Does an upper respiratory tract infection during pregnancy affect perinatal outcomes? A literature review

Association of common maternal infections with birth outcomes: a multinational cohort study

PURPOSE

It is unclear whether common maternal infections during pregnancy are risk factors for adverse birth outcomes. We assessed the association between self-reported infections during pregnancy with preterm birth and small-for-gestational-age (SGA) in an international cohort consortium.

METHODS

Data on 120,507 pregnant women were obtained from six population-based birth cohorts in Australia, Denmark, Israel, Norway, the UK and the USA. Self-reported common infections during pregnancy included influenza-like illness, common cold, any respiratory tract infection, vaginal thrush, vaginal infections, cystitis, urinary tract infection, and the symptoms fever and diarrhoea. Birth outcomes included preterm birth, low birth weight and SGA. Associations between maternal infections and birth outcomes were first assessed using Poisson regression in each cohort and then pooled using random-effect meta-analysis. Risk ratios (RR) and 95% confidence intervals (CI) were calculated, adjusted for potential confounders.

RESULTS

Vaginal infections (pooled RR, 1.10; 95% CI, 1.02-1.20) and urinary tract infections (pooled RR, 1.17; 95% CI, 1.09-1.26) during pregnancy were associated with higher risk of preterm birth. Similar associations with low birth weight were also observed for these two infections. Fever during pregnancy was associated with higher risk of SGA (pooled RR, 1.07; 95% CI, 1.02-1.12). No other significant associations were observed between maternal infections/symptoms and birth outcomes.

CONCLUSION

Vaginal infections and urinary infections during pregnancy were associated with a small increased risk of preterm birth and low birth weight, whereas fever was associated with SGA. These findings require confirmation in future studies with laboratory-confirmed infection diagnosis.

PredictPTB: an interpretable preterm birth prediction model using attention-based recurrent neural networks

BACKGROUND

Early identification of pregnant women at risk for preterm birth (PTB), a major cause of infant mortality and morbidity, has a significant potential to improve prenatal care. However, we lack effective predictive models which can accurately forecast PTB and complement these predictions with appropriate interpretations for clinicians. In this work, we introduce a clinical prediction model (PredictPTB) which combines variables (medical codes) readily accessible through electronic health record (EHR) to accurately predict the risk of preterm birth at 1, 3, 6, and 9 months prior to delivery.

METHODS

The architecture of PredictPTB employs recurrent neural networks (RNNs) to model the longitudinal patient's EHR visits and exploits a single code-level attention mechanism to improve the predictive performance, while providing temporal code-level and visit-level explanations for the prediction results. We compare the performance of different combinations of prediction time-points, data modalities, and data windows. We also present a case-study of our model's interpretability illustrating how clinicians can gain some transparency into the predictions.

RESULTS

Leveraging a large cohort of 222,436 deliveries, comprising a total of 27,100 unique clinical concepts, our model was able to predict preterm birth with an ROC-AUC of 0.82, 0.79, 0.78, and PR-AUC of 0.40, 0.31, 0.24, at 1, 3, and 6 months prior to delivery, respectively. Results also confirm that observational data modalities (such as diagnoses) are more predictive for preterm birth than interventional data modalities (e.g., medications and procedures).

CONCLUSIONS

Our results demonstrate that PredictPTB can be utilized to achieve accurate and scalable predictions for preterm birth, complemented by explanations that directly highlight evidence in the patient's EHR timeline.

Maternal Neutrophil Depletion Fails to Avert Systemic Lipopolysaccharide-Induced Early Pregnancy Defects in Mice

Maternal infection-induced early pregnancy complications arise from perturbation of the immune environment at the uterine early blastocyst implantation site (EBIS), yet the underlying mechanisms remain unclear. Here, we demonstrated in a mouse model that the progression of normal pregnancy from days 4 to 6 induced steady migration of leukocytes away from the uterine decidual stromal zone (DSZ) that surrounds the implanted blastocyst. Uterine macrophages were found to be CD206⁺ M2-polarized. While monocytes were nearly absent in the DSZ, DSZ cells were found to express monocyte marker protein Ly6C. Systemic endotoxic lipopolysaccharide (LPS) exposure on day 5 of pregnancy led to: (1) rapid (at 2 h) induction of neutrophil chemoattractants that promoted huge neutrophil infiltrations at the EBISs by 24 h; (2) rapid (at 2 h) elevation of mRNA levels of MyD88, but not Trif, modulated cytokines at the EBISs; and (3) dose-dependent EBIS defects by day 7 of pregnancy. Yet, elimination of maternal neutrophils using anti-Ly6G antibody prior to LPS exposure failed to avert LPS-induced EBIS defects allowing us to suggest that activation of Tlr4-MyD88 dependent inflammatory pathway is involved in LPS-induced defects at EBISs. Thus, blocking the activation of the Tlr4-MyD88 signaling pathway may be an interesting approach to prevent infection-induced pathology at EBISs.

Clinical Characteristics of and Preventative Strategies for Peripartum Group A Streptococcal Infections

OBJECTIVE

To describe clinical characteristics in parturients with group A streptococcal infection and suggest preventive strategies.

METHODS

We performed a retrospective review of all group A streptococci cultures from women presenting with peripartum fever or abdominal tenderness between January 2008 and May 2015 in a university hospital. Records and epidemiologic investigations of patients and staff were reviewed.

RESULTS

Thirty-seven patients with group A streptococci cultures were identified, with an incidence of one identified postpartum group A streptococcal infection per 2,837 deliveries. Eighty-nine percent of infections occurred postpartum with isolates obtained mainly from the genital tract. Symptoms for group A streptococcal puerperal sepsis were high fever and abdominal tenderness, mostly appearing within 48 hours postpartum. More than one fifth of patients (n=7) developed streptococcal toxic shock syndrome often complicated by multiorgan failure, hysterectomy, and hospitalization in the intensive care unit. There were no uniform risk factors before infection. Epidemiologic investigations suggested that only 23% of infections were nosocomially acquired and that 77% were community-acquired.

CONCLUSION

The high morbidity and the scarcity of distinct risk factors related to parturient group A streptococcal infections in the face of often community-acquired group A streptococci call for reassessing preventive strategies. These may include improved microbiological screening during pregnancy in high-prevalence areas or clinical and microbiological risk stratification in the immediate prepartum and peripartum period.

Maternal β-hemolytic streptococcal pharyngeal exposure and colonization in pregnancy

Objectives. To report the pharyngeal colonization rate of β-hemolytic streptococci and changes in the value of antistreptolysin O (ASO) and anti-DNase B serology titers during pregnancy. Methods. Healthy pregnant women were recruited and blood was drawn in each trimester. The upper limit of normal (ULN) values for ASO and anti-DNase B was calculated for each trimester. Throat swabs were collected for culture and positive cultures were further assessed for the identification of serogroup of the isolated β-hemolytic streptococcus. Results. Out of a total of 126 pregnant women, 34.1% had positive throat cultures. Group C and group G strains were isolated in 18.2% of throat cultures while group F was detected in 13.5% of cases. The rate of colonization with GAS was 1.6%. There was an overall drop in ASO titer during pregnancy while anti-DNase B titers remained relatively unchanged. ULN values of 164^IU, 157^IU, and 156^IU were calculated for ASO at the first, second, and third trimesters, respectively. Based on the ULN values, 28.6% of patients had recent streptococcal exposure. Conclusions. These results show that pregnant women act as a reservoir for spreading potentially immunogenic (groups C and G) and disease producing (group F) virulent strains of streptococci.