Eligibility criteria

The review scope was defined using an adapted setting, perspective, intervention, comparison, and evaluation (SPICE) framework [22]. We sought the perspective (P) of women giving birth, their partners and families, and healthcare providers in healthcare facilities globally (S). The interventions (I) were the use of antibiotics and antiseptics for prevention of infection during labour and birth, there was no comparison (C), and we were specifically interested in their perspectives and experiences on factors affecting use of the interventions (E) [23].

Primary qualitative, quantitative, and mixed-methods studies were eligible. For the qualitative component, studies that used both a qualitative data collection method (e.g. focus group discussions, individual interviews, observation, diaries, document analysis, open-ended survey questions) and qualitative data analysis (e.g. thematic analysis, framework analysis, grounded theory) were eligible. For the quantitative component, primary studies using an observational or interventional design (including randomised controlled trials, cohort studies, cross-sectional studies) were eligible. Mixed-methods studies were also eligible. We excluded other article types such as case reports, case series, letters, editorials, commentaries, reviews, study protocols, and conference abstracts. One study has been classified as “awaiting classification”, as no full text could be retrieved.

There was no restriction on language or country of publication. Full texts of studies published in languages other than English were translated using freely available online software (Google Translate). We included studies pertaining to any level of healthcare facility (e.g. hospitals, clinics, and primary healthcare settings). The timeframe of interest was the time from admission for childbirth until the woman’s discharge from the facility, i.e. the peripartum period during which prophylactic antibiotics or antiseptic agents would be administered to a woman by a healthcare provider.

We excluded studies on the clinical assessment, diagnosis or treatment of bacterial peripartum infections or their complications, and those reporting solely on the effectiveness, prevalence or extent of use of the specified interventions. We also did not consider other types of infection (such as viral or parasitic infections).

Findings on use of prophylactic antibiotics to prevent peripartum infections

All relevant qualitative data from 6 studies reflected the perspectives of healthcare providers. Thirteen descriptive themes were identified, grouped under four second-order themes: (1) provider beliefs about benefits and harms; (2) provider perceptions of infection risk; (3) provider preferences regarding prophylactic antibiotic regimens and administration; and (4) other factors influencing provider decision-making on prophylactic antibiotic use. Only one study reported the perspectives from pregnant women, thus narratively described below [54]. Table 2 presents the summary of qualitative findings and GRADE-CERQual assessments. Findings from quantitative evidence were mapped to the qualitative findings and are presented in Table 3. The full GRADE-CERQual evidence profile is available at S3 Table.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Summary of qualitative findings on perspectives and experiences of healthcare providers on use of peripartum prophylactic antibiotic.

https://doi.org/10.1371/journal.pone.0272982.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Summary of findings from quantitative evidence on perspectives and experiences of women and healthcare providers on use of peripartum prophylactic antibiotics.

https://doi.org/10.1371/journal.pone.0272982.t003

Provider beliefs about benefits and harms of prophylactic antibiotic use. Qualitative research found providers had mixed views on whether prophylactic antibiotics are effective and beneficial for preventing infection (low confidence) [20, 37, 44]. Provider views varied regarding the indications for use. Some routinely used antibiotic prophylaxis only for women considered high-risk, such as women undergoing emergency caesarean section or if post-operative complications occurred. Others used prophylactic antibiotics routinely for all women undergoing caesarean section (very low confidence) [20, 44]. Some providers were concerned about unnecessary antibiotic use due to the potential for unwanted side effects, overtreatment and medicalisation of birth, while others considered the risk of adverse reactions to be low, and outweighed by the risk of harm due to infection (low confidence) [20, 42, 44]. Providers had varying levels of concern about antimicrobial resistance—some prescribe less antibiotics for this reason, while others did not consider it a threat and have not changed their prescribing practices (low confidence) [20, 37, 42].

Analysis of quantitative evidence similarly found that providers weighed various benefits and risks in deciding whether to use prophylactic antibiotics [20, 43, 51]. Many had a positive attitude toward administering prophylactic antibiotics [20, 43]. However, some reported that they were more likely to administer prophylactic antibiotics for emergency caesarean section than elective caesarean section [50]. Some providers in high-income countries (USA and Canada) believed that benefits of prophylactic antibiotics outweigh its risks, while some were concerned about the impact of antibiotic use on neonatal outcomes [43, 51].

Provider perceptions of infection risk. Qualitative evidence indicated that providers may be motivated by a fear of post-operative infection and the risk of resulting blame and damage to their professional reputation, leading to a belief that erring on the side of overtreatment is preferable (very low confidence) [20, 44]. The risk of infection, and therefore the need for antibiotic prophylaxis, was considered to vary depending on environmental factors, such as local infection rates and whether adequate infection control measures were in place at their facility (low confidence) [37, 44]. Surveys of providers in high-income countries (Israel and USA) found that sometimes concerns about medico-legal risk motivated decisions to adopt particular protocols and practices for antibiotic use, supporting the qualitative finding regarding fear of blame and reputational damage [41, 52].

