Epidemiology of fatal tricyclic antidepressant ingestion: implications for management

Advances in the management of depression: implications for the obstetrician/gynecologist

The selection of an appropriate medication is important for successfully treating depression in women. Although antidepressants do not differ in their efficacy for the treatment of major depression, they do differ in their side effect profiles, toxicity, and mechanisms of action. Tricyclics, heterocyclics, and newer agents such as bupropion, fluoxetine, paroxetine, sertraline, and venlafaxine are used most commonly to treat depression in women. The dosages, side effects, indications, and precautions for these antidepressants are reviewed. Factors affecting selection of a particular antidepressant, as well as the general management of depression in women, are discussed.

Does a sodium-free buffer affect QRS width in experimental amitriptyline overdose?

STUDY OBJECTIVES

We carried out this study to determine the effects of pH alteration on QRS width with administration of tromethamine, a non-sodium-containing buffering agent, in experimental amitriptyline overdose.

DESIGN

Prospective, nonblinded trial.

PARTICIPANTS

Adult mongrel dogs.

INTERVENTIONS

Pentobarbital-anesthetized dogs were overdosed with amitriptyline 5 mg/kg followed by infusion at 1.0 mg/kg/minute until the QRS width doubled, then decreased to .5 mg/kg/minute until the end of the experiment. At two defined points of toxicity, the dose of tromethamine required to raise the pH to 7.50 +/- 4 was given. pH and QRS width at a speed of 100 mm/second were measured over a 30-minute period after each tromethamine dose. Data were analyzed with non-linear-regression analysis.

RESULTS

At toxicity 1 the mean pH was 7.32, with a QRS width of 11.6 mm. Two minutes after the tromethamine dose the pH rose to 7.51, with narrowing of the QRS width to 8.4 mm. At toxicity 2 the pH was 7.40, with QRS width of 10.6 mm. Two minutes after tromethamine, the pH rose to 7.49 and the QRS width decreased to 9.7 mm. Regression analysis showed a correlation between pH and QRS width; as pH increased, QRS width decreased (P = .0001).

CONCLUSION

Cardiac toxicity of amitriptyline overdose, as manifested by QRS widening, is reversible by pH changes alone.

Identifying toxicity risk early after antidepressant overdose

Despite the risk of life-threatening toxicities, care of most patients after antidepressant overdose (ADO) does not require the use of critical care resources. The use of emergency department (ED) clinical findings to identify ADO patients who subsequently manifested toxicity was evaluated prospectively. ADO risk assessment (ADORA) criteria included development of QRS interval > 0.10 seconds, arrhythmias, altered mental status, seizures, respiratory depression, or hypotension. Sixty-seven ADO patients were identified on presentation to a single ED and classified as low risk (LR, absence of criteria) or high risk (HR, presence of one or more criteria) based on development of criteria within 6 hours of ingestion (or ED presentation if ingestion time was not established). This system demonstrated 100% sensitivity in identifying study patients who developed significant toxicity problems. None of the 28 LR patients and 13 of 39 HR patients had subsequent complications (P < .01). No single clinical finding permitted risk classification. ADORA should identify patients who do not require further monitoring or other aggressive medical management for ADO.

Relative mortality from overdose of antidepressants

OBJECTIVE

To compare the fatal toxicities of antidepressant drugs in 1987-92.

DESIGN

Retrospective epidemiological review of prescription data of the Department of Health, Scottish Office Home and Health Department, and Welsh Health Common Services Authority (excluding data from most private general practices and most hospitals), and mortality data from the Office of Population Censuses and Surveys and General Register Office in Scotland.

SETTING

General practice, England, Scotland, and Wales.

MAIN OUTCOME MEASURES

Deaths per million prescriptions and deaths per defined daily dose.

