Addiction

Proactive Addiction Consultation for Hospitalized Patients with Opioid Use Disorder: A Pilot Study
Karol

Abstract:

Background: Opioid use disorder (OUD) negatively impacts individuals, communities, and health care systems with significant morbidity and mortality. There is a need for early intervention with hospitalized patients with OUD to improve health outcomes. Proactive consultation-liaison psychiatry, with and without electronic health record tools, is an effective way to reach more patients in need and to improve health services outcomes. The impact of proactive addiction consultation is unknown.

Objective: To describe a pilot study examining the effects of proactive addiction consultation on service delivery and clinical outcomes for hospitalized patients with opioid use disorder.

Methods: This quasi-randomized, controlled pilot study tested whether, in hospitalized patients with likely OUD, early identification and prompts to initiate consultation resulted in an increased proportion of consultations completed as compared to usual care. Secondary outcomes included initiation of medication for OUD, leaving against medical advice, emergency room utilization, and hospital readmission. An electronic health record-generated daily report was used to identify patients admitted to the hospital in the previous day with diagnoses synonymous with OUD (e.g., opioid abuse and opioid dependence) or often associated with a co-occurring OUD (e.g., bacterial endocarditis and accidental overdose). Individuals confirmed through further chart review as having a high likelihood of OUD were assigned to intervention versus treatment-as-usual in a standardized fashion. The intervention consisted of our research team suggesting an addiction consultation if the patient agreed. Primary teams for the treatment-as-usual participants were not contacted, and consultations were only completed if the primary team independently requested them.

Results: Participants randomized to the intervention arm were significantly more likely to receive an addiction consultation (54.8% vs. 16.2%, P < 0.0001). There were no significant differences between the 2 groups in secondary outcomes. Secondary analyses showed that participants who received an addiction consultation (independent of group assignment) were significantly more likely to receive medication for OUD than those who did not (61.2% vs. 29.3%, P < 0.0001).

Conclusions: A proactive model suggesting addiction consultation in patients with likely opioid use disorder resulted in an increased proportion of completed consultations as compared to usual care. Addiction consultations were associated with initiation or continuation of medication for OUD, demonstrating the value of addiction consultation services in the hospital. More research is needed to show the potential impact of addiction consultation services on health outcomes.

Keywords: health services research; opioid use disorder; population health; proactive addiction consultation; proactive consultation; substance use disorder consultation.

Annotation

The findings:

In this pilot study published in JACLP, Karol and colleagues tested whether an electronic health record (EHR)–based proactive addiction consultation strategy could improve service delivery for hospitalized adults with likely opioid use disorder (OUD). Over 40 weeks, the authors screened 1,119 admissions identified by an automated report and, after chart review, enrolled 214 patients felt to have a high likelihood of current OUD; 84 were assigned to proactive outreach and 130 to treatment as usual. The intervention consisted of the study team contacting the primary team to suggest an addiction consultation if the patient was agreeable, whereas treatment-as-usual patients received consultation only if the primary team independently requested it. The primary outcome was clearly positive: patients in the intervention arm were far more likely to receive an addiction consultation (54.8% vs 16.2%, OR 6.9, 95% CI 3.4–14.0, p<0.001). The pre-specified secondary outcomes — MOUD initiation or continuation, discharge against medical advice, 90-day emergency department use, and 90-day rehospitalization — did not differ significantly between groups. However, in a secondary analysis independent of assignment, patients who actually received an addiction consultation were substantially more likely to be initiated or continued on medication for OUD (61.2% vs 29.3%, OR 5.3, 95% CI 2.1–13.2, p<0.001), suggesting that the main measurable value of the proactive approach in this pilot was improving access to specialty addiction input, which in turn increased MOUD delivery.

