Related Experiment Video
Updated: Jul 3, 2026
The 4-vessel Sampling Approach to Integrative Studies of Human Placental Physiology In Vivo
Published on: August 2, 2017
Statistical methods for estimating the probability of spontaneous abortion in observational studies--analyzing
Reinhard Meister1, Christof Schaefer
1Department of, Mathematics, University of Applied Sciences, Technische Fachhochschule Berlin, Luxemburger Strasse 10, 13353 Berlin, Germany. meister@tfh-berlin.de
Background:
Spontaneous abortion rates are of general interest when investigating pregnancy outcome. In most studies observations are left truncated as pregnant women enter with a delay of several weeks after conception. Apart from spontaneous abortion pregnancy may end in induced abortion or live birth. These outcomes are considered as competing events (risks). Although statistical methods for handling this setting are available since more than 10 years, studies on pregnancy outcome after drug exposure usually report crude rates of spontaneous abortions, ignoring left truncation and competing risks.
Methods:
The authors propose simple methods which remove bias inherent to crude rates. The probability of spontaneous abortion is estimated using an event-history based approach for the subdistribution of competing risks that handles left truncation appropriately. Variance estimation enables the construction of approximate confidence intervals and of a simple test-statistic for comparing rates between different cohorts. The proposed methods are applied to a comparative prospective study on the association of spontaneous abortion and exposure to coumarin derivatives.
Results:
The naive analysis using crude rates gives substantially different results than those based on the proposed methods, with up to a twofold change. Correctly incorporating left truncation into the analysis may increase the variance of the estimators, relative to an ideal sample where all pregnancies are followed from the time of conception. The consequences of such truncation for study design are discussed.
Conclusion:
Combining corrections for left truncation and competing risks offers a powerful method for analyzing miscarriage risk.
Related Concept Videos
01:23Types of Biopharmaceutical Studies: Controlled and Non-Controlled Approaches
Non-controlled studies, commonly employed for initial exploration, lack a control group, rendering them susceptible to biases and external influences. In contrast, controlled...
01:25Statistical Methods for Analyzing Epidemiological Data
01:15Assumptions of Survival Analysis
01:09Censoring Survival Data
01:20Comparing the Survival Analysis of Two or More Groups
01:11Testing a Claim about Mean: Known Population SD
Estimating a population mean requires the samples to be distributed normally. The data should be collected from the randomly selected samples having no sampling bias. The sample size needed to be higher than 30, and most importantly, the population standard deviation should be already known.
In most realistic situations, the population standard deviation is often unknown, but in rare circumstances, when it...