Implementation of evidence-based practices in normal delivery care

ABSTRACT Objective: to evaluate the impact of the implementation of evidence-based practices on normal delivery care. Method: quasi-experimental, before-and-after intervention study conducted in a public maternity hospital, Amapá. Forty-two professionals and 280 puerperal women were interviewed and data from 555 medical records were analyzed. The study was developed in three phases: baseline audit (phase 1), educational intervention (phase 2) and post-intervention audit (phase 3). Results: after the intervention, there was an increase of 5.3 percentage points (p.p.) in the normal delivery rate. Interviews with the women revealed a significant increase of the presence of companions during labor (10.0 p.p.) and of adoption of the upright or squatting position (31.4 p.p.); significant reduction of amniotomy (16.8 p.p.), lithotomy position (24.3 p.p.), and intravenous oxytocin (17.1 p.p.). From the professionals’ perspective, there was a statistical reduction in the prescription/administration of oxytocin (29.6 p.p.). In the analysis of medical records, a significant reduction in the rate of amniotomy (29.5 p.p.) and lithotomy position (1.5 p.p.) was observed; the rate of adoption of the upright or squatting position presented a statistical increase of 2.2 p.p. Conclusions: there was a positive impact of the educational intervention on the improvement of parturition assistance, but the implementation process was not completely successful in the adoption of scientific evidence in normal delivery care in this institution.


Introduction
The predominant model of childbirth care in Brazil is characterized by the abusive or inappropriate use of interventions (routine amniotomy, lithotomy and intravenous infusion of oxytocin) and restriction of the parturients' rights (restriction of the presence of companions) in all stages of labor. These problems can be prevented or reduced by adopting the best scientific evidence available in obstetric care (1) . in normal-risk pregnancies (2) .
If, on the one hand, the advance of modern obstetrics has contributed to the improvement of indicators of maternal and perinatal morbidity and mortality, on the other hand, it has allowed the implementation of a model that deals with pregnancy and childbirth as if they were diseases rather than as expressions of health. As a consequence, women and newborns (NBs) are exposed to high intervention rates (1) . Therefore, the World Health Organization (WHO) proposed in 1996 some changes in labor and delivery care, including the need to rescue the idea of childbirth as a natural event, with a stimulus for the work of obstetric nurses, the use of practices based on the best scientific evidence and access to appropriate technologies for childbirth care (3) . In this perspective, the importance of health policies and practices to be grounded on the best available evidence and of translating knowledge into action, promoting the effectiveness and safety of interventions, stands out (4)(5)(6) .
However, introducing practices scientific evidencebased care practices requires more than knowledge and beliefs, since changes of behavior, overcoming barriers and filling gaps in the transfer of knowledge are required (4) . Implementation refers to the use of evidence in clinical practice, through changes in care and/or in health services (4) , and clinical audit is one of the strategies to achieve its consolidation.
Clinical audit is a quality improvement strategy that provides data on the disparities between current practice and desired performance. It is based on the assumption that practitioners modify their practices when they receive a feedback showing that they are inconsistent with the desirable performance (7) .
A usual clinical audit models is the one adopted by the Joanna Briggs Institute (JBI), which consists of three phases: baseline audit, implementation of best practices, and post-implementation audit (4) .
Although plenty research on good delivery care practices is available, there is a lack of studies on the assessment of the impact of interventions proposing the implementation of this evidence. Aspects such as implementation methodology still need to be better investigated (8)(9) .
In this sense, this study seeks to answer the following question: does the implementation of evidence-based practices modify normal delivery care?
Thus, this study was proposed to evaluate the impact of the implementation of evidence-based practices in normal delivery care.

Method
Quasi-experimental, before-and-after intervention study that followed the clinical audit process to implement evidence-based care practices, used by the JBI, composed of the phases: 1) planning and conduction of a baseline audit; 2) evaluation and discussion of results of the baseline audit and educational intervention to implement best practices with the audit team; 3) post-intervention audit, in which the same criteria of the baseline audit are measured and whose objective is to compare the differences between the results of the two audits and to verify the conformity of each audited criterion.
The study was conducted between July 2015 and March 2016 at the Mãe Luzia Women's Hospital (HMML), a reference public maternity hospital in the state of Amapá, in Macapá, which provides care to women at normal and high obstetric risk. In 2015, the monthly mean of admissions was 650 births (66.5% normal and 33.5% cesarean section) 1 .
This study used data obtained through interviews with health professionals and puerperal women and by consulting the medical records of women attended at the HMML.
In the case of health professionals, it was decided to use the population; 71 professionals were eligible.
Of these, 52 met the inclusion criteria, but 10 were not located or did not agree to participate in the study.
Thus the remaining 42 professionals participated in the baseline audit (25 nurses and 17 physicians).
In addition to these losses, 10 professionals refused 1 Data from the HMML Medical Archive and Statistical Service (SAME).
www.eerp.usp.br/rlae 3 Côrtes, CT; Oliveira, SMJV; Santos, RCS; Francisco, AA; Riesco, MLG; Shimoda, GT. or were unavailable in the subsequent phases of the study, resulting in a total of 32 professionals (20 nurses and 12 physicians) who shared in the educational intervention and post-intervention audit. The inclusion criteria of professionals were: obstetric nurse, obstetrician, nursing or medical resident and assistants of normal delivery in the HMML.
The sample size of puerperal women was defined by the prevalence test for before-and-after studies, considering the rate of oxytocin use in the Northern region of Brazil of 22.8% (2) , reducing to 10.5%.
Thus, it was estimated that at least 280 women (140 before intervention and 140 after it) were needed to reach an 80% test power and 95% confidence

