A randomised survey of the quality of antibiotics and other essential medicines in Indonesia, with volume-adjusted estimates of the prevalence of substandard medicines.

The World Health Organization warns that substandard and falsified medicines threaten public health in low- and middle-income countries. However, medicine quality surveys are often small and unrepresentative of the market, and the true scale of the problem remains unknown. We conducted a large field survey of essential medicines in Indonesia, and investigated how weighting survey results by market volume altered estimates of medicine quality. We collected 1274 samples of allopurinol, amlodipine, cefixime, amoxicillin and dexamethasone from the internet and a randomised sample of all outlet-types where medicines are sold or dispensed in seven districts across the world's fourth most populous nation. We conducted compendial testing for identity, assay, dissolution and uniformity. Samples that failed any chemical test were considered substandard. We compared raw prevalence of substandard medicines with prevalence adjusted by the national sales volume of each brand, relative to its weight among survey samples. The weighted prevalence of substandard products was 4.4%, 47% lower than the raw estimate (8.2%). Only 0.5% of samples (unweighted 1.2%) deviated from permitted limits by more than 10%. More antibiotics failed testing than other medicines (weighted prevalence 8.5 vs 3.1; raw prevalence 13.6 vs 4.9, both p<0.000). We found no relationship between quality and price; branded status; public procurement status; or outlet type. In Indonesia, unweighted survey data appeared to substantially over-estimate the health threat posed by substandard or falsified medicines. The types of sampling bias that led to over-representation of poor quality products in our survey are also indicated in other published surveys, possibly exaggerating the scale of the threat to public health posed by substandard and falsified medicines globally. Weighting survey results by sales volume likely improves robustness of estimates of medicine quality measured in field surveys.