Background

Pulmonary complications after opioid overdose include: non-cardiogenic pulmonary edema, aspiration pneumonia/pneumonitis and acute respiratory distress syndrome. Multiple mechanisms have been proposed for these complications. These complications have been described in opioid overdose both with and without Naloxone administration. The purpose of the study was to determine if higher doses of Naloxone are associated with an increased risk of pulmonary complications after opioid overdose.

Clinical Question

In patients treated with Naloxone for a presumed opioid overdose in the prehospital setting by bystanders, first responders, or EMS, is a higher dose of Naloxone associated with an increased risk of pulmonary complications?

Design

This was a retrospective, cohort study involving 1,831 patients receiving Naloxone in the prehospital setting for respiratory depression presumed to be due to an opioid overdose. The study was conducted in an urban area in the US with an all paramedic emergency services. Naloxone could also be administered by bystanders and first responders. The authors defined a total Naloxone dose of > 4.4 mg or an initial Naloxone dose greater the 0.4 mg as a high dose exposure. This maximum total dose corresponded to their current (2017) protocol maximums but not to the total maximum dose of 2.0 mg that was in place during the study period. The total dose recommended by EMS protocol was exclusive of bystander or first responder dose(s). It is unclear whether Naloxone given by bystander or first responder were included in the study totals.

Outcomes

The primary study outcome was a composite of pulmonary complications which included: pulmonary edema, aspiration pneumonia and aspiration pneumonitis. This was defined by imaging interpretation or by the treating clinician. However, imaging was only obtained in 17% of patients indicating the possibility of verification bias. The study results were reported for the composite outcome and separately for pulmonary edema.

Inclusion

  1. Patients in the prehospital setting with a presumed opioid overdose AND
  2. Treated with Naloxone by bystanders, first responders or EMS AND
  3. Transferred to an Emergency Department

Excluded

Naloxone not administered, hospital chart not available

Primary Results

1,831 patients were included in the primary analysis. 80% (1,456/1,831) were considered to be likely due to opioid overdose. The mean total Naloxone dose was 2.5 mg +- 1.5 mg. 6.1% of patients received a total dose of more than 4.4 mg and 94% of patients received an initial dose of greater than 0.4 mg.

The composite outcome of pulmonary complications occurred in 26.5% (485/1,831) of patients. This seems remarkably high. The proportion of pulmonary complications in those likely to be an opioid overdose was not reported. The vast majority of pulmonary complications were classified as aspiration pneumonia or aspiration pneumonitis (95.1% (461/485)). Pulmonary edema was a rare pulmonary complication (4.9% (24/485)).

Pulmonary complications were more common in those receiving a total Naloxone dose > 4.4 mg and an initial dose > 0.4 mg in both the univariate and multivariate analyses (Table below). 42% of those receiving more than a total dose of 4.4 mg had a pulmonary complication compared to 26% who did not. 27% of those receiving more than an initial dose of 0.4 mg had a pulmonary complication compared to 13% who did not. A sensitivity analysis, excluding those not thought to be due to opioid overdose revealed similar results in those receiving a total dose > 4.4 mg (Odds Ratio: 2.08, 95% CI (1.38, 3.14). It would have been helpful to present a subgroup analysis of those not considered to be opioid related to determine the effect of Naloxone in those without opioids as a separate reason to have pulmonary complications. An additional sensitivity analysis, those receiving a total dose of more than 2 mg of Naloxone were more likely to have a pulmonary complication than those who did not (Odds Ratio: 1.66, 95% CI (1.33, 2.06)).

There was not a statistically significant association of a total dose of Naloxone > 4.4 mg with pulmonary edema (Odds Ratio: 2.23, 95% CI (0.66, 7.60)). Nor was there a statistically significant associate of an initial dose of Naloxone > 0.4 mg with pulmonary edema (Odds Ratio: 1.51, 95% CI (0.20, 11.30)). The small number of patients with pulmonary edema (n=24) contributed to the odds ratio’s wide confidence intervals.

 

PULMONARY COMPLICATIONS  
 

Naloxone

 Absolute Risk (YES) Absolute Risk (NO) Univariate

OR (95% CI)

 Multivariate

aOR (95% CI)
Total Dose > 4.4 mg 42% 26% 2.14 (1.44, 3.88) 1.84 (1.12, 3.04)
Initial Dose > 0.4 mg 27% 13% 2.57 (1.45, 4.54) 2.02 (1.07, 3.80)

 

The regression analysis does not appear to include all of the factors that were statistically significant in the univariate analysis (e.g. emesis after Naloxone). In addition, it does not appear that endotracheal intubation, requiring bag-valve-mask ventilation or comorbid pulmonary or cardiac conditions were included in either the univariate of multivariate analyses. These are factors likely to be associated with pulmonary complications.

