who had been taught a bleach disinfection procedure consisting of two complete fillings of the syringe with bleach, followed by two complete fillings with rinse water, and discarding of the used cleaning solutions. Study participants were then evaluated for proper disinfection technique 6 to 12 months following initial training. At follow-up, a substantial proportion of the injection drug users did not perform all of the previously taught steps, and less than half completely filled the syringe with bleach or rinsed the syringe twice with bleach. These investigators also reported that compliance with disinfection methods decreased as the number of steps required increased.

Several behavioral characteristics of injection drug users have been found to be associated with use or nonuse of needle and syringe disinfection (Brown and Beschner, 1993; Latkin et al., 1992; Gleghorn et al., 1994; Edlin et al., 1993). Injection drug users who use shooting galleries, or who share needles and syringes with more than one partner, may be more likely to use bleach or alcohol for disinfection (Latkin et al., 1992). Neither injection frequency nor cocaine injection has been found to be associated with disinfectant use (Latkin et al., 1992). Injection drug users who also smoke crack cocaine report a lower rate of cleaning needles and syringes prior to injecting than nonsmokers, even though they inject less frequently (Edlin et al., 1993). Crack cocaine-smoking injection drug users are also more likely to use unsafe needles and to share rinse water, cookers, and cotton (Edlin et al., 1993).

Injection drug users who disinfect their injection equipment with bleach or alcohol are also more likely to be acquainted with other HIV risk reduction behaviors (Celentano et al., 1991). Knowledge of bleach effectiveness, condom effectiveness, and several other HIV transmission risks is higher among injection drug users with a history of bleach or alcohol disinfection. Further studies of beliefs and attitudes that differentiate bleach users from nonusers or that predict bleach use are needed to increase the adoption of effective cleaning strategies.

The National AIDS Demonstration Research (NADR) program represents a large-scale multisite prevention effort supported by the National Institute on Drug Abuse (National Institute on Drug Abuse, 1994). A total of 28 sites participated in this aspect of the NIDA initiative. This undertaking has produced what is probably the largest single database on out-of-treatment injection drug users with evaluative data on outreach program effectiveness. Brown (1994) describes the NADR "standard" intervention as typically involving at least: (1) the provision of information about and distribution of condoms and bleach by outreach and counseling staffs; (2) one session of individual educational counseling regarding AIDS transmission and strategies for AIDS risk reduction; (3) an offer to make available HIV testing; (4) posttest counseling for those accepting HIV testing;