Hi all
I am planning a study to assess the efectiveness of a Clinical Practice
Guideline on DVT prophylaxis. The CPG uses a risk classification scheme to
help in deciding which method of prophylaxis is appropriate. So, the
variability in CPG usage could be in part due to variability in Phisicians
assesment of risk. In order to measure this I have planned to perform an
agreement study between several physicians, using Clinical vignettes (short
summaries of patients with the necesary data, but presented in the style we
use to present patients).
I will use STATA with the KAP command who allows me to compare several
raters.
questions : A) How many vignettes I must include? I am thinking about two
or three for each of four risk categories. There is a sample size
calculation, apart from the ones for proportions? B) Do you know references
about using vignettes in biomedical research?
Many Thanks
Dr. Martín O´Flaherty
Servicio de Clinica Medica
Hospital Universitario Austral
Peron 1500 B1664INZ
Pilar, Buenos Aires
Argentina
|