This paper shows the success and challenges in caring for this high-risk population. Women in this study had high rates of co-occurring medical and mental health problems, high poverty rates, high rates of interpersonal violence, and high smoking rates. Despite these challenges, with the appropriate interventions and care, women with addictions can have relatively normal birth outcomes. Previous work by one of the authors showed a 4-fold increased risk of preterm delivery with methamphetamine use in this population . We definitely did not see this increase, as our preterm delivery rate was not different from the cohort from the same hospital, as well as state and national averages (12.8 and 12.3 respectively) . We did see increased preeclampsia rates, which are not surprising as asthma, chronic hypertension, and methamphetamine use are all associated with preeclampsia .
By providing a safe environment to obtain prenatal care, we ameliorated many the effects of the drug use, especially preterm delivery rates. We have shown that increased prenatal visits, which roughly translates into participation with other clinic services is associated with increased abstinence and decreased relapse postpartum, though this could be a result of selection bias. Other authors have shown this with other drugs . Quality prenatal care of at least 4 visits has been shown to significantly improve birth outcomes, decrease preterm delivery, and increase fetal birth weights . Studies with cocaine-using pregnant women demonstrate that "comprehensive care" increases the likelihood of carrying to term, having fewer complications, being drug free at delivery, and having fewer exposed repeat pregnancies .
Our experience has shown us the importance of a comprehensive approach to treating women with addiction. It is important to address all components of the woman's life, including work and family: As one of our clients said, "I began using because I was working nights. My parents don't speak English and have many health problems, so I would have to get up after just and hour or two of sleep and take them to their doctors' appointments. I used ice [MA] to have energy to do that." Addressing nutrition is very important for example, as weight gain can be a powerful trigger for relapse, especially postpartum. One woman in our clinic stated that she was very unhappy with her weight postpartum which hadn't been a problem with her past pregnancies as she "just started using again and the weight came right off."
We believe a harm-reduction approach does work with these women, as we had high rates of engagement in services. As the clinic is in a house, many women stated they felt extremely comfortable there and would come when in crisis or just "to hang out." We had very high rates of abstinence during pregnancy, despite not mandating an abstinence-only approach. We believe this approach encourage honesty about their use. The women felt very comfortable discussing their use, and many times would disclose their relapses before a positive toxicology result was obtained.
Another success is the relatively low rates of post-partum depression, being fairly comparable to population-based rates of 14-21% [53, 54], given the high incidence of antecedent depression in this population, and that a history of depression is one of the biggest risk factor in postpartum depression . Having a safe, supportive environment postpartum was cited by many women as helpful to their general wellbeing.
The low rates of HIV and Hepatitis C point to the success of other harm reduction approaches; Hawaii was the first state in the nation to have a syringe-exchange program, which has been operating since 1990 and the majority of women in Hawaii smoke MA rather than inject.
The high rates of asthma are concerning, exceeding Hawaii's traditionally elevated rates of approximately 11% of the female population , and there are multiple explanations: higher asthma rates among Native Hawaiians; extremely high smoking rates in this population, and possibly the concomitant use of albuterol with MA to potentiate its effects. The high percentage of Native Hawaiians in this clinic population points to a large health disparity, which we have described before , and needs to be addressed with community-based research models.
The protective effect of residential treatment is seen within the drug use and custody domains. This result is to be expected as the residential treatment facility is the only one on the island of Oahu that allows women to bring a child into treatment with them. Women often chose this center in order to maintain custody. This type of program has demonstrated success in other areas and should be expanded.
The effect of a history of domestic violence on poor infant outcome is somewhat surprising, but not entirely unexpected, given high levels of psychosocial stress is associated with higher cortisol levels, and worse pregnancy outcomes . Not surprising is the effect of maternal medical conditions on infant outcomes. Women should ideally enter pregnancy in optimal health, medically as well as mentally to insure the health of the infant.
While the majority of women were able to stop using drugs during pregnancy, they continued to smoke cigarettes. This is concerning given emerging data by us and others that smoking may be more harmful than MA use (at least on measurable pregnancy outcomes)  (Wright et al: The placental and pregnancy effects of methamphetamines and smoking, submitted) Chang, L. et al (personal communication, Hawaii Addictions Conference, March 19, 2010). While the numbers are too small to show any association with MA and congenital anomalies, it is reassuring that there were no large increases in these rates over baseline.
These women have high gravidity and parity, and a majority of these women have previously lost custody of at least one child. Most of the women in the "none" category were clients of a six-month residential treatment facility which permitted no sexual contact and thus thought themselves not to be at risk for pregnancy. Of the 78 women we have information on, only 17% became pregnant again within 18 months. Statewide, 21% of Medicaid women became pregnant within 15 months . We showed that women who lost custody were 2.5 times more likely to not use reliable contraception and have repeat pregnancies. Helping women maintain custody seems to encourage women to chose reliable contraception. Our low repeat pregnancy rates is a great success.
There is room for improvement in the postpartum period, however, as 22% of the women were lost to follow-up. Of the women who did follow up, there was a 13% relapse rate to heavy usage at 6 months. Addition of pediatric care will help to keep women involved with the clinic, and allow long-term follow-up of the children and their outcomes.