Pre-hospital treatment of acute poisonings in Oslo

Poisoned patients are not treated in emergency departments and hospitals only [1‐3]. In Scandinavia, poisonings related to drug abuse, especially heroin overdoses, are frequently discharged from the ambulance service without admission to hospital. Pre-hospital discharge may be due to the patients' refusal to be admitted, but can also be due to a medical judgement made by the paramedics or physicians on site. The number of patients discharged from ambulances and the outpatient clinic is difficult to determine because these patients are not recorded in hospital records, and the ambulance record databases may contain inaccurate diagnoses.

Most studies of acute poisonings focus on hospital admissions or emergency department visits, either using hospital records or by prospective inclusion of patients; but with no inclusion of poisonings treated outside hospitals. Studies based on poison information centre registers seldom give precise information regarding the poisonings [4, 5]. To study the epidemiology of all acute poisonings in a geographically defined area, prospective inclusion of patients in all health care facilities at the different health care levels treating such patients is required. To our knowledge, no such study has previously been performed.

We performed a study on all acute poisonings independent of treatment levels before discharge, with the following aims: 1) to describe the acutely poisoned patients treated by ambulance personnel and an outpatient clinic; 2) to compare patients transferred to a higher treatment level with patients discharged without transfer, and 3) to study one-week mortality after pre-hospital discharge.

We recorded 2997 episodes of acute poisonings in 2298 individuals. The population in Oslo in 2004 was 521 886, of whom 428 198 were ≥ 16 years of age [9]. This gave an overall incidence of 5.4 per 1 000, but with no correction made for poisonings in persons living outside Oslo (the catchment area).

The majority of poisoned patients treated by ambulance personnel and in the outpatient clinic were discharged without referral to higher treatment levels. These poisonings were mostly due to opiate overdoses and ethanol poisoning. Patients discharged from the ambulances were often severely poisoned regarding consciousness and respiratory depression or arrest, while those discharged from the outpatient clinic were less often comatose and less often had respiratory depression. Male patients dominated the pre-hospital discharged cases, while female patients were in the majority among the hospitalized, where pharmaceuticals were the most common toxic agents. One death due to a new overdose was recorded among the pre-hospital discharged in the first week after the poisoning.

Our findings of significant male dominance in the pre-hospital setting are supported by a previous study in Oslo, reporting 77% male patients [10], and by studies of pre-hospital opiate overdoses from Australia (67–77% males) [11, 12] and Austria (66% males) [13]. In our study, female patients were in the slight majority among the hospitalized. Gender differences regarding toxic agents are often reported, with female patients poisoned with pharmaceuticals and male patients poisoned with non-pharmaceutical agents [14], which influences the gender distribution in the hospitals [6]. Studies on deliberate self-poisonings, which are more frequently admitted to hospital, show female majorities, with 66% in a Greek study [15] and 56%–66% in a Norwegian study [16]. Although discharging males more than females from the pre-hospital setting, the multivariate logistic regression analyses showed that gender was not an independent predictor for discharge versus transferral. It also revealed that the agent taken combined with the clinical severity were the key factors for patient flow through the system. This can probably be explained by opiate overdoses being potentially reversible on site, and ethanol poisoning as a transient, most often uncomplicated situation. Furthermore, many of the poisonings by medications need more thorough observation and treatment – which often implies hospitalization.

Although opiate overdoses often present with pronounced respiratory depression and altered consciousness, reversal by naloxone makes it possible for the patients to refuse further treatment, or further treatment is considered unnecessary. Norwegian law does not permit treatment against a person's will, except in suicidal or psychotic patients. The majority of such overdoses are not assessed as suicidal, but obviously, this evaluation is difficult on site. The situation in the outpatient clinic was quite different from that in the ambulances. Most patients were brought by ambulance or were able to be transported by other means. Ambulance personnel who decided to transport the patient to the outpatient clinic instead of the hospitals have made an evaluation based on the patient's clinical condition, resulting in the selection of probably less-affected patients compared to the remaining population. The outpatient clinic had the opportunity to observe patients for some hours, and fewer needed admission to hospital.

