as well as those who did not take the drug, had an over four times higher risk of dying compared to those treated with pentavalent antimonial (p<0.001). Particularly noteworthy is the 16.36 times higher risk of death (95%CI: 11.62 – 23.03) in cases that were not treated with the specific drugs (Table 2).

Table 2

Clinical and treatment characteristics associated with deaths from Human Visceral Leishmaniasis in the Ceara State, 2007-2018.

Variables	Case (n)	Death (n)	%	RR (95%CI)	p-value
HIV Coinfection (n=3,808)
Yes	301	28	8.51	1.31 (0.89 – 1.91)	0.162*
No	3,253	226	6.50	1.0
Starting Drug (n=4,685)
Pentavalent Antimonial	2,949	8	2.77	1.0
Amphotericin B	573	96	4.35	5.18 (3.91 – 6.86)	<0.001*
Pentamidine	13	1	7.14	2.58 (0.38 – 17.25)	0.321
Liposomal Amphotericin B	618	88	12.46	4.50 (3.38 – 5.99)	<0.001
Other	183	16	8.04	2.90 (1.73 – 4.86)	<0.001
Untreated	35	29	45.31	16.36 (11.62 – 23.03)	<0.001

*Fisher’s exact Chi-square test. The difference in number between the variables is due to the ignored and blank records excluded in all variables.

The HVL case fatality rate in people who had no fever was 1.92 times higher (p<0.001) than for those who had a fever. However, the cases that manifested weakness, infection, edema, jaundice, pallor or hemorrhagic phenomena had higher and statistically significant case fatality (Table 3).

Table 3

Clinical manifestations associated with the occurrence of death from Human Visceral Leishmaniasis in the Ceara State, 2007-2018.

Clinical Manifestation	Cases (n)	Deaths (n)	%	RR (95%CI)	p-value
Fever (n=4,845)	4,280	306	6.67	0.48 (0.35 – 0.66)	<0.001
Weakness (n=4,796)	3,308	270	7.55	1.44 (1.10 – 1.87)	0.007
Weight Loss (n=4,780)	3,262	251	7.14	1.09 (0.86 – 1.39)	0.477
Cough and/or Diarrhea n= (4,734)	1,983	163	7.60	1.14 (0.93 – 1.40)	0.205
Splenomegaly (n=4,736)	3,347	244	6.79	0.90 (0.71 – 1.15)	0.407
Hepatomegaly (n=4,723)	2,908	225	7.18	1.08 (0.86 – 1.35)	0.512
Infectious Condition (n=4,585)	872	138	13.66	2.65 (2.15 – 3.27)	<0.001
Edema (n=4,666)	850	172	16.83	3.81 (3.11 – 4.67)	<0.001
Jaundice (n=4,653)	1,026	150	12.76	2.45 (1.99 – 3.01)	<0.001
Hemorrhagic Phenomena (n=4,615)	330	82	19.90	3.43 (2.73 – 4.31)	<0.001
Pallor (n=4,733)	2,845	241	7.81	1.34 (1.07 – 1.68)	0.012

The difference in number between the variables is due to the ignored and blank records excluded in all variables.

In the multivariate analysis, the following variables remain associated with risk factors for LVH deaths: age groups: from 20 – 50 years old (RR=3.05 95%CI: 1.26 – 7.38; p=0.013); from 50 – 60 years old (RR=5.54; 95%CI: 2.25 – 13.63; p<0.001); 60 years old or more (RR=8.69; 95%CI: 3.56 – 21.20; p<0.001); black-skinned people (RR=2.21; 95%CI: 1.45 – 3.35; p<0.001); jaundice symptoms (RR=1.72; 95%CI:1.38 – 2.14; p<0.001); edema (RR=2.62; 95%CI: 2.10 – 3.26; p<0.001); hemorrhagic events (RR=1.63; 95%CI:1.26 – 2.10; p<0.001); and no prescription drug intake (RR=4.03; 95%CI:2.98 – 5.46; p<0.001). On the other hand, the incidence of deaths was lower in the group that received initial treatment with the pentavalent antimonial drug (RR=0.35; 95%CI:0.27 – 0.45; p<0.001) (Table 4).

Table 4

Multivariate statistical analysis adjusted by age and final of variables associated with deaths from Human Visceral Leishmaniasis in the Ceara State, 2007 to 2018.

Variables (n=4674)	Age-Adjusted (95%CI)	Final (95%CI)
Age Group (n=4594)	-
<=1-year-old		2.12 (0.86 – 5.25)
1 – 10		1
10 – 20		1.64 (0.59 – 4.52)
20 – 50		3.05 (1.26 – 7.38) *
50 – 60		5.54 (2.25 – 13.63) **
>60 Years Old		8.69 (3.56 – 21.20) **
Education (n=1631)	2.29 (1.59 – 3.28) **	-
Black Skin Color (n=4674)	1.77 (1.10 – 2.84)*	2.21 (1.45 – 3.35) **
Jaundice (n=4853)	2.42 (1.98 – 2.94) **	1.72 (1.38 – 2.14) *
Pallor (n=4853)	1.57 (1.26 – 1.95) *	-
Edema (n=4853)	3.33 (2.72 – 4.08) **	2.62 (2.10 – 3.26) **
Hemorrhagic Events (n=4853)	2.79 (2.20 – 3.53) **	1.63 (1.26 – 2.10) **
Pentavalent Antimonial (n=4674)	0.28 (0.21 – 0.36) **	0.35 (0.27 – 0.45) **
Amphotericin B (n=4674)	2.16 (1.72 – 2.71) **	-
Liposomal Amphotericin B (n=4674)	1.34 (1.03 – 1.73) *	-
No Drugs (n=4674)	5.24 (3.93 – 6.99) **	4.03 (2.98 – 5.46) **

*0.01

*0.01