The presence of poor quality medicines in the market is a global threat on public health, especially in developing countries. Therefore, we assessed the quality of two commonly used anthelminthic drugs [mebendazole (MEB) and albendazole (ALB)] and one antiprotozoal drug [tinidazole (TNZ)] in Ethiopia. A multilevel stratified random sampling, with as strata the different levels of supply chain system in Ethiopia, geographic areas and government/privately owned medicines outlets, was used to collect the drug samples using mystery shoppers. The three drugs (106 samples) were collected from 38 drug outlets (government/privately owned) in 7 major cities in Ethiopia between January and March 2012. All samples underwent visual and physical inspection for labeling and packaging before physico-chemical quality testing and evaluated based on individual monographs in Pharmacopoeias for identification, assay/content, dosage uniformity, dissolution, disintegration and friability. In addition, quality risk was analyzed using failure mode effect analysis (FMEA) and a risk priority number (RPN) was assigned to each quality attribute. A clinically rationalized desirability function was applied in quantification of the overall quality of each medicine. Overall, 45.3% (48/106) of the tested samples were substandard, i.e. not meeting the pharmacopoeial quality specifications claimed by their manufacturers. Assay was the quality attribute most often out-of-specification, with 29.2% (31/106) failure of the total samples. The highest failure was observed for MEB (19/42, 45.2%), followed by TNZ (10/39, 25.6%) and ALB (2/25, 8.0%). The risk analysis showed that assay (RPN = 512) is the most critical quality attribute, followed by dissolution (RPN = 336). Based on Derringer's desirability function, samples were classified into excellent (14/106,13%), good (24/106, 23%), acceptable (38/106, 36%%), low (29/106, 27%) and bad (1/106,1%) quality. This study evidenced that there is a relatively high prevalence of poor quality MEB, ALB and TNZ in Ethiopia: up to 45% if pharmacopoeial acceptance criteria are used in the traditional, dichotomous approach, and 28% if the new risk-based desirability approach was applied. The study identified assay as the most critical quality attributes. The country of origin was the most significant factor determining poor quality status of the investigated medicines in Ethiopia.
