Dear all,
I will conduct a study to describe antibiotic prescribing behaviour of pediatricians especially for ARI cases (KAP study)
Also, I would like to measure prescription rate in those cases by using WHO drug use indicators (I.e. number of antibiotic per encounter)
My question is that how can I specify my sample as
Target population: pediatricians in outpatient clinics in different health system levels and their prescription for ARI
N.B. unfortunately there is no electronic health system to review the prescription.
Define the outcome of interest as % of outpatients' prescription with antibiotics. Use this for sample size calculation
Find this % from previous studies or do pilot study on 50 outpatients' prescription
@Abdel-Hady El-Gilany
Thanks a lot professor for your reply.
My question here is not about prescription sample only but pediatricians how can I sample them to conduct a questionnaire??
Say the prescription rate is 50% and sample is about 400 prescription to be reviewed. I will take many hospital so,total prescriptions is 400 or 400 from each hospital? (I.e. district ,general and teaching) and in this case I will do KAP survey for pediatricians who write those prescription or should I take sample from all pediatricians in those hospitals?
Thanks
Number of pediatricians will be limited, so can include all of them if the unit of analysis is pediatrician (no sampling). However to evaluate their practice you should examine their prescription. What is the outcome of interest in this situation? It should be % of pediatricians who prescribe antibiotics irrationally.
If the unit of analysis is the prescription (% with antibiotics) & you included many hospitals you should multiply the calculated sample size by a design effect (equal to number of hospitals to be included) to over the clustering effect of hospitals
- Badges
- Science topic
- Similar topics
- Chemistry
- Pharmacology
- Pharmaceuticals
- Drugs