- Email: [email protected]
Management of Preterm Premature Rupture of Membranes: A Comparison of Inpatient and Outpatient Care
OBSTETRICS
Management of Preterm Premature Rupture of Membranes: A Comparison of Inpatient and Outpatient Care Elisabeth Catt, MD, Rati Chadha, MD, FRCSC, FACOG, Selphee Tang, Elizabeth Palmquist, RN, MN, Ian Lange, MD, FRCSC Department of Obstetrics and Gynecology, University of Calgary, Calgary AB
Abstract
Résumé
Objectives: We sought to evaluate the safety of outpatient management of pregnancy complicated by preterm premature rupture of membranes (PPROM).
Objectifs : Nous avons cherché à évaluer la sûreté de la prise en charge ambulatoire de la grossesse compliquée par une rupture prématurée des membranes préterme (RPMP).
Study Design: We performed a retrospective cohort study of women with PPROM and a latency period of at least one week in one provincial health region between January 2007 and December 2012. We evaluated pregnancy outcomes for 133 women whose cases were managed using specialized community care and compared these with outcomes of a similar group of 122 women whose cases were managed entirely in the hospital. The primary outcome measured was the difference in the latency period between the two groups. For categorical variable outcomes, data were analyzed using chi-square tests, and continuous variable outcomes were compared using t tests.
Devis d’étude : Nous avons mené une étude de cohorte rétrospective auprès de femmes ayant connu une RPMP et une période de latence d’au moins une semaine au sein d’une région sanitaire provinciale entre janvier 2007 et décembre 2012. Nous avons évalué les issues de grossesse qu’ont connues 133 femmes ayant fait l’objet d’une prise en charge au moyen de soins communautaires spécialisés et les avons comparées aux issues qu’ont connues un groupe similaire de 122 femmes ayant fait l’objet d’une prise en charge entièrement mise en œuvre à l’hôpital. Le critère d’évaluation principal a été la différence en matière de période de latence entre les deux groupes. Les données relevant des variables nominales ont été analysées au moyen de tests de chi carré, tandis que les données relevant des variables continues ont été comparées au moyen de tests t.
Results: The median latency period for inpatients was 11 days compared with 18 days for patients in the community (P < 0.001). The most common reason for delivery was spontaneous labour (57% of inpatients and 50% of outpatients). Rates of stillbirth and neonatal mortality were similar between the two groups (3% in the inpatient group and 4% in the outpatient group). Precipitous vaginal delivery of a preterm breech infant was associated with mortality. Umbilical cord pH was < 7.10 in 5% of the inpatient group and 3% of the outpatient group. Median Apgar scores were slightly higher among the outpatient group. Conclusion: The safety of outpatient management of appropriately selected patients with PPROM is comparable with the safety of in-hospital management. Patients with PPROM and a fetus in breech presentation may not be appropriate for outpatient management, especially prior to 28 weeks’ gestation. The decision to manage a patient with PPROM on an outpatient basis must be made after careful evaluation, with a thorough discussion of the risks and benefits and with serial reassessment of patient suitability.
Key Words: PPROM (preterm premature rupture of membranes), outpatient management, inpatient management Competing interests: None declared. Received on April 2, 2015 Accepted on July 18, 2015 http://dx.doi.org/10.1016/j.jogc.2016.03.001
Résultats : La période de latence médiane a été de 11 jours dans le cas des patientes hospitalisées, par comparaison avec 18 jours dans celui des patientes traitées dans la communauté (P < 0,001). Le travail spontané a été la raison ayant le plus couramment mené à l’accouchement (57 % des patientes hospitalisées et 50 % des patientes ambulatoires). Les taux de mortinaissance et de mortalité néonatale ont été semblables dans les deux groupes (3 % dans le groupe des patientes hospitalisées et 4 % dans le groupe des patientes ambulatoires). L’accouchement vaginal précipité d’un fœtus préterme en présentation du siège a été associé à la mortalité. Le pH du cordon ombilical était < 7,10 chez 5 % des femmes du groupe des patientes hospitalisées et chez 3 % des femmes du groupe des patientes ambulatoires. Les indices d’Apgar médians ont été légèrement accrus chez les femmes du groupe des patientes ambulatoires. Conclusion : La sûreté de la prise en charge ambulatoire de patientes présentant une RPMP adéquatement sélectionnées est comparable à la sûreté de la prise en charge en milieu hospitalier. La prise en charge ambulatoire pourrait ne pas convenir aux patientes présentant une RPMP dont le fœtus est en présentation du siège, particulièrement avant 28 semaines de gestation. La décision de procéder à la prise en charge ambulatoire d’une patiente présentant une RPMP doit être prise à la suite d’une évaluation rigoureuse, laquelle doit s’accompagner de la tenue d’une discussion avec cette patiente traitant exhaustivement des risques et des avantages de cette façon de faire; la compatibilité de ce type de prise en charge
-
JOGC
-
2016
l
1
OBSTETRICS
pour la patiente en question doit également faire l’objet d’une réévaluation périodique. Copyright ª 2016 The Society of Obstetricians and Gynaecologists of Canada/La Société des obstétriciens et gynécologues du Canada. Published by Elsevier Inc. All rights reserved.
J Obstet Gynaecol Can 2016;-(-):1-8
INTRODUCTION
P
reterm premature rupture of membranes (PPROM), defined as the rupture of membranes more than one hour prior to labour and before 37 weeks’ gestational age, may result in devastating maternal, fetal, and neonatal outcomes. In Canada, preterm delivery is responsible for 60% to 80% of neonatal deaths,1 with PPROM implicated in close to one third of preterm deliveries.2 Preterm delivery is also associated with significant neonatal morbidity because the preterm infant is at higher risk of respiratory distress syndrome, intraventricular hemorrhage, necrotizing enterocolitis, sepsis, and retinopathy of prematurity. Prolongation of the latency period between rupture of the membranes and labour is not always feasible because spontaneous labour frequently ensues within one week of rupture of membranes,3 and the risks of maternal and fetal infection, placental abruption, umbilical cord compression, and cord prolapse are significantly increased in the presence of PPROM.4 Expectant management of a pregnancy complicated by PPROM prior to 34 weeks’ gestational age demands close monitoring to ensure timely detection of any signs of infection or fetal compromise.2 The time from rupture of membranes to delivery (the latency period) may extend for days, weeks, or even months. Although care of these patients has traditionally consisted of inpatient surveillance, bed shortages and high inpatient costs have led many centres to develop outpatient management strategies that simulate inpatient care. In a survey of 508 maternalfetal medicine specialists, Ramsay et al. found that 43% of respondents endorsed outpatient management for expectant management of PPROM despite limited literature on this topic.5 In Calgary, Alberta, the Antenatal Community Care Program (ACCP) facilitates outpatient care for women with high-risk pregnancies. Since 2001, this specialized home care nursing program has provided care to over 250 women with pregnancies complicated by PPROM. Similar
ABBREVIATIONS ACCP
Antenatal Community Care Program
PPROM preterm premature rupture of membranes
2
l
-
JOGC
-
2016
to inpatient care, management within the community involves reduced maternal activity levels, weekly obstetrical ultrasound assessment, and daily monitoring of maternal vital signs and fetal non-stress testing. Patients are assessed daily for any change in clinical status, such as abdominal pain, uterine contractions, vaginal bleeding, or malodorous discharge, and are immediately taken to the hospital in the presence of any concern. Further research is required to substantiate the safety of home management. Proven benefits of outpatient management include cost efficacy6 and patient satisfaction with care.7 A randomized controlled trial by Carlan et al. of 67 carefully selected women with PPROM did not show a significant difference in safety between inpatient and outpatient management.8 Conversely, a retrospective chart review of 65 inpatients meeting strict outpatient suitability criteria as defined by the study of Carlan et al. suggested that serious obstetric complications necessitating delivery within two hours of a change in clinical status occurred in 18% of study participants.9 METHODS
