Door to Discharge in 60 Minutes?

No Comments

Reviewing the HOUR with Naloxone rule.



Over the last decade, deaths from drug overdose have continued to rise in the United States. In 2017, nearly 68% of all overdose deaths involved an opioid, and though deaths from prescription opioids and heroin remained stable between 2016-2017, deaths from other synthetic opioids have continued to increase. 1

Emergency physicians are on the frontline of this epidemic as the number of opioid related ED visits continues to surge. 2 There is ongoing debate and various recommendations on the appropriate observation time in the ED following naloxone administration. Previous reviews and standard references have suggested a four- to six-hour observation time for patients with heroin or other short acting opioids 3, while others suggest a one- to two-hour observation period. 4,5

In 2000, development of the St. Paul’s Early Discharge Rule for Patients with presumed opioid overdose was published and proposed a one hour clinical decision rule based on six clinical findings including ambulation, oxygen saturation, respiratory rate, body temperature, heart rate and GCS. 6


Since then, several pre-hospital studies have evaluated the idea of “treat and release” in patients who received naloxone in the field for presumed IV heroin overdoses, but refused transport to the hospital. Several retrospective studies reported mortality rates of zero to 0.49%. 7–9 Many have argued that the relatively low mortality rates can be extrapolated to the ED setting and that brief observation following naloxone may be a safe strategy.

This brings us to The Hospital Observation Upon Reversal (HOUR) with naloxone, which attempts to further clarify the ideal observation time by validating the St. Paul’s Early Discharge Rule 10.

The HOUR Trial:

The study is a single center, prospective observational validation study of a modified version of the St. Paul’s Early Discharge Rule. It compares clinical judgement and a six-component rule to determine which patients may be safe for discharge after evaluation at one hour post pre-hospital naloxone administration.  A convenience sample of 690 adult patients ³18 years of age who received naloxone by EMS, first responders (firefighters, police), or laypersons were screened from May 2016 to September 2017.


At the time of data collection, the typical duration of observation following naloxone administration in the study institution was four hours. Patients received care at the discretion of their treating emergency medicine provider and were discharged based on the provider’s clinical judgement.

Patients were assessed by an Emergency Medicine Provider (attending, resident or advanced practice provider) one hour after the first administration of naloxone. They were asked to answer “yes” or “no” to each of the following components of the decision rule.

HOUR Decision Rule:


  • Can mobilize as usual
  • Has a normal O2 saturation (>95%)
  • Has a normal respiratory rate (>10 and <20 breaths/min)
  • Has a normal temperature (>35.0 and <37.5 C)
  • Has a normal heart rate (>50 and <100 beats/min)
  • Has a GCS score of 15

If the patient had any abnormal findings or “no” answers they were considered high risk for discharge. In addition, practitioners were asked if the patient appeared stable to be discharged based on their clinical judgement.

After patients were discharged the hospital record was retrospectively examined by blinded reviewers for the presence of adverse events (AEs). Predefined AEs included: death, repeat naloxone for respiratory rate £ 10 breaths/min or O2 saturation £ 92%, delivery of supplemental oxygen for a saturation £ 92%, assisted ventilation (including BiPAP), administration of IV inotropic agents, administration of antiarrhythmic medications for sustained tachycardia >130 beats/min, cardioversion, administration of mannitol, dialysis, administration of bicarbonate for HCO3 or CO2 <5. Unclear AEs were reviewed by a blinded board-certified emergency physician. Additionally, local county medical examiner records were queried for patient death within 48 hours.


  • A total 538 patients met inclusion/exclusion criteria.
  • Fifteen percent (82 of 538 patients) had adverse events.
  • Sixty-four percent of patients had ED stays >4 hours compared to 28.8% in the derivation study and 6.5% of patients were observed <2 hours compared to 48.5% in the derivation study.
  • Eighty-five percent of patients received intranasal naloxone compared to none in the derivation study.
  • No deaths were reported within 48 hours of presentation.
  • Similar sensitivity of the prediction rule 84.1% (95% CI=76.2%-92.1%) and provider judgement 85.4% (95% CI=77.7%-93.0%).
  • Similar specificity of the prediction rule 62.1% (95% CI= 57.6%-66.5%) and provider judgment 60.9 (95% CI= 56.3%-65.4%).
  • Negative predictive value was also similar between the prediction rule, 95.6% (95% CI= 93.3%-97.9%) and clinical judgement 95.8% (93.4%-98.1%).
  • The prediction rule failed to predict 13 (2.4%) AEs, provider judgment failed to predict 12 (2.3%) AEs and the prediction rule combined with provider judgment failed to predict 10 AEs (1.9%).
  • Clinical judgement and the decision rule failed to predict two patients that required additional doses of naloxone and one patient that required BiPAP.
  • Of the 10 AEs missed by both clinical judgment and the decision rule, five involved polysubstance ingestion with benzodiazepines, alcohol or carisoprodol and two involved methadone ingestions.

The authors conclude that the decision rule may be useful to identify patients with suspected parenteral opioid overdose who are safe for discharge at one hour after treatment with naloxone.

Bottom Line: 

Can we reliably use the HOUR rule as a decision tool to safely discharge opioid intoxicated patient’s that received naloxone in the field?

While the strengths of this study include the large sample size, clearly predefined AEs and rigorous review process for unclear AEs, there are several significant limitations. The study cohort is a convenience sample at a single center. Follow-up is limited to hospital and medical examiner records and did not include phone follow-up with patients allowing for the potential to miss return ED visits at other institutions. There is a potential for recollection bias and selection bias given that AEs were identified retrospectively, and a large number of patients were excluded from the study because a one-hour evaluation did not occur.

