Highlights
Despite an electronic health record (EHR)-embedded intervention and access to education, clinicians continued to report discomfort initiating treatment for opioid use disorder (OUD) in the hospital.
Inconsistent messaging across hospital sites, uneven access to addiction expertise, and the complexity of acute inpatient care limited adoption of the intervention.
The study suggests that successful implementation requires more than decision support tools; it also needs leadership alignment, stigma reduction, and reliable referral pathways after discharge.
Study Background
Opioid use disorder remains a major contributor to preventable morbidity and mortality in the United States, and hospitalization represents a high-risk but potentially high-yield opportunity to initiate evidence-based treatment. Medications for opioid use disorder (MOUD), including buprenorphine and methadone, reduce withdrawal, improve engagement in care, and lower opioid-related mortality. Yet in-hospital initiation remains substantially underused, even though many patients admitted with or affected by OUD would benefit from treatment started before discharge.
This implementation gap is especially important because hospitalization often brings patients into contact with clinicians, social workers, pharmacists, and discharge planners who can coordinate treatment initiation and linkage to follow-up care. Electronic health record-embedded tools are appealing because they can deliver guidance at the point of care without interrupting workflow. However, digital tools alone do not guarantee uptake. The real-world barriers to implementation, particularly in busy inpatient settings, often include stigma, limited training, uneven specialist support, and uncertainty about outpatient continuation of treatment.
The study by Calcaterra SL, Lockhart S, Binswanger IA, and Holtrop JS explored these barriers in a large Colorado health system that implemented a non-interruptive EHR-embedded OUD intervention across 12 hospitals. Rather than measuring clinical efficacy directly, the investigators sought to understand why adoption was challenging and what organizational factors shaped implementation success or failure.
Study Design
This was a qualitative implementation study conducted at 6 and 12 months after rollout of the intervention. The research team used focus groups and key informant interviews with a multidisciplinary sample of hospital-based physicians, advanced practice professionals (APPs), nurses, social workers, and pharmacists working across three Colorado hospital regions: northern, southern, and central.
In total, 61 professionals participated: 24 hospital-based clinicians, 19 nurses, 12 social workers, and 6 pharmacists. The investigators used a directed content analysis with both deductive and inductive coding approaches. This means they analyzed the data using pre-specified implementation concepts while also allowing new themes to emerge from the interviews and discussions.
The intervention itself was described as non-interruptive and embedded in the EHR, and it provided OUD treatment guidance. The study did not compare patient-level outcomes against a control group; instead, its primary endpoint was understanding adoption barriers, workflow effects, and implementation experiences from the perspective of frontline staff.
Key Findings
The analysis identified five major themes that shaped whether the intervention was used effectively and consistently.
1. Education alone did not eliminate discomfort with OUD treatment
Even after access to educational sessions and MOUD initiation protocols, many participants still felt uncomfortable treating OUD. This finding is important because it suggests that knowledge gaps are only part of the problem. Stigma, fear of precipitating withdrawal, uncertainty about dosing, concerns about managing polysubstance use, and limited experience with addiction care likely continued to influence clinician behavior. In implementation terms, this represents a classic “knowing-doing gap”: providers may know that treatment is recommended, but still hesitate to start it in practice.
2. Messaging differed across hospitals, leading to variable uptake
Participants described inconsistent health system messaging and communication. In some settings, leadership reinforced the intervention as a priority; in others, the initiative seemed less visible or less strongly endorsed. When guidance is not consistently communicated, staff may interpret the intervention as optional, secondary, or relevant only to certain teams. That variability can quickly translate into uneven use across sites, particularly in multi-hospital systems where local culture strongly shapes practice.
3. Access to addiction expertise strongly influenced treatment initiation
Differential access to addiction specialists affected whether clinicians felt able to initiate MOUD. When addiction consultation support was available, staff were more likely to move forward with treatment. When that expertise was limited or difficult to access, clinicians were more likely to defer treatment or avoid initiation altogether. This finding highlights an important implementation principle: point-of-care tools work best when paired with expert backup, especially for clinicians who do not routinely manage OUD.
4. Increased attention to withdrawal assessment had both benefits and unintended burdens
The intervention increased emphasis on assessing and documenting opioid withdrawal severity. That produced some positive effects, including greater awareness of withdrawal as a clinically meaningful syndrome and more structured documentation. However, it also created additional workflow burden and, in some cases, uncertainty about how to use withdrawal scores appropriately. This dual effect is clinically relevant. Formalizing assessment can improve standardization, but only if staff understand how to interpret the data and how it fits into treatment decisions.
