OBJECTIVETo review the re-examination sampling method and procedure of national immunization survey for the national immunization program vaccines coverage at the township level by the national level in 2013, China.

METHODSAccording to the result of immunization coverage at the township level through self-assessment by county, all townships were stratified by whether a township with ≥ 90% immunization coverage for the second dose of measles-containing vaccines (MCV2), and a township was urban or rural. After then, 2 townships for each province were randomly sampled from those strata through the surveyselect procedure of the statistics analysis system 9.2 (SAS 9.2).

RESULTS64 townships were randomly sampled from 32 provincial units for the re-examination, i.e. an urban township and a rural township each province, of which there were 38 townships with ≥ 90% MCV2 coverage and 26 townships with < 90% MCV2 coverage. There were 24 urban townships, 2 rural townships with < 90% MCV2 coverage and 8 urban townships, 30 rural townships with ≥ 90% MCV2 coverage, respectively.

CONCLUSIONSBased on the information of the self-assessment result of immunization coverage survey by counties, Stratified randomized sampling were employed for the re-examination through the Surveyselect procedure of statistics analysis system (SAS) to implement the sampling procedure. This can enhance the sampling efficiency, ensure the randomness of the sample, and make the survey much more representative and comparable.