Substance misuse: what place for women‐only treatment programs?

Abstract

The question of whether programs targeted at women are more beneficial than mixed gender programs is a key clinical and policy issue. Twenty years ago, Vannicelli pointed out that this question had not been examined using rigorous methods (Vannicelli 1984). Since then, several well designed observational studies have been conducted, but there has been only one trial, a 1989 Swedish study comparing a long-term women's residential program with mixed gender programs (Dahlgren & Willander 1989). The paper by Kaskutas and colleagues in this issue (Kaskutas et al. 2004) thus adds to a small but important body of research on the effectiveness of different treatment settings/modalities in treating women with alcohol and drug problems. It broadens the settings studied to outpatient programs, the most commonly found treatment setting in the USA, and to social model programs. Using an empirical design, Dr Kaskutas contrasts a day treatment community-based women's program with three mixed gender day treatment programs (two community-based and one hospital-based program). The community-based programs adhered to a social model program philosophy which differed from the hospital-based clinical or medical philosophy, especially in terms of staffing. How does this study add to what we know about women's programs? On the whole, findings point to more similarities than differences in outcomes between the women's and mixed gender programs. However, the multivariate analysis found higher rates of alcohol and drug abstinence in the hospital mixed gender program when all three follow-ups were included. Unfortunately, other findings are weakened by the multiple outcomes examined requiring corrections for multiple comparisons (alcohol, drug, and total abstinence; at the end of treatment, 6 months and 12 months; and comparing the women's program to each of the three mixed gender programs). Program sizes range from 22 to 42, and statistical power is limited. The differences in some baseline characteristics and outcomes may be clinically meaningful, although they are not statistically significant. For instance, type of dependence at treatment entry differed greatly: 42% of those in the women's program were both alcohol and drug dependent vs. only 17% of those in the hospital-based program. There were also large differences in abstinence rates at follow-ups that were not statistically significant due to correcting for multiple comparisons. The authors also found that the social model women's program produced outcomes similar to the mixed gender social model programs, one of which was twice as costly as the women's program. This is the key comparison, since it controls for program setting and type of staff. Given these results, a study that would compare women's programs to mixed gender programs within other treatment models, such as medical model programs is called for. Policy makers should also be cautioned to remember the potential importance of women's programs in promoting treatment access. Institute of Medicine studies (Institute of Medicine 1990; Edmunds et al. 1997) have argued that these programs may attract women to treatment who would otherwise not go. If outcomes are similar, this is a critical argument in favor of women's programs. This study suggests directions for further study. It challenges research on women's programs to move from observational to empirical designs and demonstrates that it is possible. It recommends studying different modalities, settings, and program philosophies. Because of the importance of program environment, studies also should compare residential with outpatient programs. The need for larger sample sizes is critical. These findings should spur the field to address the roadblocks discussed in the paper. Since observational studies have found better outcomes for women's than mixed gender programs, the issues of self-selection and choice should also be considered in future empirical research designs. The findings underscore the need to develop methods to study group process, as well as program environment. The questions raised are also relevant to the larger area of cultural competency. This study should generate a great deal of controversy and trigger renewed attention to even more rigorous studies of women's programs.