Between 2014 and 2018, 2,215 cases were treated with naloxone by the Oslo City Center EMS (Figure 1). Eight patients declined participation. Twenty-nine patients were excluded because they were administered naloxone by others prior to EMS attendance, and no further naloxone administration was needed.
The mean age of the patients was 38.3 years, and 77.1% were men (Table 1). The median GCS was 4, and the median respiratory rate was 7 breaths/minute. As shown in Table 1, the safe injection facility was the place of attendance in 33.5% of the patients. The remaining cases (not shown) were attended in public places (50.1%), private homes (7.3%), shelters/other facilities for people using drugs (6.5%), and other places such as hotels and public transport (2.6%).
-----------------Insert figure 1 here-----------------
-----------------Insert table 1 here-----------------
Intramuscular injection was the most common route of naloxone administration (Table 2), as 91.9% (n=2,035) of the 2,215 cases received this as their initial treatment. Only a minority of patients were treated with IV naloxone; 1.9% (n=41) were treated with IV alone, and 3.8% (n=84) were administered IV naloxone after the administration of an IM dose. A minority of patients (2.5%) were administered naloxone by other routes, such as intranasal or subcutaneous routes. The use of IV naloxone as the initial treatment became less frequent during the study period, decreasing from 50 cases in 2014 to only two cases in 2018.
Among those treated with IM naloxone (n=2035), the most common dose was 0.8 mg (56.5%), followed by 0.4 mg (39.9%). Only 3.6% (n=74) received IM naloxone in other doses.
Overall, only 15.0% (n=332) of the 2,215 cases were administered a second or third dose of naloxone. The majority (82.0%) of these 332 cases were treated with only one additional dose. Among those administered multiple doses (≥2), 51.5% received IV and 48.5% received IM naloxone.
Among the 2,215 cases, the total administered naloxone dose was 0.4 mg for 33.0% of the patients, 0.8 mg for 51.2% of the patients and more than 0.8 mg for 12.7% of the patients. 3.1% received other doses less than 0.8 mg.
This included the initial and subsequent doses of titration. Only 1.0% of patients received ≥ 2 mg naloxone in total, and the maximum dose used was 3.0 mg. The mean total dose of naloxone in patients with respiratory arrest or cyanosis was 0.8 mg.
-----------------Insert table 2 here-----------------
Naloxone dose and its associations with clinical variables
Of the 2,215 cases, 1,720 cases had a valid national identity number (Table 1). This subgroup comprised 869 individuals; 76.3% were men, and the mean age was 38.6 years. The majority of these individuals (66.0%, n=574) were only attended once. Two attendances were registered in 15.4% of these patients (n=134), while 18.5% of these patients (n=161) were attended three times or more, with a maximum of 27 attendances in the same individual.
The majority (89.0%, n=1,530) of the 1,720 patients with a valid national identity number were treated with either 0.4 mg or 0.8 mg IM naloxone. Among these patients (Model 1, Table 3), unconscious patients with GCS scores of 3 or 4 to 9 were seven- and four-times more likely to be administered 0.8 mg naloxone than those who were awake (GCS 15). Compared to patients with a respiratory rate of ≥9 breaths/minute, those with respiratory arrest or a respiratory rate of 1-8 breaths per minute were three- and two-times as likely to be treated with 0.8 mg naloxone, respectively. Furthermore, men were more than twice as likely as women to be administered a dose of 0.8 mg. Those attended at the safe injection facility were 40% less likely to receive 0.8 mg naloxone than patients treated at other locations.
-----------------Insert table 3 here-----------------
Multiple naloxone dosages and their associations with clinical variables
Overall, multiple doses (≥2) of naloxone during one EMS attendance were administered in 14.8% (n=227) of the 1,530 patients with a valid national identity number who received either 0.4 mg or 0.8 mg IM naloxone. Among these cases (Model 2, Table 4), unconscious patients with GCS scores of 3 or 4 to 9 were seventeen- and eight-times more likely to be administered multiple doses than those who were awake. Compared to patients with a respiratory rate of ≥9 breaths/minute, patients with respiratory arrest were twice as likely to be treated with multiple doses. Furthermore, men were almost twice as likely as women to receive multiple dosages. Those attended at the safe injection facility were 80% less likely to be treated with multiple dosages than patients treated at other locations. Finally, those treated with an initial naloxone dose of 0.8 mg were 60% less likely to receive multiple doses than patients treated with an initial dose of 0.4 mg naloxone.
-----------------Insert table 4 here-----------------
Transfer rates
The majority (57.1%) of the 2,215 patients were left at the scene (Table 5), 28.1% were taken to the Oslo Accident and Emergency Outpatient Clinic, 12.9% were hospitalized and 1.9% were transferred to other places. One patient was in cardiac arrest and died despite treatment with advanced cardiac life support. This patient was treated with naloxone and was therefore included in the study. In the subsample of patients left on the scene (n=1264), 50.4% were left without medical supervision, while 49.6% were left at the safe injection facility or other health services such as nursing homes. For patients left on the scene, the average time for EMS attendance was 32.7 minutes. Whether the patient was transferred from the scene following treatment was not significantly associated with the initial dose either in the univariate logistic regression analysis (OR 1.1, 95% CI 0.9-1.3), or after adjusting for individual characteristics and vital signs (AOR 1.1, 95% CI 0.9-1.5). However, patients transferred following treatment were 70% more likely to have been treated with multiple doses of naloxone both in unadjusted analysis and after adjusting for individual characteristics and vital signs (AOR 1.7, 95% CI 1.2 – 2.3).
-----------------Insert table 5 here-----------------
One-week mortality
Among the 1,720 episodes between June 1st, 2014 and December 31st, 2018 with a valid national identity number, there were 10 deaths within the first week after EMS treatment. The crude one-week mortality rate was 5.8 per 1000 episodes. None of the patients died on day 0. However, three died on day 1, another three died on day 2, and four more deaths occurred during the following five days, between days 3 and 7.
Overall, ten patients died. Seven deaths were drug-related deaths, six of which were classified as unintended poisoning and one as a suicide by way of heroin. Three patients died from natural causes: one 96-year-old nursing home patient, one patient in palliative care, and a 62-year-old complex medical patient in home care. Those who died from overdose or suicide had all been left at the scene. The overall one-week mortality rate for drug-related deaths was 4.1 per 1000 episodes and 5.5 per 1000 episodes for patients left at the scene by the EMS.
Limitations and strengths
Data collection was based on paper records, which limited the number of variables. National identity numbers were not available in 22.3% of patients. Data on clinical evaluations after naloxone administration, such as GCS scores and respiratory rate, were missing in a large proportion of the records, which made it difficult to reliably estimate the efficacy of treatment. Local guidelines recommend naloxone dosages based on the patients’ weight which was not recorded in the medical records and could not be included in analyses. There were few recorded overdoses with fentanyl or other strong synthetic opioids in Norway, and the results are therefore not necessarily generalizable to settings where fentanyl is more frequent. Linking of data with other national registers and better data on follow-up would have improved the study.
A strength of the study was the long observation period of five years. Key demographic variables in the study could be compared with previous reports in Oslo and other countries(15). The issues with missing data were handled by including missing data as a variable in the models to avoid observations being deleted listwise. Norway has unique national identity numbers, which made it possible to link the data to the National Cause of Death Registry.