Provider preferences regarding prophylactic antibiotic regimens and administration. Qualitative evidence found that providers had variable preferences regarding prophylactic antibiotic regimens. Their choice of antibiotic was affected by whether it was recommended or common practice, as well as perceptions of its effectiveness relative to other options (low confidence) [40, 44]. In deciding how many doses to prescribe, providers were influenced by locally recommended practices and personal experience. Some believed multiple dose regimens are more effective despite evidence of the effectiveness of single dose regimens (very low confidence) [20, 44]. Some providers reportedly commenced prophylactic antibiotic administration after clamping the umbilical cord for caesarean section. Reasons included to avoid passing antimicrobial agents to the baby, and to mitigate increased infection risk arising from complications or potential contamination during surgery (low confidence) [40, 44].

Quantitative evidence suggests that in practice many providers’ choice of an antibiotic is based on drug availability [38, 39, 52]. Quantitative studies indicated a variety of preferences for timing of antibiotic administration for caesarean section—some providers in France and the USA preferred pre-incision prophylaxis [47, 48, 51], while in Nigeria some preferred intra-operative administration [38], and in Thailand administration after umbilical cord-clamping [20, 44]. These preferences also varied by type of providers—paediatricians and anaesthetists were reported to prefer pre-incision prophylaxis [47, 48], yet obstetricians were reported to prefer administration after cord clamping [20]. Factors underpinning this decision included risk of maternal anaphylactic shock and the impact on newborns’ bacteriological samples and need for antibiotic therapy [48]. Some providers considered single-dose administration to be not cost-effective [20]. Providers also believed administrating antibiotics is an obstetric task and not an anaesthetist’s responsibility [45, 47], which demonstrates that lack of clarity on decision making responsibility may impact provider decision making.

Other factors influencing provider decision-making on prophylactic antibiotic use. Qualitative studies found that providers may consider cost-effectiveness for the health facility and affordability for the patient in making decisions about whether to use prophylactic antibiotics, and what agent to use (low confidence) [20, 44, 49]. Obstetricians and obstetric unit staff in questioned whether international effectiveness evidence regarding prophylactic antibiotics is applicable to their local setting, and expressed a preference for evidence from local trials (low confidence) [20, 37]. This complements the finding that infection risk is perceived to vary depending on the environment, informing providers’ perception of localised costs and benefits of antibiotics.

Providers reported their decision-making about prophylactic antibiotic use is informed by a range of written reference materials. There were mixed views on the usefulness and uptake of guidelines–for example, providers in Thailand expressed a preference for textbooks over journals (low confidence) [20, 44, 49]. Nurses in Ghana raised concerns that guidelines were not implemented in practice [49]. Thai obstetricians’ prescribing practices were highly influenced by professional norms and expectations, including pressure from colleagues and the observed practice of supervisors (very low confidence) [20, 44]. This was related to their fear of blame for adverse events but also reflected respect for supervisors’ knowledge and expertise.

Quantitative studies found providers were influenced by guidelines, regulations, journals, textbooks, teaching curriculums, and hospital policy [20, 39, 41, 52]. Despite guidelines and protocols existing at national or facility level, however, not all providers were aware of evidence regarding prophylactic antibiotic use. Those who were aware may not use antibiotics correctly in practice due to perceived inadequacy of evidence, doubts about benefits, lack of training, and absence of local guidelines or protocols regarding its use [38, 51]. Providers had mixed views on the usefulness and uptake of guidelines and policies [20, 39, 41, 52]. Similar to qualitative evidence, providers’ decisions regarding antibiotic prophylaxis are influenced to some degree by the views of others, including supervisors, specialists, senior and same-level colleagues. Providers consider the cost-effectiveness to some degree in administering antibiotics [20, 43].

Factors influencing provider decisions to use prophylactic antibiotics at birth

Findings from qualitative and quantitative evidence suggest providers’ decisions about whether and how to use prophylactic antibiotics at birth are complex and based on explicit or implicit consideration of a range of factors. We developed a framework of those factors affecting provider’s use of prophylactic antibiotics at birth using COM-B (Fig 2). We mapped factors under physical and psychological capability (Capability domain), physical and social opportunity (Opportunity domain), and how the interaction between these domains can influence provider motivation towards the behaviour of interest, i.e. appropriate use of peripartum prophylactic antibiotics. That is, when providers have improved awareness, skills and experience around correct prophylactic antibiotic use, we would expect positive changes to provider motivation in using antibiotics appropriately. Aligning social factors (such as peers, superiors and professional groups supportive of good prescribing practice) and physical factors (such as the clinical environment, as well as the availability of guidelines, policies and medicines) can also benefit motivation. These Capability and Opportunity domains can affect provider’s motivations, such as their attitudes, fears and beliefs around prophylactic antibiotic use.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Factors affecting peripartum prophylactic antibiotic prescribing behaviour.

