J Urban Health https://doi.org/10.1007/s11524-020-00452-8 Jail and Emergency Department Utilization in the Context of Harm Reduction Treatment for People Experiencing Homelessness and Alcohol Use Disorder Susan E. Collins & Silvi C. Goldstein & Bow Suprasert & Samantha A. M. Doerr & Joanne Gliane & Clarissa Song & Victoria E. Orfaly & Rddhi Moodliar & Emily M. Taylor & Gail Hoffmann # The New York Academy of Medicine 2020 Abstract People experiencing homelessness are dis- ED utilization were extracted for 6 months pre- and proportionately affected by alcohol use disorder posttreatment. Findings indicated no statistically signif- (AUD). Abstinence-based treatment, however, does icant treatment group differences on 6-month changes in not optimally engage or treat this population. Thus, jail or ED utilization (ps > .23). Exploratory analyses Harm Reduction Treatment for Alcohol (HaRT-A) showed that 2-week frequency of alcohol use was pos- was developed together with people with the lived ex- itively correlated with number of jail bookings in the perience of homelessness and AUD and community- 12 months surrounding their study participation. Addi- based agencies that serve them. HaRT-A is a compas- tionally, self-reported alcohol-related harm, importance sionate and pragmatic approach that aims to help people of reducing alcohol-related harm, and perceived physi- reduce alcohol-related harm and improve quality of life cal functioning predicted more ED visits. Future studies (QoL) without requiring abstinence or use reduction. are needed to further assess how harm-reduction treat- The parent RCT showed that HaRT-A precipitated sta- ment may be enhanced to move the needle in criminal tistically significant reductions in alcohol use, alcohol- justice and healthcare utilization in the context of larger related harm, AUD symptoms, and positive urine toxi- samples, longer follow-up timeframes, and more inten- cology tests. This secondary study tested HaRT-A ef- sive interventions. fects on more distal, 6-month pre-to-posttreatment changes on jail and emergency department (ED) utiliza- Keywords Alcohol use disorder . Alcohol treatment . tion. People experiencing homelessness and AUD (N = Harm reduction . Homelessness . Jail . Emergency 168; 24% women) were recruited in community-based department . Service utilization clinical and social services settings. Participants were randomized to receive HaRT-A or services as usual. Over four sessions, HaRT-A interventionists delivered Introduction three components: (a) collaborative tracking of participant-preferred alcohol metrics, (b) elicitation of Population-based studies conducted in the USA and harm-reduction and QoL goals, and (c) discussion of worldwide have shown that alcohol use disorder safer-drinking strategies. Administrative data on jail and (AUD) affects approximately 40% of people experienc- ing homelessness [1, 2]. This number dwarfs the 5.6% * : : : S. E. Collins ( : ) S. C. Goldstein: B.: Suprasert : prevalence of AUD observed in the general US popula- S. A. M. Doerr J. Gliane C. Song V. E. Orfaly tion [3]. This disproportionately high prevalence of R. Moodliar : E. M. Taylor : G. Hoffmann University of Washington – Harborview Medical Center, 325 AUD precipitates correspondingly high levels of health Ninth Ave, Box 359911, Seattle, WA 98104, USA and behavioral problems in people experiencing home- e-mail: [email protected] lessness. Studies on mortality rates have indicated that Collins et al. people experiencing homelessness and AUD are many AUD would be a helpful complement and could engen- times more likely to die of alcohol-attributable causes der similar improvements in alcohol and service utiliza- than the general population [4–6]. This population is tion outcomes. With this in mind, a team of researchers also disproportionately affected by crime and violence: worked in partnership with people with lived experience A USA-based study found that 77% of homeless ar- of homelessness and AUD and a community-based agen- restees report lifetime heavy episodic drinking [7]. To cy serving this population to develop harm-reduction reduce this population’s high criminal justice involve- treatment for AUD (HaRT-A) [32]. ment and health-related problems, interventions to ad- HaRT-A is a client-driven approach that emphasizes dress AUD—among other issues—are necessary. accepting people “where they’re at” in their motivation The extent of behavioral and health-related alcohol for change [19]. The HaRT-A mindset stems from the problems in this population results in disproportionately harm-reduction philosophy, which prioritizes clients’ high utilization of publicly funded services, particularly own goals over abstinence achievement and recognizes within the criminal justice