RESULTS

81.6% (1310/1606) of deaths from antidepressant overdose were due to two drugs, amitriptyline and dothiepin. The overall average of deaths per million prescriptions was 30.1. The overall rate for tricyclic drugs was 34.14 (95% confidence interval 32.47 to 38.86; P < 0.001), monoamine oxidase inhibitors 13.48 (6.93 to 22.19; P < 0.001), atypical drugs 6.19 (4.04 to 8.80; P < 0.001), and selective serotonin reuptake inhibitors 2.02 (0.64 to 4.17; P < 0.001). The numbers of deaths per million prescriptions of amoxapine, dothiepin, and amitriptyline were significantly higher than expected, while nine drugs had a significantly lower number of deaths per million prescriptions than expected. Analysis of deaths per defined daily dose showed a similar pattern.

CONCLUSIONS

Safety in overdose should be considered in risk-benefit and cost-benefit considerations of antidepressants. A switch in prescribing, from drugs with a high number of deaths per million prescriptions to drugs with a low number, could reduce the numbers of deaths from overdose. Although this form of suicide prevention can be implemented easily and immediately, its introduction needs to be considered against the higher costs of some of the newer drugs.

Cardiovascular repercussions of seizures during cyclic antidepressant poisoning

Cyclic antidepressant overdose involves a risk of generalized seizures and cardiovascular disturbances. We have conducted a retrospective study to test the hypothesis of a relationship between generalized seizures and the onset of arrhythmia, hypotension or cardiac arrest during cyclic antidepressant intoxication. Patients who had seizures after ingestion of toxic amounts of tri- or tetracyclic antidepressants were included. Limb-lead QRS complex duration and systolic blood pressure were recorded before and after seizure. Twenty-four of the 388 patients (6.2%) who were admitted to our ICU over a four-year period had seizures (2.3 +/- 2 seizures/patient). Cardiac repercussions of cyclic-induced seizure were frequent and severe. In the postictal period, broadening of the QRS duration or hypotension occurred or were exacerbated in at least 41% and 29% of cases, respectively. In three patients (12.5%), the seizure-induced cardiovascular state was life-threatening and required massive alkalinization therapy and vasopressors, and two of the three required cardiac massage or cardioversion. Prior to seizure, these three patients had severe intoxications characterized by QRS duration > or = 120 ms and systolic blood pressure < or = 80 mm Hg. The results of this work confirm the potential risk of cardiovascular deterioration after cyclic antidepressant-induced seizure and raise the question of a prophylactic approach especially towards the subgroup with unstable hemodynamic status.

Fatal seizures after flumazenil administration in a patient with mixed overdose

OBJECTIVE

To report a fatal case of refractory status epilepticus precipitated by flumazenil use in a mixed benzodiazepine-tricyclic antidepressant overdose.

CASE SUMMARY

A 39-year-old woman was brought to the emergency room (ER) in a stupor from a suspected suicidal overdose of an unknown mixture of drugs. Past medical history included seizures and psychiatric disorders managed with benzodiazepine and tricyclic antidepressants. Initial ER electrocardiogram showed a QRS interval of 136 milliseconds. The patient developed refractory seizures after being given flumazenil. Lorazepam, phenytoin, and phenobarbital were administered; however, seizures persisted for 4 hours, resulting in rhabdomyolysis, acute renal failure, severe brain damage, and death.

DISCUSSION

Flumazenil should be used with caution in patients with chronic benzodiazepine use, prior seizure history, or when a mixed overdose is suspected. Flumazenil may unmask tricyclic antidepressant-induced seizures by antagonizing the antiepileptic effect of concomitantly ingested benzodiazepine. In this patient seizures occurred within two minutes of flumazenil administration. As benzodiazepine-induced central nervous system depression is rarely life-threatening, the use of flumazenil must be balanced against potential risk.

CONCLUSIONS

Seizure risk factors should be assessed in all patients in whom flumazenil use is considered. If risk factors are present, the benefit of flumazenil use is outweighed by the potential risk. If flumazenil is used, resulting seizures may require larger doses of benzodiazepine.