We performed a retrospective cohort study of women with PPROM with a latency period of at least one week in the Calgary, Alberta, region between January 2007 and December 2012. Information was obtained for patients whose cases were managed on an outpatient basis using the Calgary ACCP patient database. Inpatient data were obtained using medical records coding to locate patients with antepartum hospital admission for PPROM with seven days or more between admission and delivery. Inclusion in the study required rupture of membranes confirmed on sterile speculum examination in the absence of labour between 20 and 34 weeks’ gestational age. Patients were assessed in one of the three obstetrical units in Calgary and were managed in high-risk antepartum care units prior to consideration of outpatient management. Patients completed a recommended PPROM protocol involving intravenous administration of erythromycin and ampicillin, followed by oral erythromycin and amoxicillin (with the exception of patients with known allergies or adverse reactions), and intramuscular betamethasone (two 12 mg doses given 24 hours apart) as an inpatient.9,10 The criteria for outpatient management included the absence of maternal fever, labour, or significant vaginal bleeding; a normal fetal heart rate pattern; a cephalic, frank breech or complete breech position; and a fixed address and telephone number in proximity to Calgary. Our criteria were similar to those used in the study of Carlan et al.,8
Management of Preterm Premature Rupture of Membranes: A Comparison of Inpatient and Outpatient Care
although that study only included women with a fetus in cephalic presentation and required confirmation of cervical dilatation of < 4 cm and a pocket of amniotic fluid measuring 2 cm 2 cm on ultrasound evaluation. Our comparison group consisted of women who met similar criteria (i.e., with confirmed rupture of membranes, absence of labour after one week or maternal fever, and a normal fetal heart rate pattern) but whose cases were managed in the hospital as a result of the assessment of their obstetrical care provider, their preference, or having a geographical location outside the ACCP limits. Cervical examination and transvaginal ultrasound examinations were not routinely performed to minimize the risk of infection ascending into the amniotic cavity. Patients with multiple pregnancies and with a known congenital anomaly in the fetus were also excluded from the study, as were patients with equivocal sterile speculum examination findings, delivery prior to viability, delivery within one week of assessment, and patients lost to follow-up. We sought to evaluate maternal, neonatal, and fetal outcomes for a group of patients whose cases were managed in the community and compare those with a similar group whose cases were managed in the hospital. Our primary outcome was a measure of latency period from the time of PPROM to delivery. We performed a chart review to gather information including baseline demographics, indication for delivery, mode of delivery, length of latency period, placental pathology, Apgar scores, neonatal umbilical cord pH, neonatal sepsis, mortality, and stillbirth. In addition, we sought to evaluate the frequency of obstetrical emergencies, such as cord prolapse, preterm delivery out of hospital, preterm vaginal breech delivery, placental abruption, or chorioamnionitis in both inpatient and outpatient groups. For our primary outcome (latency), we calculated our required sample size using a two-tailed t test with a ¼ 0.05 and power ¼ 0.8. We assumed a standard deviation of approximately 17 days for latency (the average between the outpatient and inpatient groups in the study of Carlan et al.8) and calculated that a total of 129 patients per group would allow us to detect a true difference in latency of six days between the outpatient and inpatient groups. Baseline characteristics were summarized using descriptive statistics and were compared between groups using the chisquare test for categorical variables and the t test for continuous variables. To compare outcomes between the two groups, Wilcoxon rank sum test was used for nonnormally distributed continuous variables, and chi-square or Fisher exact test was used for categorical variables. Risk differences and Hodges-Lehmann estimation of median differences, with 95% confidence intervals, were
reported for each outcome. Data were analyzed using SAS Version 9.3 (SAS Institute Inc., Cary, NC). The study received ethics approval from the Conjoint Health Research Ethics Board (University of Calgary). RESULTS
We identified 182 pregnancies of women whose care for PPROM was facilitated through ACCP during the study period. Twenty-seven percent of these women did not meet the inclusion criteria, as shown in Figure 1, and 133 were included in the analysis. For our inpatient group, 161 women with singleton pregnancies and latency periods of at least seven days whose care for PPROM was managed entirely in one of three tertiary care centres in Calgary were identified. Of these women, 24% were excluded and 122 women were included in the analysis (Figure 2). The baseline characteristics of the study population are shown in Table 1. Mean gestational age at PPROM was comparable between groups (28.3 weeks for inpatients and 28.6 weeks for outpatients). There were no statistically significant differences in gravidity and parity between the groups. The majority of women did not have a history of preterm delivery; 23% of women in the outpatient group and 13% of women in the inpatient group had a history of previous preterm delivery. In both groups more than 60% of infants were male. Both gestational age at delivery and birth weight were significantly greater in the outpatient group than in the inpatient group. For the cases managed in hospital, the median length of admission was 11 days (range 3 to 71). In the outpatient group, the median length of the initial inpatient admission was seven days (range 3 to 51), whereas the median duration under ACCP care was 10 days (range 1 to 68). Including admission at the time of delivery, 78% of patients whose cases were managed under ACCP required only one readmission to the hospital, 20% were readmitted twice, and 2% had three readmissions. The median latency period for patients whose cases were managed in the hospital was 11 days (range 7 to 71) compared with 18 days (range 7 to 77) for the outpatient group (P < 0.001) (Table 2). The majority of patients delivered as a result of spontaneous labour (57% of inpatients and 50% of outpatients). All outpatients were transferred to a hospital with an appropriate level of obstetrical and neonatal care prior to delivery, and deliveries were attended by health care providers skilled in neonatal resuscitation. Six infants in the outpatient
-
JOGC
-
2016
l
3
OBSTETRICS
Figure 1. Outpatient population flow chart
population were born by vaginal breech delivery compared with only one vaginal breech delivery in the inpatient population. However, in all the patients who had a vaginal breech delivery in the outpatient group, the delivery occurred in the hospital, with all the patients being in the hospital for at least a few hours prior to delivery. Of the patients who did not go into labour spontaneously, 43% of inpatients and 50% of outpatients underwent induction of labour or Caesarean section after reaching 34 weeks’ gestation. There was no maternal mortality. The indications for delivery and placental pathology are shown in Table 2.
There was no significant difference in rates of stillbirth and neonatal mortality between the two groups (Table 3). The majority of umbilical cord pH values were 7.10 in both groups, with 3% of the outpatient group and 5% of the inpatient group having a cord pH < 7.10. Median Apgar scores were slightly higher among the outpatient group. Three percent of infants in both groups demonstrated blood or cerebrospinal fluid culture evidence of neonatal sepsis. Of the nine neonatal deaths, seven occurred in infants with a birth weight < 1000 g. In the outpatient group, three of the infant mortalities occurred in infants delivered vaginally with the infant in a breech position; two of these occurred at 25 weeks’ gestational age, and one of the patients declined the recommended Caesarean section. Another neonatal death was the
Figure 2. Inpatient population flow chart
4
l
-
JOGC
-
2016
Management of Preterm Premature Rupture of Membranes: A Comparison of Inpatient and Outpatient Care
Table 1. Baseline characteristics of study participants Inpatient n ¼ 122
Outpatient n ¼ 133
P
Gravidity 1
32 (26%)
32 (24%)
2
25 (20%)
41 (31%)
3
65 (53%)
60 (45%)
0
53 (43%)
53 (40%)
1
34 (28%)
49 (37%)
2
35 (29%)
31 (23%)
106 (87%)
102 (77%)
16 (13%)
31 (23%)
20 to 23 weeks
15 (12%)
10 (8%)
24 to 28 weeks
48 (39%)
58 (44%)
29 to 32 weeks
49 (40%)
61 (46%)
> 32 weeks
10 (8%)
4 (3%)
0.164
Parity 0.289
Prior preterm delivery 0 1
0.036
Gestational age at PPROM
Mean (SD) gestational age at PPROM, weeks
28.3 (SD 3.6)
28.6 (SD 3.0)
Mean (SD) gestational age at delivery, weeks
30.6 (SD 3.1)
32.3 (SD 2.7)
0.147
0.395 < 0.001
Sex Male
75 (61%)
82 (62%)
Female
47 (39%)
51 (38%)
0.977
Mode of delivery Vaginal
51 (42%)
86 (65%)
Caesarean section
71 (58%)
47 (35%)
Mean (SD) birth weight, g
1599.1 (SD 561.3)
1887.8 (SD 635.5)
< 0.001 < 0.001
Presentation Vertex
74 (61%)
112 (84%)
Non-vertex
48 (39%)
21 (16%)
< 0.001
Mode of delivery with respect to fetal presentation Breech vaginal delivery
1 (1%)
6 (4%)
Cephalic vaginal delivery or Caesarean section
121 (99%)
128 (96%)
0.216
result of delayed-onset sepsis in an infant born at 31 weeks’ gestational age, and one infant delivered at 27 weeks’ gestational age with a birth weight of 850 g did not survive after umbilical cord prolapse. In the inpatient group, there were four neonatal deaths, none of which were associated with cord prolapse. In this group, the causes of death were pulmonary complications, necrotizing enterocolitis, and sepsis. The cases in which cord prolapse occurred among the outpatient group are summarized in Table 4.
