Several factors help predict labor onset, emergent outcome in PPROM

CHICAGO – A novel composite prognostic index score helped predict the onset of labor in a retrospective cohort of patients with preterm premature rupture of membranes.

In a separate study, researchers identified three independent predictors of emergent outcomes in patients with preterm premature rupture of membranes (PPROM).

The findings of both studies were presented in posters at the annual meeting of the American Congress of Obstetricians and Gynecologists.

The composite prognostic index score in the first study predicted the likelihood of labor within 12 hours while maintaining a significantly high negative predictive value in 78 patients between 24 and 34 weeks of gestation who were admitted with PPROM over a 2-year period, according to Dr. Yelena Feldman of Trihealth, Cincinnati.

In fact, differences in all variables included in the score were significant at 12 hours prior to labor, she noted.

Variables included as part of the score were deepest vertical pocket of amniotic fluid by ultrasound, fetal heart rate, changes in fetal heart rate variability, presence of decelerations, number of contractions, vaginal bleeding, and record of nursing concern. Each variable was scored at four time points prior to spontaneous labor onset (48, 36, 24, and 12 hours), and a model using the presence of dichotomous variables at 12 hours prior to labor onset was used to make the composite score.

A binary model with the outcome of 12 hours until labor onset had the best results (91.9% specificity; 51.25% sensitivity, 85% negative predictive value, and 67.8% positive predictive value).

Each variable was assigned a number of points based on its beta coefficient in the multivariate model, and the patient could be assigned a score based on the presence of these characteristics.

"The value would correspond to the risk of labor starting within 12 hours," she noted.

The cutoff score, determined by the receiver operating characteristic curves that signified the likelihood of starting labor in 12 hours, was 18, Dr. Feldman explained in the poster.

A score of 18 yielded a negative predictive value of 90.5%, a positive predictive value of 52%, and a sensitivity of 80.9%.

"This composite score may serve as a useful tool in clinical settings where patients admitted with PPROM need decisions regarding patient transfer, administering magnesium sulfate for neuroprotection, or administering a rescue dose of steroid," she concluded.

In the second study, Dr. Tripp Nelson of the Medical University of South Carolina, Charleston, found that malpresentation, bleeding, and sexually transmitted infection each predicted emergent outcomes in PPROM patients.

An admission test utilizing these three factors had 96.4% negative predictive value for emergent outcomes, 57.4% positive predictive value, 91% specificity, and 75.9% sensitivity. For the retrospective case-control study, Dr. Nelson and his colleagues identified 624 subjects, including 83 with at least one emergent outcome.

The emergent group had significantly higher rates of perinatal death and acidosis, and while bivariable comparison showed increased incidence of leukocytosis, urinary tract infection, sexually transmitted infection (STI), malpresentation, latency, vaginal bleeding, and fundal tenderness; only vaginal bleeding, STI, and malpresentation remained significant on logistic regression analysis.

Further randomized testing is needed for model validation, Dr. Nelson concluded.

The authors of both studies reported having no disclosures.

CHICAGO – A novel composite prognostic index score helped predict the onset of labor in a retrospective cohort of patients with preterm premature rupture of membranes.

In a separate study, researchers identified three independent predictors of emergent outcomes in patients with preterm premature rupture of membranes (PPROM).

The findings of both studies were presented in posters at the annual meeting of the American Congress of Obstetricians and Gynecologists.

The composite prognostic index score in the first study predicted the likelihood of labor within 12 hours while maintaining a significantly high negative predictive value in 78 patients between 24 and 34 weeks of gestation who were admitted with PPROM over a 2-year period, according to Dr. Yelena Feldman of Trihealth, Cincinnati.

In fact, differences in all variables included in the score were significant at 12 hours prior to labor, she noted.

Variables included as part of the score were deepest vertical pocket of amniotic fluid by ultrasound, fetal heart rate, changes in fetal heart rate variability, presence of decelerations, number of contractions, vaginal bleeding, and record of nursing concern. Each variable was scored at four time points prior to spontaneous labor onset (48, 36, 24, and 12 hours), and a model using the presence of dichotomous variables at 12 hours prior to labor onset was used to make the composite score.

A binary model with the outcome of 12 hours until labor onset had the best results (91.9% specificity; 51.25% sensitivity, 85% negative predictive value, and 67.8% positive predictive value).

Each variable was assigned a number of points based on its beta coefficient in the multivariate model, and the patient could be assigned a score based on the presence of these characteristics.

