ARTICLE
Prevalence of HIV infection in parturient women and coverage of hiv testing during prenatal care and delivery in Brasília, Brazil

INTRODUCTION

In the 1990s, it was proved that chemoprophylaxis done in HIV-seropositive pregnant and parturient women and in their newborns was effective to reduce HIV vertical transmission^1,2. With the need of appropriately making available medicine that is used for such chemoprophylaxis and of implementing preventive, diagnostic, and therapeutical actions recommended in national level³, it became essential to health services to know the frequency and distribution of HIV infection cases in pregnant women, and the coverage of diagnostic and prophylaxis actions with regard to opportunity and adaptation.

From Decree 993, from September 4^th, 2000⁴, the Brazilian Health Ministry (HM) established a compulsory notification of HIV infection cases in pregnant women; therefore, from that time on they had information about the occurrence of such cases.

Between 2004 and 2009, the detection rate of seropositive pregnant women in Distrito Federal (DF), calculated from the Information System of Notification Losses and Information System of Live Births, which presents the official records of compulsory notification and births, respectively varied from 0.10 to 0.15% of the live births⁵. However, a serological testing concerning Brazilian parturient women⁶ that was performed in 2004, pointed out that 0.413% of them (95%CI: 0.294-0.533) were HIV-seropositive, and it also indicated a 0.425% prevalence for the Midwest area.

We must remember that in DF there were no prevalence studies of HIV infection in pregnant women and the number of notifications with regard to such losses remained beyond estimations for Brazil and for the Midwest area, therefore a hypothesis that such prevalence between parturient women in DF was lower than the one estimated for the Midwest area and for the country was suggested.

We believe that knowledge on the prevalence of HIV infection and of coverage of actions concerning prenatal care and delivery allows a better planning of strategies for prevention, diagnosis, and prophylaxis of this increase in vertical transmission.

OBJECTIVE

To estimate the prevalence of HIV infection among parturient women, and to calculate the coverage of serology testing during prenatal care and delivery in the public health system of Distrito Federal, Brazil.

METHODS

This is a cross-sectional study, in which the variables sociodemographic data, information on prenatal care, and outcomes in HIV serology were collected from prenatal care cards and medical records of pregnant women residing in DF.

The study happened in public maternities from DF, from June 2009 to May 2010. This methodology, with secondary data, was proposed for the State and Municipal Health Secretariats by the Brazilian HM⁷.

Mothers that were still hospitalized after birth were asked without scheduling a day or a time to sign the free informed consent to agree to take part in the research, and then data were collected. Only information about self-reported educational level and race/skin color was taken from an interview with the mother, the others were taken from the pregnant women cards and hospital documents. Records about how many women refused to take part in the study could not be found. The inclusion criteria included only women who were residents in DF. For a 0.425% rate of infected pregnant women⁸, it was considered a bilateral absolute error established in 0.22% and a 5% level of significance. Thus, a minimum sample of 3,359 parturient women was indicated to be collected. Following the national recommendation to develop this study in a state level⁷, we tried to compensate the possible losses due to the inclusion of pregnant women who did not have information concerning serology by stipulating the size of the sample to be studied in 4 thousand pregnant women.

The number of participants in each public maternity was then proportionally established to that of live births residents in DF, registered in 2008, in the Information System on Live Births (SINASC), in the respective institutions. Two public maternities, Unidade Mista de Saúde de São Sebastião and Hospital da Força Aérea de Brasília, whose total was of 1.3% of births recorded in SINASC in 2008, were not included due to the small number of pregnant women care. Other two public maternities, Hospital das Forças Armadas and Hospital Regional de Santa Maria, were not included because they were closed in the beginning of the research. Parturient women from private health institutions were also not included due to operational issues. After the investigation, in the same computerized system, the number of live births by maternity during the data collection period was obtained through the verification of sample representation proportionality as to public maternities. In order to also check similarity of the proportions by age range, distributions by age range were collected both in the sample and also in the SINASC in the same period.

Diagnosis of HIV infection in Brazil can be done by conducting immunoenzymatic tests like ELISA, indirect immunofluorescence, immunoblot, Western blot, and quick tests. Herein, only positive results were considered, i.e. with two reagent stages, one through screening technique and the other through confirmatory technique (IFI, immunoblot, quick immunoblot or Western blot) or two quick tests, according to specific algorithm established by the HM for the country⁹.

Data collection was performed in a standardized form that had been previously validated by the HM.

The team of field researchers was composed initially of 19 health professionals, who represented 11 of the maternities included in the study. Among the problems concerning data collection, there were replacements of field researchers and temporary closure of some maternities. There were 24 interviewers in total. For each chosen maternity, there was an investigator in charge to orient the field researcher on how to collect information from parturient women.

Pregnant individuals who agreed to participate in the research signed the Free Informed Consent. The research was approved by the Ethics Committee in Research, from the DF Health State Secretariat, according to protocol number 389/08.

RESULTS

Distribution of the parturient women included in the study, of those that had the serology results, and of the number of live births in the research period per public hospital can be found in Table 1. Parturient women from Hospital Regional do Paranoá had lower representation in the sample, while those from Hospitais Regionais de Sobradinho and Asa Norte had a higher one.

Proportions by age range found in the sample and in the records of SINASC in the same period are in Table 2. Pregnant women aged 15 to 19 years old had lower representation in the sample.

From the 3,726 pregnant women studied, 3,627 (97.3%) presented information regarding the serological testing result for HIV. Twelve pregnant women were seropositive, representing a 0.33% prevalence coefficient (95%CI: 0.19-0.58). Prevalence by race/skin color, educational level, and age range did not show statistically significant differences with regard to the general prevalence of the sample (Table 3).

