Faculty Bibliography

Background: Patients with opioid use disorder (OUD) have improved outcomes on buprenorphine (BUP) and are twice as likely to be in treatment if BUP is initiated in the Emergency Department (ED) compared with standard referral. Our objective is to assess staff readiness for ED-initiated BUP with referral in highneed, low-resource community and academic EDs in New Hampshire and Manhattan.
Method(s): As part of an implementation study, in summer 2018, we conducted a mixed-methods formative evaluation using the modified Organizational Readiness to Change Assessment (ORCA) and rulers of readiness to begin ED-BUP on a scale of 0 (not) to 10 (completely), in EDs of one critical access hospital, one community hospital with urban/rural catchment, and one academic public safety-net hospital. ORCAs were administered electronically prior to inperson focus groups with ED physicians, PA/APRNs, nurses, social workers, and community providers. ORCAs and rulers were compared across sites and between EDs and the community providers. We conducted focus groups guided by an implementation framework to identify barriers and facilitators to the implementation of ED-BUP at each site. Findings were reported back to stakeholders, and used to develop site-specific clinical protocols and implementation strategies, along with education and resources designed to enhance uptake of ED-BUP at each site.
Result(s): ORCA completion rates at each site were 32%, 52%, and 17%. The ORCA showed that ED staff at two sites were ambivalent as to whether ED-BUP will improve treatment follow-up; one site agreed. ED staff and community providers reported respective readiness for ED-BUP with referral at each site of 3.5 (N=15) and 6.7 (N=36); 2.0 (N=11) and 6.3(N=28); 3.5 (N=28) and 9.3 (N=6). Nine focus groups of 36 stakeholders identified several barriers (lack of knowledge/experience with BUP, workflow integration, increased time/patient burden, and limited knowledge of local treatment options) and facilitators (improved patient care, a sense of moral imperative, and local champions/leadership buy-in).
Conclusion(s): A mixed-methods formative evaluation with facilitation and stakeholder input identified ambivalence in ED staff, as well as barriers and facilitators, including specific opportunities for education and facilitation, which may enhance site-specific implementation of ED-initiated BUP

BACKGROUND:Homelessness and substance use often coexist, resulting in high morbidity. Emergency department (ED) patients have disproportionate rates of both homelessness and substance use, yet little research has examined the overlap of these issues in the ED setting. We aimed to characterize alcohol and drug use in a sample of homeless vs. non-homeless ED patients. METHODS:A random sample of urban hospital ED patients were invited to complete an interview regarding housing, substance use, and other health and social factors. We compared substance use characteristics among patients who did vs. did not report current literal (streets/shelter) homelessness. Additional analyses were performed using a broader definition of homelessness in the past 12-months. RESULTS:Patients who were currently homeless (n = 316, 13.7%) versus non-homeless (n = 1,993, 86.3%) had higher rates of past year unhealthy alcohol use (44.4% vs. 30.5%, p < .0001), any drug use (40.8% vs. 18.8%, p < .0001), heroin use (16.7% vs. 3.8%, p < .0001), prescription opioid use (12.5% vs. 4.4%, p < .0001), and lifetime opioid overdose (15.8% vs. 3.7%, p < .0001). In multivariable analyses, current homelessness remained significantly associated with unhealthy alcohol use, AUDIT scores among unhealthy alcohol users, any drug use, heroin use, and opioid overdose; past 12-month homelessness was additionally associated with DAST-10 scores among drug users and prescription opioid use. CONCLUSIONS:Patients experiencing homelessness have higher rates and greater severity of alcohol and drug use than other ED patients across a range of measures. These findings have implications for planning services for patients with concurrent substance use and housing problems.

