This retrospective study compared clinical course and outcome of PPROM cases managed expectantly with prophylactic antibiotics and antenatal corticosteroids with those without prophylactic antibiotics and antenatal corticosteroids. Latency period was significantly longer in PPROM cases managed expectantly with prophylactic antibiotics and antenatal corticosteroids. Neonatal infectious morbidity was significantly lesser in cases managed expectantly with prophylactic antibiotics and antenatal corticosteroids. But maternal infectious morbidity was comparable. This finding supports the benefit of prophylactic antibiotics and antenatal corticosteroids in PPROM as shown in previous studies [8, 9].
Since 1998, cases diagnosed as PPROM in our institute were managed expectantly as per recommendation from ACOG [10, 11]. A 7-day course of parenteral ampicillin and oral therapy with amoxicillin and erythromycin was used during expectant management of PPROM hoping to prolong pregnancy and to reduce infectious and gestational age-dependent neonatal morbidity [9–11].
There have been several studies evaluating prophylactic antibiotics in PPROM [13–15]. Magwali TL et al. used co-amoxiclav in their study and found that antibiotics could prolong latency period and decreased neonatal and maternal morbidity due to sepsis . Ryo E et al. used imipenem/cilastatin sodium in their study and found that imipenem/cilastatin sodium could prolong the latency period . August Fuhr N et al. used mezlozillin in their study and found that antibiotics could prolong latency period and reduced neonatal infectious morbidity . However, there has been no study evaluating the outcome of PPROM after the recommendation to use prophylactic ampicillin and erythromycin.
The usual outcome of PPROM is labor. The latency period was longer in the study group than in the control group (3.7 vs. 1.0 days, P < 0.001) which was similar to previously studies [8, 9]. Mercer et al. performed a RCT study evaluating intravenous ampicillin and erythromycin for 48 hours followed by oral amoxicillin and erythromycin base for 5 days vs. placebo in PPROM. They found that antibiotics in the PPROM group had a longer median time to delivery than the placebo group (6.1 vs. 2.9 days, P < 0.001) . Kenyon et al. performed a RCT evaluating erythromycin, co-amoxiclav, both, or placebo given four times daily for 10 days or until delivery. They found that the use of erythromycin was associated with prolongation of pregnancy in PPROM . This suggests that antibiotics can suppress or prevent clinically significant intrauterine infection and shorten latency.
In present study, we found a significant increased number of patients who did not deliver within 48 hours and 7 days in study group when compared with a control group (64.7% vs. 31.4%, P < 0.001 and 29.4% vs. 7.8%, P = 0.002, respectively). This was similar to previous reports [8, 9]. Mercer et al. found a significantly decreased number of women assigned to antibiotics compared with placebo delivered within 48 hours (27.3% vs. 36.6%, P = 0.03) and delivered within 7 days (55.5% vs. 73.5%, P = 0.001) . Kenyon et al. demonstrated significantly fewer women on erythromycin alone delivered within 48 hours than did those on placebo (30.5% vs. 40.7%, P < 0.0001) .
The maternal infectious morbidity such as chorioamnionitis, metritis and wound infection was not different between groups in the present study. This is in contrast to previous studies [8, 9]. Mercer et al.  found that the antibiotic group had a lower incidence of clinical amnionitis when compared with placebo (23.0% vs. 32.5%; P = 0.01). But, the incidence of postpartum endometritis was similar regardless of antibiotic treatment (11.0% vs. 11.5%; P = 0.85). Kenyon et al.  found that the use of both erythromycin and co-amoxiclav (5.0% vs. 8.4%, P = 0.001) and any antibiotic (6.2% vs. 8.4%, P = 0.008) was associated with significantly less uterine infection than use of placebo.
Neonatal infectious morbidity was significantly decreased in the antibiotic group in the present study. Mercer et al. found that maternal antibiotic therapy was associated with reductions in the incidence of neonatal pneumonia and sepsis in the group B streptococcus negative cohort . A broad spectrum of aerobic and anaerobic bacteria and mycoplasmas have been implicated as causative agents for intrauterine infection in PPROM [16, 17]. Thus, broad spectrum antibiotics (ampicillin plus erythromycin) are more beneficial in this condition.
The limitation of this study was retrospective study. We cannot control the confounding factors. Thus, bias may have been introduced. However, the demographic characteristics between groups in present study were not different. Further RCT comparing different regimens of antibiotics should be conducted.