Objective Infection is likely to contribute to the complex aetiology of preterm birth (PTB). It can be detected using white blood cell count and C-reactive protein (CRP). However, nearly 10% of women have subclinical infection. Laboratory analysis has demonstrated that vaginal interleukin-6 (IL-6) is correlated with PTB. The authors aimed to investigate a bedside test in this context.

Method Vaginal secretions were collected from women at high risk of PTB. Samples were incubated for 20 min then analysed by the bedside optical reader (IL-6 in pg/ml). Maternal and neonatal infectious markers and pregnancy outcome were recorded.

Results In the 80 women investigated, IL-6 was able to predict PTB, latency to gestation and maternal infection with some efficacy. It was a poor predictor of neonatal infection. Women with visible fetal membranes had significantly higher IL-6 concentrations than those with closed cervices (p=0.002). All of those with visible membranes and a high IL-6 (>50 pg/ml) had a PTB (n=12) compared to half of those with a low IL-6 (<50 pg/ml, n=1). No women with a low IL-6 delivered at <24 weeks' compared to 83% (n=10) of those with high IL-6. In women with prelabour premature rupture of membranes (PPROM), high IL-6 was associated with more extreme PTB's (57% vs 0%, p=0.19) and delivery within 7 days (71% vs 50%, p=0.09).

Conclusion IL-6 may be useful in guiding the difficult management of patients with visible membranes and PPROM, for example the potential benefit of a cervical cerclage and antibiotics.