Extracorporeal circulation in the management of severe tricyclic antidepressant overdose

Extracorporeal circulation is a technique that provides precise control of circulation, oxygenation, temperature, and blood composition in patients suffering from cardiopulmonary failure. The investigators present the case of a near fatal tricyclic antidepressant overdose that failed to respond to standard therapy but was resuscitated using extracorporeal circulation.

Cyclic antidepressant overdose: a review of current management strategies

Cyclic antidepressant (CA) overdose can produce life-threatening seizures, hypotension, and dysrhythmias. It accounts for up to half of all overdose-related adult intensive care unit admissions and is the leading cause of death from drug overdose in patients arriving at the emergency department alive. Several factors contribute to the significant morbidity and mortality associated with CA overdose. First, CAs are widely prescribed and are dispensed to patients at increased risk for attempting suicide. Second, drugs of this class generally have a low therapeutic toxic ratio. Third, in the majority of fatal cases, the patient dies before reaching a hospital. Finally, and of greatest significance for the clinician, the presenting signs and symptoms of CA overdose may be missed by the physician, even in cases of severe toxicity. Therefore, CAs must be considered early in any case of suspected overdose, and all such cases should be managed as potentially fatal ones. The following case demonstrates the current approach to the patient with significant CA toxicity.

Experimental amitriptyline poisoning: treatment of severe cardiovascular toxicity with cardiopulmonary bypass

STUDY OBJECTIVE

To compare cardiopulmonary bypass (CPB) with more conventional therapy in the treatment of severe amitriptyline poisoning.

DESIGN

Prospective, randomized, controlled, laboratory investigation.

INTERVENTIONS

Profound cardiovascular toxicity was induced in 20 anesthetized Yorkshire swine (72 +/- 8.3 kg) by amitriptyline infusion at 0.5 mg/kg/min. Ventilation was adjusted to keep arterial pH at 7.50 +/- 0.05 and the PCO2 at 35 mm Hg. The swine were randomized in a 1:1 ratio to one of two groups, CPB or control. Both groups received amitriptyline infusion until they experienced near-lethal toxicity, defined as a systolic blood pressure below 30 mm Hg for one minute. The control group was then given supportive treatment, including IV fluids, sodium bicarbonate, vasopressors, and standard pharmacologic (advanced cardiac life support) interventions. Control animals failing to respond to supportive measures after five minutes were given open-chest cardiac massage for 30 minutes or until the return of spontaneous circulation. The CPB group received only mechanical support by CPB for 90 to 120 minutes. No sodium bicarbonate, antiarrhythmics, or cardiotonic agents were provided to the CPB group during this resuscitation.

RESULTS

All 20 animals experienced cardiac conduction delays, dysrhythmias, and progressive hypotension within 30 minutes of receiving IV amitriptyline at 0.5 mg/kg/min. The ten swine receiving CPB as treatment for cardiovascular toxicity were able to completely correct the dysrhythmias, cardiac conduction abnormalities, and hypotension produced by the amitriptyline; however, only one of ten control animals could be resuscitated (P = .0001). Nine of ten swine treated with CPB were easily weaned off bypass without any pharmacologic intervention; however, one required norepinephrine to be weaned. All 11 resuscitated swine were able to be salvaged.

CONCLUSION

CPB improved survival in our swine model of severe amitriptyline poisoning.

Greater toxicity in overdose of dothiepin than of other tricyclic antidepressants

