Chorioamnionitis (CA) is a feature of intra-amniotic infection and is significantly associated with preterm, prolonged rupture of membranes (PPROM) 1 and extended duration of labor, positive smoking and alcohol status during pregnancy, and group B Streptococcus (GBS) colonization of the lower genital tract 2. Complicating as many as 40–70% of PPROM births 3, CA is thought to be caused primarily by ascending microbes from the lower genital tract 4. CA can lead to significant adverse outcomes for the fetus, including cerebral palsy, preterm birth, sepsis, pneumonia, respiratory distress, and death 5,6.
As the early detection of CA has been demonstrated to improve fetal outcomes 7, it is essential to identify the development of CA early to minimize the risk of adverse events to the fetus. Although laboratory-based tests such as C-reactive protein, white cell count, and interleukin-6 may aid in the diagnosis of CA, recent reviews remain skeptical as to their clinical application 8,9. Physicians often have to rely upon clinical signs to diagnose patients with CA. Maternal fever and tachycardia, in particular, are common clinical signs in the presence of CA 1 and can be used to develop a more systematic approach to the early identification of CA.
The Modified Early Warning Scores (MEWS) system 10 is a multiparameter track and trigger system whereby seven variables, including temperature and heart rate, are scored according to their degree of derangement (Table 1). MEWS has been shown to improve early recognition and response to adult deteriorating patients and improve their outcomes 11; this system may also have the potential for use as a practical bedside tool for the accurate and early prediction of CA.
Therefore, the aim of this study was to investigate MEWS as an early prediction tool for the development of CA following PPROM so as to inform appropriate and timely management and reduce the risk of adverse events to the fetus.
Patients and methods
The Australian Capital Territory (ACT) Health and Research Ethics Committee approved this study but waived the need for patient informed consent, guided by the National Health and Medical Research Council guidelines for audit.
The Canberra Hospital is a teaching hospital of the Australian National University. The Canberra Hospital serves as the tertiary referral center for high-risk pregnancies and neonatal intensive care for the ACT and the surrounding New South Wales region, attending to more than 2500 deliveries each year.
A retrospective observational study was undertaken of all patients who underwent a PPROM with placental histology from September 2008 until January 2011, identified from an existing obstetric database. Criteria for exclusion were set as absence of placental histology results, age younger than 18 years, absence of patient records, and being outside the gestational age range of 20 weeks and 0 days to 36 weeks and 6 days (36.87 weeks).
Data to be provided at the time of admission to the obstetric unit included demographic data, gestational age, and vital sign measurements. Risk factor information included labor duration and GBS colonization of the genital tract as well as smoking and alcohol status. MEWS was recorded at three predefined time points: upon obstetric admission; between 18 and 30 h before delivery; and within 6 h before delivery. The highest MEWS attained by the patient within 72 h before delivery (72 h MEWS) was also recorded and used as the fourth time point. If a particular vital sign was not recorded it was assumed to be normal for the purpose of calculating MEWS. In addition, if more than one set of recordings qualified for a given time point the earlier of the two was included.
Seventy-five gestational-age-appropriate patients were identified with placental histology in association with PPROM from September 2008 to January 2011. After excluding patients younger than 18 years (one patient) and those with missing records (one patient), 73 patients remained, who were divided into two cohorts on the basis of their placental histology: 61 patients with CA (CA+) and 12 patients without CA (CA−).
Statistical analyses were performed using SPSS 17.0 (SPSS Inc., Chicago, Illinois, USA). Descriptive statistics are presented using median and interquartile ranges. The Mann–Whitney U-test was used for group comparisons of demographics and MEWS, whereas comparisons of binomial proportions between the two groups were performed using Fisher’s exact test. Significance was set at P values less than 0.05. A receiver operating characteristic (ROC) curve was used to examine the diagnostic attributes of each MEWS time point. The most appropriate MEWS cutoff for further analysis was then defined using the Youden index (the point of the ROC curve with the highest sensitivity + specificity − one result 12) for any time point, which showed a significant ROC result.
Between the two groups there were no significant differences in maternal or gestational age, total labor duration, or risk factors including smoking or alcohol intake during pregnancy. GBS colonization rate records were completed in only 34% of cases and thus no meaningful conclusions about GBS status could be made. The CA− group had a shorter length of hospital stay compared with the CA+ group [4 (3.0–6.0) and 7 (4.0–10.8) days, respectively; P=0.002] and shorter duration of membrane rupture before delivery [53.3 (26.3–77.9) and 93 (45.9–196.5) h, respectively; P=0.025].
Modified Early Warning Scores
The median MEWS did not exceed 0 in either group upon admission, at 24 h before delivery, or within the 6 h before delivery and was not significantly different between groups. There was, however, a significant difference in 72 h MEWS between the two groups [CA+: 2 (1–2); CA−: 1 (0–2); P=0.007]. For CA+ patients this peak MEWS occurred at a median of 15 (5.25–44) h before delivery.
