Is it possible to use a previous prevalence rate of tuberculosis in a neighboring country to calculate the sample size for my study in my country, knowing that there is no information in my country about the prevalence of this disease that I can use to calculate the sample size to be used?
Prevalance rate of tuberculosis of a particular place is indeed much more than reported taking into account the occurrence of undiagnosed undiagnosed and latent TB cases. Rapid sensitive cost-effective diagnostics as forwarded by Dr Dipak Kumar Chattopadhyay must be conducted first to report undiagnosed tuberculosis cases. Then by adding number of freshly diagnosed cases the accurate prevalence rate might be assayed.
Yes, it is reasonable. To estimate the sample size required to estimate a prevalence wuth specified precision, you need an initial guess as to the likely value. In this field use of data from a similar country is a recognized method.
see attached for method.
Yes, it is possible to use a previous tuberculosis (TB) prevalence rate from a neighboring country to calculate the sample size for your study, provided the epidemiological and demographic contexts are sufficiently comparable.
In estimating sample size one makes the best guess. Using data from a neighboring country should improve the quality of that guess, but socioeconomic differences between countries could still lead to a poor choice.
An alternative is to process as many samples as there is room for in the budget. The larger your sample size the narrower are your confidence intervals about your prevalence estimate. Pay close attention to demographics and how different risks in different sub-populations will interact with your sampling methodology. Depending on current prevalence, consider group testing (pooling samples), but realize that pooling can alter the detection accuracy of your disease test in several ways. Group testing only works if disease prevalence is low because group testing requires that at least one of the pooled samples tests negative. For the analysis of pooled samples there is a program PooledInfRate that is available in Excel or a better version in R. It is through the United States Centers for Disease Control and written by Biggerstaff.
Yes, it is possible to use the prevalence rate of tuberculosis from a neighbouring country to estimate the sample size for your study if no local data is available, but with caution.
- Badges
- Science topic
- Similar topics
- Medicine
- Epidemiology