Kicking People Out of Addiction Treatment: An Update and Commentary (Izaak Williams and William White)


Kicking Image Orlando Rosu

The practice of administratively discharging people from addiction treatment, primarily for confirming their diagnosis (via continued alcohol or other drug use) or violating rules with little nexus to addiction recovery, has a long history within modern addiction treatment. The authors' calls to examine and alter this practice began a decade ago and have continued to the present (See suggested reading list below). This brief report outlines the conclusions drawn from our latest analysis of data regarding administrative discharges from addiction treatment in the United States.

Data on discharge status from addiction treatment in the United States is limited, but is available for the years 2002-2011 within the Substance Abuse and Mental Health Services Administration's Treatment Episode Data Set (TEDS). The number of states and jurisdictions reporting within the TEDS database has grown through these years from 23 to 49, with the number of patient records increasing from 792,513 to 1,922,385. In analyzing the TEDS discharge status data for 2002-2011, several clinically important findings are evident.

Trends in Treatment Completion Rates: The majority of people admitted to addiction treatment in the U.S. do not successfully complete treatment. The addiction treatment completion rate (discharge with staff approval) was 40.6% in 2002, reached its peak of 47.5% in 2006, and was 43.7% in 2011.

Trends in Treatment Non-completion: Those not completing treatment fall into seven categories: dropped out, terminated by the facility (administrative discharge, AD), transferred, incarcerated, death, other, and unknown. Between 2002 and 2011, drop-out rates showed little change with an average of 24.4% between 2002-2011, with a high of 26.6% and a low of 22.2%. Transfer rates increased from 8.6% in 2002 to 15.2% in 2011--reflecting greater levels of care available and a trend toward stepped care--assertive linkage between multiple levels of care in response to changing needs of the patient. While death and other categories have remained stable from 2002 to 2011, terminations due to incarceration have slowly risen over this time span to 2.4% of all discharges.

Trends in Administrative Discharge (AD) Rates: AD rates have ranged from 15.9% the first reporting year (2002), dropped to 8.2% in 2003, and have ranged in all subsequent years between a low of 6.3% (2008) and a high of 7.9% (2005), with a most recent (2011) reported level of 7.3%. The classification of discharge status and reporting may be administratively influenced, as indicated by states with very large treatment systems (e.g., California) that report no ADs or seemingly underreport AD data (e.g., in 2010 Arizona reported 6 terminations out of 17,452 discharges).

AD by Treatment Modality: AD rates vary considerable across service modalities, with AD rates lowest in the briefest service interventions (detoxification and hospital residential) and highest in those involving sustained care (long-term residential, opioid replacement therapy, and outpatient). It appears the longer the service relationship, the greater the probability of being administratively discharged from addiction treatment.

Demographic Characteristics: ADs are not evenly distributed across patient identifiers. When service type at time of discharge is considered, AD appears most starkly gender slanted toward men in long-term residential treatment and hospital inpatient treatment. At the aggregate level, across all service modalities, women and men are at roughly equal risk of AD irrespective of referral source; however, in controlling for age, the most pronounced gender difference in AD is seen in patients under the age of 20, with male patients at considerably higher risk for AD. Patients aged 18-29 have a higher risk for AD than all other age groups. African Americans are at significantly higher risk of AD than are other racial groups entering addiction treatment. A lower level of educational achievement--less than a high school diploma or its equivalent--a socioeconomic class indicator--significantly increase one's likelihood of AD, as does the status of no income.

AD Discharge and Number of Prior Treatment Episodes: The probability of an AD discharge decreases in tandem with an increase in prior treatment episodes until the fifth prior treatment episode, at which the risk of termination via AD increases?a trend suggesting that AD in this group may serve as a proxy for problem severity, complexity (co-occurring psychiatric illness), and chronicity. The probability of AD also increases in tandem with increased frequency of drug use in the 30 days prior to admission?another indicator that those with the most severe, complex and chronic problems are overrepresented within such discharges.

Three major conclusions can be drawn from this brief data summary.

  1. That more than half of people admitted to addiction treatment do not complete treatment is highly disturbing, particularly given efforts during the years reviewed to expand choice in addiction treatment and enhance therapeutic engagement via greater use of staged change models and widespread training in motivational interviewing.
  1. In spite of some agitation for reform, the AD rate from addiction treatment has not substantially declined in the U.S. According to the TEDS data, 1,071,091 patients admitted to addiction treatment between 2002 and 2011 were administratively discharged--more than 126,000 in the last available reporting year (including patients who were admitted multiple times in the same reporting year). This is disturbing for several reasons.

First, the use of administrative discharge as punishment (e.g., extruding patients from care for exhibiting symptoms of the disorder for which they are being treated or for rule violations unrelated to the treatment of that condition) is unprecedented in the larger health care system.

Second, there is no scientific evidence that ADs, or so-called "therapeutic discharges" have any therapeutic value as a motivational fulcrum for recovery-related behavioral change. In fact, we would suggest that practice contributes to further clinical deterioration (e.g., escalation of problematic drug use, criminal offending, incarceration, etc.) and re-enmeshment in drug and criminal subcultures at the exact time the patient is in greatest need of a recovery-enriched social environment. Reports from families of overdose deaths immediately following administrative discharge of their family member suggest the potential increased risk of mortality linked to ADs.

Third, the fact that AD decisions may inordinately target African Americans and persons of low socioeconomic standing, as well as those persons in greatest need of treatment those with highest problem severity, complexity and chronicity and the lowest recovery capitalIs particularly disturbing. The existing rate of failure to complete addiction treatment, in general, and the AD rate, in particular, is an indicator of inadequate assessment and level of care placement, weak therapeutic alliance, problems of countertransference, and racial and social class conflicts that exist across addiction treatment modalities and programs. The AD rate remains unchanged, in part, because accrediting and regulatory authorities have also failed to hold programs accountable for this critical benchmark of quality.

  1. Any addiction treatment systems reform effort must address the low treatment completion and high AD rates. Given the relationships between treatment duration, adult discharge status, and long-term recovery outcomes, enhancing treatment engagement and completion rates must become one of our highest priorities within the addiction treatment field. A wide spectrum of changes in clinical practices will be required to achieve that goal.

Further Reading

White, W., Scott, C., Dennis, M. & Boyle, M. (2005) It's time to stop kicking people out of addiction treatment. Counselor, 6(2), 12-25. See

White, W. (2014). Stop Kicking People out of Addiction Treatment. Posted April 4, 2014, /Blog/Posts/62/William-White/2014/4/Stop-Kicking-People-Out-of-Addiction-Treatment/blog-post/

Williams, I. L. (in press). Moving clinical deliberations on administrative discharge beyond moral rhetoric to empirical ethics: A call for research. Journal of Clinical Ethics.

Williams, I. L. (in press). Is administrative discharge an archaic or synchronic program practice?The empirical side of the debate. The Online Journal of Health Ethics.

Williams, I.L. & Taleff, M.J. (in press). Key arguments in unilateral termination from addiction treatment: A discourse of ethical issues, clinical reasoning, and moral judgments. Journal of Ethics in Mental Health.

William, I.L. & Taleff, M.J. (2015). Sex, romance, and dating in treatment recovery: Ethical reflections and clinical deliberations on challenging addiction decision making. Ethics in Mental Health, Open Volume, (1), 1-7.