We are really good at what we do. We make hundreds of important decisions every shift. Sometimes we try to help ourselves by creating decision rules to simplify those decisions. Ideally, decision rules would be better than clinical judgement, but as we discussed last week with pediatric head injuries, they rarely are. This week we look at another paper comparing clinical judgement to a decision rule. How will they fare in the context of opioid overdose?
Clemency BM, Eggleston W, Shaw EW, et al. Hospital Observation Upon Reversal (HOUR) With Naloxone: A Prospective Clinical Prediction Rule Validation Study. Academic emergency medicine. 2018; PMID: 30592101 [free full text]
This is a prospective observational study looking to validate a previously defined decision rule. (Christenson 2000)
A convenience sample of adult patients (18 years and older) who arrived at the emergency department after being treated with naloxone.
The HOUR decision rule.
- Clearly defined adverse events included death, repeat naloxone for respiratory depression, supplemental oxygen (for hypoxemia), assisted ventilation, IV inotropes, antiarrhythmics for sustained tachycardia, cardioversion, administration of mannitol, dialysis, and administration of bicarbonate for a bicarb level less than 5.
- There were a number of scenarios that were defined as unclear adverse events, such as the administration of naloxone without evidence of respiratory depression, IV antibiotics, administration of activated charcoal, and any unscheduled surgery. These all had guidelines to determine whether they were truly adverse events.
538 patients were included, out of 690 screened. Overall, 82 patients (15.4%) had an adverse event. There were no deaths within 48 hours.
The easiest summary is that the numbers for both the HOUR rule and clinical judgement are very similar. The sensitivity of both was about 85% and the specificity 61%. A combination of clinician judgement and the HOUR rule, such that you had to pass both to be considered safe for discharge, was not clinically different.
The HOUR rule would have missed 13 adverse events, clinician judgement 12, and a combination 10. Three of the adverse events appeared to be clinically important (2 patients received more naloxone and 1 was put on BiPAP).
There are a number of methodological weaknesses with this paper, including incorporation bias, selection bias, and potentially incomplete follow-up of patients, but I am not going to go through them all here. We did a full critical appraisal of this paper as part of the SGEM Hot Off the Press series, which includes a great discussion with the lead author Brian Clemency explaining why various methodological choices were made.
The one key issue that will impact how this paper influences your practice is their choice of outcome. They use a very broad definition of adverse events that includes events directly related to opioid use, such as apnea, but also things like IV antibiotics which are unlikely to be related to this specific overdose. By using a composite outcome, they also combine really important (patient oriented) outcomes like death, with surrogate or disease oriented outcomes of questionable importance. They seemed to be trying to catch everything that can possibly go wrong with an IV drug user, rather than focusing on outcomes from the overdose itself.
Learning evidence based medicine from Dr. Jerry Hoffman on the EMA tapes, one of the things I heard over and over again was that decision rules can be helpful if they ask very simple questions, like does this patient have a broken ankle, but are doomed to fail as they increase in complexity. In an opioid overdose population, I would like to see a rule that asks something like: what are the chances that this patient becomes apneic in the next 4 hours? That is a relatively simple, objective question that a decision tool could help with. It would then be the responsibility of the physician to do a thorough history and physical to consider other possible complications. I think this rule fails by overreaching. However, if you only look at the truly important outcomes of death, apnea, and respiratory distress, this rule performs better, with only 3 misses.
Much like last week’s paper on pediatric head injury, this paper demonstrates the value of clinician judgement. Physicians were just as accurate as the decision rule. Developing a decision tool that beats clinician judgement is incredibly difficult. (Schriger 2017) Although we like the comfort of objective scores, we need to be more comfortable relying on our incredible training. A doctor’s job is to make tough decisions.
So how will this paper affect my practice? I won’t be using this decision tool. It was no better than clinical judgement, and there are too many weaknesses to suggest widespread use without more research. I think this paper can be useful to help learners understand the factors we consider when deciding to discharge patients after an overdose, although it overlooks a number of other factors that I think are important, such as the specific opioid ingested, the route of ingestion, comorbidities, and social factors such as a safe place to go. I think it is reasonable to discharge patients with normal exams after 1-2 hours of observation in the emergency department, however I am very cautious with longer acting drugs like methadone, especially when taken orally, if multiple drugs were ingested, or the ingested agent is unknown.
Maybe more important is learning how to use naloxone. When I use naloxone, patients don’t wake up. I titrate naloxone to respiratory rate and oxygen saturation, not GCS. I start with low doses (0.04 mg IV) and titrate up. As a result, my opioid overdose patients are not awake and asking to go home. When they do wake up, it isn’t because of the naloxone, but rather because their opioid has been metabolized, so I know they are safe to go home.
I won’t be using this rule in practice. Ideally, I let patients wake up on their own by using low doses of naloxone titrated to respiratory status. If a larger dose of naloxone has been used, I will use clinical judgement when making the decision the discharge, which includes the components of this rule (vital signs and ability to walk), in addition to other important factors like the type of opioid, route of ingestion, patient’s comorbidities, and social situation. In addition to considering if the patient is safe in the short term, it is also essential to consider their long term health and provide options for detox, counselling, information about needle exchange programs, and take home naloxone kits.
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;7:1110–8.
Schriger DL, Elder JW, Cooper RJ. Structured Clinical Decision Aids Are Seldom Compared With Subjective Physician Judgment, and Are Seldom Superior Annals of Emergency Medicine. 2017; 70(3):338-344.e3.
I’d love to hear your thoughts on this paper, or the management of opioid overdose in general. You can leave comments below, but if you leave them on the SGEM website, the lead author of the paper will be able to respond. You can also comment on Twitter using the hashtag #SGEMHOP. Join the conversation!