"The value would correspond to the risk of labor starting within 12 hours," she noted.

The cutoff score, determined by the receiver operating characteristic curves that signified the likelihood of starting labor in 12 hours, was 18, Dr. Feldman explained in the poster.

A score of 18 yielded a negative predictive value of 90.5%, a positive predictive value of 52%, and a sensitivity of 80.9%.

"This composite score may serve as a useful tool in clinical settings where patients admitted with PPROM need decisions regarding patient transfer, administering magnesium sulfate for neuroprotection, or administering a rescue dose of steroid," she concluded.

In the second study, Dr. Tripp Nelson of the Medical University of South Carolina, Charleston, found that malpresentation, bleeding, and sexually transmitted infection each predicted emergent outcomes in PPROM patients.

An admission test utilizing these three factors had 96.4% negative predictive value for emergent outcomes, 57.4% positive predictive value, 91% specificity, and 75.9% sensitivity. For the retrospective case-control study, Dr. Nelson and his colleagues identified 624 subjects, including 83 with at least one emergent outcome.

The emergent group had significantly higher rates of perinatal death and acidosis, and while bivariable comparison showed increased incidence of leukocytosis, urinary tract infection, sexually transmitted infection (STI), malpresentation, latency, vaginal bleeding, and fundal tenderness; only vaginal bleeding, STI, and malpresentation remained significant on logistic regression analysis.

Further randomized testing is needed for model validation, Dr. Nelson concluded.

The authors of both studies reported having no disclosures.

CHICAGO – A novel composite prognostic index score helped predict the onset of labor in a retrospective cohort of patients with preterm premature rupture of membranes.

In a separate study, researchers identified three independent predictors of emergent outcomes in patients with preterm premature rupture of membranes (PPROM).

The findings of both studies were presented in posters at the annual meeting of the American Congress of Obstetricians and Gynecologists.

The composite prognostic index score in the first study predicted the likelihood of labor within 12 hours while maintaining a significantly high negative predictive value in 78 patients between 24 and 34 weeks of gestation who were admitted with PPROM over a 2-year period, according to Dr. Yelena Feldman of Trihealth, Cincinnati.

In fact, differences in all variables included in the score were significant at 12 hours prior to labor, she noted.

Variables included as part of the score were deepest vertical pocket of amniotic fluid by ultrasound, fetal heart rate, changes in fetal heart rate variability, presence of decelerations, number of contractions, vaginal bleeding, and record of nursing concern. Each variable was scored at four time points prior to spontaneous labor onset (48, 36, 24, and 12 hours), and a model using the presence of dichotomous variables at 12 hours prior to labor onset was used to make the composite score.

A binary model with the outcome of 12 hours until labor onset had the best results (91.9% specificity; 51.25% sensitivity, 85% negative predictive value, and 67.8% positive predictive value).

Each variable was assigned a number of points based on its beta coefficient in the multivariate model, and the patient could be assigned a score based on the presence of these characteristics.

"The value would correspond to the risk of labor starting within 12 hours," she noted.

The cutoff score, determined by the receiver operating characteristic curves that signified the likelihood of starting labor in 12 hours, was 18, Dr. Feldman explained in the poster.

A score of 18 yielded a negative predictive value of 90.5%, a positive predictive value of 52%, and a sensitivity of 80.9%.

"This composite score may serve as a useful tool in clinical settings where patients admitted with PPROM need decisions regarding patient transfer, administering magnesium sulfate for neuroprotection, or administering a rescue dose of steroid," she concluded.

In the second study, Dr. Tripp Nelson of the Medical University of South Carolina, Charleston, found that malpresentation, bleeding, and sexually transmitted infection each predicted emergent outcomes in PPROM patients.

An admission test utilizing these three factors had 96.4% negative predictive value for emergent outcomes, 57.4% positive predictive value, 91% specificity, and 75.9% sensitivity. For the retrospective case-control study, Dr. Nelson and his colleagues identified 624 subjects, including 83 with at least one emergent outcome.

The emergent group had significantly higher rates of perinatal death and acidosis, and while bivariable comparison showed increased incidence of leukocytosis, urinary tract infection, sexually transmitted infection (STI), malpresentation, latency, vaginal bleeding, and fundal tenderness; only vaginal bleeding, STI, and malpresentation remained significant on logistic regression analysis.

Further randomized testing is needed for model validation, Dr. Nelson concluded.

The authors of both studies reported having no disclosures.