Weir Jean E, Jones Carrie
SIL Clinic, Ukarumpa, Papua New Guinea.
P N G Med J. 2008 Sep-Dec;51(3-4):155-9.
Rapid survey methodologies are widely used for assessing immunization coverage in developing countries, approximating true stratified random sampling. Non-random ('convenience') sampling is not considered appropriate for estimating immunization coverage rates but has the advantages of low cost and expediency. We assessed the validity of a convenience sample of children presenting to a travelling clinic by comparing the coverage rate in the convenience sample to the true coverage established by surveying each child in three villages in rural Papua New Guinea. The rate of DTF immunization coverage as estimated by the convenience sample was within 10% of the true coverage when the proportion of children in the sample was two-thirds or when only children over the age of one year were counted, but differed by 11% when the sample included only 53% of the children and when all eligible children were included. The convenience sample may be sufficiently accurate for reporting purposes and is useful for identifying areas of low coverage.
快速调查方法在发展中国家被广泛用于评估免疫接种覆盖率，近似于真正的分层随机抽样。非随机（“便利”）抽样被认为不适用于估计免疫接种覆盖率，但具有成本低和便捷的优点。我们通过将前往流动诊所的儿童便利样本中的覆盖率与在巴布亚新几内亚农村三个村庄对每个儿童进行调查所确定的真实覆盖率进行比较，评估了该便利样本的有效性。当样本中儿童比例为三分之二或仅统计一岁以上儿童时，便利样本估计的三联疫苗（DTF）免疫接种覆盖率与真实覆盖率相差在10%以内，但当样本仅包括53%的儿童以及包括所有符合条件的儿童时，两者相差11%。该便利样本对于报告目的而言可能足够准确，并且有助于识别低覆盖率地区。
