All GMS Samples were kept in sealed containers under refrigeration at 4-10°C All samples of ergometrine and ampicillin injections were tested immediately, irrespective of sample category. For the other 11 drugs, the facility samples which were "endpoint only" (unmatched by GMS samples) were sent for testing immediately; matched facility samples (longitudinal series) were stored under refrigeration together with reference samples until completion of sampling from facilities. All paired samples for a particular drug were then sent for testing according to a schedule advised by the laboratory. Samples were stored under the same conditions at the laboratory.
All testing was carried out at the Zimbabwe Regional Drug Control Laboratory (ZRDCL) in Harare. For laboratory test methods, source of method and assay limits see Annex 1.1 - 1.13. Wherever possible, robust methods (BP, USP methods) were used. Retinol assay was carried out by a ZRDCL in-house method. Three separate measurements were done on each of the samples. Results indicated a high precision which allowed the mean value to be taken. In two cases, an outlier value was excluded and the laboratory reported a mean based on two values.
For verification a small number of samples of each drug (blind selection) were concurrently or subsequently tested in the Chemical Control Laboratory of the National Board of Health in Denmark (see Annex 2). A larger number of retinol samples was later included for verification of the assay method (which then proved reliable).