I want to report the prevalence of an eye disease in subjects aged 20 and more drawn from patients visiting an eye out-patient department. Is it appropriate in such a study to report 95% conference interval of the prevalence rates? Or is the latter reported only when the study is done in a population?
It would be appropriate if you obtained a RANDOM sample from your patients' population. The results would apply to your patients' population.
Thank you for the answer. The selection of my patient has been this way. Patients were randomly selected on 3 days of the week. Can you suggest the most appropriate method to randomly select patients coming for the first time in a clinic when there are no pre-existing list of their names
Maybe like this. Suppose that you expect that you are going to see 500 patients in a month period (your population) and you want to study 100. You can ask excel to generate 100 random numbers between 1 and 500 (without repetition) and then select the corresponding patients as they show up sequentially to the clinic. Of course this will give you the prevalence for that particular month, which may be different than the prevalence at another month (depends on the disease).
Starting from the assumption that you are interesting to present the prevalence oh the disease in group of patients aged 20 of more years it means that you are reporting the prevalence of the disease in specific stratum. So it is necessary to think about stratified sample size. So simple random sampling is not appropriate solution. You should think about designee effect and inflation factor for sample size calculation. It is necessary to have an assumption about the structure of entire population and differences between different strata. Let say, if you have 100 patients in which you have let say 18% above 20 years of age, your sample size should be proportional to that relations in it would not be the same if have let say 30% above 20%. Could you tel me the size of the total population of patient, the size of those aged 20 years and more and how big was the sample size?
HI. Thanks for your interest. The sample collection has been already done. The inclusion were people aged 20 and above. 570 subjects were included. The age distribution is
20-29: 136; 30-39:25;40-49:27; 50-59:213; 60-69:158; 70 and more:11
Obviously the data is not well distributed. So I have taken an interval of 20-35 years; 36-50 years ..... 15 years interval. Obviously we had not planned well due to lack of knowledge on how to randomly collect the sample as age wise strata.
Hi All. Need some help again. I need an online calculator/excel format/others to calculate Prevalence in age groups adjusted for sex and prevalence in sexes adjusted for age. I am not able to work it out.
Slavoljub Stanojevic Muhammad M Hammami
- Badges
- Science topic
- Similar topics
- Education
- Higher Education
- Conferences