Maternal and Perinatal Outcomes Associated with Intrapartum Antibiotic Regimens in Women with Prolonged Membrane Rupture and Unknown Group B Streptococcus Status: A Retrospective Comparative Study Open Access

The duration between rupture of membranes (ROM) and delivery, known as the latency period, has been shown to be positively related to the risk of developing chorioamnionitis [1]. This risk is compounded by factors such as nulliparity [2], the presence of bacterial vaginosis [3], group B Streptococcus (GBS) colonization [4], and maternal conditions such as obesity [5].

For prevention of GBS-related early onset sepsis (EOS), the American College of Obstetrics and Gynecology (ACOG) recommends intravenous penicillin as intrapartum antibiotic prophylaxis (IAP), and intravenous ampicillin as an acceptable alternative [6]. For women with low-risk penicillin allergies, cefazolin is preferred. Current guidelines recommend penicillin allergy testing as many childhood allergies either resolve or were never true allergies. When the risk of anaphylaxis is high, clindamycin can be used, but only if the GBS isolate is confirmed as sensitive to clindamycin [6]. However, concerns have been raised about increasing GBS resistance to clindamycin [7]. Studies from the USA reported resistance rates of up to 33% [7]. In a multicenter study, EOS rates were higher among women who received second-line (macrolides, clindamycin, or vancomycin) rather than first-line IAP (ampicillin or penicillin); however, maternal and microbiological outcomes were not addressed [8].

When GBS colonization status is unknown, antibiotic sensitivities are also unknown, and IAP is indicated in women with risk factors for GBS EOS, such as those with prolonged ROM ≥18 h at term [6]. According to our department protocol, women with ROM ≥18 h who have unknown GBS colonization status and reported penicillin allergy receive clindamycin. In this study, we aimed to compare maternal, neonatal, and microbiological outcomes among women with prolonged ROM ≥18 h and unknown GBS colonization status according to the IAP regimen, ampicillin vs. clindamycin.

This retrospective study included women hospitalized in a tertiary university-affiliated hospital between March 2020 and May 2024. The study protocol was approved by the Ethics Committee of Galilee Medical Center, Nahariya, Israel (Approval No.: 0068-18-NHR). Informed consent was not obtained due to the retrospective design.

Included were singleton (≥37 weeks) deliveries with ROM ≥18 h at term and unknown GBS status. Exclusion criteria were known positive and negative GBS colonization, missing information such as unknown ROM time, intrauterine fetal death, and major congenital anomalies.

The diagnosis of ROM relied on clinical symptoms and physical examinations, such as a reported history of fluid leakage and visible pooling of amniotic fluid observed during a sterile speculum exam. If pooling was not apparent, the AmniSure immunochromatography test was employed to detect trace levels of placental alpha-microglobulin-1 protein in vaginal fluid to confirm ROM. Once confirmed, ROM onset was documented according to the patient’s account. According to our department protocol, women with prolonged ROM ≥18 h and unknown GBS colonization status received intrapartum ampicillin 2 gr q.i.d. Penicillin allergy was defined based on patients’ report. According to our protocol, women who reported penicillin allergy received clindamycin 900 mg t.i.d.

Intrapartum fever was defined as a single maternal oral temperature of ≥39.0°C or two measurements between 38.0°C and 38.9°C taken 30 min apart [9]. Clinical chorioamnionitis was diagnosed in the presence of intrapartum fever plus one of the following: leukocytosis>15,000/mm³, fetal tachycardia or foul-smelling amniotic fluid. Chorioamnionitis was treated by ampicillin 2 g q.i.d. and gentamicin 240 mg q.d. [9]. Postpartum endometritis was diagnosed based on body temperature ≥38°C in the absence of any other cause, together with an associated clinical finding such as pelvic tenderness, purulent lochia, tachycardia, or abdominal discomfort. At neonatal intensive care unit (NICU) admission, EOS workup consisted of a blood culture obtained from neonates with risk factors for EOS, according to the Kaiser Permanente electronic calculator [10]. EOS was defined as sepsis at 7 days from birth. Invasive ventilator support was defined as either conventional mechanical ventilation or high-frequency ventilation. Non-invasive ventilation was defined as either continuous positive airway pressure bubbles or nasal high-flow treatment.

The primary maternal outcome was the rate of clinical chorioamnionitis, compared between women treated with ampicillin vs. clindamycin. The primary neonatal outcome was the rate of NICU admission. Secondary maternal outcomes comprised intrapartum fever, maternal sepsis, delivery method, puerperal endometritis, antibiotics in the postpartum, and postpartum length of stay. Secondary neonatal outcomes were 5-min Apgar score of ≤7, umbilical cord pH ≤7.1, hypoxic brain injury, EOS, respiratory distress syndrome, meconium aspiration syndrome, ventilation support (invasive and non-invasive), and NICU length of stay.

