Analysis of the Space and Time Distribution of Syphilis in Brazil: Cases of Acquired Syphilis From 2010-2018 and Congenital and Syphilis in Pregnant Women From 2001-2017

Background. In Brazil, the notication of acquired syphilis (AS), congenital syphilis (CS) and syphilis in pregnant women (SiP) is compulsory. Notication data provided by the Ministry of Health (MoH) in combination with the mapping of vulnerable geographic areas is essential to forecasting possible outbreaks and more effectively combating infection through monitoring. We aim evaluated the space and time distribution of reported cases of CS and SiP in Brazil (2001 to 2017), as well as an estimation of cases of AS (2010 to 2018). Methods. A retrospective ecological study was carried out using secondary surveillance data obtained from the Brazilian National Notiable Diseases Information System (SINAN) database, considering all reported cases of CS and SiP between 2002 to 2017, as well as MoH epidemiological bulletin data regarding cases of AS between 2010 to 2018. Epidemiological characteristics and time trends were analyzed using joinpoint regression models and spatial distribution (three-year moving averages), considering microregions or states/macroregions as units of analysis. Results. A total of 188,630 (5.4/100,000) CS, 235,895 SiP (6.3/100,000) and 479,731 cases of AS (27.4/100,000) were reported during the periods studied. The epidemiological prole of Brazil indicates most reported CS cases occurred among ‘mixed-race’ newborns who were diagnosed within seven days after birth and whose mothers had received prenatal care. Regarding SiP, most cases were among women who self-reported ‘mixed-race’, were aged 20-39 years, had up to eight years of formal education and were diagnosed with primary or latent syphilis. Overall, rates of AS rose around 400% from 2010 to 2018. Nearly all microregions reported at least one case of CS and SiP. From 2012 to 2016, CS cases increased signicantly in almost all Brazilian states, most notably in the South, Southeast, and Central-West macroregions, from 2001-2018 and the relative risk of SiP increased around 4,000%. Conclusions. Considering the epidemiological scenario of the infection in Brazil, it is necessary to enhance preventive, control and eradication measures.

contaminated body uids [6][7][8]. In vertical transmission from mother-to-child (CS), following direct contact with the bacterium by the mother, infection spreads to the fetus hematologically, predominantly via the transplacental route [9]. Most of the signs and symptoms of the disease, e.g. tissue damage, arise from the in ammatory reaction to infection [10].
Despite the existence of diagnostic tests and effective antibiotic treatment, increasing numbers of syphilis cases registered in Brazil re ect the fragility of the public health system. In accordance with Brazilian Health Regulations (Ordinances 542 for CS − 1986, 33 for SiP -2005, and 2,472 for AS − 2010), con rmed syphilis infection requires compulsory noti cation to contribute to incidence/epidemiological investigations. The noti cations sent to public health authorities provide data to the Ministry of Health for monitoring and to predict potential outbreaks [11]. Although national estimates facilitate international comparisons, these cannot solve heterogeneity at the geographical level where public health actions usually take place. Using ner scales to monitor disease variation and to identify high-risk communities is a critical aspect to develop targeted interventions towards the reduction of the burden of communicable diseases [12].
Globally, rising numbers of syphilis cases have also been reported in the United States, Canada, Europe, Russia and China, which has drawn attention to a public health crisis linked to the synergy between the epidemiology and biology of syphilis and the human immunode ciency virus [13]. In Latin America, Asia and Africa, recent increases in case frequency and the size of affected geographic areas have marked a new era of infection transmission, seriously burdening limited-resource public health systems [14].

Ethical statements
As this study was based on secondary data, and all presented information is in the public domain, none of the described variables allowed for individual identi cation. In 2016, a new resolution published by the Brazilian National Health Council abrogated the need to seek approval from any Institutional Review Board for studies using publicly available secondary data that does not provide individually identi able information (http://conselho.saude.gov.br/resolucoes/2016/reso510.pdf).

Study area
This study was performed in Brazil, the fth largest country in the world, with a total area of 8,515,767 km 2 . The Federative Republic of Brazil is comprised of 26 states and a Federal District, which were grouped into ve macroregions (Central-West, North, Northeast, South, and Southeast-see Fig. 1 [11].

