Using the Peanut Exercise Ball to Reduce Cesarean Section Rates: A Randomized Controlled Trial

This study is designed as a prospective, randomized controlled trial that will compare the
use of the peanut exercise ball in laboring patients compared with no use of the exercise
ball. The investigators plan to include all pregnant women over the age of 18 who present to
VCU Labor and Delivery in labor or for induction of labor with a fetus in cephalic
presentation. They will have the peanut shaped exercise ball placed between the maternal legs
in either the supine or upright position during labor. The peanut ball can be seen at this
link, which is similar to the ball the investigators use on Labor and Delivery
(http://www.amazon.com/Isokinetics-Inc-Brand-Peanut-Ball/dp/B0026N2JWG). We use various sizes
of this durable, non-latex exercise ball. There is a small size (for women ,5'3" that is
40inches), medium size for women 5'3" to 5'"7 (50inches) and a large for women >5'7"
(60inches). The length of time using the peanut ball will be quantified by charting in OB
TraceVue software (which is currently standard practice with peanut ball use).

Each patient will be assigned a number in order starting at 1. Each patient will then be
randomized using a block randomization scheme with pre-assigned envelopes in the labor and
delivery team room starting at number 1. Each participant will have an envelope pulled in
order of their number assignment when admitted to labor and delivery and consent obtained.

Exclusion criteria for this study include women with multifetal gestations, fetus in breech
or transverse position, musculoskeletal abnormalities in the mother which might preclude them
from proper positioning on the peanut ball.

This study is a prospective, two-armed randomized controlled trial (RCT) with interim
analyses planned after 50% and 75% of the patients have completed the study. Using nTerim
2.0, 2000 total subjects (1000 per group) will be needed to have a study powered at 80% to
detect a difference in the C-section rate if the treatment and control groups have C-section
rates of 11.25% and 15%, respectively, using a one-sided test at the 0.05 level. The primary
outcome will be the observed incidence of cesarean section delivery in the two populations of
randomized, laboring women.

Bayesian stopping rules, for expected success and futility, are to be added to this study.
Bayesian stopping rules have advantages over other traditional types of interim analyses.
Specifically, they can be used for Bayesian or frequentist (e.g. traditional) trial designs
without requiring the statistical analysis or study design to be altered. Additionally, the
study is not penalized for additional 'looks' at the data, as is the case in traditional
interim analyses. A drawback of this strategy is that the methods used to analyze the data
(e.g. frequentist and Bayesian) have different interpretations and may not come to the same
conclusion.

The investigators plan on conducting three interim analyses, after sample sizes of 500, 1000,
and 1500 have been achieved. At each look, the predicted probability of trial success,
assuming the full 2004 subjects are accrued, will be calculated. The predictive probability
is defined as the probability that the unanticipated C-section rate for the Peanut Ball (PB)
group is lower than that of the control (C), as measured assuming the full 2004 subjects are
accrued and conditional on the data obtained from the subjects that the study has already
accrued. If this probability is small (e.g. < 0.10), then the study will stop for futility.
If the predicted probability with the current sample size is large, meaning it is likely that
the Peanut ball group has a lower unanticipated C-section rate, then the study will be
stopped for expected success. Predicted probabilities of 0.95, 0.90, and 0.90 for the interim
analyses at 500, 1000, and 1500 subjects will be used to determine expected success,
respectively. If neither of these conditions are met, the study will continue to accrue
subjects until a maximum of 2004 subjects.

The calculation of the predictive probabilities are slightly different for futility and trial
success. For the futility analysis, the posterior distribution of an unanticipated C-section
will be will be specified as a beta distribution with Xig +1 and Nig - Xig +1 degrees of
freedom (Beta(Xig +1, Nig - Xig +1)), where Xig denotes the total number of unanticipated
C-sections for the group g (PB, C) for assessment i (1,2,3) from the total sample size in
each group (Nig). This distribution comes from the commonly used beta-binomial posterior
distribution using and uninformative (uniform) prior. For each treatment group and interim
analysis, an unanticipated C-section rate will be simulated from the respective posterior
distribution (pigj) and, based on this simulated value, a simulated number of additional
unanticipated C-sections will be calculated. This process will repeat itself 10,000 times,
and for each iteration, the total estimated rate of unanticipated C-sections will be
calculated. The posterior probability for futility, P(pi,PB+0.0375 total 2004 subjects will be calculated by summing up the instances where pi,PBj+0.0375 where pi,gj is the predictive probability of an unanticipated C-section after full enrollment
for simulation j. The offset, 0.0375, is calculated from the hypothesized difference from the
original power analysis (0.15-0.1125=0.0375)

A similar process will be used for efficacy, however, the predictive probability will be
calculated from the inequality P(pi,PB
The investigators will collect the data, perform data entry and review charts to ensure
accuracy of information. Medical Record Numbers will be collected with data collection to
ensure that we can adequately review the chart of the patient for data collection.The
investigators will also collect information on mode of delivery (vaginal, operative vaginal,
csection), gravidity, parity, age, bmi, epidural anesthesia (yes/no), induction vs
spontaneous labor, gestational age at delivery, dilation at admission, fetal position at
delivery, birth weight, cord ph, Apgars, dystocia, length of active phase of labor and second
stage of labor.

Inclusion Criteria:

- Inclusion Criteria for this study include pregnant women over the age of 18 who
present to VCU Labor and Delivery in labor or for induction of labor with a fetus in
cephalic presentation.

Exclusion Criteria:

- Exclusion criteria for this study include women with multifetal gestations, fetus in
breech or transverse position, musculoskeletal abnormalities in the mother which might
preclude them from proper positioning on the peanut ball.