While the original derivation study did not exclude patients based on type of opioid exposure, 86% of the original cohort reported using heroin. This validation study did not exclude patients with enteral exposure or polysubstance overdose and did not report on the overall prevalence of reported heroin vs. other opioid use.

The majority of AEs were in cases that involved polysubstance overdose or long acting opioids, and the authors correctly discuss the potential hazard of applying the rule to these patients.  From a toxicologic perspective, patients with oral overdose or overdose on long acting opioids are at high risk of recurrent symptoms after naloxone and should not be considered for short observation. Methadone has a half-life of up to 59 hours and there have been case reports of delayed, recurrent toxicity up to eight hours after last naloxone administration from illicit hydrocodone tablets contaminated with fentanyl. 11

Of note, when compared to the initial St. Paul’s Early Discharge Rule, patients in this study received much larger doses of naloxone (average of 3.1mg) likely in part due to increasing prevalence of intranasal preparations of naloxone, though it is unclear if high doses of IV or IM naloxone were also given.

This leads to two other important factors to consider, the role of naloxone in opioid overdose and naloxone’s half-life. Ideally naloxone should be dosed to reverse respiratory depression symptoms, not to fully reverse the opioid effect, potentially putting patients into precipitated withdrawal. Intravenous and IM naloxone has a half-life of about 60 minutes and IN naloxone may have a half-life of up to two hours.

Naloxone doses as low as 0.04-0.1mg IV may be sufficient to reverse respiratory effects of opioid overdose. Large doses of 2-4mg IV may potentially obscure the risk of recurrent toxicity. For example, if a patient is given 2mg of naloxone, but would have responded to a 0.1mg dose, it may take four- to five-hours to see recurrent symptoms based on half-life.

When viewed from a patient centered approach, longer observation may allow for more than just avoidance of adverse effects. More time in the ED allows providers to address other harm reduction strategies, provide access to social services and discuss assisted treatment (MAT) with buprenorphine.

Finally, 10 AEs, three which were potentially life-threatening, were missed in patients where the clinical decision rule was applied in conjunction with clinical judgment. While this study lends some support to others that suggest there may be a role for shorter observation periods in IV heroin overdose, the findings are not practice changing. Further investigation is warranted to explore which patients if any, may be safely discharged after short observation. It is essential that we continue to evaluate each patient individually and recognize that all opioid overdoses should not be treated with a one-size-fits all approach.


  1. Scholl L, Seth P, Kariisa M, Wilson N, Baldwin G, Release E. Drug and Opioid-Involved Overdose Deaths — United States , 2013 – 2017. 2019;67:2013-2017. doi:
  2. Vivolo-Kantor AM, Seth P, Gladden RM, et al. Vital Signs : Trends in Emergency Department Visits for Suspected Opioid Overdoses — United States, July 2016–September 2017. MMWR Morb Mortal Wkly Rep. 2018;67(9):279-285. doi:10.15585/mmwr.mm6709e1
  3. Boyer EW. Management of Opioid Analgesic Overdose. N Engl J Med. 2012;367(2):146-155. doi:10.1056/NEJMra1202561
  4. Clarke SFJ, Dargan PI, Jones AL. Naloxone in opioid poisoning: Walking the tightrope. Emerg Med J. 2005;22(9):612-616. doi:10.1136/emj.2003.009613
  5. Scheuermeyer FX, DeWitt C, Christenson J, et al. Safety of a Brief Emergency Department Observation Protocol for Patients With Presumed Fentanyl Overdose. Ann Emerg Med. 2018;72(1):1-8.e1. doi:10.1016/j.annemergmed.2018.01.054
  6. Christenson J, Etherington J, Grafstein E, et al. Early discharge of patients with presumed opioid overdose: Development of a clinical prediction rule. Acad Emerg Med. 2000. doi:10.1111/j.1553-2712.2000.tb01260.x
  7. Levine M, Sanko S, Eckstein M. Assessing the Risk of Prehospital Administration of Naloxone with Subsequent Refusal of Care. Prehospital Emerg Care. 2016;20(5):566-569. doi:10.3109/10903127.2016.1142626
  8. Wampler DA, Molina DK, McManus J, Laws P, Manifold CA. No deaths associated with patient refusal of transport after naloxone-reversed opioid overdose. Prehospital Emerg Care. 2011;15(3):320-324. doi:10.3109/10903127.2011.569854
  9. Vilke GM, Sloane C, Smith AM, Chan TC. Assessment for deaths in out-of-hospital heroin overdose patients treated with naloxone who refuse transport. Acad Emerg Med. 2003;10(8):893-896. doi:10.1197/aemj.10.8.893
  10. Clemency BM, Eggleston W, Shaw EW, et al. Hospital Observation Upon Reversal (HOUR) With Naloxone: A Prospective Clinical Prediction Rule Validation Study. Acad Emerg Med. 2018. doi:10.1111/acem.13567
  11. Sutter ME, Gerona RR, Davis MT, et al. Fatal Fentanyl: One Pill Can Kill. Acad Emerg Med. 2017;24(1):106-113. doi:10.1111/acem.13034


Dr. Moore is an Assistant Professor of Clinical Emergency Medicine and a Medical Toxicologist at Los Angeles County Hospital + University of Southern California. She is the course director for the medical toxicology elective at USC Keck School of Medicine, co-founder of Women in Toxicology (WiT), and creator and co-host of the podcast Tox in Ten. Follow on Twitter: @ElissaMoore3

Leave A Reply