The present study reports a high proportion of opiate overdoses not transferred to a higher treatment level after ambulance treatment (78%) and outpatient clinic treatment (93%). Previously, 85% of opiate overdoses treated with naloxone were reported as not being transferred from ambulances in Oslo [10] (this subgroup comprised 72% in the present study). From most other countries, lower numbers were reported, but with large variations. In a Californian study, 166/609 (27%) were not transferred from the ambulances, and 12/442 (2.7%) were admitted from the Emergency Department (ED) [3]. Two studies from Austria reported 59/308 (19%) [13] and 295/1087 (27%) [17] not transferred, while an Italian study reported 52/124 (42%) not transferred to hospital [18]. However, an Australian study reported 324 (84%) not transferred of 388 naloxone-responders [19]. The higher proportion of non-transferred opiate overdoses in Oslo compared to most other studies is probably not due to differences in severity. It could reflect a more liberal tradition to either accept the wish not to be transported to hospital or to consider further referral as unnecessary. An Australian study across different jurisdictions showed that differences in overdose management, e.g. naloxone dosing, may affect the transportation and hospitalization rates [12].

Delayed complications, such as recurrence of opiate toxicity and occurrence of heroin-related non-cardiogenic pulmonary oedema (NCPE), are important in the assessment of whether patients were allowed to refuse transfer and observation. The half-life of naloxone is shorter than that of many opioids. Recurrence has been found in several studies, ranging from 31% to all admitted opiate overdoses [2, 20]. The procedure in Oslo is to give one IM dose of naloxone before the IV doses, in order to prolong the naloxone effect. This study has no data on recurrence, as 78% of opiate overdoses treated in ambulances were not transferred. NCPE is found in about 2% of opioid overdoses treated in EDs [21, 22], and is most frequently evident on arrival or within an hour. A study of pre-hospital heroin overdoses found pulmonary oedema in 0.8% [18]. In our study, 12% of the hospitalized opiate overdose cases (1% of all opiate overdose cases in the study) were mechanically ventilated [7], and presumably some of this was due to pulmonary oedema. As a result of delayed problems, recommendations for the observation time for an apparently uncomplicated opioid overdose reversed by naloxone have been made, but vary between one [23, 24], two to three [22, 25], and up to eight hours [26]. However, studies on opioid toxicity recurrence and NCPE are mainly performed in ED- or hospitalized patients, which are selected populations compared to patients not transferred from ambulance. Nevertheless, one has to assume that there may be a risk associated with the limited observation time offered by ambulance personnel when patients refuse further treatment. We have no data on the length of observation made by ambulance personnel, but a previous study in Oslo reported a mean observation time of 8 minutes (range 1–30 minutes) [10].

Drug abusers are difficult to follow up. Consequently, it is challenging to study complications other than death and readmissions among pre-hospital-treated overdoses. A study of early discharge of opiate overdose cases from an ED showed no readmissions or deaths [22], and studies on opiate overdose cases not transferred from ambulances found no subsequent life-threatening events [2] and no deaths due to the overdose [1]. This corresponds with our findings of only one death (which was due to a new overdose) in the first week after the poisoning among patients with known identity: 91% of all (88% of opiate) poisonings treated outside hospital.

Ethanol poisoning cases were the other large group of patients not transferred from the ambulances or outpatient clinic, most of which were observed in the outpatient clinic. In this study, 21% of ethanol poisoning cases were transferred to hospital from the outpatient clinic, which is more than the 12% admitted from the ED in an American register study of equivalent poisonings [27]. The Oslo Emergency Ward represents a lower treatment level than most hospital-based EDs, and hence transfers more patients to hospital. Ethanol poisoning is usually not life threatening. However, beside complications due to impaired consciousness, there is a risk of cardiac arrhythmias with heavy drinking [28]. A recent study of patients with high serum ethanol levels showed ECG changes associated with increased risk of arrhythmias [29]. Such risks should be taken into account when assessing ethanol-intoxicated patients. However, when in a stable clinical condition, early discharge of these patients is less controversial compared to in other poisonings.

In addition to the risk of complications, one can advocate for longer observation to channel patients into anti-addiction therapy. A French study showed that 80% of acute ethanol poisonings had elevated γ-glutamyltransferase and carbohydrate-deficient transferrin values, indicating harmful alcohol consumption over some time, suggesting that these patients should be offered treatment for alcoholism [30].

Poisoned patients were often discharged from pre-hospital treatment levels. These poisonings were more often due to opiate and alcohol abuse than in the hospitalized patients, and more than two-thirds were male patients. Patients discharged from ambulances were often severely poisoned as regards consciousness and respiratory depression, whereas patients discharged from the outpatient clinic were less often comatose and less often had respiratory depression. Almost half of those discharged from an ambulance received an antidote. The pre-hospital treatment of these poisonings appears safe with regard to short-term mortality.