Epi in the News

1. Cross-sectional Descriptive Study. Looking at the prevalence of FASD in adopted children (both suspected and diagnosed children)

2. Sampling Bias: The sample was not randomly selected and therefore you could have disproportionate number of people with children with FASD or without. Each adoptive parent of the population should ideally have an equal chance of participating in the survey. You could get a selection bias from this as either more or fewer parents would participate. Parents with children having FASD may feel inclined to participate in a research that concerns their child. Conversely, parents may be unwilling to participate in a survey that might involving disclosing sensitive health information of the child (protection)

3. 8% were diagnosis with FASD = Total number of adopted children diagnosed with FASD according to FASD diagnostic criteria in the sample population / Total Number of adopted children in the sample population (~5000)

17% has suspected FASD = Total number of adopted children with features of FASD but not fulfilling FASD diagnostic criteria in the sample population / Total Number of adopted children in the sample population (~5000)

I am not too sure what questions the survey included but a few concerns would be:
- Reporting Bias where adoptive parents are unwilling to fully disclose the healthy status of their children or if they had signs and symptoms of FASD like behavioural disorders or developmental delays ==> Underreporting of cases
- Verification bias due to healthcare practitioners not having sufficient knowledge to make an accurate diagnosis ==> Underreporting of cases

4. Choose a random sample of participants. If resource permitted and information is available, to ensure that there is a standard of diagnostic criteria for FASD used to assess all children.

5. A randomised control trial: Intervention Group = strictly no alcohol Control Group = No interventional restriction or permission (Encouraged to manage pregnancy as they would with current advise that normally takes place at antenatal visits). Monitor alcohol use through surveys. There are some tests that might be able to detect alcohol levels beyond 48 hours. Monitor short and long term side effects for pregnancy in two groups and follow up on the developmental status of their children

https://www.newscientist.com/article/dn4662-booze-tests-reveal-all-about-your-drinking/#:~:text=A%20battery%20of%20new%20tests,monitor%20patients%20with%20alcohol%20problems.

Would appreciate feedback!!

1.Cross-sectional design (observational, descriptive)

2. There was no mention of random sampling in the article. Since there was no mention of how the sampling was done, there could be an overestimation of the actual prevalence of FADS among adopted children especially if they were chosen from orphanages in areas where there is a prevalence of alcoholism as compared to other areas. Secondly, there could have been a non response bias created due to the absence of some adopters who may not have seen the essence of the study and a response bias from those who perceived they had children affected by FASD.

3.
The case definitions may have included a history of certain amounts of alcohol consumption during some particular trimesters of pregnancy, and either behavioural deviations from the norm and a means of assessing their intelligence. However, since there are no comprehensive records for alcohol consumption by birth mothers, some who may be wrongly characterised as cases .

a. %children with diagnosis of FASD= (no of children adopted by the nearly 5000 adopters who met the FASD diagnosis criteria / total number of children adopted by nearly 5000 adopters) * 100%
b.%children suspected to have diagnosis = (no of children adopted by the nearly 5000 adopters who had symptoms of FASD but did not meet the FASD diagnosis criteria / total number of children adopted by nearly 5000 adopters) * 100%

Since the response to the survey was by adopted parents of the adopted children there may have been a misclassification of cases depending on whether certain behaviours were depicted to actually be as a result of neurodevelopmental problems and ticked on the poll or their ability to recall particular answers to certain questions on the survey leading to measurement bias. This may have either led to an overestimation or underestimation of actual cases or overestimation of those suspected as they may have other neurodevelopmental or behavioural problems.
4. I would read on FASD from secondary sources and study the research done in scotland. From that, I'd design a case-control study which would take random samples of children who passed to adoption centres all over the country, then they'd be assessed by licensed professionals, after isolating true cases there'd be a need to go back in "time " to look at maternal records for alcohol consumption.

5.
i would stratify different pregnant women who are similar in all other characteristics including antenatal care except the average units of alcohol they consume a week during into 0 units, 1-7 units, 7-14 units, 14-21units and > 21 units based on the recommended weekly dose of alcohol for women. This stratification will be done separately for women in their first second or third trimester. Their offspring will be followed up from birth to about 7 years and physical characteristics and behavioural characteristics will be measured by well trained personnel.