ex2.6

It is desired to perform a quality control audit of laboratory data from a large clinical trial for purposes of estimating the proportion of laboratory values in the database that are invalid. There are 394 patients in the clinical trial, and each person has had from 60 to 200 laboratory determinations during the course of the trial. The sampling plan chosen was to select a random sample of 10 patients and for each sample person take a random sample of 10 laboratory determinations, which would then be checked against the patient medical record for accuracy.

The following table shows the total number of laboratory determinations for the 10 sample patients described in Exercise 2.6 as well as the number of in- valid determinations among the 10 determinations sampled: Total Laboratory Total Invalid Laboratory Subject Values Values in 10 Sampled 103 123 93 200 128 165 -COOO -O NO 132 189 176 180 Based on these data, estimate the proportion of laboratory values that are in- valid using all of the data in the above table