I am interested to carry out a study to detect prevalence of a disease in animal population. I am interested to include animals from various farms to have better epidemiological information.
I am interested to know how to calculate the total sample size as well as number of samples to be collected from each farm (as all farms do not have same population) so that the study will be statistically significant.
Again how to calculate the sample size if I am interested to know the prevalence of more than one diseases in my study population. As all diseases have different prevalence rate, so sample size will vary for each disease. how to finalize one sample size that will fit for all the diseases I am interested to screen?
Your task is to estimate a proportion (statistically speaking).
Generally, it is not complicated to determine the sample size required to get an estimate with a specified precision. There are even simple online-tools for this job:
http://epitools.ausvet.com.au/content.php?page=1Proportion
It is sufficient to determine the sample size for the lowest prevalence. The other prevalences will then be estimated with more than sufficient precision, but this should not be a problem.
In your case the solution might be more complicated, when the assumption of a homogeneous population / distribution of der dieseased animals is not reasonable. The prevalence may vary from farm to farm and from country to country. It may depend on the kind of the fram or something else. If this is the case, then you will need some guess (or preleminary data or experience) how much the prevalence varies between farms etc. You will then need to simulate data according to these covariables and determine the required sample size by simulation.
Your task is to estimate a proportion (statistically speaking).
Generally, it is not complicated to determine the sample size required to get an estimate with a specified precision. There are even simple online-tools for this job:
http://epitools.ausvet.com.au/content.php?page=1Proportion
It is sufficient to determine the sample size for the lowest prevalence. The other prevalences will then be estimated with more than sufficient precision, but this should not be a problem.
In your case the solution might be more complicated, when the assumption of a homogeneous population / distribution of der dieseased animals is not reasonable. The prevalence may vary from farm to farm and from country to country. It may depend on the kind of the fram or something else. If this is the case, then you will need some guess (or preleminary data or experience) how much the prevalence varies between farms etc. You will then need to simulate data according to these covariables and determine the required sample size by simulation.
For better study and statistical point of view at least collect 6 samples of a type. Put your need for diagnostic assays ti be done, and calculate the total sample size accodringly, You may need 2-5 ml of serum samples for epidemiologivcal studies. Do random sampling and select different age indivuals in your study. For examples six cubs, six juveniles, six adults for blood collection.
I think the EPIINFO can provide an answer to your questions as mentioned by Javier Millan.
I have read a very nice paper from the Radiology Journal, Volume 230, Number 3 Estimation of Sample Size for Complex Study Designs by John Eng. It may help
I was tasked to quantify the prevalence of a plant disease in some insects. There were a number of problems. 1) The test was expensive, about $10/test. 2) Depending on location prevalence ranged from
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size