DISCUSSION
PPROM is encountered in 2.0% to 3.5% of pregnancies, and the optimal management for patients with this condition remains unclear. Given the paucity of data supporting the safety of outpatient management, the clinical practice guidelines from the American College of Obstetricians and Gynecologists recommend inpatient management from the time of rupture of membranes until delivery.13 Even though this approach is prudent, it poses significant difficulty for the patient with a high-risk pregnancy who is at the same time removed from her family and home environment. In addition to this, the logistics of providing inpatient care for patients who remain clinically well for weeks at a time are challenging from a resource management perspective. Inpatient care often consists of little more than daily non-stress testing, recording vital signs every four hours, and weekly ultrasound examinations for the stable patient with PPROM, and yet costs approximately $1860 at Foothills Medical Centre (L. Hickey, Unit Manager Labour and Delivery, Foothills Medical Centre, personal communication, February 2016). In comparison, ACCP care costs on average $200 per day (E. Palmquist, Area Manager of ACCP, personal communication, February 2014). The benefit of inpatient management is the ability to intervene quickly and to have neonatal support available in the event that preterm labour, placental abruption, chorioamnionitis, or cord prolapse should occur. Overall, our study found that maternal and neonatal outcomes were similar in the inpatient and outpatient groups. Median Apgar scores were higher at both one and five minutes in the outpatient group, although this was confounded by the more advanced gestational age at delivery seen in the outpatient group. Rates of chorioamnionitis were higher on pathologic evaluation of the placenta in the inpatient group, and pathologic evidence of placental abruption was comparable between the two groups. All of the outpatients who developed clinical evidence of chorioamnionitis or placental abruption were readmitted to the hospital for close observation and treatment prior to the occurrence of any adverse events. Although mean gestational age at the time of rupture of membranes was comparable between the inpatient and outpatient groups (28.3 weeks and 28.6 weeks, respectively), the mean gestational age at delivery in the outpatient management group was 32.3 weeks. In comparison, the mean gestational age at delivery in the inpatient management group was 30.6 weeks, a difference that was statistically significant (P < 0.001). The more advanced gestational age at delivery and longer latency period
-
JOGC
-
2016
l
5
OBSTETRICS
Table 2. Maternal outcomes Inpatient n ¼ 122
Outpatient n ¼ 133
Median latency (IQR), days
11 (IQR 12) Range 7 to 71
18 (IQR 24) Range 7 to 77
Chorioamnionitis on placental pathology With funisitis
78 (64%) 66 (85%)
63 (47%) 46 (73%)
4 (3%)
5 (3%)
Spontaneous labour
69 (57%)
66 (50%)
Non-labour
53 (43%)
67 (50%)
Planned delivery at 34 weeks’ GA
21 (40%)
36 (54%)
Other
32 (60%)
31 (46%)
13 (11%)
10 (8%)
Abruption on placental pathology
Difference (95% CI)
P < 0.001
6 (4 to 9) 17% (29% to 5%) 12% (25% to 2%)
0.008 0.090 > 0.999
0% (5% to 4%)
Reason for delivery
Non-reassuring fetal status Cord prolapse
1 (1%)
5 (4%)
13 (11%)
7 (5%)
Abruption
4 (3%)
8 (6%)
Other
1 (1%)
1 (1%)
Chorioamnionitis
0.268
7% (19% to 5%)
0.124
14% (4% to 32%)
IQR, interquartile range.
observed among the outpatient population may reflect inherent differences between the two groups. For example, patients with atypical non-stress test results, subclinical chorioamnionitis, vaginal bleeding, advanced cervical dilatation, and anhydramnios would have been more likely to be managed as inpatients because the study was not randomized, and patients were assessed by obstetricians prior to being considered appropriate for outpatient management. Differences noted in our study included a higher prevalence of PPROM prior to 24 weeks’ and after 32 weeks’ gestational age in the inpatient population, whereas the outpatient group had more patients with PPROM in the 24 to 28 weeks’ and 28 to 32 weeks’ gestational age ranges. Very early PPROM is known to be associated with worse neonatal outcomes. Furthermore, the outpatient group had a higher prevalence of previous preterm
delivery, which may have affected our outcome measures if these patients were more involved in clinical decisionmaking regarding the timing of delivery as a result of their previous experiences. In our study, 57% of deliveries in the inpatient group and 50% in the outpatient group were a result of spontaneous labour. All patients were transferred to the hospital at the earliest sign of labour, and no deliveries occurred out of the hospital. Only one patient in the inpatient management group delivered a preterm breech infant vaginally compared with six patients in the outpatient group. Precipitous vaginal delivery of a preterm breech infant was associated with three of the five neonatal deaths in the outpatient population. In all three cases, delivery occurred after the patient had been in the hospital for at least two
Table 3. Fetal outcomes Inpatient n ¼ 122 Umbilical cord pH < 7.10 Stillbirth/neonatal mortality
5 (5%) (missing ¼ 17) 4 (3%)
Outpatient n ¼ 133 3 (3%) (missing ¼ 13) 5 (4%)
P
Difference (95% CI)
0.478
2% (7% to 3%)
> 0.999
0% (4% to 5%)
Median (IQR) Apgar score at 1 minute
6 (IQR 4) Range 0 to 9
7 (IQR 3) Range 0 to 9
< 0.001
1 (0 to 1)
Median (IQR) Apgar score at 5 minutes
8 (IQR 2) Range 2 to 9 (missing ¼ 3)
9 (IQR 1) Range 0 to 9
< 0.001
1 (0 to 1)
Blood or CSF culture positive sepsis
4 (3%) (missing ¼ 1)
Need for pressors
20 (16%)
IQR, interquartile range; CSF, cerebrospinal fluid.