As to HIV testing, only 22.1% of the studied pregnant women performed the two HIV tests needed during gestation, 57.0% had only one, and 18.3% carried out one at delivery. A 2.7% rate was found for women that did not have the test results in any moments (Table 4).

At delivery, 74.9% of the pregnant women had an indication to do the exam, because there were no records of a result in the last quarter of gestation. 7.2% of the parturient women did not take the test in labor although there was an indication (Table 4).

Proportion of parturient women that had the mother’s card at delivery was of 94.6%. Only 0.8% reported not having the card. The others did not take it or the information was not collected (Table 4).

Coverage to carry out the two HIV tests during gestation varied according to educational level, age range, number of prenatal care appointments, and quarter when the prenatal care started. It was lower in pregnant women with lower educational level, in younger ones, in those who started late the prenatal care and had few consultations. In the race/skin color variable, there were no statistically significant differences of coverage of two HIV tests in pregnancy (Table 5).

DISCUSSION

Prevalence of HIV seropositivity among parturient women at delivery in the sample did not present a statistically significant difference in comparison to that estimated for Brazil. This finding indicates that a sub-notification of HIV seropositive pregnant women may exist in DF, since the records of compulsory notification show a highly lower prevalence than the one found in this and in other studies^8,10.

It was not possible to achieve the initial estimation of 4 thousand parturient women for the sample size, due to lack of people to gather data in some maternities throughout the research period. However, the final number was higher than the minimum limit established to calculate the prevalence according to established parameters.

Representation differences in the sample happened more due to alterations in the demand of maternities than to study losses. The number of parturient women who had their deliveries in Hospital Regional do Paranoá, for example, changed from 1,645 in 2008 to 2,711 in 2009, a 65% raise. Such differences cause some caution as to the result extrapolation for the population of parturient individuals from the public health system. The public maternities that were not included in this sample took care of only 3.5% of the parturient women from public institutions in the DF. Some of them provide services for specific populations, like those from military hospitals that receive servers of the Armed Forces and the Unidade Mista de São Sebastião, which performs only normal deliveries.

On the other hand, the proportion of pregnant women that had some record of HIV testing results was pretty high, therefore the absence of serology results records little influenced on the prevalence calculation.

The analysis of prevalence by race/skin color, educational level, and age range variables did not show any statistically significant differences; however, such analysis was limited due to the reduced size of the sample, which was not sized to carry out prevalence stratified analyses.

As to the number of tests, the most frequent situation was the performance of only one serology testing in pregnancy, so a new test was performed at delivery. More than half of the parturient women were in this situation.

The proportion of pregnant women with indication of serology testing at delivery, due to lack of serology result in the last quarter of gestation, including those that did not perform the test in any previous moments, was very high. Most pregnant women took the test at delivery, but the opportunity of prophylaxis of vertical transmission in the prenatal care was missed. Furthermore, the great demand for testing in maternities overloads the services and may compromise the quality of the result, especially due to late availability after the child’s birth. Quick tests in Brazilian maternities have some implantation issues, mainly concerning the availability of fast results and the appropriate intervention¹¹.

Even parturient women who had several prenatal care appointments and those who initiated prenatal care in the first quarter had low coverage of taking two HIV tests during gestation, which are recommended by the Decree from the Health State Secretariat of DF Government (SES-DFG) number 37/08¹². For those who initiated prenatal care late and had few appointments, coverage with two tests was even lower. The parturient women with lower educational level and the younger ones also had a very low coverage. It is possible that an important condition that created low coverage may be the precarious socioeconomic condition of the parturient, which can cause several other inter-connected factors, such as late prenatal care beginning, with few consultations, low educational level, and pregnancy at a younger age range. National and local experiences have showed this influence and reinforce the condition of higher social vulnerability in the poorest social classes^{8,13,14,15,16}. Such finding indicates the need of health services that prioritize a solution for the problems associated with the access of such population to health services.

Low coverage of the serology testing found among the youngest women also indicates the need of specific strategies to increase prevention knowledge, opportunities, and options in this populational range^17,18,19.

Since this research works with a survey of secondary data, which were obtained from the mother’s card and from administration records, the possible failures concerning the notes in such documents should be considered as the limitations of this study.

There is a strong possibility that some parturient women have reported a false district address to the health services, since they actually reside in other states, mainly around the DF. A study carried out in 2002 estimated in 3.6 and 8.0% the proportion of mothers that stated living in DF at delivery, but they actually lived in other states²⁰.

Parturient women from the public health system correspond to 78.9% of DF residents²¹, however, since SINASC data indicate that the sociodemographic profile of parturient women who had their children in private maternities is very different from that of public services²¹, it was considered that it would not be appropriate to extrapolate the result from this study for all the parturient women who reside in the district region.

CONCLUSION

Prevalence of HIV seropositivity among the studied parturient women was not significantly different from that estimated for Brazil, therefore it is in accordance with the possibility that there may be a sub-notification of HIV seropositive pregnant women in DF.

HIV serology testing coverage in pregnancy in DF with two tests is low.

Prevention guidelines of HIV vertical transmission with respect to the access to previous diagnosis are not being fully complied, because a very high proportion of pregnant women at labor arrive with an indication to take the quick test, both because they do not present the HIV test result from the first quarter or they do not have the result written in their cards.

The main recommendations from this study concern structure of the health system in order to overcome aspects that difficult the access to tests; the efficient flow of laboratorial results; and care of health professionals, especially obstetricians, nurses and pharmaceuticals, to the importance of fully following the screening protocol for HIV in the prenatal care service.

Efficient access to services, correct prophylactic managements, epidemiological surveillance, and other actions that reduce HIV vertical transmission, according to the protocols approved by the HM and the World Health Organization, without a doubt will reflect upon future studies that aim at estimating the HIV prevalence in pregnant women.