Statistical analysis
The distribution of cases of AS, CS and SiP were evaluated using spatial analysis methods and geoprocessing techniques. Data on CS and SiP were analyzed according to the microregion in which noti cation occurred to allow for greater precision in intra-and inter-regional differences, and also to aid in revealing speci c areas in which to target potential intervention efforts. Cases of AS were analyzed on a state and macroregion level. To minimize potential interference from random uctuations that could occur in time-series studies, distribution maps were constructed using three-year moving averages from 2001 to 2017 for CS and SiP, and between 2010 and 2018 for AS cases; annual noti cation rates was calculated for 558 Brazilian microregions and ve macroregions and 26 states plus the Federal District, respectively. The infection rate, expressed as the number of infected individuals per 100,000 inhabitants, was used to constructed thematic maps. Trends changes in annual noti cation rates were calculated using the joinpoint regression model and expressed as Annual Percentage Change (APC) with 5% signi cance (p < 0.05). Digital maps were obtained from the IBGE cartographic database in shape le (.shp) format, then reformatted and analyzed using QGIS version 3.10 (Geographic 140 Information System, Open Source Geospatial Foundation Project. http://qgis.osgeo.org). This software package was used for data processing, analysis and the presentation of cartographic data. A checklist (see Additional le 1) is provided according to the Strengthening the Reporting of Observational studies in Epidemiology (STROBE) guidelines.