6
l
-
JOGC
-
2016
4 (3%) 17 (13%)
> 0.999 0.413
0% (5% to 4%) 4% (12% to 5%)
Management of Preterm Premature Rupture of Membranes: A Comparison of Inpatient and Outpatient Care
Table 4. Cases of umbilical cord prolapse Location of cord prolapse
Details of presentation(all transported to hospital as an emergency)
Outcome (all underwent emergency Caesarean section)
Case 1
Home
Taken to hospital for prolapse Fetal heart rate normal on arrival Cephalic presentation
Gestational age: 34 weeks Apgar scores 8/9 Cord pH 7.23 Birth weight 2010 g Neonatal survival
Case 2
Home
Taken to hospital for prolapse Fetal heart rate details unknown Cephalic presentation
Gestational age: 27 weeksApgar scores 3/5/6 Cord pH not available Birth weight 1140 g Neonatal survival
Case 3
Home
Taken to hospital for prolapse Fetal heart rate atypical with decelerations Cephalic presentation
Gestational age: 27 weeks Apgar scores 1/4 Cord pH not available Birth weight 850 g Neonatal death few days later
Case 4
Unclear
Taken to hospital in spontaneous labour, cord prolapse noted on vaginal examination Fetal heart rate normal on arrival Breech presentation
Gestational age: 29 weeks Apgar scores 6/8 Cord pH 7.16 Birth weight 1290 g Neonatal survival
Case 5
Home
Taken to hospital for prolapse Fetal heart rate normal on arrival Cephalic presentation
Gestational age: 33 weeksApgar scores 7/9 Cord pH 7.29 Birth weight 1975 g Neonatal survival
days, and one patient declined the recommended Caesarean section. The neonatal mortalities were likely in part related to the prematurity of these neonates, with birth weights less than 1000 g. The other neonatal death in the outpatient group was due to delayed-onset neonatal sepsis, an outcome which would have been unchanged whether the management was on an inpatient or outpatient basis. Interestingly, in three cases in which umbilical cord prolapse was recognized among the outpatient population, the infants were vigorous at birth, a finding which may relate to a lack of cord compression by the fetal presenting part because of the small size of these infants. Although breech presentation is believed to predispose an infant to cord prolapse,14 cord prolapse was seen mainly among cephalic presentations. Again, this is likely a result of the small fetal size in combination with the increased prevalence of cephalic presentation. The higher prevalence of breech vaginal deliveries among the outpatient population and the association of preterm breech vaginal delivery with neonatal death seen among this population raise concerns about the safety of outpatient management for women with a fetus in breech presentation. Although all deliveries occurred in the hospital, there would have been some delay in obstetrical assessment because of the need to transport the patient. Time constraints could necessitate vaginal breech delivery rather than Caesarean section if at presentation the
presenting part was significantly low in the pelvis or if a woman’s urge to push was too great. The second-largest number of infants were delivered once they had reached 34 weeks’ gestational age. Twenty-eight percent of outpatients and 18% of inpatients reached this milestone, and they had excellent outcomes regardless of inpatient or outpatient management. Clinical evidence of chorioamnionitis was an indication for delivery in 13 inpatients and seven outpatients, and placental pathology showed evidence of chorioamnionitis in 64% of inpatients and 47% of outpatients. This may represent a baseline difference between the two groups because criteria for transfer to the hospital among the outpatient population included any early signs of chorioamnionitis. Despite these differences, rates of culture-proven neonatal sepsis were equal between the two populations. Placental abruption was the indication for delivery in eight outpatients and four inpatients. Clinical placental abruption was diagnosed more in the outpatient group, but placental pathology confirmed abruption in only four patients in each group. Our ability to draw conclusions about the overall safety of outpatient management from this study is limited by the retrospective cohort design. Our inpatient population may have been a higher-risk group at baseline, given that the
-
JOGC
-
2016
l
7
OBSTETRICS
preference of the obstetrical care provider partially determined patient selection for inpatient or outpatient management, rather than randomization. Our inpatient and outpatient groups may have also differed in terms of comorbidities because these were not controlled for. In addition, although 34 weeks was generally used as the gestational age at which the risk of continuing expectant management outweighed the risk of prematurity, several pregnancies continued beyond this. In the outpatient group, several patients continued on to 36 weeks’ gestation as a result of combined patient and care provider preference. The higher Apgar scores and birth weights in the outpatient population reflect this difference. As expected, the greater number of infants delivered by Caesarean section and in breech presentation in the inpatient group reflects an increased preference to monitor the breechpresenting fetus on an inpatient basis in cases of PPROM. Importantly, our study lacked statistical power to detect a difference in less frequent outcomes such as neonatal mortality, neonatal sepsis, cord prolapse, chorioamnionitis, and placental abruption. CONCLUSION
Our findings highlight the importance of careful selection of candidates for outpatient management of PPROM and of serial re-evaluation of suitability for such management if any change occurs, including a change in fetal presentation. In addition, patients must be readmitted to the hospital at the earliest sign of labour, vaginal bleeding, or infection. In our population, 22% of outpatients with PPROM required one or more readmissions to the hospital for closer monitoring, and the threshold to admit patients if any clinical evidence of infection or abruption developed was low. Ultimately, the decision whether to manage the case of a patient with PPROM (especially with a breech presentation prior to 28 weeks’ gestational age) on an outpatient basis needs to be individualized and made carefully after a thorough discussion of the risks and benefits of outpatient management.11,12
8
l
-
JOGC
-
2016
REFERENCES 1. Joseph KS, Kramer MS, Marcoux S, Ohlsson A, Wen SW, Allen A, et al. Determinants of preterm birth rates in Canada from 1981 through 1983 and from 1992 through 1994. N Engl J Med 1998;339:1434e9. 2. Mercer BM. Preterm premature rupture of the membranes. Obstet Gynecol 2003;101:178e93. 3. Dale PO, Tanbo T, Bendvold E, Moe N. Duration of the latency period in preterm premature rupture of the membranes. Maternal and neonatal consequences of expectant management. Eur J Obstet Gynecol Reprod Biol 1989;30:257e62. 4. Caughey AB, Robinson JN, Norwitz ER. Contemporary diagnosis and management of preterm premature rupture of membranes. Rev Obstet Gynecol 2008;1:11e22. 5. Ramsey PS, Nuthalapaty FS, Lu G, Ramin S, Nuthalapaty ES, Ramin KD. Contemporary management of preterm premature rupture of membranes (PPROM): a survey of maternal fetal medicine providers. Am J Obstet Gynecol 2004;191:1497e502. 6. Turnbull DA, Wilkinson C, Gerard K, Shanahan M, Ryan P, Griffith EC, et al. Clinical, psychosocial, and economic effects of antenatal day care for three medical complications of pregnancy: a randomised controlled trial of 395 women. Lancet 2004;363:1104e9. 7. Turnbull DA, Wilkinson C, Griffith EC, Kruzins G, Gerard K, Shanahan M, et al. The psychosocial outcomes of antenatal day care for three medical complications of pregnancy: a randomised controlled trial of 395 women. Aust N Z J Obstet Gynaecol 2006;46:510e6. 8. Carlan SJ, O’Brien WF, Parsons MT, Lense JJ. Preterm premature rupture of membranes: a randomized study of home versus hospital management. Obstet Gynecol 1993;81:61e4. 9. Ellestad SC, Swamy GK, Sinclair T, James AH, Heine RP, Murtha AP. Preterm premature rupture of membrane management-inpatient versus outpatient: a retrospective review. Am J Perinatol 2008;25:69e74. 10. Yudin MH, van Schalkwyk J, Van Eyk N, Boucher M, Castillo E, Cormier B, et al. Antibiotic therapy in preterm premature rupture of the membranes. J Obstet Gynaecol Can 2009;31:868e74. 11. Lewis DF, Brody K, Edwards MS, Brouillette RM, Burlison S, London SN. Preterm premature rupture membranes: a randomized trial of steroids after treatment with antibiotics. Obstet Gynecol 1996;88:801e5. 12. Society of Obstetricians and Gynaecologists of Canada. SOGC clinical practice guideline No. 233: antibiotic therapy in preterm premature rupture of the membranes. J Obstet Gynaecol Can 2009;31:863e7. 13. American College of Obstetricians and Gynecologists. Practice bulletin No. 139: premature rupture of membranes. Obstet Gynecol 2013;122:918e30. 14. Huang JP, Chen CP, Chen CP, Wang KG, Wang KL. Term pregnancy with umbilical cord prolapse. Taiwan J Obstet Gynecol 2012;51:375e80.
