Introduction Opioid overdose is commonly treated by prehospital emergency services and the majority of the patients are discharged immediately after treatment and a short observation period. There is a minor risk for rebound opioid toxicity and other life-threatening conditions might occur after such episodes. The authors describe the short-term outcome and identify risk factors for death within 48 h after prehospital treatment of opioid overdose in Copenhagen, the capital of Denmark.
Methods Data on all cases of opioid overdose treated by the medical emergency care unit between 1994 and 2003 were recorded prospectively. Risk factors for death within 48 h after initial medical emergency care unit contact were analysed in a multivariable logistic regression analysis.
Results The authors recorded 4762 episodes of opioid overdose, covering 1967 unique identified patients. A total of 78 patients (8.4%, 95% CI 7.0 to 10.4) died within 48 h in the period 1999–2003, and 85% (66/78) of these had cardiac arrest and died. The authors found age >50 years and overdose during the weekend significantly associated with 48-h mortality. Gender, former episodes of opioid overdose, time of the day, month or year were not significantly associated with increased mortality.
Conclusions The author found a 48-hours mortality of 8.4%. Advanced age and opioid overdose in the weekends were significant risk factors. Release on scene after treatment was associated with a very small risk.
- Prehospital care
- clinical management
- drug abuse
- acute medicine-other
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Opioid overdose is a serious condition that is commonly treated by prehospital emergency services. Most patients recover quickly after being ventilated and receiving an opioid antagonist, but the prognosis is influenced by the high frequency of drug abuse and poor socioeconomic status.1 ,2 Admission to hospital after treatment may allow careful observation but drug addicts are quite often not motivated for hospitalisation and some medical emergency systems, including ours, allow patients to be discharged immediately at the physician's discretion after a short observation period. The group of drug abusers appreciates this policy, but there is a risk of recurrence toxicity and other morbidity.1 ,3–6 We have previously described three cases of likely fatal rebound toxicity in this group of drug addicts over 10 years.6 In the present study, we aimed to describe short-term outcome and to identify risk factors for death within 48 h after prehospital treatment of opioid overdose in a major European city.
Our physician-manned Mobile Emergency Care Unit (MECU) is dispatched to all cases of cardiac arrest, respiratory arrest or unconsciousness in the city of Copenhagen, the capital of Denmark. The MECU prospectively recorded data related to all contacts, including—if possible—the Danish social security number, in a dedicated MECU database, whose construction was approved by the Danish Data Protection Agency.
The MECU physician is a specialist in anaesthesiology who is assisted by a purposely-trained acute life support provider. Opioid overdose is diagnosed by the physician based on the clinical presentation, the case history, including the information received from bystanders and the dispatch centre. Treatment includes ventilation with 100% oxygen and naloxone 0.8 mg intravenously and 0.4 mg given either intramuscularly or subcutaneously. Patients may be released if substantial and lasting improvement is observed over 15–20 min, focusing on level of consciousness, spontaneous ventilation and circulation, as previously described.2 ,6
Data was collected during a 10-year period 1994–2003. We included all cases of opioid overdose treated by the MECU in Copenhagen. The MECU database records contain details on demographics, clinical presentation and treatment. The individuals in this database who could be identified by their Danish social security number were crosschecked with the Central Personal Registry (21st August 2009) to assess vital status and time of death, if relevant. The data that is used in the analysis consist of the first opioid overdose episode of unique individuals for which a valid social security number was available in the 5-year period 1999–2003; data from the period 1994–1998 is used to determine if the person had had a previous opioid overdose treated by MECU.
We report continuous data using median with 5–95 percentiles while categorical data are reported as counts with percentages.
Risk factors for death within 48 h after initial MECU contact were assessed in a multivariable logistic regression analysis including age, gender, date and time for treatment, and if previous episodes of opioid overdose had been recorded. Data analysis and statistical evaluation were performed using R language and environment for statistical computing.
Informed consent is, according to Danish law, not required for studies based on information in existing databases that are approved by the Danish Data Protection Agency.
A total of 4762 episodes of opioid overdose were recorded between 1994 and 2003. Patients were identified in 3245 (68.1%) cases, of which 75.8% were men and the median age was 34.2 years. A total of 29.6% of patients had more than one episode, varying from 2 to 30 over the 10-year period, and the corresponding number of unique individuals was therefore only 1967. In the group treated in the period 1999–2003, we identified 929 unique individuals, and 174 (18.7%) of these had already been treated for opioid overdose by the MECU during 1994–1998.
Among these 929 individuals, 66 (7.1%) died on scene, 253 (27.3%) were hospitalised, and 608 (65.6%) were released on scene after treatment of their first episode in the period 1999–2003. In two individuals, the MECU treatment was not recorded.
At MECU arrival, 70 (7.5%) patients had cardiac arrest and 66 of them died on scene, while four obtained return of spontaneous circulation (ROSC). A total of 78 patients (8.4%, 95% CI 7.0 to 10.4) died within 48 h of contact, including the individuals who died on scene. Thus, 12 patients died subsequently within 48 h, and seven of them had been discharged after treatment, corresponding to a mortality of 1.2% (95% CI 0.5% to 2.4%) in that group (figure 1).
