1 A casecontrol study was performed to determine whether hea

1. A case-control study was performed to determine whether head injury was associated with an increased risk of brain tumors in children. 200 cases with brain cancer were identified from the state cancer registry and 200 controls were recruited from the same neighborhoods where the cases lived. The mothers of the children completed a questionnaire that asked them to describe their child’s past history of head injury. The investigators found that the mothers of the children with brain tumors reported a past head injury for 70 of the cases while a past history of head injury was reported in 30 of the controls. What type of bias was likely to have influenced the findings of this study, and why? What can be done to minimize this type of bias?

Solution

The main defect in this sample is the sample taken from a particular area.

Both control group and treatment group were selected from the same area which comprises a negligible region compared to total world size.

Also the nature of questions is not given. If the questions comprise about all details relevant to the study and the mothers have adequate knowledge to answer them, then supplying questionairre would be effective. Otherwise, in case of illiterate mothers, who get others help may not give sometimes the true facts as in their mind. Also the nature of questions should be clear and not be vague or ambiguous.

The other main bias is the sample size of 200 compared to whole population of head injuries persons. 200 is not representing the population as a whole.

Steps to minimise the bias:

Data should be collected from different countries of diversified climates, nature of living, region, etc. to come to a conclusion about this injury with tumour.

Next is increasing the sample size adequately by taking a large number of samples.

Third is forming questions which are very technically prepared to get full facts about the child medical history without ambiguity and in regional languages of their mothers.