Statistical Analysis
The sample size was calculated based on a stillbirth rate in Delhi
(2.9%) obtained from an earlier study. Presuming that through
ultrasound ,the stillbirth rate can be reduced by 30% a sample size of
2700 would have statistical significance with 80% power at a 5% level
of significance. We analysed the data in the study cohort according to
the defined subcategories of FGR, SGA and AGA. Next, we compared the
outcome variables of study cohort to official data submitted to Delhi
government over the same duration using STATA software version
12.0(2011) Continuous variables were tested for normality assumption
using Kolmogorov-Smirnov test. We calculated descriptive statistics for
normally distributed data and compared the mean values using
(ANOVA)/Student’s t-independent test, as appropriate. We presented
categorical variables as frequency and percent values and compared using
Chi-square/Fishers exact test. Unadjusted risk ratio and 95% confidence
limits were calculated to identify risk variables for various outcomes.
For all statistical tests, a two-sided probability of P<0.05
was considered for statistical significance.