1. In a study to determine the incidence of chronic disease, 150 people were exa

Question

1. In a study to determine the incidence of chronic disease, 150 people were examined at the end of the follow-up period. Twelves cases were found, giving a cumulative incidence rate of 8% (check if my calculation is correct). Fifty other members of the initial cohort could not be examined, 20 of them because they died. a) What source of bias should be of concern in this example (2 points) b) What will be the incidence rate if all people who were lost to be followed had the disease and what will be the incidence rate if all people who were lost to be followed did not have the disease. (Show your denominator and numerator in the calculation (4 points) c) Based on these calculations, do you think the 8% incidence rate is valid (i.e., accurate) Why yes or why not? (2 points)

Q2. What bias would you suspect in a survey of the prevalence of depression in the elderly population of a city, based on selecting the random sample of members listed in several seniors clubs in the city? Explain and how it will affect your estimate. (2 points)

Q3. You conduct a randomized controlled trial with vitamin B-12 and placebo and their effect on depressive symptoms. You randomly allocated in a double-blind study to two groups recruiting from GPs. The difference in depressive symptoms between the two groups after 1 year of treatment was not statistically different but you find out that 30% of women were noncompliant with the vitamin B-12 tablets. If you omit them from the analysis the results of the study were significant. What analysis was done first and what was done second – what would you use and why?

Explanation / Answer

In the question 1, we are given

150 people were examined at the end of the follow-up period. Twelves cases were found, giving a cumulative incidence rate of 8%. Fifty other members of the initial cohort could not be examined; 20 of these 50 could not be examined because they died.

And we are asked to state source of bias.

First thing, its a cohort study. So we need to choose the study participants very carefully. Failing to which we may lead to incorrect incidence rate. So, one of the potential bias is the Selection Bias.

Also, being a cohort study, it include long follow up period. Thus, it may happen that we lost some patients to follow up or they die while study is ongoing. And the result again will be incorrect incidence rate. Thus, another potential bias is Survival Bias.

In the next part b) It is given,

b) What will be the incidence rate if all people who were lost to be followed had the disease and what will be the incidence rate if all people who were lost to be followed did not have the disease.

Answer:

In the question, it was mentioned that the researches examined 150 patients out of which 12 patients found with disease. Thus, the disease under study has incidence rate (12/150)*100 = 8%.

Further note that, Fifty other members of the initial cohort could not be examined; 20 of these 50 could not be examined because they died.

So basically, the initial cohort was with 200 subjects. Out of which 50 were lost to follow up.

Thus, if we suppose all the 50 subjects who were lost to follow up had disease, then the total number of subjects with disease will become 12 + 50 = 62 (12 were already detected with disease out of 150)

Thus, the incidence rate would be (62/200)*100 = 31%

Now, suppose all subjects who were lost to follow up do not have disease. Thus, in the cohort, out of 200 subjects only 12 got detected with disease. (12 from 150 subjects and 0 from 50 subjects)

Thus, the incidence rate would be (12/200)*100 = 6%

c) Based on these calculations, do you think the 8% incidence rate is valid (i.e., accurate) Why yes or why not?

When we have a cohort of 200 subjects out of which 25% (50 subjects) are lost to follow up, then it a huge lost to follow up ratio. Researcher should design a study where follow up period is relatively less so that s/he can minimize this loss of data. Such a loss in data would not yield a correct incidence rate. As we have seen in the part b above, including lost to follow up subjects as diseased or not, the incidence rate varies a lot from 6% to 31%. Which according to me is not a perfect picture of disease one would like to see. Thus, With this calculations 8% incidence rate will not be called as accurate incidence rate.

Related Questions

Navigate

• Browse (All)
• Browse 1
• Subjects

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.