Epidemiological studies have implicated dothiepin in a greater number of self-poisoning deaths than would be expected from its use. We have prospectively assessed the clinical toxicity of dothiepin and other tricyclic antidepressants (TCAs) in overdose. We followed-up consecutively admitted patients with TCA poisoning managed by our department between January, 1987, and August, 1992. 75 patients had taken dothiepin, 101 amitriptyline, 83 doxepin, and 61 other TCAs. Death after TCA poisoning is rare nowadays, so we used intermediate outcome measures--general seizures, tachyarrhythmias, sedation, and QRS width on the electrocardiogram. 15 patients had seizures and 7 tachyarrhythmias. When we excluded patients who had taken more than one TCA, general seizures were more likely after dothiepin than after other TCAs (9/67 vs 5/220) as were arrhythmias (4/67 vs 3/220). Rates of other complications were similar. The dothiepin group had ingested a larger dose, attributable to the larger average tablet strength, than patients who took other TCAs. The odds ratio for seizures with dothiepin versus other TCAs was 6.7 (95% Cl 2.2-20.7) unadjusted and 7.1 (2.2-23.2) after adjustment for sex, age, and ingested dose. The corresponding odds ratios for arrhythmias were 4.6 (1.0-21.1) and 3.4 (0.7-16.3). Dothiepin in overdose seems to be proconvulsant. Patients with only minor sedation and normal limb-lead QRS width may still have major complications. Consideration should be given to the use of other antidepressants in patients at risk of seizures or suicide. Regulatory authorities should review the need for a 75 mg strength tablet of any TCA.

Acute chemical pancreatitis associated with a tricyclic antidepressant (clomipramine) overdose

A case of acute chemical pancreatitis and associated prolonged ileus following an acute overdose of the tricyclic antidepressant clomipramine by an adult is reported. Pancreatitis is a rarely-reported serious complication of antidepressant overdose and may lead to prolonged ileus and extended hospitalization.

Analysis of indications for intensive care unit admission. Clinical efficacy assessment project: American College of Physicians

OBJECTIVE

To formulate recommendations for the development of intensive care unit (ICU) admission policies.

DESIGN

Literature review of published reports over the period 1966 to 1991 pertaining to admission criteria for intensive care or coronary care units (CCUs).

PATIENTS

Studies identifying patients least likely to benefit from ICU or CCU admission were analyzed. Patient populations of interest included adults (> or = 18 years of age) with medical conditions possibly requiring intensive care; trauma patients were excluded.

MEASUREMENTS AND MAIN RESULTS

Of 970 articles identified as being pertinent to intensive care, only two case-control studies used the direct method of measuring the effect of ICU intervention on mortality. No studies were found that compared outcomes of low-risk patients treated in a CCU vs those treated in alternative hospital locations, and none identified patients with a very high probability of a bad outcome.

CONCLUSIONS

The use of decision-making models for ICU and CCU admissions must be tested in prospective, randomized clinical trials. Critical care units and ICUs should be studied separately. Existing studies of early discharge from CCUs need to be summarized and evaluated. The triaging of ICU patients to alternative hospital locations needs to be evaluated, as do existing predictive models for early triage decision-making.

Proarrhythmia of nonantiarrhythmic drugs

Proarrhythmic effects of nonantiarrhythmic drugs have not been as extensively studied or reported compared with the effects of antiarrhythmic drugs. The proarrhythmic incidence of many of these agents is not accurately known. In some instances, the facilitation of arrhythmias may be the result of compounding clinical factors. Many agents, however, share structural similarities to antiarrhythmics and manifest the same arrhythmic tendencies. Many reports of proarrhythmia may represent toxic rather than proarrhythmic effects, and in vitro studies to elicit the underlying mechanisms may be warranted for the more common drugs. This report summarizes reported arrhythmic effects of a variety of commonly utilized nonantiarrhythmic drugs. The incidence and mechanism of the proarrhythmia is not always clear. The clinician, however, should be aware of reported events to appropriately diagnose and treat the arrhythmia.

Transport management of the patient with acute poisoning

Poisoning in children is a common clinical problem encountered by pediatricians, general practitioners, and emergency room physicians. Poisoning in children less than 5 years of age is usually accidental, whereas, in young adults, any disparity between expected history and clinical findings should suggest poisoning. It is imperative that the treating physician expeditiously recognize, begin treating, and plan to transfer, when indicated, by specialized pediatric transport team the critically ill poisoned child to a tertiary care facility.

