Prevalence and associated predictors of HIV and syphilis among women, south of Brazil

To investigate the general prevalence of sexually transmitted infections (STIs) and the factors associated with syphilis and HIV in women treated in the Integrated Health System (SINAS). Cross-sectional database study with 6.993 patients treated at a referral center in the state of Paraná, Brazil, from January 2013 to December 2015. The overall prevalence was 2.39% for syphilis and HIV, being 0.65 and 0.22% for pregnant women and 3.07 and 0.78% for non-pregnant women, respectively. After adjusting the variables in the logistic regression model, the main predictors associated with syphilis were the age groups 41 to 59 and 60 to 83 years old (p <0.001), non-pregnant women (p <0.01), marital status married (p <0.001), drug use (p <0.01) and previous STIs (<0.001). For HIV, the predictors were city of residence (<0.001), reason for search (p <0.01), reported previous STTs (p <0.001). The population served had a higher prevalence of syphilis when compared to the international panorama, while the HIV values were lower. The predictors associated with STIs were non-pregnant, heterosexual women, aged between the second and fourth decade of life, higher education and who did not use condoms regularly. This study presents relevant information for the evaluation and development of measures that prevent, control and are related to the diagnosis and treatment of the investigated STIs.


Introduction
Sexually transmitted infections (STIs) are pathologies that have affected the world population for a long time and constitute a significant collective health problem. Due to the ease of contagion and magnitude of these pathologies, STIs need to be prioritized as public health problems, making it very important to interrupt the transmission chain and prevent new occurrences through basic actions in collective health (Lima, et al., 2003;Ministério da Saúde, 2006).
The World Health Organization (WHO) estimates that around 12 million cases of STIs occur annually in Brazil (World Health Organization, 2001;Aquino, et al., 2008). Under-reporting is still high and close to 200.000 cases per year, especially as a result of the search for alternative treatments. It is estimated that 70% of affected individuals do not use health services directly, making it difficult to obtain accurate data.
It is estimated that 1 million cases of syphilis occur per year among pregnant women, and early detection and treatment of these women and their partners is advocated, once the infection can be transmitted to the fetus, with serious implications . The elimination of congenital syphilis and vertical HIV transmission has been addressed as a priority in the Latin American and Caribbean regions .
The occurrence of STIs in women, especially during pregnancy, increases maternal and fetal risks (Lima & Viana, 2009). However, the detection of the infection with human immunodeficiency virus (HIV) and syphilis during early pregnancy provides an opportunity for intervention (Inagaki, et al., 2009). Considered as a priority, early detection becomes essential for the adoption of prophylactic measures in the case of pregnant women, such as the use of drugs to reduce the risk of vertical transmission (Volmink, et al., 2002).
In this context, knowledge on regional prevalence and risk factors associated with patients is of paramount importance to improve control measures. It can trigger an epidemiological characterization, and alert for a wide diagnostic screening during the prenatal period, allowing early actions in order to prevent vertical transmission and minimizing the harm to fetal health (Costa, et al., 2010;World Health Organization, 2010). In this study, we investigated the prevalence and predictors associated with female patients (pregnant or non-pregnant) with Syphilis and HIV attended at a referrence center in Southern Brazil.

Type, period, and place of study
This is a cross-sectional, secondary data base study conducted between January and June 2016 with female patients.
The database consisted of all patients who sought care in the public service at a referrence center located in the municipality of Paranavaí (latitude: 23º 04' 23" S, longitude: 52º 27' 55" W), belonging to the 14th Health Regional Agency (RS) of Paraná (PR), Brazil. This Regional Agency attends 28 municipalities, and a population of 259.377 inhabitants.

Variables of interest
Predictor variables: Gender, age, schooling, marital status, city of origin, pregnant or non-pregnant, number of sexual partners, condom use, STIs carrier in the last year.
Variable outcome: Positive diagnosis for STIs.

Inclusion and exclusion criteria
Data from female patients (pregnant or not) treated at SINAS between the years of 2013 and 2015 and belonging to one of the municipalities attended by the 14th Regional Health Agency of Paraná were included in this study. The following medical records were excluded: male patients; patients with diseases other than STIs; patients who have acquired the disease before 2013 and after 2015; and patients who did not have the diagnosis determined.

Diagnosis
All the examinations were performed according to the current norms of the Brazilian Ministry of Health. The tests were performed by trained nurses from the Integrated Health Care System, Paranavaí PR, according to the manufacturers' instructions. For HIV diagnosis, a blood sample was collected using HIV TEST BIOEASY for qualitative detection of antibodies isotypes (IgG, IgM, IgA), specific to HIV-1, including subtype O, and HIV-2, simultaneously in serum, plasma, and whole human blood. For the diagnosis of syphilis, blood samples were collected and analyzed using the immunochromatographic assay (qualitative detection of antibodies of all isotypes, IgG, IgM, IgA) for Treponema pallidum (TP), which detects the presence of syphilis antibodies in the human body, through the immunoassay.

