Promoting evidence-based clinical practice for women in abortion situations after the implementation of a surveillance network: a situational analysis in a Brazilian university hospital

Nelio N. Veiga-Junior (  neliojunior@hotmail.com ) University of Campinas (UNICAMP) School of Medicine Caroline Eugeni University of Campinas (UNICAMP) School of Medicine Beatriz D. Kajiura University of Campinas (UNICAMP) School of Medicine Priscilla B. F. Dantas University of Campinas (UNICAMP) School of Medicine Caroline B. Trabach University of Campinas (UNICAMP) School of Medicine Aline A. Junqueira University of Campinas (UNICAMP) School of Medicine Carina C. Nunes University of Campinas (UNICAMP) School of Medicine Luiz F. Baccaro University of Campinas (UNICAMP) School of Medicine

Conclusion the use of MVA increased after the installation of a surveillance network for good clinical practices. Being part of networks that encourage the use of evidence-based practices is an opportunity for health facilities to increase access to safe abortions.

Background
Unsafe abortion is a public health concern and one of the main causes of maternal mortality worldwide [1]. It has a higher incidence in developing countries, occurring in 37 out of every 1000 pregnant women compared to 27 out of every 1000 pregnant women in developed countries [2]. It is estimated that 208 million women become pregnant every year [2], and approximately 25% of all pregnancies end in abortion worldwide [3]. Among these, 22 million unsafe abortions are performed annually, leading to 47,000 maternal deaths due to complications [2]. In Brazil, where the law ensures the right to terminate a pregnancy only in cases of sexual violence, risk of maternal death, and fetuses with anencephaly, the avoidable morbidity and mortality persists [4]. It is estimated that one in ve women up to 40 years of age has had an illegal abortion, with almost half of them having been hospitalized for complications. With this, it is estimated that every year, 200 women die in the country because of unsafe abortions [5,6]. Furthermore, as most women hospitalized for complications from unsafe abortions do not have health insurance, the annual expenditure to the government is estimated at approximately US$10 million [7].
Among abortion-related complications, there are some related to the clinical condition itself, such as an incomplete abortion with consequent massive hemorrhage [8], and those associated with unsafe induced abortion, such as cervical laceration, uterine perforation, and infection [6]. However, in women undergoing abortions, some complications occur as a result of the therapeutic procedures performed for uterine evacuation. Among these procedures, the one with the highest frequency of possible complications is cervical dilation followed by curettage [8]. Uterine perforation is the most common complication associated with cervical dilation. It occurs in approximately 2% of the cases [10] and is more frequent at higher gestational ages [9]. Another unwanted outcome after curettage is uterine synechiae. Studies estimate that approximately 90% of the cases of severe uterine synechia occur as a result of curettage performed to treat pregnancy complications, such as incomplete abortion [11,12]. According to the Brazilian Ministry of Health, post-abortion curettage is the third most common obstetric procedure performed in public health facilities [13].
The choice of the uterine evacuation method is an important aspect in the provision of health services. Manual vacuum aspiration (MVA) and management of medical abortion (MA) should be used to treat women with complications associated with spontaneous and unsafe abortions. MVA for gestational losses of up to 14 weeks is the surgical technique recommended by the World Health Organization (WHO) and the International Federation of Gynecology and Obstetrics [1]. It is a procedure that can be performed in an average of 3-10 min at outpatient clinics with analgesics or local anesthesia [14]. Uterine evacuation induced exclusively with pharmacological drugs to terminate pregnancy is an alternative to avoid the risks associated with a surgical procedure. It has similar e cacy as that of surgical abortion in cases of early pregnancy [15], is safe and effective, and has been proven to be acceptable in low-resource settings [16].
According to the WHO, a standard of care is to identify priorities for providing information on evidence-based practice [17]. Surveillance systems can be used to improve the quality of care and help countries evaluate and monitor programs aimed at preventing unplanned pregnancies and maternal morbidity and mortality. Through data collection, abortion surveillance programs can monitor changes in clinical practice patterns and provide the data necessary to manage health policies [17,18]. In a previous study, our research group demonstrated how a surveillance system was able to increase contraceptive use before hospital discharge in women who were hospitalized for abortion complications [19].
When an abortion is performed by a professional trained in providing adequate technical or ethical assistance, the procedure is very safe [4]. Choosing a method of uterine evacuation is essential for the safe management of abortion conditions [13]. Dilatation and curettage should be avoided whenever possible and replaced by other methods [8,4]. The aim of this study was to evaluate the frequency of use of MVA and exclusively MA, and investigate the associated factors after the installation of a surveillance network of good practices in a university hospital in the southeastern region of Brazil.