Swabs of the chorioamniotic membrane were collected. Following delivery, the chorioamniotic space was accessed with sterile gloves and sampled with sterile swabs [11]. The microbiological results were assessed by an infectious disease specialist. Specific species such as Coagulase-negative staphylococci, Staphylococcus hominis, and diphtheroids, were classified as contaminants. Due to the laboratory’s routine practices, Mycoplasma spp. were not cultured and thus not reported. The results of the included cultures were categorized into six major groups: gram-negative Enterobacteriaceae, GBS, other Streptococci spp., anaerobes, enterococcus faecalis, and negative cultures.

At our tertiary medical center, the rate of clinical chorioamnionitis in women with prolonged ROM is 8.4% [12] Clindamycin use has been shown to be associated with an increased risk of clinical chorioamnionitis, with an adjusted odds ratio (aOR) of 5.09. Based on these data, and assuming a power of 80%, the required sample size was calculated as at least 1,039.

Continuous variables were compared using either the independent sample t test or the Mann-Whitney test. Categorical variables were analyzed using Pearson’s chi-squared test or Fisher’s exact test. Relative risks and their 95% confidence intervals were calculated for maternal and neonatal outcomes. Multiple comparisons of chorioamniotic swab results between the study groups were corrected using the Bonferroni correction. A multivariate logistic regression model aimed to predict clinical chorioamnionitis controlled for antibiotic regimen (clindamycin vs. ampicillin), nulliparity, cervical ripening by catheter balloon, induction of labor by oxytocin, artificial ROM, meconium staining, and delivery week ≥41. A multivariate logistic regression model aimed to predict NICU admission controlled for antibiotic regimen (clindamycin vs. ampicillin), nulliparity, cervical ripening by catheter balloon, induction of labor by oxytocin, artificial ROM, meconium staining, clinical chorioamnionitis, birthweight <2,500 g, delivery week ≥41, and umbilical cord pH ≤7.1. Statistical analysis was performed using IBM SPSS Statistics for Windows, version 27.0 (IBM Corp., Armonk, NY, USA).

Among 1,507 women with unknown GBS status and prolonged ROM ≥18 h, 1,418 were treated with ampicillin, while 89 reported penicillin allergy and were treated with clindamycin. The characteristics examined did not differ significantly between the groups (online suppl. Table 1; for all online suppl. material, see https://doi.org/10.1159/000546792).

The maternal outcomes are presented in Table 1. For the clindamycin compared to the ampicillin group, the rates were higher of intrapartum fever (28.1% vs. 4.1%, p < 0.001), clinical chorioamnionitis (14.6% vs. 2.3%, p < 0.001), and maternal sepsis (2.2% vs. 0.3%, p = 0.011). In both cases of maternal sepsis within the clindamycin group, GBS was identified as the causative organism. No GBS sepsis was identified in the ampicillin group. Cesarean delivery, puerperal endometritis, and postpartum antibiotic administration were also higher among the women who received clindamycin. There was no significant difference in hospitalization length of stay.

Neonatal outcomes are shown in Table 1. For the clindamycin compared to the ampicillin group, the rates were higher of 5-min Apgar scores ≤7 and umbilical cord pH ≤7.1. NICU admissions were also more frequent following clindamycin use, as were the instances of respiratory distress syndrome, antibiotic administration in the NICU, and both invasive and non-invasive ventilation. Birthweight, sepsis workup, and NICU hospitalization length showed no significant differences between the two groups. Four neonates were diagnosed with brain hypoxic injury: two had received clindamycin and two had received ampicillin. Both who received clindamycin and one of those who received ampicillin were diagnosed with chorioamnionitis. The other woman who received ampicillin was diagnosed with uterine rupture.

The bacterial distributions of the chorioamniotic swab cultures according to the IAP regimen are shown in Figure 1. GBS-positive chorioamniotic swabs were more prevalent following clindamycin than ampicillin treatment (19.1% vs. 1.1%, p < 0.001). Although no significant differences were observed for other bacterial categories, the incidence of positive cultures was higher in the clindamycin than the ampicillin group (p = 0.046).

Table 2 presents the results of the multivariate logistic regression model for predicting clinical chorioamnionitis and NICU admission. Clindamycin compared with ampicillin (aOR [p < 0.001], and artificial ROM (p = 0.031) were independently associated with clinical chorioamnionitis. Clindamycin compared with ampicillin (p < 0.001), the presence of meconium staining (p < 0.001), clinical chorioamnionitis (p = 0.009), and umbilical cord pH ≤7.1 (p < 0.001) were significantly associated with increased odds of NICU admission.

Among women at term with ROM ≥18 h and unknown GBS colonization status, clindamycin as a second-line therapy compared with first-line ampicillin was associated with worse maternal and neonatal outcomes. Rates were higher of clinical chorioamnionitis, maternal sepsis, intrapartum cesarean delivery, puerperal endometritis, Apgar score ≤7 at 5 min, umbilical cord pH ≤7.1, hypoxic brain injury, NICU admission, respiratory distress syndrome, and ventilation support. GBS isolates in chorioamniotic membrane cultures were more prevalent among those treated with clindamycin than with ampicillin.