Strengths

  • Large sample size: n=1,831 (1,446 due to likely opioid overdose)
  • Kappa (0.59) for assessment of opioid overdose vs non-opioid overdose.
  • Sensitivity analyses: Total dose > 2.0 mg, Initial dose > 0.4 mg, excluding conditions other than opioid overdose and excluding patients with missing data

Limitations

  • Unclear if total doses inclusive of bystander and first responder given Naloxone potentially over-estimating the rate of pulmonary complications
  • CXR Obtained: 17% (Potential verification bias)
  • Pulmonary Edema: n=23 (n=18 in patients considered to be Opioid OD)
    • Unable to perform regression analysis due to small sample size and despite many differences in potential confounders in those with and without pulmonary edema (Table 1)
    • Unable to draw conclusions about risk of pulmonary edema from this study
  • Regression Analysis
    • Unclear if bag-valve-mask ventilation, co-ingestants or comorbidities (in particular lungs and heart co-morbidities) were included as potential predictors in the univariate or multivariate analysis of pulmonary complications
    • Unclear if all significant predictors of pulmonary complications identified in the univariate analysis (Table 2) were included in the regression analysis (Table 4)
  • 5% rate of pulmonary complications seems remarkably high (Selection bias?)
  • Results presented for Opioid+Non-Opioid and for Opioid alone but not for Non-opioid alone. It would have been interesting to see if Naloxone is associated in pulmonary complications in non-opioid patients

Other Issues

Differences in the patient populations, EMS response time, Naloxone dosing protocols and rates of required airway assistance should be considered in determining the generalizability of this study’s results to our patient population. The low number of cases of pulmonary edema (n=16 due opioid overdose) limits the conclusions that can be made about the association between Naloxone dosing and pulmonary edema. 26.5% (485/1,831) of patients receiving Naloxone had a pulmonary complication. This seems remarkably high. The proportion of pulmonary complications in those who received Naloxone and likely to be an opioid overdose was not reported.

The number needed to harm (NNH) for pulmonary complications in patients receiving a total dose of Naloxone > 4.4 mg was 6.3 (NNH = 1/ARD = 1/(26%-42%) = 1/(0.16 = 6.3)). For every 6.3 patients treated with a total dose of Naloxone > 4.4 mg, 1 additional patient would have a pulmonary complication.

The number needed to harm (NNH) for pulmonary complications in patients receiving an initial dose of Naloxone > 0.4 mg was 7.1 (NNH = 1/ARD = 1/(13%-27%) = 1/(0.14) = 7.1). For every 7.1 patients treated with an initial dose of Naloxone > 0.4mg, 1 additional patient would have a pulmonary complication.

Author's Conclusions

“In summary, changing patterns of opioid usage, as well as the dissemination of naloxone to non-EMS personnel, have created a new paradigm for treating of opioid overdose in which increased dosage of naloxone has become the norm. In this study, we aimed to assess the effect of this evolution and found that pulmonary complications were more likely to occur in patients receiving a higher total dose of naloxone in the out-of-hospital arena. However, this relationship may or may not be causal. There is also an association between pulmonary complications and providing naloxone by the intranasal route, but the nature of this link is not known and requires further investigation and verification. Given the overall magnitude of opioid-overdose-related morbidity, and in particular pulmonary complications, there is significant potential for harm reduction from optimizing treatment strategies with an updated approach. Therefore, future studies should prospectively evaluate the effect of limiting naloxone dosing.”

Our Conclusions

The authors correctly acknowledge that the association of higher doses or Naloxone with pulmonary complications demonstrated in this study may not represent a causal relationship. While a dose-response relationship is part of the Bradford Hill Criteria for causality (see appendix), other criteria should also be met. In addition, a potential confounding variable is that the degree of respiratory depression may be associated with an increased risk of pulmonary complications. This can be illustrated by the relationship between the depth of sedation and the risk of aspiration. It does not appear as if the need for bag-valve-mask ventilation, endotracheal intubation or comorbid conditions were analyzed as possible predictors or pulmonary complications. Finally, it is unclear whether Naloxone given by bystander or first responder were included in the study totals. Early airway management should be emphasized in order to reduce pulmonary complications due to aspiration.

Potential Impact To Current Practice

Early airway management should be emphasized in order to reduce pulmonary complications due to aspiration. Patients should be assessed for potential aspiration after ED presentation.

Read More

Link: PEMCAR iBook (Apple version)

Link: PEMCAR iBook (PDF)