Strategies For Strengthening High Risk Pregnancy Care In Primary Healthcare Services

Primary Healthcare has expanded from implementing healthcare networks, an organization form of services with centrality to primary care services. Due to the extension and heterogeneity of Brazilian municipalities, several organization forms of primary care services existed. In the city of São Paulo they were organized by three modalities of Basic Health Units. Thus, this study aimed to evaluate the care continuity in high-risk pregnancy care in different models of the basic health units. Methods: Data analysis defined performing prenatal care in primary care even after referral to the high-risk pregnancy service as the dependent variable. The independent variables were the sociodemographic and organizational aspects of health services. Logistic regression was used as the statistical technique. Results:

answer the question: "How do sociodemographic characteristics and types of PHC services influence the care continuity between PHC and Specialized Care (SC) in the different types of UBS in the care of high-risk pregnancy?".

Methods
A cross-sectional evaluative study, being a clipping from the multicenter study entitled "Inquiry on the functioning of primary healthcare and access to specialized care in Brazilian metropolitan areas -AcesSUS", which had the municipality of São Paulo as its study site and aimed to evaluate the care continuity in high-risk pregnancy care in different basic health unitmodels.

Study Object
The selection of the study object was based on using the concept of tracing condition, pointed out as an effective strategy 13 to evaluate the quality of health services in facing complex system situations by choosing to investigate a health problem which presents prevalence with a clear/definite diagnosis, and for which the health disease process and its management have known patterns.
The tracing condition chosen was High Risk Pregnancy (HRP). This condition is present in 15% of all pregnancies 14 , which has several characteristics and risk factors clearly and objectively listed which pre-dispose a pregnant woman to HRP, such as age under 15 years or over 35 years, unsafe marital status, low education, exposure to occupational risks, previous reproductive history of abortion, preterm delivery, gestational diabetes and pre-existing clinical conditions of hypertension, heart disease, and gynecopathy among others.
Given the existence of any of the characteristics or risk factors, it is recommended to evaluate the gestational risk stratification of these women and to confirm the risk. Based on this stratification, diagnostic resources and intervention procedures follow standardized recommendations through predefined protocols and guidelines. As it is a complex health condition, the health system is challenged to offer care in an articulated manner between different levels of complexity.

STUDY POPULATION
The study population consisted of pregnant women who were met in the waiting room of the specialized outpatient clinics (SC) of the city of São Paulo during the month of September 2016 and who were mandatorily referred by Primary Health Care (PHC).

SAMPLE SELECTION
The sampling process followed multiple stages. In the first step for population sizing, we considered the estimated number of medical consultations for high-risk pregnant women performed in each of the specialized care services, using the Integrated Care Management System (SIGA) of the municipality of São Paulo as the data source, from January to October 2015, totaling a monthly average of 623.2 calls.
In the second stage, we chose to estimate proportions to calculate the sample size, assuming a confidence level of 95% (z = 1.96), and a general sampling error of 0.004. The sample model was defined to allow the analysis of the three types of UBSexisting in the city of São Paulo (family, traditional and mixed health units) in order to disaggregate the data according to the three models The third stage was based on the number of women to be included in the sample and the average number of women attended weekly in each service (stratum). The inclusion criterion was to have been forwarded by the PHC through a municipal regulation system. A total data collection time of 27 days was estimated, comprising the 14 sample services. Considering the dynamics of the data collection stage, a sample weight ranging from 3.144286 to 5.50000 was assigned to each of the sample's specialized services.

INSTRUMENT AND DATA COLLECTION
The primary data collection was applied to users of the specialized level service, and was performed by applying a questionnaire composed of 49 semi-structured questions, elaborated and validated by experts (https://www.fcm.unicamp.br/acessus/metodologia/inquerito-amostragem/questionarios/gestacao-de-alto-risco). The questionnaire was about sociodemographic conditions, use of primary care services, specialized care and support systems (laboratory diagnosis and imaging).
Data collected from sociodemographic conditions were: age; years of study; self-declared skin color; which PHC service is responsible for referral and access to health insurance. For PHC services, the gestational period was surveyed when prenatal care began; if she was or remained in prenatal care (PN) in PHC; intervals between PN consultations; which professionals performed care; if the community health agent (CHA) of the territory was known and if at any time during the pregnancy had received home visit (HV).
Data collected from Specialized Care were: if they knew the reason for being in the at-risk prenatal service; time of pregnancy when she was referred; referral professional; days elapsed between referral and first consultation at SC; frequency of consultations and negotiation of the form of delivery with the medical professional.
Furthermore, the use and access to medicines during pregnancy and request and access to laboratory and imaging tests (ultrasound, cardiotocography or others) were investigatedfor the logistic and support systems. The survey was conducted by previously trained interviewers, and covered 14 SC services for high-risk pregnancy in the city of São Paulo from all health regions, out of a total of 16 services which are referrals to the municipal health system.

DATA ANALYSIS
Of the 49 survey questions, the dependent variable was extracted from the question "Mrs. (xxx), did you follow prenatal care at the station, center or health unit before coming to this service?", selecting the answer as the dependent variable "I am still doing prenatal care at the station, center or health unit", calling the continuity of care between PHC and SC services as a representation for the study.
The independent variables were subdivided from two perspectives; the first was about the sociodemographic characteristics of pregnant women, assuming that youths under 15 years old and women over 35 years old are vulnerable to develop at-risk pregnancy, as well as those with less than seven years of formal education, and having black/brown skin.
The second set of independent variables was selected to identify the care characteristics of PHC services, using the following variables: "knows the CHA", "received HV", "Gestational Age which started the PN" and the elapsed time between "referral and first consultation at SC".
Data were tabulated in an Excel spreadsheet and a preliminary description of variables was performed with support of the Stata version 14.0 program (StataCorp LP, CollegeStation, United States), while a logistic regression test was also applied in order to identify likelihood of association of continuity of care.