The safety of antidepressants

It is no exaggeration to say that, in the 30 years since antidepressants were introduced, they have revolutionised the management of depression, thus fulfilling, at least for depressive illness, Paracelsus' prophetic wish. We now have a group of drugs which can induce remission in a disease which is disabling and not infrequently fatal. Despite intensive research, however, our understanding of the mechanism of action of these drugs is still largely hypothetical. Even our methods of evaluation are sufficient only to enable one to decide whether an antidepressant drug is effective or ineffective; there is no experimental method which can rank them in an overall order of potency, and perhaps this should not be the aim, since there must be subgroups of depressed patients who will respond differently to any one agent.

Severe amitriptyline overdose: relationship between toxicokinetics and toxicodynamics

The clinical features and toxicokinetics of amitriptyline were studied in nine patients with severe amitriptyline poisoning. Amitriptyline and amitriptyline metabolites were studied in plasma, red blood cells, and cerebral spinal fluid. Eight patients were intubated and six required assisted ventilation. Two patients had ventricular arrhythmias, three patients convulsions and two were hypotensive. All complications developed within four hours of admission. Early in the course of the intoxication the QRS duration correlated with plasma, unbound and red blood cell nortriptyline concentration. The QRS duration also correlated with unbound but not the plasma amitriptyline concentration. The level of consciousness correlated with the plasma and unbound amitriptyline both in alpha and beta phase and with red blood cell amitriptyline in alpha phase. There was no correlation between nortriptyline concentration and level of consciousness. No correlation between coma grade or QRS duration and cerebral spinal fluid concentration of amitriptyline was found. There was no correlation between any hydroxymetabolite in blood or cerebral spinal fluid and QRS duration or coma grade. The beta half-life for amitriptyline was shorter for two patients with high concentrations of hydroxymetabolites. Although intubated, neither patient required assisted ventilation or developed complications. Because of the wide range of concentrations of amitriptyline and amitriptyline metabolites observed between individuals, it is not possible to predict outcome based on a single tricyclic antidepressant concentration.

Amitriptyline and amitriptyline metabolites in blood and cerebrospinal fluid following human overdose

The toxicokinetics of amitriptyline were studied in nine patients admitted to hospital in Matthew-Lawson Coma Scale grade III-IV after an estimated ingestion of 1-5 g amitriptyline. Gastric lavage was performed and 50 g activated charcoal were given orally. Venous blood samples were taken on admission and at 1, 2, 4, 8, and 24 h, and in some patients at 36 and 48 h after admission. Arterial blood samples were taken at 1, 4, 8, and 24 h after admission. Lumbar punctures were performed 1 h after admission in 8 patients and again 4 h later in 5 patients. A urine sample was screened for other drugs. The bound and unbound fraction of amitriptyline and its metabolites nortriptyline, E and Z forms of 10-OH-amitriptyline and nortriptyline were analyzed in plasma, whole blood, red blood cells, and cerebrospinal fluid using an HPLC technique. The T1/2 alpha and T1/2 beta for amitriptyline were 1.5 - 3.1 and 15 - 43 h respectively. The rate of elimination of amitriptyline was not dose-dependent. The arteriovenous differences in the total amitriptyline+nortriptyline concentration were maximal in patients admitted soon after intake of drugs. Amitriptyline concentrations in cerebrospinal fluid were quantitatively similar to the unbound amitriptyline concentration in blood. The highest cerebrospinal fluid amitriptyline concentration was 506 nmol/L. There were large individual differences in plasma, blood and cerebrospinal fluid concentrations between different individuals. Repeated quantitative analysis of amitriptyline and its metabolites is unlikely to contribute to the clinical management of most patients with amitriptyline overdose.

