There is little knowledge about the composition and cocaine content of street cocaine, nor about what users know about it.
373 cocaine users were face to face interviewed between May and December 2006 about the last sample of cocaine they had consumed and residual amounts of the substances actually used were analysed using gas phase chromatography coupled to mass spectrometry (GC-MS). Users rated the perceived quality of their product ("good", "average", "poor"), its "estimated percentage of cocaine" and any cutting agents it contained. Price, quantity, place of purchase (street, dealer's premise, appointment), mode of administration (sniffing, injection, smoking) and the supposed nature of the sample (natural, synthetic, no distinction ever made) were also reported. Perceived quality was modelled using multivariate multinomial regression.
The median cocaine content was 22%. Altogether, 343 samples contained cocaine, among which 75% contained at least one adulterant. The most frequently occurring were phenacetin (54% of the samples), caffeine (17%), paracetamol (14%), diltiazem and lidocaïne (11%). Users showed relatively poor discrimination concerning cocaine purity, and only 12% reported at least one of the detected adulterants. The major determinants of their perception of cocaine quality were: place of purchase, natural origin, price per gram, actual cocaine content and mode of administration.
The composition of street cocaine is largely unknown to users. Users' perceptions of cocaine quality are based partly on false beliefs and certain administration modes. This may contribute to favouring very risky practices. The effects of adulterants on users' health should be investigated.