ArticlePDF Available

384: 37: a mobile preterm birth calculator

Authors:
  • HRV4Training
RESULTS: 162 women met inclusion criteria, 78 in the azithromycin
group and 84 in the erythromycin group. There was no difference in
the median latency from PPROM to delivery between groups (azi-
thromycin: 5.86 days, IQR 3.12-12.05 vs. erythromycin: 6.37, IQR
3.59-10.93; p¼0.67). There was a higher rate of cesarean section
(48.8% vs. 29.5%; p¼0.01) and positive neonatal blood cultures
(13.6% vs. 4.1%; p¼0.04) in the erythromycin group, but overall
rates of neonatal sepsis were low in this cohort. There were no
signicant differences in the other secondary outcomes studied.
CONCLUSION: There is no difference in latency to delivery when a
single oral dose of azithromycin 1g is substituted for erythromycin in
the standard antibiotic regimen used in singleton pregnancies
complicated by PPROM between 23 and 33 6/7 weeks gestation. The
ease of administration and cost-effectiveness of azithromycin make it
a favorable option when compared to erythromycin and further
prospective study to validate these results is warranted.
384 37: a mobile preterm birth calculator
Marco Altini
1
, Julien Penders
1
, Eric Dy
1
, Deirdre Lyell
2
1
Bloom Technologies, San Francisco, CA,
2
Stanford University, Stanford, CA
OBJECTIVE: We sought to create and validate a mobile preterm birth
calculator, which estimates a womans relative risk for preterm birth
(PTB).
STUDY DESIGN: Using data on four million pregnancies, of which
11% resulted in PTB, published by the National Center for Health
Statistics and collected in 2013, we applied Bayesian models for
generalized linear regression to estimate the risk of delivery before 37
weeks based on maternal characteristics. The characteristics included
in the model included maternal demographics, behaviors and ob-
stetric history (level of education, ethnicity, age, BMI, height,
smoking status, payment source, plurality, prior terminations, prior
live births, interval since last pregnancy, previous preterm delivery,
previous C-section, pre-pregnancy complications), and pregnancy
characteristics (weight gain, prenatal care, complications). We used
60% of the data (2.4 million delivery outcomes) to build the
calculator using published PTB risk estimates for the above char-
acteristics, and we tested the calculator using the remaining 40% (1.4
million delivery outcomes).
RESULTS: We found that our model estimates a womans relative risk
of PTB with a sensitivity of 0.27, specicity of 0.98, PPV of 0.54 and
NPV of 0.93. Conclusions: Our calculator demonstrated relatively
high specicity and NPV, and signicantly lower sensitivity and PPV
in predicting full-term birth. This calculator may be useful to phy-
sicians in risk stratication, and could be used as the basis for risk
quantication in future prospective research.
CONCLUSION: Our calculator demonstrated relatively high specicity
and NPV, and signicantly lower sensitivity and PPV in predicting
full-term birth. This calculator may be useful to physicians in risk
stratication, and could be used as the basis for risk quantication in
future prospective research.
385 D-lactic Acid and Matrix Metalloproteinase-2
(MMP-2) levels in vaginal secretions predict cervical
length in mid-trimester pregnant women
Antonio Fernandes Moron
1
, Julie Leizer
2
, Evelyn Minis
2
,
Alan Hatanaka
1
, Marcelo Santucci
1
, Stephanno Sarmento
1
,
Rosiane Mattar
1
, Ann Marie Bongiovanni
2
, Ester Sabino
3
,
Marilza Vieira Rudge
4
, Aluisio Segurado
3
, Iara M Linhares
3
,
Steven S Witkin
2
1
Federal University of Sao Paulo- UNIFESP, Sao Paulo, Brazil,
2
Weill Cornell
Medicine, New York, NY,
3
University of Sao Paulo Medical School-USP, Sao
Paulo, Brazil,
4
Univ Estadual Paulista-UNESP, Sao Paulo, Brazil
OBJECTIVE: Short cervical length is one of the best predictors of
preterm birth. We evaluated whether compounds in vaginal uid
from mid-trimester pregnant women correlated with cervical length.
STUDY DESIGN: In this ongoing study vaginal secretions were ob-
tained before ultrasound evaluation and the soluble fraction assayed
by colorimetric assay for D- and L-lactic acid and ELISA for MMP-2,
ajog.org Poster Session II
Supplement to JANUARY 2017 American Journal of Obstetrics &Gynecology S231
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.