Methods
The multicentric network MUSA -Women in Abortion Situations -is a network created by the Latin American Center for Perinatology (CLAP) to improve care for women undergoing abortions in Latin America and the Caribbean [20].
It includes several hospitals, called sentinel centers, which periodically send their data regarding the pregnancy cycle for registration in the Perinatal Computerized System (SIP), a software developed by CLAP that helps health facilities register data related to pregnancy and epidemiologic monitoring. Our institution, University of Campinas Women's Hospital (UNICAMP) is a tertiary referral hospital for cases of complications related to pregnancy in municipalities in the region and experiences an average of 250 births and 20 cases of rst trimester pregnancy loss per month. Our hospital has been a sentinel institution of the MUSA Network since July 2017. The hospital follows the laws of Brazil regarding the legal termination of pregnancy, in which abortion is allowed only in cases of risk of maternal death, sexual violence, and fetal anencephaly [4].
The sentinel centers of the MUSA Network regularly provide information on maternal morbidity in early pregnancy loss, termination methods for uterine evacuations, incidence of complications related to pregnancy termination, incidence of preoperative antibiotic use, and prescription of contraception before hospital discharge. Through SIP, it is possible to carry out epidemiological monitoring and comparisons between different sentinel centers over time. Representatives from each sentinel center also hold regular online meetings to discuss the data collected, conduct scienti c discussions on the topic of women's health in abortion situations, and encourage good clinical practices for safe abortion.

Data collection
This cross-sectional study was conducted between July 2017 and November 2020. The inclusion criteria were women admitted for abortion due to any cause and women of any age group who visited our hospital. The exclusion criteria included women with bleeding during pregnancy who did not have a con rmed abortion and women with ectopic or molar pregnancies. All participants who agreed to participate in the project signed a free and informed consent form. The research ethics committee of our institute approved this study (approval number CAAE: 93060618.9.1001.5404).
Dependent variables MVA MA with misoprostol: when there was no need for further uterine evacuation methods

Independent variables
The independent variables were age, education, marital status, living status, health records, number of pregnancies, number of births, number of abortions, body mass index (BMI), active smoking, illegal drug use, alcohol use, planned pregnancy, pregnancy resulting from contraceptive failure, date of admission for abortion, abortion for legal reasons, gestational age, presence of any complications, and admission data.

Sample size
This was a convenience sample, including all women who ful lled the inclusion criteria and who signed the informed consent form from 07/01/2017 to 11/16/2020. We calculated the power of the sample using the estimate in a descriptive study with a categorical variable, setting the level of alpha signi cance or type I error at 5%, and the sample error at 5%. We calculated that the power of the sample to estimate the prevalence of MVA and medical methods for uterine evacuation were 99.4% and 97.3%, respectively.

Statistical analysis
Initially, a descriptive analysis of the data was performed. The calculated continuous variables were the mean, standard deviation, median, interval, and quartile. Relative frequencies were calculated for categorical variables.
The Cochran-Armitage trend test was performed with quarterly analysis to assess whether there was a change in the rates of the use of MVA and MA. The chi-square (categorical variables) and Mann-Whitney (continuous variables) tests were performed to assess the factors associated with performing MVA and MA. Finally, to assess the independent factors associated with performing MVA and MA, two models of multiple logistic regression were constructed, with stepwise criteria for selecting variables. The level of signi cance was set at 5%. The software used for the analyzes was The SAS System for Windows (Statistical Analysis System, version 9.2. SAS Institute Inc., 2002-2008, Cary, NC, USA).

Results
During the study period, 474 women with a mean age of 30.01 years (± 7.48) and a median age of 30  years were included. The mean gestational age was 11.03 weeks (± 3.56), and 30.38% of women did not have

