sented with a potential delivery at the very edge of viability. However because of our limited ability to diagnose clearly the preterm labour caused by an infection ...
c o R R E s P O N D E N c E 1055 membranes, the cervix was 3-4 cm dilated and there was some watery fluid in the posterior fomix. It is of course possible to have a 'hindwater leak' but there was no furthercharacterisation of this fluid nor ultrasound evidence of diminished or absent liquor. Initial vaginal swab and urine cultures were negative and her C-reactive protein (CRP) and white cell count were normal. Most women with chorioamnionitis have raised CRP and maternal CRP concentration correlates well with histological chorioamnionitis inpatients with pPROM (sensitivity 88% and specificity 96%)2.While some women may have chorioamnionitis with minimal clinical signs, a policy of prolonged antibiotic and steroid therapy at 23 weeks of gestation is questionable. It is unclear whether the antibiotics were instituted as treatment or prophylaxis. We presume treatment, since it would be unusual to give intravenous antibiotics as prophylaxis for pPROM, the diagnosis of which was patently insecure and based on soft clinical evidence. The meta-analysis of pPROM trials quoted by Kyle and Turner found no benefit of antimicrobial therapy as regards perinatal mortality3 despite prolongation of the latent period and reduction of neonatal infectious morbidity. Superinfection by resistant organisms was suggested as a possible explanation3.The best antibiotic, optimal route, dosage and duration of therapy are as yet unknown. The practice of intermittent speculum examinations and vaginal swabbing for culture is widespread despite the risk of introducing lower bowel organisms into the uterus. Recent evidence shows a poor correlation between lower genital tract cultures and organisms implicated in chorioamnionitis". It has been suggested that amniocentesis or cordocente~is~ might be considered to evaluate pPROM cases where infection is suspected. Although Pseudomonas was isolated from the vagina and the placenta in the case reported, this does not equate to histologic chorioamnionitis. Moreover no mention of a postmortem examination or refusal of it was made. We also question the practice of nursing patients with bulging membranes and presumed pPROM with the head tilted down. While it may seem logical to remove pressure from the cervix it also potentially creates a stagnant pool of a rich culture medium in the posterior fomix in which pathogens can thrive. This case report illustrates the difficulties we all face when presented with a potential delivery at the very edge of viability. However because of our limited ability to diagnose clearly the preterm labour caused by an infection and because of the risks of superinfection, routine antibiotic therapy in the absence of clinically apparent infection (especially at 23 weeks of gestation or less) is unjustified. We agree that the prolonged antibiotic therapy, the weekly administration of steroids commenced at 23 weeks of gestation and the intermittent speculum examinations combined to result in the overwhelming nosocomial infection suffered by this woman and her baby.
*Austin H. N. Ugwumadu, *Isaac T. Manyonda & **Phillip E. Hay Departments of *Obstetrics and Gynaecologyand **Genitourinary Medicine, St George k Hospital Medical School, London
References Kyle P, Turner DPJ. Chorioamnionitis due to Pseudomonas aeruginosa: a complication of prolonged antibiotic therapy for premature rupture of membranes. Br JObstet Gynaecoll996; 103: 181-182. Crowley P. Antibiotics for preterm prelabour rupture of membranes. In: Enkin MW, Keirse MJNC, Renfrew MJ, Neilson JP, editors. Pregnancy and ChildbirthModule. Cochrane Database of Systematic Reviews, 1993: review no. 04391. Hawrylyshyn P, Bernstein P, Milligan JE et al. Premature rupture membranes: The role of C-reactive protein in the prediction of chorioamnionitis. Am J Obstet Gynecoll983;147: 240-246. Carroll SG, Papaioannou S, Ntumazah IL, Philpott-Howard J, Nicolaides KH. Lower genital tract swabs in the prediction of intrauterine infection in preterm prelabour rupture of membranes. Br JObstet Gynaecoll996: 103; 54-59. Sir, I read with interest the article by Kyle and Turner (Vol 103, February 1996)' regarding a case of Pseudomonas ueruginosa causing
0 RCOG 1996 Br J Obstet Gynaecol 103, 1049-1056
chorioamnionitis. As stated in their article, Pseudomonas is ubiquitous in nature and colonisation is increased by hospitalisation and antibiotic therapy. Acetic acid in a 0 5 % to 5% solution has been used to eliminate Pseudomonas from bums and soft tissue wounds2. The use of acetic acid 1.0 treat wound infections is not a new idea, but it is simple and inexpensive. Recently, I had a high risk patient who had a previous stillbirth and neonatal death at 22 and 24 weeks, respectively. The first pregnancy was complicated by ruptured membranes and chorioamnionitis and the second pregnancy by definite cervical incompetence. As a result a cervical suture was inserted in the third pregnancy and at 18 weeks she complained of a yellow-green discharge. A high vaginal swab showed Pseudomonas aeruginosa. She had not been on prophylactic antibiotics. As she was systemically well, I was reluctant to treat her with parenteral antibiotics. In view of her past history and the presence of a foreign body (i.e. cervical suture), I consulted an infectious diseases physician for advice. He mentioned the use of acetic acid on infected wourtds and suggested douching as a possible solution. This was carried out using a 3% solution, as used for colposcopy, and cleared the Pseudomonas after two applications. The pregnancy continued without further problems and the only other organism cultured throughout the pregnancy was Candida albicans. She had a vaginal delivery of a 2800 g baby girl at 37 weeks. In summary, the use of acetic acid vaginally for the treatment of Pseu,domonas associated vaginitis is reported. Whether this would be appropriate in the presence of ruptured membranes is unclear, but this information may be of use to colleagues faced with a symptomatic vaginal discharge caused by Pseudomonas. This may be of particular interest if the incidence of superinfection increases with greater use of prophylactic antibiotics as predicted by Kyle and Turner.