We showed that maternal outcomes during delivery and the postpartum might be compromised in those who received second-line therapy as compared with first-line IAP. In their secondary analysis of a randomized trial, McCoy et al. [13] reported lower rates of clinical chorioamnionitis and peripartum infectious morbidity among women treated with IAP for GBS positivity than among those who were GBS negative and not treated. Although we included only women with unknown GBS status, the rate of GBS colonization reported in Israel is 21% [14]. We report a significant difference in the rate of maternal infections according to the IAP regimen. In our multivariate analysis, clindamycin compared to ampicillin was associated with a 7.7-fold higher rate of clinical chorioamnionitis. Similarly, in a meta-analysis of randomized controlled trials involving 7,671 patients with preterm premature ROM [15], following clindamycin compared to penicillin-based regimens or clindamycin in combination with gentamicin, the risk of maternal infections was higher. The higher rate of maternal GBS sepsis observed in the clindamycin group is concerning. A previous study [16]reported that GBS accounts for up to 21% of all cases of maternal sepsis, with prolonged ROM identified as a significant risk factor, underscoring the importance of appropriate antibiotic prophylaxis in this population. Our observation of a higher rate of cesarean delivery in the clindamycin group concurs with a previous study [17]. This might be related to increased rates of fetal distress due to intra-amniotic infection or dysfunctional labor caused by the infection’s effects on myometrial contractility [18].

The rate of brain hypoxic injury in the ampicillin group (0.2%) was similar to that reported in term deliveries [19]. The rate among those treated with clindamycin (2.2%) was higher. Higher rates of encephalopathy or the need for therapeutic hypothermia were reported among deliveries complicated by intrapartum fever. Accordingly, the combination of intrapartum fever and fetal acidosis was reported as particularly detrimental, with encephalopathy rates reaching 12.5% [18, 19]. Our sample size was underpowered to detect a difference in EOS rates. However, a retrospective study found similar rates of GBS EOS among women who were GBS-colonized, between those who received second-line antibiotics, including clindamycin, and those who did not receive antibiotics [8]. We report associations of clindamycin IAP with the incidence of respiratory distress syndrome, and with the use of both invasive and non-invasive ventilation. This contrasts with a previous finding of no significant differences between antibiotic regimens [8].

The rate of positive chorioamniotic swab cultures, and specifically GBS-positive cultures was higher in the clindamycin than the ampicillin group [19]. Among women who underwent an emergent cesarean section, we previously reported an association of positive uterine cultures with prolonged ROM ≥18 h [20]. Elsewhere, Enterobacteriaceae was identified as the leading pathogen in chorioamniotic swabs, following both ampicillin and clindamycin antibiotic regimens [12, 21]. The higher rates of GBS-positive isolates among women who received clindamycin reflect the increasing rate of clindamycin resistance among GBS isolates [4]. According to a 2019 CDC report, clindamycin-resistant GBS strains cause 40% of GBS-related infections [22].

Given the increased rates of maternal and neonatal complications associated with clindamycin IAP, efforts should be made to minimize its use. Only 3–7% of pregnant women with a self-reported penicillin allergy were reported to be confirmed as allergic upon testing [23]. Our findings support ACOG guidelines that recommend allergy testing to guide appropriate antibiotic selection. Cost-effectiveness analyses support this approach in pregnant patients [24]. First-generation cephalosporins may be safely used by those at low risk for anaphylaxis [6]. Additionally, universal GBS screening may reduce the need for IAP in women with prolonged ROM, and further decrease the reliance on clindamycin.

A strength of this study is the inclusion of chorioamniotic swab cultures, together with the results of maternal and neonatal outcomes. The study’s limitations include its retrospective design, single-center setup, relatively small number of women who received clindamycin, and lack of penicillin allergy skin testing. Studying other second-line antibiotics like cefazolin was beyond the scope of this study. Rapid intrapartum screening was not available at our institution. Combining rapid intrapartum screening with selective IAP was previously suggested as the most effective strategy to decrease GBS-related EOS [25]. Another limitation is that our results only apply to the specific population of women with ROM ≥18 h and unknown GBS colonization status.

In this retrospective cohort study of women with unknown GBS status and prolonged ROM, the prophylactic use of clindamycin compared to ampicillin was associated with higher rates of maternal infectious morbidity and adverse neonatal outcomes. These findings underscore the importance of minimizing clindamycin use when possible, particularly given concerns about GBS resistance.

The study protocol was reviewed and approved by the Ethics Committee of Galilee Medical Center, Nahariya, Israel (Approval No. 0068-18-NHR). Due to the retrospective nature of the study, the requirement for written informed consent was waived by the same Ethics Committee.

The authors have no conflicts of interest to declare.

This study was not supported by any sponsor or funder.

R.A.S.: Conceptualization, Data curation, Investigation, Writing – original draft, Project administration, Supervision, Writing – review and editing. S.O.: Methodology, Formal analysis, Software, Data curation, Writing – original draft, Writing – review and editing. G.G.: Conceptualization, Data curation, Resources, Visualization, Writing – review and editing. L.L.: Validation, Methodology, Formal analysis, Visualization, Writing – review and editing. M.F.W.: Conceptualization, Data curation, Investigation, Writing – original draft, Project administration, Supervision, Writing – review and editing.

The datasets generated and/or analyzed during the current study are not publicly available due to ethical restrictions but are available from the corresponding author on reasonable request.