Strength and weaknesses:

The study's main strength is that it moves the proactive consultation-liaison model into the addiction space using a pragmatic, real-world hospital workflow rather than a highly controlled research platform. It combined automated EHR case-finding with manual chart adjudication, tested a concrete implementation strategy, and included a contemporaneous treatment-as-usual comparison group. The paper is also clinically useful because it demonstrates that a proactive addiction model can be integrated into an existing addiction consultation service with relatively modest operational changes, and it captures an important process outcome — consult completion — that is often overlooked in addiction implementation research. The weaknesses are equally important. This was a pilot, single-center, quasi-randomized study rather than a true randomized trial; assignment was constrained by service-capacity limits, and there were baseline imbalances including outpatient MOUD prescription and bipolar disorder. Critically, the intervention required active and repeated outreach by the research team — including follow-up pages and secure chats when primary teams were unresponsive — raising meaningful questions about scalability outside a research context that the paper addresses only briefly. The hospital also already had an ED-based OUD screening alert active for both groups, which likely reduced between-group separation and may have enriched the sample for patients with less overt OUD. The EHR report carried a very high false-positive burden — only 214 of 1,119 screened admissions were ultimately deemed likely to have OUD — highlighting limited specificity and substantial manual review requirements. Outcomes were restricted to one health system, OUD was inferred from chart review rather than confirmed by formal diagnostic interviews, and the study was underpowered to detect changes in harder clinical endpoints such as AMA discharge, readmission, or acute-care utilization.

Relevance:

This study is especially relevant for consultation-liaison psychiatrists because it operationalizes two of the four core elements of the APA's proactive CL framework — systematic case-finding and proactive individualized intervention — specifically within the addiction space, an application that had not previously been described. In many hospitals, patients with OUD remain underrecognized unless they present with dramatic withdrawal, overdose, or injection-related complications. This paper suggests that EHR-enabled surveillance combined with clinician-to-team outreach can markedly increase addiction consultation uptake even in patients who might otherwise be missed. That finding matters because the consultation itself — not merely identification — was strongly associated with MOUD initiation or continuation, one of the most evidence-based inpatient interventions available for OUD. At the same time, the study is a reminder that improving consult penetration is not the same as improving downstream utilization outcomes, at least not in a small pilot with limited follow-up capture. For CL psychiatrists building addiction services, the paper supports investing in proactive workflows, EHR-based identification, and integrated addiction consultation teams, while also underscoring the need to measure broader outcomes that matter in hospital psychiatry: withdrawal and pain management quality, stigma reduction, linkage to outpatient treatment, naloxone distribution, length of stay, patient experience, and post-discharge mortality and retention in care.

Stigmatizing Substance Use Terminology in Grant Abstracts Following High-Level Language Guidance
Eschliman

Abstract:

No abstract available 

Annotation (unstructured):

In this JAMA Network Open research letter, Eschliman and colleagues quantify how often stigmatizing substance-use language appeared in 6,065 National Institute on Drug Abuse (NIDA)-funded grant abstracts from fiscal years 2013–2023 and whether use changed after two high-level language interventions: the 2017 White House Office of National Drug Control Policy guidance and NIDA's 2021 "Words Matter" guidance. The authors tracked 9 stigmatizing terms, distinguishing language about substance use behavior (e.g., abuse, habit) from language describing people who use substances (e.g., user, addict, abuser, alcoholic, drunk). Notably, one of the most extreme terms — "junkie" — never appeared in any abstract across the entire study period, suggesting that the most egregious language has already been largely expunged from formal scientific writing, and that what persists is more subtle and institutionally embedded.

The overall signal was encouraging but incomplete: the proportion of abstracts containing any stigmatizing term fell from 53.8% in 2013 to 25.5% in 2023, with abuse remaining the most common term across years, followed by user; terms describing behavior were consistently more common than terms describing people. In regression models, use of any stigmatizing term declined by an average of 8% per year (RR 0.92, 95% CI 0.92–0.93), and the decline accelerated between 2017 and 2021 relative to the pre-2017 period, consistent with some responsiveness to federal guidance, though the observational design and ecological timing of these periods preclude causal conclusions. However, there was no further acceleration after 2021, and by 2023 1 in 4 grant abstracts still contained stigmatizing terminology.

The study is methodologically straightforward and credible as an observational signal, but it examines only one funding agency's abstracts, a limited set of predefined terms, and cannot determine whether wording changes reflect deeper shifts in conceptualization or stigma. For consultation-liaison psychiatrists, the letter is a useful reminder that stigma is embedded not only in bedside language but also in the scientific and funding infrastructure that shapes how substance use is studied, framed, and ultimately treated. Language reform appears to matter, but institutional follow-through beyond guidance alone remains necessary, including, as the authors themselves note, structural steps such as the proposed renaming of NIDA to the National Institute on Drugs and Addiction; the field has more work to do.