Early skin-to-skin contact and risk of late-onset-sepsis in very and extremely preterm infants.

Background: To evaluate the association between exposure to early skin-to-skin contact (SSC) and incidence of late-onset sepsis (LOS) in extremely and very preterm infants.

Methods: Observational study using the national population-based EPIPAGE-2 cohort in 2011. A propensity score for SSC exposure was used to match infants with and without exposure to SSC before day 4 of life and binomial log regression used to estimate risk ratios and CIs in the matched cohort. The primary outcome was at least one episode of LOS during hospitalization. Secondary outcomes were the occurrence of any late-onset neonatal infection (LONI), LOS with Staphylococcus or Staphylococcus aureus, incidence of LOS and LONI per 1000 central venous catheter days.

Results: Among the 3422 included infants, 919 were exposed to early SSC. The risk ratio (RR) for LOS was 0.86 (95% CI, 0.67-1.10), for LONI was 1.00 (95% CI, 0.83-1.21), and for LOS with Coagulase-negative Staphylococcus or Staphylococcus aureus infection was 0.91 (95% CI, 0.68-1.21) and 0.77 (95% CI, 0.31-1.87). The incidence RR for LOS per-catheter day was 0.87 (95% CI, 0.64-1.18).

Conclusion: Early SSC exposure was not associated with LOS or LONI risk. Thus, their prevention should not be a barrier to a wider use of SSC.

Impact: Kangaroo Mother Care decreased neonatal infection rates in middle-income countries. Skin-to-skin contact is beneficial for vulnerable preterm infants but barriers exist to its implementation. In a large population-based study using a propensity score methods, we found that skin-to-skin contact before day 4 of life was not associated with a decreased risk of late-onset-sepsis in very and extremely preterm infants. Early skin-to-skin contact was not associated with an increased risk of any late-onset-neonatal-infection, in particular with staphylococcus. The fear of neonatal infection should not be a barrier to a wider use of early skin-to-skin contact in this population.