Crack, Powdered Cocaine, Both or Neither: A Generalized Logit Analysis

Background: Few studies have explored characteristics of individuals who report use of both crack and non-injected powdered cocaine within the same time frame.

Methods: The Risk Behavior Assessment (RBA; 1993) was used to elicit information on  past month, past 48-hours and drugs used before/during sex drug use in a sample of out-of-treatment drug users (N = 8538).  Past 30 day, past 48 hour, and before/during sex drug use were coded into 1) use of crack (smokable cocaine) only, 2) powdered cocaine use only, 3) both drugs used during the timeframe and 4) neither drug used.  Generalized logit analysis was used to develop models for each timeframe (48-hour and 30-day).  For all models the reference group was using neither drug.

Results: Respondents who reported use of both crack and powdered cocaine (N=306) included both men (n=207) and women (n=99).  The 48-hour, 30-day, and before/during sex models were almost identical.  Those who reported using both crack and cocaine were more likely to: be Black (OR=1.65, CI 1.27, 2.14); bisexual (OR=1.85, CI 1.39, 2.47); trade sex for drugs (OR=4.88 CI 3.73, 6.39); to be homeless (OR=3.50, CI 2.66, 4.62), have higher number of days used alcohol in the past month (OR=1.09, CI 1.08, 1.10); and to have had gonorrhea (OR=1.42, CI 1.07, 1.88). 

Conclusions:  The general logit model highlights factors associated with use of both crack and powdered cocaine within the same timeframe, including homelessness, and being a member of an ethic or sexual minority, and sex trading.