Management of Preterm Premature Rupture of Membranes: A Comparison of Inpatient and Outpatient Care Elisabeth Catt, MD, Rati Chadha, MD, FRCSC, FACOG, Selphee Tang, Elizabeth Palmquist, RN, MN, Ian Lange, MD, FRCSC Department of Obstetrics and Gynecology, University of Calgary, Calgary AB
Abstract
Résumé
Objectives: We sought to evaluate the safety of outpatient management of pregnancy complicated by preterm premature rupture of membranes (PPROM).
Objectifs : Nous avons cherché à évaluer la sûreté de la prise en charge ambulatoire de la grossesse compliquée par une rupture prématurée des membranes préterme (RPMP).
Study Design: We performed a retrospective cohort study of women with PPROM and a latency period of at least one week in one provincial health region between January 2007 and December 2012. We evaluated pregnancy outcomes for 133 women whose cases were managed using specialized community care and compared these with outcomes of a similar group of 122 women whose cases were managed entirely in the hospital. The primary outcome measured was the difference in the latency period between the two groups. For categorical variable outcomes, data were analyzed using chi-square tests, and continuous variable outcomes were compared using t tests.
Devis d’étude : Nous avons mené une étude de cohorte rétrospective auprès de femmes ayant connu une RPMP et une période de latence d’au moins une semaine au sein d’une région sanitaire provinciale entre janvier 2007 et décembre 2012. Nous avons évalué les issues de grossesse qu’ont connues 133 femmes ayant fait l’objet d’une prise en charge au moyen de soins communautaires spécialisés et les avons comparées aux issues qu’ont connues un groupe similaire de 122 femmes ayant fait l’objet d’une prise en charge entièrement mise en œuvre à l’hôpital. Le critère d’évaluation principal a été la différence en matière de période de latence entre les deux groupes. Les données relevant des variables nominales ont été analysées au moyen de tests de chi carré, tandis que les données relevant des variables continues ont été comparées au moyen de tests t.
Results: The median latency period for inpatients was 11 days compared with 18 days for patients in the community (P < 0.001). The most common reason for delivery was spontaneous labour (57% of inpatients and 50% of outpatients). Rates of stillbirth and neonatal mortality were similar between the two groups (3% in the inpatient group and 4% in the outpatient group). Precipitous vaginal delivery of a preterm breech infant was associated with mortality. Umbilical cord pH was < 7.10 in 5% of the inpatient group and 3% of the outpatient group. Median Apgar scores were slightly higher among the outpatient group. Conclusion: The safety of outpatient management of appropriately selected patients with PPROM is comparable with the safety of in-hospital management. Patients with PPROM and a fetus in breech presentation may not be appropriate for outpatient management, especially prior to 28 weeks’ gestation. The decision to manage a patient with PPROM on an outpatient basis must be made after careful evaluation, with a thorough discussion of the risks and benefits and with serial reassessment of patient suitability.
Key Words: PPROM (preterm premature rupture of membranes), outpatient management, inpatient management Competing interests: None declared. Received on April 2, 2015 Accepted on July 18, 2015 http://dx.doi.org/10.1016/j.jogc.2016.03.001
Résultats : La période de latence médiane a été de 11 jours dans le cas des patientes hospitalisées, par comparaison avec 18 jours dans celui des patientes traitées dans la communauté (P < 0,001). Le travail spontané a été la raison ayant le plus couramment mené à l’accouchement (57 % des patientes hospitalisées et 50 % des patientes ambulatoires). Les taux de mortinaissance et de mortalité néonatale ont été semblables dans les deux groupes (3 % dans le groupe des patientes hospitalisées et 4 % dans le groupe des patientes ambulatoires). L’accouchement vaginal précipité d’un fœtus préterme en présentation du siège a été associé à la mortalité. Le pH du cordon ombilical était < 7,10 chez 5 % des femmes du groupe des patientes hospitalisées et chez 3 % des femmes du groupe des patientes ambulatoires. Les indices d’Apgar médians ont été légèrement accrus chez les femmes du groupe des patientes ambulatoires. Conclusion : La sûreté de la prise en charge ambulatoire de patientes présentant une RPMP adéquatement sélectionnées est comparable à la sûreté de la prise en charge en milieu hospitalier. La prise en charge ambulatoire pourrait ne pas convenir aux patientes présentant une RPMP dont le fœtus est en présentation du siège, particulièrement avant 28 semaines de gestation. La décision de procéder à la prise en charge ambulatoire d’une patiente présentant une RPMP doit être prise à la suite d’une évaluation rigoureuse, laquelle doit s’accompagner de la tenue d’une discussion avec cette patiente traitant exhaustivement des risques et des avantages de cette façon de faire; la compatibilité de ce type de prise en charge
-
JOGC
-
2016
l
1
OBSTETRICS
pour la patiente en question doit également faire l’objet d’une réévaluation périodique. Copyright ª 2016 The Society of Obstetricians and Gynaecologists of Canada/La Société des obstétriciens et gynécologues du Canada. Published by Elsevier Inc. All rights reserved.
J Obstet Gynaecol Can 2016;-(-):1-8
INTRODUCTION
P
reterm premature rupture of membranes (PPROM), defined as the rupture of membranes more than one hour prior to labour and before 37 weeks’ gestational age, may result in devastating maternal, fetal, and neonatal outcomes. In Canada, preterm delivery is responsible for 60% to 80% of neonatal deaths,1 with PPROM implicated in close to one third of preterm deliveries.2 Preterm delivery is also associated with significant neonatal morbidity because the preterm infant is at higher risk of respiratory distress syndrome, intraventricular hemorrhage, necrotizing enterocolitis, sepsis, and retinopathy of prematurity. Prolongation of the latency period between rupture of the membranes and labour is not always feasible because spontaneous labour frequently ensues within one week of rupture of membranes,3 and the risks of maternal and fetal infection, placental abruption, umbilical cord compression, and cord prolapse are significantly increased in the presence of PPROM.4 Expectant management of a pregnancy complicated by PPROM prior to 34 weeks’ gestational age demands close monitoring to ensure timely detection of any signs of infection or fetal compromise.2 The time from rupture of membranes to delivery (the latency period) may extend for days, weeks, or even months. Although care of these patients has traditionally consisted of inpatient surveillance, bed shortages and high inpatient costs have led many centres to develop outpatient management strategies that simulate inpatient care. In a survey of 508 maternalfetal medicine specialists, Ramsay et al. found that 43% of respondents endorsed outpatient management for expectant management of PPROM despite limited literature on this topic.5 In Calgary, Alberta, the Antenatal Community Care Program (ACCP) facilitates outpatient care for women with high-risk pregnancies. Since 2001, this specialized home care nursing program has provided care to over 250 women with pregnancies complicated by PPROM. Similar
ABBREVIATIONS ACCP
Antenatal Community Care Program
PPROM preterm premature rupture of membranes
2
l
-
JOGC
-
2016
to inpatient care, management within the community involves reduced maternal activity levels, weekly obstetrical ultrasound assessment, and daily monitoring of maternal vital signs and fetal non-stress testing. Patients are assessed daily for any change in clinical status, such as abdominal pain, uterine contractions, vaginal bleeding, or malodorous discharge, and are immediately taken to the hospital in the presence of any concern. Further research is required to substantiate the safety of home management. Proven benefits of outpatient management include cost efficacy6 and patient satisfaction with care.7 A randomized controlled trial by Carlan et al. of 67 carefully selected women with PPROM did not show a significant difference in safety between inpatient and outpatient management.8 Conversely, a retrospective chart review of 65 inpatients meeting strict outpatient suitability criteria as defined by the study of Carlan et al. suggested that serious obstetric complications necessitating delivery within two hours of a change in clinical status occurred in 18% of study participants.9 METHODS