As the only time point that demonstrated a significant difference between groups and a significant area under the curve (AUC=0.801, P<0.01), 72 h MEWS was selected for further analysis (Table 2). On the basis of the Youden index, an MEWS of at least 2 was selected as the optimal diagnostic cutoff and was shown to have a sensitivity of 57.4% [95% confidence interval (CI): 44.1–70.0] and specificity of 83.3% (95% CI: 51.6–97.4) as well as positive and negative likelihood ratios of 3.44 (95% CI: 0.95–12.43) and 0.51 (95% CI: 0.35–0.75), respectively. An overall 80.3% of observations were present within this time point.
Derangements in heart rate and systolic blood pressure alone accounted for 83.6% of recorded MEWS points across the two groups within the 72 h MEWS time point and, when reviewed separately from other MEWS criteria at the same cutoff of greater than or equal to 2, were shown to have a sensitivity and specificity of 50.8% (95% CI: 37.7–63.9) and 83.3% (95% CI: 51.6–97.4), respectively; however, the AUC was not found to be significant with these two scores alone. Similarly, heart rate and temperature alone provided 65.3% of recorded MEWS points within the 72 h MEWS time point but also demonstrated an insignificant AUC.
Patients who developed CA showed a significantly higher MEWS in the 72 h before delivery and, using a cutoff greater than or equal to 2, MEWS demonstrated a high specificity but lower sensitivity with a median lead time of 15 h. Thus, an MEWS of at least 2 is able to identify patients with CA and, although it is likely to lead to some false-negative results, an MEWS system can certainly facilitate the early diagnosis of CA.
Although tachycardia and temperature have been previously indicated as sensitive markers of CA 1, and indeed tachycardia and systolic blood pressure accounted for a large portion of the observed MEWS across both groups, neither of these two vital sign pairs alone provided a significant AUC and thus cannot be described as an effective diagnostic modality. It is therefore the multiparameter nature of MEWS that supplies its diagnostic credibility.
There are a number of limitations to this study. Small sample size limits statistical power, as demonstrated through wide CIs, and larger studies are required to clarify our findings. Furthermore, dependence upon placental histology being consistently performed (even in the absence of direct clinical suspicion) could potentially bias results, and, although it is Canberra Hospital policy that all cases of PPROM be analyzed histologically, the possibility that some low-risk cases were not examined exists. In addition, the single-center nature of the study potentially limits the generalization of such findings.
Whereas other papers have examined the use of MEWS as a screening tool for CA in a general obstetric population 13, this is, to the best of the authors’ knowledge, the first paper to describe the validation of MEWS in the early prediction of CA in a population of women who had undergone PPROM. Future studies in an independent population need to be undertaken to determine the accuracy of an MEWS of at least 2 in predicting CA following PPROM, but from this initial study an MEWS system appears to have great potential as a bedside diagnostic aid in this situation.
Conflicts of interest
There are no conflicts of interest.
1. Soper DE, Mayhall CG, Froggatt JW, Baker ME, Hopwood H, Nesbitt T, et al..Characterization and control of intraamniotic infection in an urban teaching hospital.Am J Obstet Gynecol1996;175:304–310.
2. Tita ATN, Andrews WW.Diagnosis and management of clinical chorioamnionitis.Clin Perinatol2010;37:339–354.
3. Yoon BH, Romero R, Moon JB, Shim S-S, Kim M, Kim G, Jun JK.Clinical significance of intra-amniotic inflammation in patients with preterm labor and intact membranes.Am J Obstet Gynecol2001;185:1130–1136.
4. Newton ER.Preterm labor, preterm premature rupture of membranes, and chorioamnionitis.Clin Perinatol2005;32:571–600.
5. Morales WJ, Washington SR III, Lazar AJ.The effect of chorioamnionitis on perinatal outcome in preterm gestation.J Perinatol1987;7:105–110.
6. Wu YW, Escobar GJ, Grether JK, Croen LA, Greene JD, Newman TB.Chorioamnionitis and cerebral palsy in term and near-term infants.J Am Med Assoc2003;290:2677–2684.
7. Gibbs RS, Duff P.Progress in pathogenesis and management of clinical intraamniotic infection.Am J Obstet Gynecol1991;164Pt 11317–1326.
8. Van de Laar R, van der Ham DP, Oei SG, Willekes C, Weiner CP, Mol BWJ.Accuracy of C-reactive protein determination in predicting chorioamnionitis and neonatal infection in pregnant women with premature rupture of membranes: a systematic review.Eur J Obstet Gynecol Reprod Biol2009;147:124–129.
9. Genc MR, Ford CE.The clinical use of inflammatory markers during pregnancy.Curr Opin Obstet Gynecol2010;22:116–121.
10. Morgan RJM, Williams F, Wright MM.An early warning scoring system for detecting developing critical illness.Clin Intensive Care1997;8:100.
11. Mitchell IA, McKay H, Van Leuvan C, Berry R, McCutcheon C, Avard B, et al..A prospective controlled trial of the effect of a multi-faceted intervention on early recognition and intervention in deteriorating hospital patients.Resuscitation2010;81:658–666.
12. Youden WJ.Index for rating diagnostic tests.Cancer1950;3:32–35.
13. Lappen JR, Keene M, Lore M, Grobman WA, Gossett DR.Existing models fail to predict sepsis in an obstetric population with intrauterine infection.Am J Obstet Gynecol2010;203:573.e1–573.e5.