Discussion
Our study showed an increase in the use of MVA and maintenance of MA rates since the establishment of a surveillance network of good practices for safe abortion (MUSA Network) in a university hospital. Being admitted in 2020 and having a lower gestational age were factors associated with increased use of MVA. A higher level of education was the only factor associated with a greater use of MA.
The choice of the uterine evacuation method in uences the incidence of both short-and long-term adverse events. Efforts to replace uterine curettage with alternative methods are recommended in the WHO manuals [8]. Adverse events from curettage can occur immediately, such as in cases of uterine perforation. One study evaluated 706 women who underwent laparoscopic sterilization shortly after the rst trimester uterine curettage. Among them, the surgical team suspected that some uterine perforation could have occurred in 0.28% of the cases; however, the rate of perforation veri ed during laparoscopy was 1.98%, which was seven times higher than that expected [15].
Complications of curettage can also be diagnosed. In a study that evaluated women who underwent hysteroscopy after 12 months of spontaneous abortion (86% performed curettage), the prevalence of synechiae was 19.1% [21].
In addition to changes in the menstrual cycle, such as amenorrhea, it was estimated that 7-40% of women with uterine synechiae were infertile [22,23].
Currently, MVA is considered the technique of choice for surgical uterine evacuation in pregnancies of up to 12 to 14 weeks. It is a quick procedure that can be performed in outpatient clinics using less complex anesthetic procedures [8,13]. It has been emphasized that the replacement of uterine curettage by MVA reduces the mortality rate from 1.23-0.07% [24]. A systematic review that analyzed complications related to MVA showed that less than 5% of women experienced hemorrhage without blood transfusion, less than 0.1% presented with uterine perforation or bleeding requiring blood transfusion, less than 0.5% were hospitalized, and only 3% had repeated aspirations, with no maternal deaths reported [25]. These results show that the use of MVA instead of curettage should play a major role in the strategy of improving care during abortions and should be encouraged in all health facilities [26].
The WHO suggests that whenever possible, the uterine curettage procedure with a rigid instrument should be replaced by MVA [8]; however, curettage is still widely used in Brazilian hospitals [13]. According to a national mixed methods study in Brazil, only 45% of women used MVA in legal abortion services [27]. In Honduras, an initiative undertaken to increase the use of MVA at the expense of curettage has not achieved the desired success. The main obstacles cited for an increase in the use of MVA were the lack of training, lack of adequate methods to control pain, and the reluctance of some physicians to abandon the use of traditional curettage [28]. A qualitative study of MVA utilization in Malawi showed that the lack of training and limited human resources are not the only factors preventing the increase in MVA use [29]. The authors report that addressing staff relationships and power dynamics that negatively impact MVA usage is equally important and that performing regular team meetings can improve communication between cadres and promote teamwork and performance [30,31]. In the present study, we found a signi cant trend toward an increase in the use of MVA after the installation of a surveillance network in which one of the initiatives is to hold regular team meetings, which highlights the role of initiatives that promote changes in clinical practices for patient bene t.
Among the alternatives to surgical procedures, MA is considered an effective procedure, with success rates between 75% and 90% [32,33,34]. In the present study, MA was performed in 11.54% of cases, and there was no increase in use even after the establishment of a surveillance network for good practice. For comparison and contextualization purposes, in the United States, uterine evacuation performed exclusively with medications was used in approximately 40% of all abortions in 2018, with most patients being up to 9 weeks pregnant [35]. In Brazil, misoprostol is used for uterine emptying in hospitalized patients [4]. In comparison to surgical procedures, MA takes longer to complete [8], leading to a longer hospital stay; thus, it may not be the rst choice for women or doctors. Higher levels of education in women in abortion situations may increase the possibility of uterine evacuation using medication. Follow-up after MA is based on self-recognition of signs and symptoms [3,15] Brazilian study showed that among Brazilian medical residents in Gynecology and Obstetrics, knowledge of MA is limited [37]. Inadequate training of physicians caring for patients experiencing abortion can be an obstacle to the use of safe techniques for uterine evacuation. Our study highlighted that a surveillance network of good clinical practices can play an important role in medical education and help improve the quality of care provided to women.

Conclusion
MA is still underutilized in our hospital, possibly because of the fear of complications in women who would be less capable of recognizing signs and symptoms of alarm. However, the use of MVA signi cantly increased after the installation of a surveillance network for good clinical practices (MUSA Network). Being part of networks that encourage the use of evidence-based clinical practices is an opportunity for health facilities to set their priorities, evaluate outcomes, and implement changes in health policies to increase access to safe abortions. Continuous monitoring of therapeutic indicators and continuous medical education can help reduce abortion-related morbidity and help improve the healthcare needs and rights of women.

Declarations
Ethics approval and consent to participate: All participants who agreed to participate in the project signed a free and informed consent form. The research ethics committee of our institute approved this study (approval number CAAE: 93060618.9.1001.5404). All procedures were performed in accordance with relevant guidelines and regulations.

Consent to publication:
Not applicable Availability of data and materials: The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
Competing Interests : None of the authors have any competing interest.
Funding information: None