https://doi.org/10.1371/journal.pone.0272982.g002

Women’s perceptions of antibiotics use during caesarean section. Only one qualitative study reported the perspectives of Danish pregnant women on antibiotics during caesarean section [54]. Overall, women’s decisions on whether or when to receive antibiotics were related to concerns about the wellbeing of her baby and herself. Women had varied opinions on when they preferred to receive them—some women were concerned about possible impacts on their baby, the lack of scientific evidence around antibiotics, and they perceived most infections to not be serious. Some preferred receiving antibiotics after cord clamping, or not at all, to avoid risk to their baby. Other women, however, preferred receiving antibiotics pre-caesarean to minimise the risks of harm to themselves and their baby, feeling that they need to be well in order to take care of their baby. Women also described having limited knowledge about prophylactic antibiotics during caesarean section and desired more information. Many women reported trusting their healthcare providers’ judgement, even if it differed to their preference.

Findings on use of antiseptic agents to prevent peripartum infections

Only three studies (one qualitative, two quantitative) considered use of antiseptic agents for vaginal cleansing and surgical skin preparation [19, 38, 55]. As too few studies were available for a meaningful synthesis, findings are reported narratively.

Weckesser et al explored women’s perspectives on caesarean section recovery and experiences of infection prevention in conjunction with the PREPS trial of vaginal cleansing with chlorhexidine immediately before caesarean section in the UK [19]. Prior to the trial, women expressed confusion about the purpose of vaginal cleansing with antiseptic agents. Once the rationale of reducing infection (endometritis) was explained, women perceived vaginal cleansing positively as an “upgrade” to standard practice. Women also considered that a detailed explanation of what the procedure involved in advance would likely be important to ensure its acceptability. One study from Nigeria considered the perspective of healthcare providers on antiseptic use for skin preparation before caesarean section [38]. A preference for specific antiseptic agents was due to health providers’ beliefs about its benefits, and some influence of guidelines. Health providers’ non-adherence to antiseptic use guidelines was reportedly due to lack of supervision, training, inadequate supplies, absence of facility-level policies or protocols to help implement guidelines, doubt about benefits, perceived lack of clinical evidence, and lack of examples or directives from senior colleagues. Changing practice to adhere to guidelines was attributed to experience with infection cases, medico-legal events, and provider’s change of beliefs about a specific regimen’s effectiveness. Jakes et al conducted an implementation study on vaginal preparation for women undergoing caesarean section, during which 20 women completed a questionnaire on their experience [55]. No women reported abnormal or discoloured vaginal discharge, vaginal irritation, pain, or concerns about discolouration of the baby’s scalp. Only one woman declined vaginal preparation during the implementation cycle.

Strengths and limitations

This is the first systematic review of currently available evidence on how women, partners, families and providers perceive and experience the use of antibiotic antiseptic agents for infection prevention at birth. Strengths of this review include a comprehensive search strategy, adherence to a pre-specified review protocol (including duplicate screening, extraction, critical appraisal and GRADE-CERQual assessments), as well as combining evidence from qualitative and quantitative data. However, the modest number of eligible studies limited our ability to draw strong conclusions. Furthermore, some studies had serious limitations regarding adequacy of evidence and relevance to global settings, resulting in low to very low confidence assessments. While quantitative evidence broadly supported the qualitative findings, the overall evidence base remains relatively limited and further research is required.

Implications for practice, policy and research

In order to prevent death and disease due to peripartum infections, evidence-based guidelines are needed to optimise the use of prophylactic antibiotics and antiseptics around the time of childbirth. Consideration of factors we identified, like provider capability (such as their skills, experience and knowledge), their motivations and their clinical environments, are needed for to optimise strategies to improve prophylactic antibiotic use. These findings are useful in developing evidence-based guidelines, particularly in understanding stakeholder’s views, acceptability, feasibility and implementability of an intervention [33, 35]. For example, findings from this review have informed forthcoming updates of living WHO recommendations related to peripartum antibiotic and antiseptic use [64].

However, this review emphasises that our understanding of how women, their partners and families perceive use of antimicrobial agents in the peripartum period is limited. This gap should be addressed to ensure that women’s voices are included in how maternity care is delivered. Additional research is also needed to better understand how providers balance consideration of infection risk, the side effects of antibiotic use, and antimicrobial resistance as both a patient-specific and public health concern. There is a need for greater understanding of providers’ attitudes towards guidelines on antibiotic use, and barriers to guideline implementation in limited-resource settings, noting that most studies were from high-resource settings. With only three studies identified on antiseptic agents, further research on this intervention is also a priority. An improved evidence base will provide researchers and policymakers with further insights regarding why antibiotics and antiseptics may be misused in some settings, and inform the development of more effective implementation strategies to address these issues.