We performed a retrospective cohort study of women with PPROM with a latency period of at least one week in the Calgary, Alberta, region between January 2007 and December 2012. Information was obtained for patients whose cases were managed on an outpatient basis using the Calgary ACCP patient database. Inpatient data were obtained using medical records coding to locate patients with antepartum hospital admission for PPROM with seven days or more between admission and delivery. Inclusion in the study required rupture of membranes confirmed on sterile speculum examination in the absence of labour between 20 and 34 weeks’ gestational age. Patients were assessed in one of the three obstetrical units in Calgary and were managed in high-risk antepartum care units prior to consideration of outpatient management. Patients completed a recommended PPROM protocol involving intravenous administration of erythromycin and ampicillin, followed by oral erythromycin and amoxicillin (with the exception of patients with known allergies or adverse reactions), and intramuscular betamethasone (two 12 mg doses given 24 hours apart) as an inpatient.9,10 The criteria for outpatient management included the absence of maternal fever, labour, or significant vaginal bleeding; a normal fetal heart rate pattern; a cephalic, frank breech or complete breech position; and a fixed address and telephone number in proximity to Calgary. Our criteria were similar to those used in the study of Carlan et al.,8
Management of Preterm Premature Rupture of Membranes: A Comparison of Inpatient and Outpatient Care
although that study only included women with a fetus in cephalic presentation and required confirmation of cervical dilatation of < 4 cm and a pocket of amniotic fluid measuring 2 cm 2 cm on ultrasound evaluation. Our comparison group consisted of women who met similar criteria (i.e., with confirmed rupture of membranes, absence of labour after one week or maternal fever, and a normal fetal heart rate pattern) but whose cases were managed in the hospital as a result of the assessment of their obstetrical care provider, their preference, or having a geographical location outside the ACCP limits. Cervical examination and transvaginal ultrasound examinations were not routinely performed to minimize the risk of infection ascending into the amniotic cavity. Patients with multiple pregnancies and with a known congenital anomaly in the fetus were also excluded from the study, as were patients with equivocal sterile speculum examination findings, delivery prior to viability, delivery within one week of assessment, and patients lost to follow-up. We sought to evaluate maternal, neonatal, and fetal outcomes for a group of patients whose cases were managed in the community and compare those with a similar group whose cases were managed in the hospital. Our primary outcome was a measure of latency period from the time of PPROM to delivery. We performed a chart review to gather information including baseline demographics, indication for delivery, mode of delivery, length of latency period, placental pathology, Apgar scores, neonatal umbilical cord pH, neonatal sepsis, mortality, and stillbirth. In addition, we sought to evaluate the frequency of obstetrical emergencies, such as cord prolapse, preterm delivery out of hospital, preterm vaginal breech delivery, placental abruption, or chorioamnionitis in both inpatient and outpatient groups. For our primary outcome (latency), we calculated our required sample size using a two-tailed t test with a ¼ 0.05 and power ¼ 0.8. We assumed a standard deviation of approximately 17 days for latency (the average between the outpatient and inpatient groups in the study of Carlan et al.8) and calculated that a total of 129 patients per group would allow us to detect a true difference in latency of six days between the outpatient and inpatient groups. Baseline characteristics were summarized using descriptive statistics and were compared between groups using the chisquare test for categorical variables and the t test for continuous variables. To compare outcomes between the two groups, Wilcoxon rank sum test was used for nonnormally distributed continuous variables, and chi-square or Fisher exact test was used for categorical variables. Risk differences and Hodges-Lehmann estimation of median differences, with 95% confidence intervals, were
reported for each outcome. Data were analyzed using SAS Version 9.3 (SAS Institute Inc., Cary, NC). The study received ethics approval from the Conjoint Health Research Ethics Board (University of Calgary). RESULTS
We identified 182 pregnancies of women whose care for PPROM was facilitated through ACCP during the study period. Twenty-seven percent of these women did not meet the inclusion criteria, as shown in Figure 1, and 133 were included in the analysis. For our inpatient group, 161 women with singleton pregnancies and latency periods of at least seven days whose care for PPROM was managed entirely in one of three tertiary care centres in Calgary were identified. Of these women, 24% were excluded and 122 women were included in the analysis (Figure 2). The baseline characteristics of the study population are shown in Table 1. Mean gestational age at PPROM was comparable between groups (28.3 weeks for inpatients and 28.6 weeks for outpatients). There were no statistically significant differences in gravidity and parity between the groups. The majority of women did not have a history of preterm delivery; 23% of women in the outpatient group and 13% of women in the inpatient group had a history of previous preterm delivery. In both groups more than 60% of infants were male. Both gestational age at delivery and birth weight were significantly greater in the outpatient group than in the inpatient group. For the cases managed in hospital, the median length of admission was 11 days (range 3 to 71). In the outpatient group, the median length of the initial inpatient admission was seven days (range 3 to 51), whereas the median duration under ACCP care was 10 days (range 1 to 68). Including admission at the time of delivery, 78% of patients whose cases were managed under ACCP required only one readmission to the hospital, 20% were readmitted twice, and 2% had three readmissions. The median latency period for patients whose cases were managed in the hospital was 11 days (range 7 to 71) compared with 18 days (range 7 to 77) for the outpatient group (P < 0.001) (Table 2). The majority of patients delivered as a result of spontaneous labour (57% of inpatients and 50% of outpatients). All outpatients were transferred to a hospital with an appropriate level of obstetrical and neonatal care prior to delivery, and deliveries were attended by health care providers skilled in neonatal resuscitation. Six infants in the outpatient
-
JOGC
-
2016
l
3
OBSTETRICS
Figure 1. Outpatient population flow chart
population were born by vaginal breech delivery compared with only one vaginal breech delivery in the inpatient population. However, in all the patients who had a vaginal breech delivery in the outpatient group, the delivery occurred in the hospital, with all the patients being in the hospital for at least a few hours prior to delivery. Of the patients who did not go into labour spontaneously, 43% of inpatients and 50% of outpatients underwent induction of labour or Caesarean section after reaching 34 weeks’ gestation. There was no maternal mortality. The indications for delivery and placental pathology are shown in Table 2.
There was no significant difference in rates of stillbirth and neonatal mortality between the two groups (Table 3). The majority of umbilical cord pH values were 7.10 in both groups, with 3% of the outpatient group and 5% of the inpatient group having a cord pH < 7.10. Median Apgar scores were slightly higher among the outpatient group. Three percent of infants in both groups demonstrated blood or cerebrospinal fluid culture evidence of neonatal sepsis. Of the nine neonatal deaths, seven occurred in infants with a birth weight < 1000 g. In the outpatient group, three of the infant mortalities occurred in infants delivered vaginally with the infant in a breech position; two of these occurred at 25 weeks’ gestational age, and one of the patients declined the recommended Caesarean section. Another neonatal death was the
Figure 2. Inpatient population flow chart
4
l
-
JOGC
-
2016
Management of Preterm Premature Rupture of Membranes: A Comparison of Inpatient and Outpatient Care
Table 1. Baseline characteristics of study participants Inpatient n ¼ 122
Outpatient n ¼ 133
P
Gravidity 1
32 (26%)
32 (24%)
2
25 (20%)
41 (31%)
3
65 (53%)
60 (45%)
0
53 (43%)
53 (40%)
1
34 (28%)
49 (37%)
2
35 (29%)
31 (23%)
106 (87%)
102 (77%)
16 (13%)
31 (23%)
20 to 23 weeks
15 (12%)
10 (8%)
24 to 28 weeks
48 (39%)
58 (44%)
29 to 32 weeks
49 (40%)
61 (46%)
> 32 weeks
10 (8%)
4 (3%)
0.164
Parity 0.289
Prior preterm delivery 0 1
0.036
Gestational age at PPROM
Mean (SD) gestational age at PPROM, weeks
28.3 (SD 3.6)
28.6 (SD 3.0)
Mean (SD) gestational age at delivery, weeks
30.6 (SD 3.1)
32.3 (SD 2.7)
0.147
0.395 < 0.001
Sex Male
75 (61%)
82 (62%)
Female
47 (39%)
51 (38%)
0.977
Mode of delivery Vaginal
51 (42%)
86 (65%)
Caesarean section
71 (58%)
47 (35%)
Mean (SD) birth weight, g
1599.1 (SD 561.3)
1887.8 (SD 635.5)
< 0.001 < 0.001
Presentation Vertex
74 (61%)
112 (84%)
Non-vertex
48 (39%)
21 (16%)
< 0.001
Mode of delivery with respect to fetal presentation Breech vaginal delivery
1 (1%)
6 (4%)
Cephalic vaginal delivery or Caesarean section
121 (99%)
128 (96%)
0.216
result of delayed-onset sepsis in an infant born at 31 weeks’ gestational age, and one infant delivered at 27 weeks’ gestational age with a birth weight of 850 g did not survive after umbilical cord prolapse. In the inpatient group, there were four neonatal deaths, none of which were associated with cord prolapse. In this group, the causes of death were pulmonary complications, necrotizing enterocolitis, and sepsis. The cases in which cord prolapse occurred among the outpatient group are summarized in Table 4.