Results
The comparative analysis between phases 1 and 3 showed that there was an increase of 5.3 percentage points (p.p.) in the normal delivery rate (Table 1).
In the interview with the puerperal women,  (Table 2). The interviews with the professionals revealed a statistical decrease of oxytocin prescription or administration after the intervention. No statistical differences were seen in the other practices (Table 3).
In the analysis of medical records, there were a large number of records in the information topic. After the intervention, there was a statistically significant difference in the performance of amniotomy and in the use of lateral position during delivery (Table 4).

Discussion
This quasi-experimental, before-and-after intervention study was based on the clinical audit model and sought to assess the impact of the implementation of evidence-based practices on normal delivery care. Few studies have explored this subject in the obstetric area despite its great clinical and academic relevance.
The care model adopted in Brazilian hospitals results in the exposure of women, especially those with habitual obstetric risk, to unnecessary interventions that lack evidence to justify their use (2) . Thus, the protocol of this research defined some practices and maternal outcomes that represent the use of the best evidence in childbirth care recommended by the WHO. In the HMML, such changes were fundamental in view of the framework found in the baseline audit, that is, high rates of interventions without scientific recommendation or even considered iatrogenic.
Although this study has resulted in improved clinical practice, scientific evidence-based childbirth care practices has not been fully implemented, possibly because this is a complex and continuous process that involves changes and overcoming of barriers at the individual and institutional levels, as it has been indicated by other researchers (4,11) .
The scientific literature points out several factors that hinder the implementation of evidence-based clinical practice. At the organizational level, the main barriers are lack of time, inadequate facilities and lack of support (12) . In the hospital where this research was conducted, workload of professionals was detected, as well as a small number of beds inconsistent with the high demand of deliveries, and lack of physical infrastructure.
The implementation of evidence could have been more successful if there were an organizational context that supported evidence-based practice.
At the individual level, barriers include lack of knowledge about research methods and results and negative attitude towards evidence-based practice (12) .
Added to this is the resistance of some health professionals who cannot break up with the current paradigm of childbirth care (13) , probably because they were trained in a time before the launching of humanization policies and evidence-based practice.
Moreover, medical education does not yet focus on the training of professionals to provide comprehensive, quality and humanized care, but it is rather inclined to reproduce the use of interventionist practices (13)(14) .
The competition imposed by other health priorities, the scarcity of resources, the lack of motivation to implement and sustain the changes in the practice of care and the ineffective dissemination of the results are factors that contribute to the resumption of the previous practice after a research intervention (11) .
There are a large number of strategies that can contribute to an effective implementation of changes in the clinical practice. These are based on different theories about human behavior, professional change and organizational performance. The literature suggests that real and sustainable changes can be achieved by combining these different approaches (15) .
After lack of physical structure, lack of human and material resources in health institutions and lack of institutional support and guidelines for implementation of the Law on Companions (14) .
In The second review concluded that there is no evidence that amniotomy is associated with shortened cervical dilatation, cesarean section rate, maternal satisfaction, and Apgar score at the fifth minute (20) .
Therefore, this procedure should not be routinely adopted as part of parturient care.
As for delivery position, there was a significant were adopted by all parturient women and lithotomy was no longer used, a finding supported by statistical significance (p = 0.001) (21) .
In humanized care, women are encouraged to use their freedom to choose their position at labor and delivery. However, in Brazil, the lithotomy position continues to be used during the expulsive period by the majority of parturient (2) .
Study shows that, when women adopt vertical positions, the physical and psychological benefits include shorter duration of labor, fewer interventions and less severe pain, and greater satisfaction with childbirth (22) . the scientific evidence in the service where this research was conducted.
Directed pushing and Kristeller maneuver were also significantly reduced after the educational intervention.
When comparing the results of the interviewed mothers with those of the professionals, the data on these practices reveal partial agreement, since only directed pushing had a statistically significant decrease.
Regardless of complete cervical dilatation, stimulating the parturient to force, preventing her from obeying her own impulses, thus disrespecting the physiology of childbirth, is a frequent practice in maternities (3) . The National Guideline on Parturition Care recommends spontaneous pushing during the expulsive period in women without analgesia, and avoidance of directed pushing (1) .
A recent literature review that assessed maternal and neonatal morbidity associated with the type of pushing used during the expulsive period found that the groups did not differ in perineal lesions, episiotomies or type of delivery. Only one study found a higher Apgar score in the fifth minute and better umbilical artery pH in the spontaneous pushing group. The study concluded that the low methodological quality of the studies and the differences between the protocols do not justify recommendations on any type of pushing (25) .
