The Emergency Department as an Opportunity for Naloxone Distribution

A Systematic Review

Alexander H. Gunn, MBS, EMT-B; Zachary P.W. Smothers, MBS, EMT-B; Nicole Schramm-Sapyta, PhD; Caroline E. Freiermuth, MD; Mark MacEachern, MLIS; Andrew J. Muzyk, PharmD, MHPE


Western J Emerg Med. 2018;19(6):1036-1042. 

In This Article


Five articles out of the 2,286 we identified met the inclusion criteria and had direct relevance to the naloxone distribution from the ED setting. The included articles varied in study design from randomized clinical trial (1) to prospective cohort studies (2), retrospective qualitative analysis (1), and descriptive study (1).

Across the studies, there is variation in the methods of implementation and evaluation of ED take-home naloxone programs. These methods of implementation included grant-funded counselors available to perform the intervention, medical student volunteers to screen patients in the ED, electronic health record (EHR) alerts that notified providers of eligible patients, and a physician's assistant (PA) with training in addiction medicine. The methods of evaluation included two studies that examined the rate of prescribing take-home naloxone, two that followed up with patients to determine effectiveness of the intervention, and one that examined the amount of time between the intervention and the next EHR-recorded opioid overdose.

In the three studies that attempted patient follow-up, the rate of successful follow-up was low, which limits the evidence for effectiveness. Authors attributed the poor follow-up to social and economic factors of the patient population, including that a majority of enrolled patients were homeless or living in impermanent housing. In the included studies, there is evidence that distributing take-home naloxone from the ED has the potential of harm reduction; however, the uptake of the practice remained low. Barriers to implementation included time allocated for training hospital staff and the burden that distribution and counseling place on ED workflow.

Banta-Green et al.[19]

This randomized clinical trial identified 241 adults at risk for opioid overdose in two hospital EDs and placed participants to either overdose education with a brief behavioral intervention and take-home naloxone, or usual care. Participants were identified through EHR review or staff referral and the majority of participants were male, white, non-Hispanic, homeless, unemployed, and more than half had used opioids every day of the previous month. The 30-minute intervention was conducted by interventionists with a master's degree who had basic training in motivational interviewing.

The primary outcome was the number of opioid-related events recorded in the EHR following the intervention for the intervention and control group. The authors found no significant difference in the number of opioid events between the control and intervention group as well as no significant difference in the time to the first overdose between the groups. The authors concluded that the null findings may have been the result of the low housing security in their study population and that more intensive interventions may have been necessary to have substantial impact on opioid overdoses. The study did not report self-reported overdoses or the use of naloxone administration due to low follow-up rates. Finally, the authors suggested that due to the constraints of timing and space in the ED, a more concise overdose and naloxone training may be sufficient and congruent with the population-level benefit in mortality rates in communities with greater rates of naloxone distribution.

Barbour et al.[20]

This prospective cohort study included 24 patients at risk of opioid overdose. In the ED, two medical students trained in harm reduction identified patients with an opioid- or overdose-related chief complaint. Participants completed a brief survey, and the medical students then delivered education in overdose reversal and naloxone usage, which took approximately 15 minutes per participant. The treating physician prescribed naloxone to eligible patients, which could be filled after discharge.

While 71 patients at risk of opioid overdose presented to the ED during this study and 43 were interested in the study, only 24 were included. For 16 eligible participants, the treating physician refused to prescribe naloxone and as a result they were excluded. Seven of the 24 patients enrolled in the study were successfully contacted for the three-month follow-up. Of these seven patients, only two had filled their prescription despite none of the other participants reporting obstacles to obtaining naloxone. The authors concluded that the greatest barrier to take-home naloxone in the ED was physician resistance. The authors believed that the high number of patients whose physician would not prescribe naloxone emphasizes the need to improve physician education about harm reduction. Another identified barrier was the pharmacy policy that prevented the ED from providing take-home naloxone directly at discharge.