5. Caring for patients with OUD remained challenging in busy hospital environments
Participants emphasized the complexity of inpatient care for patients with OUD. Competing demands, short lengths of stay, medical instability, behavioral health comorbidity, discharge timing, and limited outpatient linkage options all made treatment initiation harder. In other words, the intervention was introduced into a system already under strain. A well-designed EHR tool cannot fully compensate for fragmented care pathways or limited time on the wards.
Interpretation and Clinical Relevance
This study provides a nuanced reminder that implementation success depends on the broader care ecosystem, not just on the tool itself. The EHR intervention likely improved visibility of OUD treatment options and encouraged more systematic assessment, but it did not fully overcome structural and cultural barriers. The findings are consistent with broader implementation science: interventions are more likely to succeed when they are supported by leadership, aligned with workflow, reinforced over time, and connected to practical resources such as consultation services and discharge follow-up.
From a clinical standpoint, the study reinforces several priorities. First, training should not be a one-time event; it should be repeated and reinforced. Second, hospitals need clear and consistent messaging that MOUD initiation is an expected component of evidence-based inpatient care. Third, addiction consultation services or other expert support should be readily available, particularly in systems where generalists provide most inpatient care. Fourth, discharge linkage matters. Clinicians are more likely to initiate treatment when they know patients have a realistic path to continue care after hospitalization.
The study also underscores the role of stigma. Although not always explicit in clinical documentation, stigma can shape clinician comfort, influence assumptions about patient motivation, and reduce willingness to address OUD directly. Training that addresses stigma alongside prescribing mechanics may be more effective than education focused narrowly on protocols.
Expert Commentary
One of the most valuable aspects of this study is that it shifts the conversation from “Did the EHR tool work?” to “What does a hospital need in order for an EHR tool to work?” That distinction matters. In implementation research, low adoption often reflects system-level barriers rather than the absence of clinical benefit. MOUD is effective, but its use depends on the ability of clinicians and hospitals to operationalize it.
The multi-site design is also informative because it reflects real-world heterogeneity. In large health systems, practice change rarely spreads uniformly. Local leadership, unit culture, staffing patterns, and specialty support all vary, and those differences can either amplify or dilute a central intervention. The reported variation across hospitals suggests that implementation strategies must be tailored locally even when the intervention itself is standardized.
There are limitations to keep in mind. Qualitative findings are not intended to estimate effect size or prove causality. The sample, while multidisciplinary, came from a single large Colorado health system, which may limit generalizability to smaller hospitals or systems with different staffing models. Participant views may also reflect the experiences of staff who were available and willing to engage, introducing possible selection bias. Still, the depth of the thematic analysis offers actionable insight that quantitative outcome studies alone often miss.
These findings also align with current clinical and public health priorities. National efforts increasingly emphasize hospitalization as a critical touchpoint for initiating OUD treatment and improving continuity after discharge. The practical challenge is that initiation requires not only willingness to prescribe, but also workflow support, patient engagement, and access to ongoing care. This study helps explain why many systems continue to struggle despite having protocols on paper.
Conclusion
This qualitative assessment shows that an EHR-embedded, non-interruptive OUD intervention can surface important opportunities for improving inpatient addiction care, but it is not a stand-alone solution. Persistent clinician discomfort, inconsistent leadership messaging, variable access to addiction expertise, workflow burden, and difficult discharge logistics all limited adoption across 12 Colorado hospitals.
The most actionable lesson is that increasing in-hospital MOUD initiation requires a systems approach: consistent leadership endorsement, repeated education, stigma reduction, accessible addiction consultation, and robust outpatient linkage resources. For hospitals seeking to improve OUD care, the intervention should be viewed as one component of a broader implementation strategy rather than the endpoint itself.
Funding and ClinicalTrials.gov
The abstract provided does not specify funding information or a ClinicalTrials.gov registration number.
References
1. Calcaterra SL, Lockhart S, Binswanger IA, Holtrop JS. A Qualitative Assessment of an Electronic Health Record-Embedded Intervention to Increase In-Hospital Opioid Use Disorder Treatment Initiation. J Gen Intern Med. 2026 Jun 15. PMID: 42298203.
2. SAMHSA. Medications for the Treatment of Opioid Use Disorder. Treatment Improvement Protocol (TIP) 63. Updated guidance on evidence-based MOUD use.
3. American Society of Addiction Medicine. The ASAM National Practice Guideline for the Treatment of Opioid Use Disorder: 2020 Focused Update.
4. Wakeman SE, Barnett ML. Primary care and the opioid-overdose crisis — buprenorphine myths and realities. N Engl J Med. 2018;379:1-4.
5. Huhn AS, Dunn KE. Why aren’t physicians prescribing more buprenorphine? J Subst Abuse Treat. 2017;78:1-7.