DISCUSSION
PPROM is encountered in 2.0% to 3.5% of pregnancies, and the optimal management for patients with this condition remains unclear. Given the paucity of data supporting the safety of outpatient management, the clinical practice guidelines from the American College of Obstetricians and Gynecologists recommend inpatient management from the time of rupture of membranes until delivery.13 Even though this approach is prudent, it poses significant difficulty for the patient with a high-risk pregnancy who is at the same time removed from her family and home environment. In addition to this, the logistics of providing inpatient care for patients who remain clinically well for weeks at a time are challenging from a resource management perspective. Inpatient care often consists of little more than daily non-stress testing, recording vital signs every four hours, and weekly ultrasound examinations for the stable patient with PPROM, and yet costs approximately $1860 at Foothills Medical Centre (L. Hickey, Unit Manager Labour and Delivery, Foothills Medical Centre, personal communication, February 2016). In comparison, ACCP care costs on average $200 per day (E. Palmquist, Area Manager of ACCP, personal communication, February 2014). The benefit of inpatient management is the ability to intervene quickly and to have neonatal support available in the event that preterm labour, placental abruption, chorioamnionitis, or cord prolapse should occur. Overall, our study found that maternal and neonatal outcomes were similar in the inpatient and outpatient groups. Median Apgar scores were higher at both one and five minutes in the outpatient group, although this was confounded by the more advanced gestational age at delivery seen in the outpatient group. Rates of chorioamnionitis were higher on pathologic evaluation of the placenta in the inpatient group, and pathologic evidence of placental abruption was comparable between the two groups. All of the outpatients who developed clinical evidence of chorioamnionitis or placental abruption were readmitted to the hospital for close observation and treatment prior to the occurrence of any adverse events. Although mean gestational age at the time of rupture of membranes was comparable between the inpatient and outpatient groups (28.3 weeks and 28.6 weeks, respectively), the mean gestational age at delivery in the outpatient management group was 32.3 weeks. In comparison, the mean gestational age at delivery in the inpatient management group was 30.6 weeks, a difference that was statistically significant (P < 0.001). The more advanced gestational age at delivery and longer latency period
-
JOGC
-
2016
l
5
OBSTETRICS
Table 2. Maternal outcomes Inpatient n ¼ 122
Outpatient n ¼ 133
Median latency (IQR), days
11 (IQR 12) Range 7 to 71
18 (IQR 24) Range 7 to 77
Chorioamnionitis on placental pathology With funisitis
78 (64%) 66 (85%)
63 (47%) 46 (73%)
4 (3%)
5 (3%)
Spontaneous labour
69 (57%)
66 (50%)
Non-labour
53 (43%)
67 (50%)
Planned delivery at 34 weeks’ GA
21 (40%)
36 (54%)
Other
32 (60%)
31 (46%)
13 (11%)
10 (8%)
Abruption on placental pathology
Difference (95% CI)
P < 0.001
6 (4 to 9) 17% (29% to 5%) 12% (25% to 2%)
0.008 0.090 > 0.999
0% (5% to 4%)
Reason for delivery
Non-reassuring fetal status Cord prolapse
1 (1%)
5 (4%)
13 (11%)
7 (5%)
Abruption
4 (3%)
8 (6%)
Other
1 (1%)
1 (1%)
Chorioamnionitis
0.268
7% (19% to 5%)
0.124
14% (4% to 32%)
IQR, interquartile range.
observed among the outpatient population may reflect inherent differences between the two groups. For example, patients with atypical non-stress test results, subclinical chorioamnionitis, vaginal bleeding, advanced cervical dilatation, and anhydramnios would have been more likely to be managed as inpatients because the study was not randomized, and patients were assessed by obstetricians prior to being considered appropriate for outpatient management. Differences noted in our study included a higher prevalence of PPROM prior to 24 weeks’ and after 32 weeks’ gestational age in the inpatient population, whereas the outpatient group had more patients with PPROM in the 24 to 28 weeks’ and 28 to 32 weeks’ gestational age ranges. Very early PPROM is known to be associated with worse neonatal outcomes. Furthermore, the outpatient group had a higher prevalence of previous preterm
delivery, which may have affected our outcome measures if these patients were more involved in clinical decisionmaking regarding the timing of delivery as a result of their previous experiences. In our study, 57% of deliveries in the inpatient group and 50% in the outpatient group were a result of spontaneous labour. All patients were transferred to the hospital at the earliest sign of labour, and no deliveries occurred out of the hospital. Only one patient in the inpatient management group delivered a preterm breech infant vaginally compared with six patients in the outpatient group. Precipitous vaginal delivery of a preterm breech infant was associated with three of the five neonatal deaths in the outpatient population. In all three cases, delivery occurred after the patient had been in the hospital for at least two
Table 3. Fetal outcomes Inpatient n ¼ 122 Umbilical cord pH < 7.10 Stillbirth/neonatal mortality
5 (5%) (missing ¼ 17) 4 (3%)
Outpatient n ¼ 133 3 (3%) (missing ¼ 13) 5 (4%)
P
Difference (95% CI)
0.478
2% (7% to 3%)
> 0.999
0% (4% to 5%)
Median (IQR) Apgar score at 1 minute
6 (IQR 4) Range 0 to 9
7 (IQR 3) Range 0 to 9
< 0.001
1 (0 to 1)
Median (IQR) Apgar score at 5 minutes
8 (IQR 2) Range 2 to 9 (missing ¼ 3)
9 (IQR 1) Range 0 to 9
< 0.001
1 (0 to 1)
Blood or CSF culture positive sepsis
4 (3%) (missing ¼ 1)
Need for pressors
20 (16%)
IQR, interquartile range; CSF, cerebrospinal fluid.