Professionals justify their guiding long and directed pushing with the purpose of shortening the expulsive period (3) . However, the systematic review that investigated the interference of directed pushing in the expulsive period concluded that this action resulted in no effect in the length of this period and in the rates of perineal trauma when compared to spontaneous pushing (26) . Therefore, the woman should be encouraged to follow her own impulses.
Regarding the Kristeller maneuver, an observational study was performed in Egypt with 8,097 women undergoing normal labor to verify the effects of this maneuver. The authors found that, despite the shorter duration of the second period, there was a significant increase in the risk of severe perineal lacerations, uterine rupture, dyspareunia and urinary incontinence 6 months postpartum. In the NB, there were shoulder dystocia, increased risk of Apgar scores below seven in the fifth minute, fetal sequelae such as hypoperfusion and cerebral palsy (27) .
The practices and maternal outcomes identified in the interviews with professionals showed that, after the intervention, all of them reported "always or most of the times" to allow women to choose their companions, but this result had was not statistically significant, because almost all professionals already did so before the intervention. Similar results were found in the interviews of the women. Thus, it can be inferred that the educational intervention improved this practice, ratifying the maintenance of scientific evidence.
Regarding amniotomy, although there was a decrease in the number of professionals who reported they "always or most of the times" adopt this procedure, the result was not significant. This finding differs from that obtained in the interviews with the puerperal women, probably due to the small number of professionals interviewed.
It is worth mentioning that, because amnioscopes were not available in the studied scenario, amniotomy is used as a method to assess the appearance of amniotic fluid, which may have contributed to the maintenance of high rates and the difficulty in changing the care practice. Administration of oxytocin, whether or not associated with early amniotomy, should not be routinely performed in cases of labor with good progression (1) .
the higher dosage of the drug significantly reduced the duration of delivery (± 3.5h) and cesarean rates, and increased vaginal delivery. The study concluded that there is insufficient evidence to recommend the use of high doses of oxytocin in women with slow progression of labor, and recommended further research to evaluate this effect (29) .
There was a non-significant reduction in orienting or encouraging directed pushing and Kristeller maneuver after the educational intervention. The reduction of these interventions, although satisfactory, did not allow the audit criterion initially defined to be achieved. After the educational intervention, the record of the presence of companions increased, although without statistical significance, probably because this information was still absent in more than 90% of the medical records. On the other hand, all the professionals interviewed in the phase 3 reported that they "always or most of the times" allow the presence of companions, and more than 90% of the puerperae reported having been allowed to have a companion during labor.
As for the practice of amniotomy, it was found that there was a decrease of records in the medical charts after the intervention, with a significant difference. This result is similar to the reports of puerperae, but differs from those of the professionals. It is worth clarifying that only when the amniotic sac was described as intact in the medical record at the moment of admission without reference to artificial rupture until delivery is that nonrealization of amniotomy was noted. Furthermore, in the post-intervention audit, the missing records of this practice more than doubled compared to the baseline audit.
The registration of membrane integrity is an important factor in the evolution of labor, in contrast with the negative repercussions of artificial rupture on maternal and fetal health. There is evidence that undesirable effects result from this intervention, including increased early deceleration of fetal heart rate and a higher risk of fetal and puerperal infection (1) .
Thus, it is essential to encourage the recording of this information in the medical record.
Regarding the positions adopted in the expulsive period, the records in the charts showed that, after the educational intervention, there was a decrease in the lithotomy position, while upright/squatting position increased, with a significant difference. This finding corroborates the reports of the puerperal women, but not those of the professionals, probably, due to the small number of professionals interviewed.
The records in the medical records concerning the prescription of oxytocin during labor revealed that this practice was proportionally reduced after the educational intervention, but without statistical difference. In the post-intervention audit, it was found that the use of oxytocin in the period of dilatation was less frequency in all sources of data audited, i.e., interviews with puerperal women, professionals and medical records. It should be noted that for this variable, the information was available in the medical records.
These findings are in line with a Palestinian study that used the methodology to implement better evidence in normal childbirth care to investigate possible changes in practices adopted by professionals. The authors demonstrated a lower frequency of some important practices, including liberal use of oxytocin and artificial rupture of membranes after the intervention, with statistically significant difference (30) .
The absence of pre-defined audit criteria in the JBI for the implemented practices was an important limitation in this study. Other limitations were the high frequency of missing data on the practices in the medical records, the non-randomized collection of puerperae to be interviewed, and the difficulty to recruit professionals.
Our findings not only bring contributions to the knowledge of professionals who assist labor and delivery, but also eliminate empirical, routine and unnecessary care measures for parturient women and improve clinical practice.

Conclusion
Our results allow us to infer that the methodology