Devries et al.[21]

This descriptive study of a healthcare systemwide quality improvement project describes a multisite, interdepartmental effort to increase take-home naloxone access for patients at risk for opioid overdose. This widespread initiative included the development of prescribing guidelines, educational materials for providers, EHR alerts and order sets, and the inclusion of all types of naloxone in standard pharmacy stock. In the ED, a medical student screened patients for opioid-overdose risk and eligibility for take-home naloxone. Once identified, providers would prescribe take-home naloxone and had the option of billing private insurance when available or the use of internal funds to cover the cost of naloxone for patients that were un- or under-insured.

Across the health system, the education program conducted 13 training sessions in eight departments. In the ED, specifically, 40 of the 98 physicians and 40 of the 184 nurses completed the training. In 2015, the ED had zero prescriptions for take-home naloxone and from May 2016 to September 2016, they prescribed 46 take-home naloxone kits. Of all the naloxone prescriptions, 43% were intramuscular, 53% were intranasal, and 4% were naloxone auto-injectors. The EHR alert led to a prescription for take-home naloxone 14% of the time. The authors emphasized the need for more-targeted EHR alerts to increase the rate of prescriptions and avoid alert fatigue. The study results showed that take-home naloxone programs can be initiated at large, multisite health systems and, specifically, within the ED.

Drainoni et al.[22]

This study retrospectively examined the uptake of nasal naloxone distribution in the ED following the implementation of a new policy encouraging the intervention. The study team supplemented this data with qualitative interviews of the ED staff. In the eight months prior to policy implementation, 8% of ED patients at risk for opioid overdose received take-home naloxone kits. The low distribution rate was attributed to a variety of factors, including lack of knowledge of the intervention. In addition to broader distribution of naloxone, the new policy meant that take-home naloxone kits were available 24 hours a day. Despite this, in the eight months following the policy initiation, only 7% of ED patients with the same overdose risks received take-home naloxone in the ED. Despite the low uptake, the qualitative interviews with ED staff revealed strong philosophical acceptance of the intervention. The barriers to implementation identified from interviews included logistical workflows, ambiguous staff roles, and lack of education.

The authors concluded that the successful implementation of a naloxone distribution in the ED setting is largely driven by factors other than acceptance by providers. The specific recommendations for establishing implementation included the following: creating a focused target population with a high degree of risk to initiate the innovation; developing training to engage providers in overdose prevention and harm reduction; and identifying at least one clinical champion from each role in the ED.

Dwyer et al.[23]

This prospective cohort study included 415 patients who were at risk for opioid overdose. A PA approached those patients to provide education about overdose risks as well as how to recognize and respond to an overdose. Of this group, 359 received opioid education only and 56 received opioid education and naloxone. The delivered opioid education and naloxone distribution took five minutes. Each kit cost 55 dollars for two atomized 2 mg naloxone vials; these were funded by the Massachusetts Department of Public Health. One year following the ED visit, these patients were contacted for a telephone survey.

Fifty-one of the original group of patients completed the survey: 37 patients who had received opioid education and naloxone, and 14 who received opioid education only. Of those who completed the survey, over half (53%) had witnessed an overdose since their ED visit. Moreover, within the group that witnessed an overdose, the majority (65%) called 911 and nearly all (93%) stayed with the victim. Of those who received a naloxone kit within the surveyed group, 16% reported using their kit to successfully reverse a witnessed overdose, which is consistent with previous reports of take-home naloxone programs distributed in the community.[4]

The study authors concluded that the ED is a promising opportunity for opioid overdose harm reduction and naloxone distribution to laypersons. While the results of the study demonstrated the potential for the ED setting, this study was limited by its low follow-up interview enrollment. Only 12% of the patients who received either intervention completed the survey; however, over 50% of the group that received naloxone participated in the survey.

Implementation Considerations

The variability of the implementation methods across the studies highlights the need for future research to determine the most effective practices. The following categories are general themes for implementation considerations: (1) Identification of personnel; (2) education for providers and staff, (3) EHR integration; (4) patient identification methods; (5) funding for take-home naloxone; and (6) method of dispensing take-home naloxone. Table 2 contains detailed explanations for these implementation considerations.