Results
Annual Percentage Change (APC) in syphilis cases

Discussion
We performed a systematic space and time distribution analysis of reported cases of AS, CS and SiP in Brazil, and also described the distribution of infection throughout the country during the respectively studied periods. Overall, our results indicate the existence of a distinct trend for each form of syphilis, with an increasing tendency in case numbers over time. In fact, data reveal the number of cases has been growing in recent years, suggesting that syphilis remains a public health concern in the country.  [18]. The strategy to combat sexually transmitted infections (STIs) from 2016 to 2021 prioritizes the elimination of congenital syphilis by implementing comprehensive syphilis screening and treatment among pregnant women, as well as in speci c populations, with a goal of reducing the global incidence of syphilis by 90%, with 50 or fewer cases of CS per 100,000 live births in 80% of the countries worldwide by 2030 [19]. Furthermore, in comparison to data published in international studies, the noti cation rates of CS are signi cantly higher in Brazil than those reported in other countries [18]. It is important to note that changes in the epidemiologic infection pro le in recent years are likely associated with: a) increased testing coverage enabled by the adoption of rapid diagnostic tests, which therefore allowed for the expanded identi cation of incident cases throughout the country, and b) the shortage of benzathine penicillin, which, since 2014, has affected Brazil, as well as other countries, due to a de cit of raw materials required for its production. Consequently, this shortage contributed to increases in untreated or inadequately treated syphilis in pregnancy, which directly led to higher rates of CS [20,21]. On the other hand, surveillance data also show that the general pro les of those infected contain reports of high numbers of sexual partners, unprotected sex, recreational drug use and the use of sex-oriented social networking apps [17].
The present study also investigated the epidemiological pro le of syphilis in both maternal and newborn populations in Brazil. A similar prevalence (~ 50%) was observed between both male and female cases of CS, which stands in accordance with a study conducted in the south of Brazil, which found no signi cant differences in noti cation rate according to sex. This can be explained by the fact that CS is transmitted vertically and that infection transmissibility is in uenced by the mother's infection stage and fetal exposure [22]. With respect to race/ethnicity, the majority of reported CS cases occurred among newborns identi ed by the family as 'mixed-race'. Other data have also shown that the noti cation rate of congenital syphilis tends to be signi cantly higher in black or 'mixed-race' children. In fact, several studies have linked cases of CS with family history, including the racial classi cation of children's parents, with high numbers of pregnant women self-identifying as 'mixed-race' or black [22][23][24][25].
Herein, most cases of CS were diagnosed less than seven days after birth, and most mothers reported receiving prenatal care. It has been reported that CS diagnosis usually occurs within seven days after birth (typically between the 1 st and 2 nd day of life) [23,25]. In addition to prematurity, newborns usually present classic signs and symptoms of infection soon after birth, including low birth weight, anemia, jaundice, respiratory distress, visceromegaly, congenital malformations, serosanguinous discharge and rhinitis, skin lesions, heart disease and/or hearing loss [23,26,27]. Interestingly, it was found that despite the predominance of prenatal care in 70+% of the CS cases investigated herein, signi cantly high numbers of cases of CS were nonetheless reported throughout the country. While prenatal care is becoming more commonplace in Brazil, this rate still remains below the recommendations established by the Ministry of Health, which advocate that prenatal care must be properly provided to all pregnant women [28]. It is known that the noti cation rate of syphilis is considered to be an important indicator of accessibility and prenatal care quality [23]. However, despite the expansion of diagnosis and treatment in Brazil, increases in the number of cases indicates shortcomings in the efforts designed to control and prevent this STI [15]. At the same time, health authorities have also attributed increases in noti cation rates to the success of public health actions in improving detection rates. Nevertheless, actions designed to improve health care access for pregnant women have performed poorly in terms of CS prevention [21]. One of the main purposes of prenatal care is to assist women in a quali ed and humanized way beginning in the early stages of pregnancy, adopting early screening procedures coupled with timely interventions [28]. Early diagnosis and treatment of SiP, ideally before the 20 th week of pregnancy, can reduce CS-related cases, such as miscarriages, stillbirths, and infant deaths [29]. It is therefore important that all pregnant women be tested at the rst prenatal visit scheduled in the rst trimester of pregnancy, with repeat testing performed at around 28 weeks and in the beginning of the third trimester in order to promptly implement appropriate therapy if necessary [30]. Information regarding CS vertical transmission should be provided to pregnant women at the onset of prenatal care, and physicians must inform patients regarding the risks and consequences of the disease to the mother and her fetus [24].
An established diagnosis does not guarantee adhesion to appropriate treatment by the patient. The late onset of symptoms (mostly detected in the third trimester), the interruption of and/or low attendance in prenatal examinations, di culties in diagnosis, a lack of information regarding infection and unsafe sexual practices have been reported as risk factors for the development of syphilis [32][33][34]. Furthermore, errors in antibiotic dosage have also been of concern. Other key factors, such as conjugal in delity, the absence of partners at prenatal appointments and reluctance in adhering to treatment protocols have also been reported by patients [27,31].
Regarding the sociodemographic pro le of pregnant women observed herein, most cases were identi ed among women aged 20-39 years who self-identi ed as 'mixed-race', had up to eight years of formal education and were mainly diagnosed with primary syphilis. However, it worth noting that many records contained missing race/ethnicity classi cation data in P1, which can be considered as a bias in the interpretation of our results. Indeed, the data presented herein corroborate other studies that identi ed a signi cant correlation between these sociodemographic characteristics and SiP [15,[21][22][23][24][25][26][27]29,30,[32][33][34][35][36][37][38]. We suggest that distinct strategies are required to reach more vulnerable populations and to minimize inequalities that enable greater access to health services. Poverty prompts speci c vulnerabilities, whether behavioral or brought on by de ciencies in health services, such as prenatal care access and quality, which are also signi cantly associated with SiP. Brazilian social inequality in health supports the hypothesis that the prevalence of SiP is associated with a lower socioeconomic status [25,32,34].
Several child and maternal factors have been associated with increased risk and vulnerability to CS, such as race/ethnicity, socioeconomic status and maternal age [23]. In fact, STI and SiP were found to be more strongly associated with women who dropped out of school, self-identi ed as 'black' or 'mixed-race', were under 20 years of age or between 20 and 30 years old [30], had limited access to quality health services, preventive and educational programs or received assistance at public health care units, but without adequate prenatal care [25,32,34]. The association between younger mothers and CS strengthens the hypothesis that low-income adolescents are more vulnerable to STI and are at higher risk of teen pregnancy, as some authors have attributed this behavior to emotional and general immaturity [30,38]. A low education level is also considered a marker of greater risk exposure due to unawareness regarding the importance of prevention. In some settings in Brazil, higher frequencies of syphilis were diagnosed during the 2 nd and 3 rd trimesters of pregnancy, possibly related to (I) delayed initiation of prenatal care and (II) substandard quality of obstetric care [36]. These observations rea rm the importance of early syphilis detection in pregnant women, as well as the availability of appropriate treatment women and their partners [36]. SiP control programs should place greater focus on these more vulnerable populations [32,34], especially considering that the lack of or inadequacies in public sexual education policies for younger individuals was associated with decreased condoms use in casual sexual relations in recent years [25].
The present study identi ed groups of municipalities with high rates of AS, CS, and SiP in Brazil. Between 2010 and 2018, higher numbers of AS cases were recorded throughout most of the country, especially in some southern and northern states. In addition, almost all microregions of the country reported a higher intensity of CS and SiP infections. Considering the epidemic pro le of Brazil, we call attention to some priority areas in which intervention, such as appropriate patient management and effective infection control measures, could prove bene cial. The observed variations in noti cation rates among the municipalities may be the result of a decline in the underreporting of cases or re ect problems in local health systems, such as a lack of access to specialized services. Importantly, incomplete reporting hinders the elaboration of preventive strategies by policymakers, resulting in ineffective epidemiological surveillance [26,39]. It is evident that the Brazilian healthcare system will continue to be challenged by this scenario, as despite government investment in awareness campaigns, the circumstances remain far from ideal. Low adherence to treatment among patients and their partners is a main obstacle that must be overcome. Insu cient social awareness regarding prevention and treatment re ects the urgent need for educational policies aimed at preventing congenital infections [25,26] in Brazil, especially in the affected macroregions and microregions identi ed in this study.

Conclusion
We conclude that despite the existence of control and awareness programs for STIs, current measures have proven ineffective in decreasing the noti cation rates of AS, CS and SiP in Brazil. Additionally, the underreporting of registered cases throughout the country results in biased data analysis. Considering the epidemiological pro le of infection in Brazil, our results highlight the need to enhance preventive and control measures for eradication of syphilis. Abbreviations