Logistic regression analysis revealed that advanced age and opioid overdose in the weekend were significant risk factors for 48-h mortality. There was no significant association between 48-h mortality and gender, month, year or if previous episodes had been recorded (table 1).
We found that prehospital opioid overdose was associated with a mortality of 8.4% within 48 h, and 85% of the deaths (66/78) were on scene. Age >50 years and episodes in the weekends were significant risk factors in those individuals who could be identified, with ORs around four, roughly corresponding to a fourfold increase in risk.
Approximately 30% could not be identified and that is unfortunate because this group is probably at a high risk of complications after treatment.
We previously reported that on scene mortality after opioid overdose decreased significantly in the period 1994 to 2003. Furthermore, we also analysed the long-term prognosis previously with follow-up until 2009.2 Here a strikingly different pattern was found, as compared with 48-h mortality. Episodes occurring in the summertime and a history of opioid overdose were associated with a higher long-term mortality but weekday was not important in that analysis. Increased age was a significant risk factor for both short-term and long-term mortality.2 Hence, long-term mortality may likely relate to socioeconomic factors and lifestyle, such as previous episodes of opioid overdose. Death within 48 h seems to be more common in the weekends, which may indicate that less experienced users and recreational users are more likely to abuse opioids in the weekends. One other study has found increased frequency of opioid overdoses on weekends but only in a subgroup of juveniles.7 Multidrug use is common and a known risk factor for fatal intoxication.8 ,9 This might be partly responsible for the mortality rate within 48 h. Other factors such as comorbidities might also play a role. Our risk factor analysis only included those treated by the MECU, meaning that someone called an ambulance. Still, the ORs should be valid as the analysis focuses on those factors that had an impact on risk of mortality, and that may be applied to other groups having a different baseline risk as well.
Release on scene carries a risk of rebound toxicity, which is, however, very low according to our own data, where 0.13% died because of rebound toxicity within 48 h after being discharged on scene.6 The most important concern after discharge is probably that some individuals take a further dose of opioid. The majority of the deaths occurred on the scene and most had cardiac arrest when the MECU physician arrived. Resuscitation was only successful in four out of 70, suggesting that hypoxia had been long lasting since ROSC can often be obtained after hypoxic cardiac arrest in young subjects. Ingestion of other drugs can also explain why advanced life support does not result in ROSC.
In our study the median age for overdose was 34 years, similar to some previous studies.1 ,10 Most other studies report a younger median age.7 ,11–17 Of all the cases in this study, 75.8% were men which corresponds well with former studies with percentages between 66% and 85%.1 ,4 ,7–13 ,15 ,17 It is interesting that we found no significant difference in mortality according to gender, which is in contrast to the reported higher mortality in men found in former studies.9 ,16
In Denmark, methadone is widely used in treatment of opioid abuse or as an opioid maintenance programme, and all fatal intoxications in drug addicts are medicolegally examined. In 2007, the prevalence of fatal intoxications was 6.92 per 100 000 inhabitants and opioids were responsible for 92% of the deaths. In 1991, approximately one-third of the fatal opioid overdoses were caused by methadone and this number was increasing in 2002, and in 2007 more than half of the deaths were caused by methadone. The number of drug addicts in methadone treatment in Denmark has more than doubled from 1995 to 2007. This might explain the increasing number of fatal methadone overdoses. The highest frequency of fatal overdoses in Denmark in 2007 occurred in the age group 35–39 years but in the other Scandinavian countries it was in the age group 25–29 years. The reason for this could be the more liberal use of methadone in Denmark.8 Methadone is shown to reduce the mortality, when the addicts follow the methadone treatment.18 ,19 This corresponds well with our finding of an advanced age as a significant risk factor for fatal overdose.
The common use or treatment with methadone may partly explain the surprising result that we did not find a difference in mortality according to time of the day the overdose takes place, which former reports have found.7 ,11 ,12 ,14 We did not find a difference in mortality according to month and year.
Strengths of this study include the long study period of 10 years in which all cases of opioid overdose that were treated by the MECU were recorded. Furthermore, the possibility of follow-up on identified patients was because of the unique Danish personal identification number carried by all Danish citizens.
Limitations of this study are that only 68.1% of the patients treated by the MECU were identified. Reasons were that some patients were foreigners and some refused to or were unable to identify themselves. Many opioid overdose patients are sceptical of authorities but confidence in MECU was increasing over the study period and more patients identified themselves.2
We found a mortality of 8.4% within 48 h after prehospital treatment of opioid overdose. Significant risk factors were ≥50 years of age and overdose during the weekend. Release on scene after treatment was associated with a very small risk.
Secretary Gitte Brofeldt, Copenhagen MECU.
Competing interests None.
Ethics approval Informed consent is, according to Danish law, not required for studies based on information in existing databases that are approved by the Danish Data Protection Agency.
Provenance and peer review Not commissioned; externally peer reviewed.