Date of Award

5-30-2016

Document Type

Dissertation

Degree Name

Master of Medicine (MMed)

First Supervisor/Advisor

Prof. Marleen Temmerman

Second Supervisor/Advisor

Dr. A. Mwaniki Mukaindo

Third Supervisor/Advisor

Dr. Geoffrey Omuse

Department

Obstetrics and Gynaecology (East Africa)

Abstract

Background: There is no consensus on the potential role of inflammatory markers in identifying chorioamnionitis in women with Preterm Pre-labour Rupture of Membranes (PPROM) or in predicting Early Onset Neonatal Sepsis (EONS) in their neonates.

Objectives: To perform a quantitative review on the accuracy of maternal C reactive protein (CRP), Procalcitonin (PCT) and Interleukin 6 (IL6) in the diagnosis of Histological Chorioamnionitis and/or Funisitis (HCA/Funisitis) and their role in the prediction of EONS in PPROM.

Methods: MEDLINE, EMBASE and The Cochrane Library databases were searched from inception to October 2015, for studies where these markers were assessed against a reference standard of HCA/Funisitis or outcome of EONS in PPROM. Two reviewers independently performed screening, data extraction and quality assessments. The Quality Assessment of Diagnostic Accuracy Studies 2(QUADAS-2) and the Quality in Prognostic Studies (QUIPS) tools were used to assess methodological quality. Hierarchical summary receiver operating characteristic (SROC) models were used in the diagnostic review. In the prognostic review, unadjusted Odds Ratios (ORs) were pooled in a random effects meta-analysis.

Results: The diagnostic review included 14 studies reporting 361 episodes (47.4%) of HCA/Funisitis in 761 participants, median prevalence 41% (IQR 36-53). The pooled indices for CRP at the commonest cut-off of 20mg/L (5 studies, 252 participants) were sensitivity 59% (95% CI 48-69), specificity 83% (95% CI 74-89), Likelihood Ratio positive (LR+) 3.45(95% CI 2.24-5.30) and Likelihood Ratio negative (LR-) 0.50(95% CI0.38-0.64 ). The sensitivity, LR+ and LR- for CRP at all cut-offs (11 studies, 570 participants) and at a selected specificity of 80% were 55%, 2.75 and 0.56 respectively. Indices for IL6 at a specificity of 80% were sensitivity 62%, LR+ 3.1 and LR- 0.48. No pooled indices were derived for PCT as included studies were few.

The prognostic review included 7 studies with 332 participants and 97 episodes of EONS, median prevalence 26% (IQR 26-34). The pooled unadjusted OR for studies evaluating CRP at the commonest cut-off of 10mg/L (4 studies, 161participants) was 2.79 (95%CI 1.33-v 5.88, p 0.007). No pooled estimates were obtained for PCT and IL6 as included studies were few. Included studies were mainly prospective cohort design but were of poor quality.

Conclusions: There is insufficient evidence to support use of CRP, PCT or IL6 in maternal blood for the diagnosis of HCA/Funisitis in PPROM and prediction of EONS in PPROM.

Recommendations: We do not recommend the routine use of maternal CRP, PCT or IL6 singly in the management of PPROM. There is need for good quality prospective cohort studies to better assess the role of these biomarkers in PPROM.

Share

COinS