I adopted a two-stage sampling strategy to conduct a study on patient medical expenses. Instead of going to the households,
We decided in the first stage to select health institutions and in the second stage 50 patients in each institution selected in the first stage of sampling.
We can evaluate the inclusion probability for the primary sampling units, but we have doubts to do for the patients.
Knowing that on the day of the team’s visit, the number of patients attending the institution may be less than the number of patients to be investigated per institution (50 patients);
At this time a second day of visit is necessary to complete the number of patients.
We do not know if each patient corresponds to a unique household.
In this case, how to determine the inclusion probability of a patient? Knowing that if the institution does not have a lot of attendance on the day of the visit, we survey all the patients present. Otherwise, from the consultation register, we draw the 50 patients using the systematic sampling method.