Introduction
The researchers’ aim was to determine the relationship between the application of Assisted Reproductive Technology (ART) and occurrence of structural congenital defects in humans in reference to the US population. The researchers embarked on a case-control study of birth defects using data from the National Birth Defects Prevention. The study was also a population –based and multicenter research. Therefore, the paper is relevant to the key issue or topic.
Internal validity
Study question
The study question relates to the relationship between ART and the number of birth defects recorded every year in the US. The researchers noted that despite a number of studies carried out in the recent past to investigate the same study question, a number of factors limit the knowledge available, including the use of small numbers of cases, lack of the appropriate control samples, confounding as well as heterogeneity of the study population. Thus, using a nationwide focus, the case-control study attempts to determine the relationship between ART and congenital birth defects.
Selection of subjects
The researchers selected both the case and control samples from comparable populations. In particular, the researchers focused on data supplied by the national birth defects prevention study (NBDPS), an ongoing case-control research on the nationwide level seeking to investigate genetic and environmental risk factors in birth defects. The researchers focused on 10 states, from where the data was obtained. Cases were infants with birth defects identified through the normal birth defect surveillance systems. The defects were identified from hospital records. The researchers used cases on live-born, pregnancy terminators as well as fetal deaths. Control sample included infants who were live-born without congenital defects, randomly selected from the locations where cases were also collected. Therefore, the same exclusion and inclusion criterion was used for both the case and control populations.
Secondly, the researchers used interviews for both populations, focusing on mothers to the infants identified as sample populations. Interviews were carried out on each individual mother, resulting into a response rate of 67.2% for the control and 70.5% for the cases. Quite evidently, this rate is acceptable in a case-control study. It is also evident that the right collection procedures and criteria for data were used.
The study focused on making comparisons between the case and the control populations. The researchers used birth defects to establish that controls were non-cases and cases were non-control populations. This was possible because the surveillance methods used in hospitals are well established.
Assessment
The researchers used measures to prevent the prior knowledge of primary exposure interfering with the ascertainment of cases. For instance, they ensured that information from the hospital records on each infant in both populations was not recorded straight into the interview questions. They also ensured that the individuals assigned the work of data collection were blinded from the information about the defect statuses of the populations.
The measure of exposure in the case-control study provides evidence of being valid, standard and reliable. The researchers’ main exposure of interest was the use of Assisted Reproductive Technology by each of the mothers. The exposure was also defined as “the use of any treatment on the mother to conceive the pregnancy, where the handling of both the sperm and the embryo was done in a medical way’. Specifically, these treatments were ICSI, zygote or gamete intrafallopian transfer and IVF. Mothers to the control population were “unexposed” only if they reported No to the use of any medication or procedures to aid in conception, either by the mother or the father.
Confounding
The researchers identified maternal age as a confounding factor. Therefore, it was included as a continuous variable in each of the models. In addition, all factors associated with the birth defects and ART were identified as the potential confounders. Nevertheless, the researchers state that confounding factors were not necessarily associated with congenital defects. Thus, they considered a number of confounding factors such as center of study, maternal race, parity, history of miscarriages income, level of education, smoking by the mothers, use of alcohol 1month prior to conception, use of folic acid supplements and body mass index.
Results: Statistical analysis
The researchers worked with a sample of 5,008 controls and 13,586 as the case sample. Mothers of these infants were interviewed. However, 39 participants were excluded due to missing data on fertility, while 876 were excluded due to the use of treatments other than ART. Thus, the study population was 17,679.
Data analysis was carried out using SPSS, emphasizing on measures of central tendency and dispersion. Tables were used as the main data analysis methods in the study. In addition, confidence has been provided, providing evidence of a well-developed case-control approach to the study question.
The results indicated that art was used by 230 (2.4%) and 51 (1.1%) of the case and control mothers respectively. In addition, 16 cases and 5 control had have used ICSI, while 27 case and 9 control mothers had used donor egg, sperms or embryo. In addition, 45 case and 10 control mothers respectively used frozen eggs.
The study used a confidence interval CI and P-value to determine the reliability of the estimates. If the expected number in the cells was less than five, the researchers used Fisher’s exact test for estimating the confidence interval. A standard confidence interval of 95% was used for the cells. With a confidence interval of 95%, all the variables satisfied the test. For instance, singletons with heart defects had a confidence range of 1.1-4.0, 1.5-6.1 for VSD and 1.2-7.0 for ASD.
A P-value was used to determine the trial occurred due to chances presented by confounding factors. With a P-value of less than 0.00, it was found that most of the confounders did not affect the test. However, the P-value was relatively high for previous live births (0.003), previous miscarriages (0.415), BMI (0.706), alcohol use (0.509) and smoking (0.005), indicating that they are the most potential confounders in the test.
Overall assessment of the study
The researchers attempted to use several measures to minimize the risk of bias as well as effect of confounding factors. For instance, they were able to identify all the possible confounding factors associated with birth defects, which may have an impact on the occurrence of birth anomalies. Most of these factors are associated with the mothers, their lifestyles and behavior prior to conception or during pregnancy. In addition, the researchers have used several exclusion criteria to exclude cases and controls that had the potential to bring bias. For instance, the excluded samples whose clinical information was missing as well as those who had used other treatments to aid pregnancy other than ART.
The results of the study indicate that a significant association between ART and a number of birth defects, including ASD, VSD, esophageal atresia as well as anorectal atresia. The researchers noted that the increase in the use of ART in the US between 1996 and 2004 might explain why some of the prevalence of some of these defects has increased in the same period. In addition, the researchers acknowledge a study with almost similar results, which was carried out in Sweden two years before this study was completed.
Taking into account clinical considerations, the methodology used and the statistical analysis of data in this article, it is evident that the researchers have shown a clear evidence of the association between exposure to ART and occurrence of a number of birth defects in the US. Thus, the results of this study can be directly applied to the patient groups using the guidelines developed within the study.
Conclusion
This critical appraisal finds that the study by Reefhuis, Honein, Schieve, Correa, et al (2008) is an important source of information for additional studies as well as development of guidelines or knowledge in practice. It also finds that the study was carried out in the right protocol, especially with respect to the use of case-control approach. It is also worth noting that the study topic and question is relevant to the current situation in medicine, given that the increased use of ART in the US and the world is evident, but the possible impact on pregnancies and the infant is not well known.
Reference
Reefhuis, J., Honein, M. A., Schieve, L. A., Correa, A., et al. (2008). Assisted reproductive technology and major structural birth defects in the United States. Human reproduction, 24(2), 360-366.