Results
It is possible to identify prevalence and similar distribution among the three types of services from the results obtained from the total of 688 pregnant women interviewed for the sociodemographic variables, pregnant women aged 16 to 35 years old and schooling up to complete high school (61.22%), resembling the overall average (Table 1).

Discussion
Among the three types of UBS, the EFS had the lowest chances of continuity of care after referral to SC for both the variables of age and education. Surveillance strategies and health education in the territories are essential, since there is a tendency of repetition of pregnancy in adolescents and chances of reduced years of schooling and less opportunity for access to income 15,16,17 . Another important action to be taken by the PHC services is family planning aimed at the adolescent public, considering that 23% of pregnancies were scored as planned, with a positive association (OR 1.92) of desire to become pregnant in women with less than eight years of studying 17 and highlighting the increase in the number of pregnant women under 14 years of age 15 .
The results show that women of black and brown color are more likely to establish continuity of care between PHC and SC in ESF (OR 1.99) and mixed (OR 1.

25) units. Non-white women in the United
States were more likely to receive general health counseling on HIV testing, breastfeeding, and alcohol and tobacco use than white-skinned women 18 .
Although the results suggest that the ESF and mixed units have greater opportunity to perform care for pregnant women of black or brown color, 19 it is pointed out that these groups of pregnant women are unable to perform the recommended routine examinations, they have more commute during childbirth, have a reduced number of prenatal consultations 20 and overall have higher rates of inadequacy for prenatal care 21,22 . This situation reinforces the discussions presented that expanding access to prenatal care was not sufficient for inducing quality care 23 , and although EFS and mixed units are able to strengthen the continuity of care between services, they may not be able to effect care for pregnant women.
Regarding the set of sociodemographic characteristics of pregnant women (age, education and color), the EFS units were generally the ones which most reinforced the disparities in the continuity of care among vulnerable groups. Health actions in PHC services, especially here in the EFS units, need to be more careful in being guided considering the conditions of social vulnerability of the territories in order to reduce health inequities.
Among the characteristics related to health services, the absence of home visits (HV) was shown as a reducing factor of the chances of continuity of care. The HV provides the opportunity to expand the knowledge of users' life context 24 , and has the ability during the pregnancy and puerperal period to reduce the prevalence of low birth weight and increase the prevalence of exclusive breastfeeding in the first six months 25 .
The HV is shown as a successful action in maternal and child healthcare, but some challenges must be put in the agenda seeking improvement. The visits last less than 5 minutes 26 , a relatively short period to develop health promotion and disease prevention actions; also, the HVs focus on diseases or the health status of the user, with low focus on health education issues 26 .
The HV should be performed by all professionals who make up the PHC teams, but there is a leading role for the work of Community Health Agents. After reformulating the Brazilian National Primary Care Policy 27 , the CHA no longer constitutes a mandatory member of the teams, thus weakening a potent form of PHC care. Inclusion of CHAs in teams is essential for strengthening primary care.
The results point to the prevalence of early prenatal intake in the three types of UBS, however there is a greater chance of effective continuity of care in the EFS and mixed units when intake occurs late.
Late prenatal intake reduces the chances of risk identification and timely intervention of complications 19,28 , triggering alarm in health teams and moving various resources such as HV and CHA to maintain surveillance. This situation may be the explanatory hypothesis for the findings.
Sending PHC referrals to SC in a timely manner is a challenge for healthcare networks, when over 30 days proved to be a variable capable of reducing the chances of continuity of care in high-risk pregnancy. Strategies such as the use of clinical practice and regulation protocols widely known by health professionals, and strengthening strategies such as matrix support and the use of health technologies can reduce the demand and the long waiting time for specialized care 29,30 .
For the characteristics of the services, the ESF and mixed units presented similar results, but there are necessary advances in the three UBS modalities, such as strengthening the HV, maintenance and expansion of the CHA in the health teams, and improving the PHC resoluteness to reduce referrals and the time taken for effective service by the specialized services.

Study Limitations
As a limitation of the study, the adopted statistical analytical model did not follow the consensus in the literature and the selection of independent variables for the logistic regression test were included in order to enable representing personal characteristics and services which may directly or indirectly influence continuity of care and not based on pre-test significance for each individual variable (chisquared test). It is noteworthy that this strategy was adopted because it is an evaluative study of health services, and was not conducted from the epidemiological perspective.
The results, however representative of the sample, cannot be extended to other populations, since the statistical significance to assess relationships has not been confirmed, but they collaborate in order to highlight weaknesses in the healthcare in the studied sample.

Conclusion
The ESF and mixed units are located in the territories in order to reduce inequities in access to healthcare; a fact represented by the prevalence of black and brown women in high-risk pregnancy For the set of sociodemographic variables (age, education and color), traditional-type units are more likely to maintain continuity of care in the low-educated population than the EFS for the black/brown population. Working the territory with all types of UBS units according to their vulnerabilities is essential, considering as many variables as possible in an attempt to reduce the selective offer of continuity of care.
Home visits and community health agents were effectivein the set of service characteristic variables, increasing the chances of continuity of care. Such care tools should be strengthened in the different UBS models in order to enhance the performance of PHC services.

Declarations
Ethics approval and consent to participate

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests" in this section.

Funding
This work was supported for Portuguese to English translation from the University of São Paulo.

Authors' contributions
NA conducted the analyses of data, interpreted the data and drafted the manuscript. OYT contributed to the interpretation of data and critically revised the manuscript. All authors read and approved the final manuscript.