and emergency healthcare sys- any movement toward reducing harm and improving tems. Studies suggest that about one-third of people QoL as positive steps on client-defined pathway to experiencing homelessness and AUD visit the ED every recovery [20]. HaRT-A providers embody a harm- year [8, 9], and alcohol intoxication was cited by one of reduction heart-set, which is a compassionate, transpar- these studies as the most common reason for ED visits in ent, and advocacy-oriented way of being with clients. this population [8]. In fact, more severe biological (i.e., After helping clients understand the HaRT-A rationale, breath alcohol concentration) and self-report (heavy drink- providers introduce three HaRT-A components: (a) col- ing episodes) indices of alcohol use were associated with a laborative tracking of participant-preferred alcohol out- greater likelihood for ED visits [8]. The strong and syner- comes, (b) elicitation of participants’ harm-reduction gistic intercorrelations among homelessness, AUD, and and QoL goals, and (c) discussion of safer-drinking incarceration are also well-documented [10, 11]. strategies. A recent, 3-month, two-arm randomized con- Given the extent of this population’s use of the crim- trolled trial (N = 168) showed that compared to services- inal justice and emergency healthcare systems and their as-usual control participants, HaRT-A participants associated costs, there has been a call for more engaging evinced significant and clinically meaningful reductions and effective means of addressing AUD in this popula- on peak alcohol use, AUD symptoms, and alcohol- tion. In particular, harm-reduction approaches have been related problems as well as increased confidence for viewed favorably by people experiencing homelessness engaging in alcohol harm reduction [32]. and AUD and the providers that serve them [12–18]. As The aim of the present, secondary study was to applied to alcohol use, harm reduction refers to a broad further test the HaRT-A effect on criminal justice and spectrum of compassionate and pragmatic approaches emergency healthcare utilization as measured by num- that seek to reduce alcohol-related harm without requir- ber of jail bookings and ED visits, respectively. It was ing abstinence or use reduction [19, 20]. expected that compared to services-as-usual control par- In the homeless population, harm-reduction ap- ticipants, HaRT-A participants would evince statistical- proaches applied to date have primarily comprised ly significant reductions in number of jail bookings and community-level interventions, such as Housing First ED visits. Additional, exploratory analyses were con- (i.e., the provision of immediate, permanent, low-barri- ducted to test predictors (i.e., sociodemographics, sub- er, non-abstinence-based housing [12, 21–25]) and stance use frequency, and indices of physical and mental managed alcohol programs (i.e., programs that prescribe health-related QoL) of service utilization. and medically supervise alcohol use in the context of housing or emergency shelter [14–17, 26–30]). Re- search studies have shown that such community-level harm-reduction interventions are associated with de- Methods creased contacts with the criminal justice and emergen- cy and safety-net healthcare systems [22, 24, 28, 30, 31]. Trial Design Given the effectiveness of community-level harm-re- duction approaches in this population, it would stand to The design of the parent study was a parallel, two-arm, reason that individual-level harm-reduction treatment for 3-month, non-blinded, randomized controlled trial of Jail and Emergency Department Utilization in the Context of Harm Reduction Treatment for People... HaRT-A compared to a services-as-usual control condi- changing drinking behavior to reduce its “negative side tion in a sample of homeless people with AUD. effects,” where 0 = not at all important and 10 = totally important. Readiness rulers have been shown to be Participants associated with alcohol outcomes and have evinced adequate psychometric properties [35–38]. This single Participants were 168 people experiencing homeless- item was used as a predictor in exploratory analyses. ness and AUD who were recruited for the parent study Frequency of smoking, alcohol, and other drug use from three community-based healthcare and social ser- was assessed using the psychometrically sound Addic- vice agencies in Seattle, Washington (for more details, tion Severity Index – 5th Edition [39]. Participants were see parent study [32]). Participants were at least 21 years asked to self-aggregate and report on the total number of of age, met criteria for an AUD according to the DSM-5, days on which they used alcohol, cigarettes, and other and met federal criteria for homelessness for at least