The ability of RBT to retain participants in drug treatment is consistent with prior studies of RBT in non-pregnant patient.e.g.,17
Retention in drug treatment has been associated with improved longer-term treatment outcomes, including drug abstinence, improved maternal-child relationships, and employment.24,25
The dramatically shorter treatment retention in the Usual Care condition affirms the efficacy of RBT as a part of a comprehensive treatment program. RBT’s efficacy in improving treatment retention may be centered around its focus on having every interaction with each participant be one that is positively reinforcing to the participant. RBT has a strength-based focus and methods for breaking large goals (e.g., drug abstinence, obtaining a job) into small, concrete steps that are tracked thereby increasing participant’s accountability for her behavior and provided multiple opportunities each week for positive reinforcement from staff about her progress. Finally, RBT’s proactive outreach and assistance of participants though compiled social service systems are seen positively by the participants and may each be factors in the increased retention.
In terms of substance use disorder treatment outcomes, as expected, relative to UC, RBT showed efficacy in increasing the number of days participants lived in recovery housing.
These results suggest that recovery housing may be beneficial to pregnant patients recovering from substance use disorders by removing stimuli associated with drug use while providing monitoring and social support for abstinence from substances; however, it is not always viewed as an attractive option by all substance-abusing pregnant women due to the rules and restrictions imposed by the recovery house lifestyle. These results are also somewhat consistent with previous research in non-pregnant participants that has also shown that drug-free housing promotes treatment retention,e.g.,26
but not necessarily drug abstinence.17,27
Future studies to explore other models of supported housinge.g.,16
that might better retain patients and increase efficacy of drug use abstinence while being more attractive to individuals recovering from addiction may be of considerable interest.
Although at first look, the lack of significant differences between opiate and cocaine abstinence rates in RBT versus UC appears in contrast to previous comparisons of RBT and usual care conditions in non-pregnant patients. However, all prior studies compared RBT against a very brief and minimal ’standard care’ condition of referrals to other community treatment programs. Thus, although heroin and cocaine drug abstinence rates with RBT were not significantly different from the Usual Care condition in the present study, the fact that both groups showed significant improvements with time and relatively low rates of positive urine test results at delivery speaks to the relative success of an intensive usual comprehensive care approach in producing favorable drug use outcomes in pregnant women, thereby decreasing the ability to uncover significant differences between RBT and UC.
Of central importance to this study are the neonatal delivery-related outcomes. There was a statistically significant difference between the two conditions in the length of hospitalization after birth. Specifically, neonates of mothers who received RBT were discharged from the hospital 1.3 days earlier than neonates of mothers who received Usual Care. While the exact reasons driving the shorter hospitalization are elusive at this point, it may be the case that RBT improved certain unmeasured lifestyle factors (e.g., improved nutrition as nutritious meals are provided in the recovery houses) or RBT may have resulted in more subtle reductions in the quantity and/or frequency of drug that were not detected in this study. Finally, a reduction in hospitalization days of 31% as observed in this study with RBT relative to UC has the potential to amount to substantial reductions in health care costs to society.
The present sample of drug-dependent pregnant patients has, on average, characteristics comparable to similar such participants in other published studies. For example, the average patient characteristics include mean age in the mid-childbearing age, with a majority being unemployed, unmarried, and poly-drug users including nicotine.8,28,29
Because the maternal characteristics of the present sample are similar to those characteristics of mothers in other reports, these similarities support the ability to generalize from these findings to the larger population of pregnant women treated for substance use disorders.
Limitations to this study deserve comment. First, a broader array of maternal treatment outcomes measured over a longer period of time as complements to the delivery and one-month outcomes might have proven informative. However, because of the somewhat exploratory nature of this study for pregnant women, a focus on one-month outcomes seemed the most appropriate for both the behavioral maternal measures and the neonatal outcomes. Second, the fact that 43% of potential participants declined to participate, and 17% failed to enter treatment following randomization suggests that the treatment options may not have been overly attractive to these women. However, of the 43% who declined to participate, 51% (22% of the total pool of potential participants) did so because they were not interested in the study, while only 30% (13% of the total pool of potential participants) were not interested in recovery housing. These numbers suggest that recovery housing in and of itself may not be an unattractive feature of a treatment intervention, and did not play a critical role in declining participation. Moreover, although it is disappointing that 30 (39/128) of the participants who consented to participate dropped out prior to learning their treatment assignment, the population of pregnant women, who are abusing multiple substances are quite fluid and this drop-out rate may reflect the need for different treatment regimens (e.g., more medication and non-medication comfort measures and/or change in treatment intensity) to better address the detoxifying pregnant patient’s needs. Third, it might be argued that the present study is one in which contingency management wasn’t effective in producing behavior change. However, it is also the case that the UC participants showed significant positive change in a number of treatment outcome measures, suggesting that the intensive comprehensive usual care was itself a reasonably effective treatment approach. Further, the reward of rent paid to the recovery house may not have represented the critical components of effective contingency management (i.e., a tangible reward that is highly valued by the participant and is provided quickly after the behavior is emitted). As such, future studies of RBT should examine the use of rewards that are more diverse, of higher magnitude and tailored to the participant. Fourth, although it cannot be discounted that the efficacy of RBT is due simply to more time and attention provided to participants in the RBT condition, yet having a time and attention matched control group would have minimized our ability to examine the outcomes from RBT in the proper perspective – understanding the efficacy of abstinence-contingent recovery house placement combined with specialized individual counseling.
Finally, because a no-care condition was avoided due to ethical constraints, caution needs to be exercised in interpreting any change from baseline to one-month follow-up reflecting a treatment impact rather than simply a maturation effect. Nonetheless, this controlled study adds additional support to the literature demonstrating the efficacy of comprehensive care in engendering drug abstinence and the efficacy of RBT added to that care to reduce the length of neonatal hospitalization following birth.
There were several strengths to the present study. First, the prospective random assignment to conditions boosts the strength with which conclusions can be drawn. Second, the lack of negative birth outcomes (i.e., lack of prematurity) in both groups demonstrates the utility of a comprehensive care model for pregnant women. Finally, all treatment, obstetrical care, deliveries, and neonatal observations were performed within a single hospital and by one group of experienced medical practitioners. In summary, the results of this randomized study suggest that Reinforcement-Based Treatment integrated into a rich array of comprehensive care treatment components is a promising approach to increase maternal retention in treatment and days in recovery housing as well as reduce neonatal length of hospital stay after birth.