Stephen G. Cook Wesley Medical Centre, Auchenfowel;Brisbane, Australia
References Kyle P, Turner OPJ. Chorioamnionitis due to Pseudomonas aeruginosa: a complication of prolonged antibiotic therapy for premature rupture ofmembranes. BrJObstet Gynaecol1995;103: 181-183. 2 Sloss JM, Cumberland N, Milner SM. Acetic Acid used for the elimination of Pseudomonas aeruginosa from burn and soft tissue wounds. J Royal Army Medical Corps 1993; 139: 49-5 1. 3 Taylor K. Treatment of Bacillus pyocyaneus infection. JAMA 1916; 67: 1598-1 599. 1
Sir, We thank both correspondents for their interest in our case report and the useful comments provided. In response to Ugwumadu et al. we should explain more clearly that intravenous antibiotics were commenced as initial treatment because of the patient's initial presentation of feeling unwell, flushed, temperature 37.8"C, and a purulent vaginal discharge. We were very surprised when the patient improved clinically and all the preliminary investigations used to diagnose chorioamnionitis returned negative. Nevertheless, to be cautious, it was thought appropriate to complete an oral course of antibiotics as a prophylactic measure. It was at this stage that the steroids and 'head-down tilt' position were introduced now that the probability of increasing gestation was much greater. We acknowledge that the use of the 'head-down tilt' position in a woman with a dilated cervix and bulging membranes in an attempt to prevent passive cervical dilatation by an anti-gravity position has never been put to the test of a randomised trial. Furthermore, these authors raise a pertinent point that, potentially, the intervention could be harmful if a stagnant pool of fluid, secondary to spontaneous rupture of membranes (SROM), sits in the posterior fornix. The final overwhelming Pseudomonas infection was confirmed at postmortem examination. Histology of the placenta and membranes showed inflammatory chartges consistent with acute chorioamnionitis and microscopy of all the .fetal organs showed gram-negative micro-organism mfiltration. Fluid aspirated from the fetal skin pustules grew I? aeruginosa.
1056 c o R R E s P O N D E N C E That eradication of associated with a foreign body using topical 3% acetic acid is possible is an interesting comment from Cook but it is unlikely that such an intervention would be useful in cases such as ours, because not only was there no cervical suture, but also because the interval between detection of asymptomatic colonisation and overwhelming sepsis was so short. Furthermore, in cases of suspected SROM, vaginal ‘douches’ could promote ascending fetal infection. We again emphasise that the management dilemmas faced with SROM, potential chorioamnionitisand preterm labour at the limits of viability are difficult and this case exhibits one complication associated with prolonged antibiotic therapy.At present none of the trials of prophylactic antibiotic treatment for preterm rupture of membranes, and their meta-analyses, show an improvement in perinatal survival and therefore further large randomised clinical trials are required.
Phillippa Kyle Queen Charlotte S and Chelsea Hospital, London
David Turner Norfolk and Nonvich Hospital, Nonvich
%in to twin blood transfusion in a dichorionic pregnancy without the oligohydramniospolyhydramnios sequence Sir,
King et al.’ reported in this journal transfusion of blood from one twin to the other in a dichorionicpregnancy without the oligohydramnios-polyhydramnios sequence. We have now seen a dichorionic twin pregnancy with the oligohydramnios-polyhydramnios sequence. A 34 year old woman with a previous obstetric history of six uneventful term pregnancies was referred at 18 weeks of gestation with a twin pregnancy because of oligohydramnios in one sac. On detailed ultrasound examination, oligohydramnios in the sac of fetus 1 with polyhydramnios in the sac of fetus 2 were noted. Biometry agreed with her menstrual dates and fetal anatomy and umbilical artery pulsatility index were normal in both babies. Twin-
to-twin transfusion syndrome was suspected. At 21 weeks of gestation the woman presented with a small amount of painless vaginal bleeding and, on ultrasound examination, although the measurements were within the normal ranges, there was now an 18% difference in the estimated fetal weights. There was almost no amniotic fluid in sac 1 (donor) and the fetus was ‘stuck’ to the placenta. There was severe polyhydramnios in sac 2 (recipient). The bleeding settled within 24 hand the woman was dischargedhome. One week later she was readmitted with a history of ruptured membranes. The following day she lost large amounts of amniotic fluid and developed clinical features of chorioamnionitis. She went into spontaneous labour and delivered stillborn twin boys. Permission for a postmortem examination was declined. The babies weighed 540 g and 620 g, respectively, and no external anomalies were seen. The placenta was single (diamniotic, dichorionic) and weighed 371 g. The naked-eye appearances suggested chorioamnionitis. Both umbilical cords had three vessels and their placental insertions were eccentric. Examination of the vessels of the chorionic plate showed vascular anastomoses between the two circulations. This case confirms the previous report of functional vascular anastomoses in the placentas of dichorionic twins. However, unlike King et al.’, the vascular connections in this case were associated with the polyhydramnios-oligohydramniossequence. These findings add suipport to the notion that zygozity rather than chorionicity is important in the clinical management of twin pregnancy.
Juan 4;. Rodriguez, Helen Porter, Gordon M. Stirrat & Peter W. Soothill Fetal Medicine Research Unit, University of Bristol. St Michael S Hospital, Bristol
Reference 1
King AD, Soothill PW, Montemagno R, Young MP, Sams V, Rodeck CH. Twin to twin blood transfusion in a dichorionic pregnancy without the oligohydramnios-polyhydramnios sequence. Br J Obstet Gynaecol1995; 102: 334335.
0 RCOG 1996 Br J Obstet Gynaecol 103, 1049-1056