6
l
-
JOGC
-
2016
4 (3%) 17 (13%)
> 0.999 0.413
0% (5% to 4%) 4% (12% to 5%)
Management of Preterm Premature Rupture of Membranes: A Comparison of Inpatient and Outpatient Care
Table 4. Cases of umbilical cord prolapse Location of cord prolapse
Details of presentation(all transported to hospital as an emergency)
Outcome (all underwent emergency Caesarean section)
Case 1
Home
Taken to hospital for prolapse Fetal heart rate normal on arrival Cephalic presentation
Gestational age: 34 weeks Apgar scores 8/9 Cord pH 7.23 Birth weight 2010 g Neonatal survival
Case 2
Home
Taken to hospital for prolapse Fetal heart rate details unknown Cephalic presentation
Gestational age: 27 weeksApgar scores 3/5/6 Cord pH not available Birth weight 1140 g Neonatal survival
Case 3
Home
Taken to hospital for prolapse Fetal heart rate atypical with decelerations Cephalic presentation
Gestational age: 27 weeks Apgar scores 1/4 Cord pH not available Birth weight 850 g Neonatal death few days later
Case 4
Unclear
Taken to hospital in spontaneous labour, cord prolapse noted on vaginal examination Fetal heart rate normal on arrival Breech presentation
Gestational age: 29 weeks Apgar scores 6/8 Cord pH 7.16 Birth weight 1290 g Neonatal survival
Case 5
Home
Taken to hospital for prolapse Fetal heart rate normal on arrival Cephalic presentation
Gestational age: 33 weeksApgar scores 7/9 Cord pH 7.29 Birth weight 1975 g Neonatal survival
days, and one patient declined the recommended Caesarean section. The neonatal mortalities were likely in part related to the prematurity of these neonates, with birth weights less than 1000 g. The other neonatal death in the outpatient group was due to delayed-onset neonatal sepsis, an outcome which would have been unchanged whether the management was on an inpatient or outpatient basis. Interestingly, in three cases in which umbilical cord prolapse was recognized among the outpatient population, the infants were vigorous at birth, a finding which may relate to a lack of cord compression by the fetal presenting part because of the small size of these infants. Although breech presentation is believed to predispose an infant to cord prolapse,14 cord prolapse was seen mainly among cephalic presentations. Again, this is likely a result of the small fetal size in combination with the increased prevalence of cephalic presentation. The higher prevalence of breech vaginal deliveries among the outpatient population and the association of preterm breech vaginal delivery with neonatal death seen among this population raise concerns about the safety of outpatient management for women with a fetus in breech presentation. Although all deliveries occurred in the hospital, there would have been some delay in obstetrical assessment because of the need to transport the patient. Time constraints could necessitate vaginal breech delivery rather than Caesarean section if at presentation the
presenting part was significantly low in the pelvis or if a woman’s urge to push was too great. The second-largest number of infants were delivered once they had reached 34 weeks’ gestational age. Twenty-eight percent of outpatients and 18% of inpatients reached this milestone, and they had excellent outcomes regardless of inpatient or outpatient management. Clinical evidence of chorioamnionitis was an indication for delivery in 13 inpatients and seven outpatients, and placental pathology showed evidence of chorioamnionitis in 64% of inpatients and 47% of outpatients. This may represent a baseline difference between the two groups because criteria for transfer to the hospital among the outpatient population included any early signs of chorioamnionitis. Despite these differences, rates of culture-proven neonatal sepsis were equal between the two populations. Placental abruption was the indication for delivery in eight outpatients and four inpatients. Clinical placental abruption was diagnosed more in the outpatient group, but placental pathology confirmed abruption in only four patients in each group. Our ability to draw conclusions about the overall safety of outpatient management from this study is limited by the retrospective cohort design. Our inpatient population may have been a higher-risk group at baseline, given that the
-
JOGC
-
2016
l
7
OBSTETRICS
preference of the obstetrical care provider partially determined patient selection for inpatient or outpatient management, rather than randomization. Our inpatient and outpatient groups may have also differed in terms of comorbidities because these were not controlled for. In addition, although 34 weeks was generally used as the gestational age at which the risk of continuing expectant management outweighed the risk of prematurity, several pregnancies continued beyond this. In the outpatient group, several patients continued on to 36 weeks’ gestation as a result of combined patient and care provider preference. The higher Apgar scores and birth weights in the outpatient population reflect this difference. As expected, the greater number of infants delivered by Caesarean section and in breech presentation in the inpatient group reflects an increased preference to monitor the breechpresenting fetus on an inpatient basis in cases of PPROM. Importantly, our study lacked statistical power to detect a difference in less frequent outcomes such as neonatal mortality, neonatal sepsis, cord prolapse, chorioamnionitis, and placental abruption. CONCLUSION
Our findings highlight the importance of careful selection of candidates for outpatient management of PPROM and of serial re-evaluation of suitability for such management if any change occurs, including a change in fetal presentation. In addition, patients must be readmitted to the hospital at the earliest sign of labour, vaginal bleeding, or infection. In our population, 22% of outpatients with PPROM required one or more readmissions to the hospital for closer monitoring, and the threshold to admit patients if any clinical evidence of infection or abruption developed was low. Ultimately, the decision whether to manage the case of a patient with PPROM (especially with a breech presentation prior to 28 weeks’ gestational age) on an outpatient basis needs to be individualized and made carefully after a thorough discussion of the risks and benefits of outpatient management.11,12
8
l
-
JOGC
-
2016
REFERENCES 1. Joseph KS, Kramer MS, Marcoux S, Ohlsson A, Wen SW, Allen A, et al. Determinants of preterm birth rates in Canada from 1981 through 1983 and from 1992 through 1994. N Engl J Med 1998;339:1434e9. 2. Mercer BM. Preterm premature rupture of the membranes. Obstet Gynecol 2003;101:178e93. 3. Dale PO, Tanbo T, Bendvold E, Moe N. Duration of the latency period in preterm premature rupture of the membranes. Maternal and neonatal consequences of expectant management. Eur J Obstet Gynecol Reprod Biol 1989;30:257e62. 4. Caughey AB, Robinson JN, Norwitz ER. Contemporary diagnosis and management of preterm premature rupture of membranes. Rev Obstet Gynecol 2008;1:11e22. 5. Ramsey PS, Nuthalapaty FS, Lu G, Ramin S, Nuthalapaty ES, Ramin KD. Contemporary management of preterm premature rupture of membranes (PPROM): a survey of maternal fetal medicine providers. Am J Obstet Gynecol 2004;191:1497e502. 6. Turnbull DA, Wilkinson C, Gerard K, Shanahan M, Ryan P, Griffith EC, et al. Clinical, psychosocial, and economic effects of antenatal day care for three medical complications of pregnancy: a randomised controlled trial of 395 women. Lancet 2004;363:1104e9. 7. Turnbull DA, Wilkinson C, Griffith EC, Kruzins G, Gerard K, Shanahan M, et al. The psychosocial outcomes of antenatal day care for three medical complications of pregnancy: a randomised controlled trial of 395 women. Aust N Z J Obstet Gynaecol 2006;46:510e6. 8. Carlan SJ, O’Brien WF, Parsons MT, Lense JJ. Preterm premature rupture of membranes: a randomized study of home versus hospital management. Obstet Gynecol 1993;81:61e4. 9. Ellestad SC, Swamy GK, Sinclair T, James AH, Heine RP, Murtha AP. Preterm premature rupture of membrane management-inpatient versus outpatient: a retrospective review. Am J Perinatol 2008;25:69e74. 10. Yudin MH, van Schalkwyk J, Van Eyk N, Boucher M, Castillo E, Cormier B, et al. Antibiotic therapy in preterm premature rupture of the membranes. J Obstet Gynaecol Can 2009;31:868e74. 11. Lewis DF, Brody K, Edwards MS, Brouillette RM, Burlison S, London SN. Preterm premature rupture membranes: a randomized trial of steroids after treatment with antibiotics. Obstet Gynecol 1996;88:801e5. 12. Society of Obstetricians and Gynaecologists of Canada. SOGC clinical practice guideline No. 233: antibiotic therapy in preterm premature rupture of the membranes. J Obstet Gynaecol Can 2009;31:863e7. 13. American College of Obstetricians and Gynecologists. Practice bulletin No. 139: premature rupture of membranes. Obstet Gynecol 2013;122:918e30. 14. Huang JP, Chen CP, Chen CP, Wang KG, Wang KL. Term pregnancy with umbilical cord prolapse. Taiwan J Obstet Gynecol 2012;51:375e80.