Ethical Aspects
The study was authorized by SINAS and approved by the Ethics Committee on Research in Human Beings of the State University of Maringá (COPEP/UEM), with registration number 1.375.279 in 12/18/2015. In this work, there was no contact with the patients, and the data were collected directly from the medical records of Integrated Health Care System in Paranavaí.

Typing and Verification of Data Consistency
The information was entered in the database of the software Epidata 3.

Results
We analyzed data from 6.993 female patients who sought care in the period between January 2013 and December 2015. The general prevalence of STIs was 2.39% (167/6993). Among the patients attended in the study period, 0.59% (41/6993) had a positive test for HIV, and 1.80% (126/6993) were positive for syphilis ( Figure 1). The prevalence of positivity during pregnancy was relatively low, for both HIV (0.22%) (7) and syphilis (0.65%) (21). Among the 2.674 non-pregnant women, the positivity was 0.78% (21) for HIV and 3.07% (82) for syphilis.  Among the 6.993 women attended, 5.889 reported being pregnant or not, that is 1.104 women chose not to report it.
The socio-demographic characteristics of the bivariate of the women studied in the study showed that HIV had an association with the age group 21-40 years old (OR 0.2; p=0.016), while syphilis had an association with the age groups 21-40 years old (OR 2.2; p<0.01), 41-59 years old (OR 5.5; p<0.001) and 60-83 years old (OR 12.3; p<0.001) ( Table 3).
The prevalence of HIV and syphilis was higher in non-pregnant women, with 75%, (OR 3.9; p <0.001) for HIV, and 79.61% for syphilis (OR 5.4;p <0.001). Women with married marital status had a high positivity rate for syphilis (OR 3.0; p<0.01) when compared to the separated marital status. The level of schooling ≥8 years had higher positivity for HIV (OR 2.8; p<0.01) and syphilis (OR 2.6; p<0.001) in this study. The lowest prevalence for HIV occurred in the city of Paranavaí (OR 0.1; p<0.001). Patients using any type of drug had higher positivity for syphilis 53.23% (OR 2.4; p<0.001) when compared with those who reported no use (Table 3). p<0.01), while condom use reported as "sometimes" showed significant results for both HIV 33.33% (OR 3.8; p<0.01) and syphilis 20.72% (OR 3.5; p<0.01) ( Table 4).  0.677 0.037* a Adjusted for gestational age and blood donor. b Adjusted by gestational age, blood donor and type of exposure. OR = odds ratio; CI = confidence interval. b Adjusted by Zone. *Statistically significant at P < 005. The values with unit zero, could not be included in the logistic regression. Source: The authors.
Multiple logistic regression analysis was conducted to define the impact of different predictors of HIV and syphilis diseases. We can highlight that patient in the age groups of 21-40 (OR 2.5, p = 0.031), 41-59 (OR 5.2, p <0.001) and 60-83 (OR 11.3; p <0.001) years old showed higher prevalence of syphilis. Non-pregnant women were more likely to be positive for the diseases in this study (HIV (OR 3.5, p = 0.023) and syphilis (OR 2.5, p <0.01)). Women with married marital status had a larger positivity rate for syphilis (OR 2.4; p<0.001) when compared to the separated marital status (Table 3).
The level of schooling ≥8 years had higher positivity for HIV (OR 3.0; p=0.011) and syphilis (OR 1.6; p=0.030) in this study. For the predictor city, patients living in Paranavaí showed a greater significant prevalence of HIV (OR 0.1; p <0.001). Regarding drug use, women who reported having used some types of drugs were more likely to have syphilis (OR 1.9, p <0.01) ( Table 3).
The adjusted variables still yielded other relevant results. The predictor seek reason for examinations showed that for HIV risk gradients, prevention (OR 2.7; p=0.044), medical request (OR 4.6; p< 0.01) e other health services (OR 13.4; p<0.001) grew. For syphilis, the estimated risk rates were for medical request (OR 2.8; p< 0.001). Patients who had other STIs showed strong association with the presence of HIV (OR 7.9; p<0.001) and syphilis (OR 3.9; p<0.001) with patients in this study ( Table 4). Type of sexual partner with adjusted analysis values showed that women who have a relationship with women and men had a higher risk of syphilis positivity (OR 6.2; p = 0.020), while the analysis of condom use brought greater positivity for those who did not use any type (OR 3.3; p = 0.016) for HIV and used sometimes (OR 2.2; p = 0.037) for syphilis (Table 4).