Background: Substance use (SU) is common among ED patients, with 1 in 10 ED users having an alcohol or drug use disorder. ED patients also have high levels of social needs such as homelessness and food insecurity. Yet, little research has examined how such social determinants of health (SDOH) intersect with SU among ED patients. In this study, we compared the prevalence of several SDOH among ED patients who did and did not screen positive for unhealthy alcohol and drug use. Methods: We surveyed a random sample of ED patients at a NYC public hospital from Nov 2016-Sept 2017. Eligible patients were >=18 years old, medically/psychiatrically stable, not in prison/police custody, and spoke English or Spanish. RA shifts occurred during all days of the week and hours of the day. RAs administered a 20-40 minute survey with validated single-item screeners for unhealthy alcohol and drug use and questions on self-reported past year social needs from national surveys or prior studies. We compared prevalence of SDOH by SU screening status in bivariate analyses with chisquare tests. Results: About half of patients (52.0%) approached were ineligible, primarily because they were medically unfit, intoxicated, or in prison/police custody. 2,396 of 2,925 eligible patients participated (81.9%); 76 duplicate patient records were removed, leaving a final sample size of 2,321 patients. Nearly one-third (32.3%) screened positive for unhealthy alcohol use and 21.8% for any drug use. Regarding SDOH, rates among patients overall vs. those with unhealthy alcohol use vs. those with drug use were: 1) homelessness 13.8%, 18.7% (X2 p<0.01 for difference between those who did vs. did not screen positive), 25.8% (p<0.01); 2) housing instability 25.2%, 29.5% (p<0.01), 35.9% (p<0.01); 3) food insecurity 50.9%, 56.3% (p<0.01), 63.4% (p<0.01); 4) inability to meet essential expenses 40.8%, 45.9% (p<0.01), 52.7% (p<0.01); and 5) unemployment 43.3%, 45.1% (p=0.23), 55.1% (p<0.01). Conclusion: Rates of homelessness and other social needs were high among ED patients in this study, suggesting the importance of considering SDOH in emergency medicine practice. We add to prior literature by showing that these needs were even higher among patients who screened positive for SU. These findings are important, as patients' significant comorbid social needs may affect the success of ED-based efforts to address substance use

Study Objectives: Despite the high prevalence of unhealthy alcohol use among emergency department (ED) patients, screening and intervention is rarely integrated into care. Streamlining assessment procedures could boost adoption as time and competing priorities are commonly cited barriers. A single-item screening question (SISQ) has been validated (Saitz et al., 2009, 2010); however, further assessment is then needed for risk stratification to inform the intervention. The 10-item Alcohol Use Identification Test (AUDIT) is often used but may not be practical for the ED. We hypothesized that use of categorical, instead of dichotomous, response options for the SISQ would provide information on drinking severity and eliminate the need for further assessment to guide intervention. Methods: Using a non-targeted, systematic approach, we screened and assessed a purposive sample of English-or Spanish-speaking adults from an urban, public ED between the hours of 8am to 12am on each day of the week. We asked each participant the SISQ, "How many times in the past 12 months have you had [X] or more alcoholic drinks in a day?" (where X is 5 for men and 4 for women). Patients with positive responses (ie, any heavy use) were asked to select the closest frequency of use among the following categories: "less than monthly," "monthly," "weekly," or "daily or almost daily," followed by AUDIT. We tested our a priori hypothesis that heavy drinking frequency dichotomized to [less than or equal to monthly] vs. [weekly or more] could discriminate between patients with AUDIT scores of 0-15 (zones 1-2) and 16-40 (zones 3-4), to determine who should receive brief advice vs. brief intervention. We calculated the sensitivity, specificity, likelihood ratios, area under the receiver operating curve (AUROC), and performed k-fold cross-validation using k=5 folds. Results: 1310 of 4281 patients (30.6%) screened positive for unhealthy drinking. Among those, 72.3% had an AUDIT score < 15; 27.7% had a score >15. Figure 1 shows the positive relationship between AUDIT score and heavy drinking frequency. Using a frequency of weekly or more to identify patients with an AUDIT score of > 15 had a sensitivity of 89.3%, specificity of 74.0%, positive likelihood ratio of 3.43, negative likelihood ratio of 0.14, AUROC of 0.90 (95% CI: 0.88-0.92), and cross-validation AUROC of 0.85 (95% CI: 0.82-0.87). These relationships held across age, sex, and race/ethnicity. When AUDIT scores of < 15 were imputed for patients who screened negative on the SISQ, the AUROC and cross-validation AUROC rose to 0.98 (95% CI: 0.97-0.98) and 0.97 (95% CI: 0.96-0.97). Conclusions: In this sample of ED patients, the categorized frequency of heavy drinking reported in the single-item screening question was excellent at discriminating AUDIT scores at a cut-off of 15, and thus at determining intervention needs. [Figure Presented]