Discussion
The World Health Organization (WHO) estimated that in 2012, there were 357 million new STIs cases, with an international prevalence of STIs of around 4.5% worldwide (Torrone, et al., 2018). It is estimated that 1.8 million people were infected with HIV in 2016. Syphilis affects HIV (OR 3.5, p = 0.023) and syphilis (OR 2.5, p <0.01). Many studies have found similar and other rates (Newman, et al., 2015;World Health Organization, 2016;Gomes, et al., 2017;Dionne-Odom, et al., 2018;Bennani, et al., 2017;Olugbenga, et al., 2018;Saindou, et al., 2012;Ginindza, et al., 2017) varying according to the region and type of study. In Latin America and the Caribbean, it is estimated that between 166.000 and 344.000 children are born with congenital syphilis annually (Ministério da Saúde, 2006;World Health Organization, 2001;Aquino, et al., 2008;. In our study the prevalence of STIs was of 2.39%. The prevalence of syphilis observed in this study in non-pregnant women was of 3.07%, which is close to the 4.3% prevalence in women found by Gomes et al. (2017) (Bennani, et al., 2017).
The HIV prevalence in women reported in this study was 0.78%. A study with HIV positive women in Nigeria also reported low prevalence (3%) (Olugbenga, et al., 2018), and in a study in the Indian Ocean, no prevalent case of HIV infection was detected (Saindou, et al., 2012). However, among women of reproductive age in Swaziland, the prevalence was 42.7%, according to Ginindza, et al., (2017). Different prevalence rates among countries with different socio-demographic characteristics is consistent in research of this nature, and they may also reflect, partly, the effect of syndromic STIs treatment performed by health services over the years (Yahya-Malima, et al., 20018).
In multivariate analysis of risk factors for predictors, data demonstrated an increasing risk gradient according to age.
The age group 21-40 years old had an estimated risk for syphilis more than twice the risk for the age 13-20, while for the age groups 41-59 and 60-83 years old the risk was approximately five times and eleven times, respectively, higher than the group 13-20. The prevalence of positivity was also concentrated in the age group 20-34 years in the studies by Oliveira et al. (2014) (20-34 years) and Gomes et al. (2017), with a higher positivity in the age group of 25-34 years.
Pregnancy is a phase in which women need medical follow-up (prenatal examinations), and this monitoring becomes a protection for them. After modeling the adjusted variables of this study, the data analyzed evidenced that the predictor nonpregnant had a significant prevalence when associated to the presence of HIV and syphilis, with a higher risk for syphilis (OR=2.5; p<0.01) compared to HIV (OR=3.5; p<0.023).
Syphilis had an increasing tendency during the period from 2013 to 2015, while for HIV the relevant increase was from 2013 to 2015. Regarding the year 2015, we believe that one of the reasons for the small number of cases is underregistration in the system. This is a problem presented with some frequency by secondary databases, and it is the absence or sub-registration of data, which hinders the integrity of the analyses (Scheidell, et al., 2018). Patients with married marital status had a higher prevalence of positivity for HIV (48.7%) and syphilis (38.71%), and married patients with positivity for syphilis had twice the risk when compared to patients with widow marital status.
Few studies are available involving the detection of STIs in women, comparing pregnant or non-pregnant women (Costa, et al., 2010;Oliveira, et al., 2014;Scheidell, et al., 2018). It is important to highlight that the studies used to compare our findings belonged to similar referrence centers. Early diagnosis, combined with knowledge of the physical and social characteristics of the affected population, such as their lifestyle, age, gender, profession, among others, has been highly effective in preventing new cases through educational measures for the specific population, thus preventing the development of severe cases that evolve to malignancy and death (Miranda, et al., 2018;Conceição, et al., 2009;Liell, et al., 2009). Many aspects regarding the pathology and impact of STIs infections in public health are still poorly known, evidencing the importance of studies directed to epidemiology (World Health Organization, 2001).

Conclusion
In conclusion, a significant frequency of syphilis infection was demonstrated in women attending the Integrated Health Care System in southern Brazil. The syphilis infection showed a significant association with the predictors older than 45 years, non-pregnant, married marital status, use of drugs, and reported previous STI. For HIV the predictors were city of residence, seek reason, and previous STI reported. Epidemiological studies can help in the implementation of adequate practices for the prevention and detection of infected women, as well as reduce the spread of the infection in the community.
Also, it is important to highlight that this type of study contributes to improve the quality of life, and to reduce public health costs.