Self-poisonings with psychopharmacological agents in Zagreb

1. A 5-year retrospective study of 1743 episodes of self-poisoning in 1608 adult patients, was undertaken to determine the frequency of self-poisoning with psychopharmacological agents, the frequency of each toxic agent usage and the epidemiological characteristics of the poisoned population. 2. The annual incidence of self-poisoning in Zagreb was 38.7 per 100,000 adult inhabitants. The anxiolytic-antipsychotic-antidepressant usage ratio was 8.8:1.5:1.0 in total self-poisoning, whereas it was 4.9:1.5:1.0 in patients with a previous history of self-poisoning. The use of anxiolytics decreased and the use of antidepressants increased during the 5 years. 3. The female-male ratio was 3.8:1 and 77% of self-poisoned persons were aged 15-44 years. The incidence of self-poisoning was 74% higher in the urban than in the rural population. The proportion of self-poisoned persons was 204% higher among the unemployed than in the population as a whole. There were 76% more repetitors among highly educated and 43% fewer repetitors among students than in the general population. The use of antipsychotics was 120% higher in males than in females. The use of antidepressants was 107% higher in the highly educated. 4. The increased use of antidepressants in self-poisoning is an alarming signal and caution in their prescription seems to be indicated.

Changes in arterial oxygen tension after gastric lavage for drug overdose

1. Changes in arterial oxygen tension after gastric lavage were investigated in a prospective study of 55 patients admitted for drug overdose. 2. A significantly greater decrease was observed in smokers compared to non-smokers. 3. A correlation was found between the tricyclic antidepressant serum level and the observed oxygen tension decrease. 4. The usefulness, as well as the safety of the lavage procedure, in managing tricyclic antidepressants overdoses is controversial.

Reversible brain death. A manifestation of amitriptyline overdose

Tricyclic antidepressants are known to cause central nervous system depression. However, a generalized depression of brainstem functions is rarely reported. We report a patient in deep coma with complete absence of brain-stem reflexes after she had taken a large quantity of amitriptyline. With continuous supportive treatment, she eventually regained all neurologic function and made a full recovery.

Tricyclic antidepressant poisoning

OBJECTIVE

To review poisoning with tricyclic antidepressants.

DATA SOURCE

English language literature search using Australian Medlars Service (1977-1989), manual search of journals and review of bibliographies in identified articles.

STUDY SELECTION

Approximately 250 articles, abstracts and book chapters were selected for analysis.

DATA EXTRACTION

The literature was reviewed and 93 articles were selected as representative of important advances.

DATA SYNTHESIS

The major features of overdose are neurological, cardiac, respiratory and anticholinergic. Life-threatening complications develop within six hours of overdose or not at all. All patients seen within six hours of overdose should have their stomachs emptied. All patients should receive activated charcoal. Coma, convulsions, respiratory depression and hypotension are treated with standard resuscitation techniques and drugs. Treat patients with significant cardiotoxicity or cardiac arrest with alkalinisation by sodium bicarbonate or hyperventilation, aiming for an arterial pH of 7.45-7.55. Lignocaine is used for ventricular arrhythmias. Other antiarrhythmic drugs are contraindicated (Class 1A, Class 1C), potentially lethal (Class II), of no benefit (phenytoin) or of unproven efficacy (Class III and Class IV). Physostigmine has no role at all. Haemodialysis and haemoperfusion are of no benefit.

CONCLUSION

The death rate of those who reach hospital is 2%-3%. Most of these deaths are cardiac in origin, and are caused by direct depression of myocardial function rather than cardiac arrhythmias.

Plasma catecholamine levels in cyclic antidepressant overdose

Cyclic antidepressant overdose is a major cause of morbidity and mortality in self-poisoned patients. The major cause of mortality with cyclic antidepressant overdose is cardiotoxicity. We determined plasma catecholamine levels in 41 symptomatic acute overdose patients to identify interactions between QRS duration (a marker for cardiotoxicity) and a presumed hyper-adrenergic state. Using a linear multivariable regression analysis, QRS duration correlated with the presence of cyclic antidepressant, plasma norepinephrine levels, the ratio of norepinephrine to epinephrine level, and pulse rate (p less than 0.001, r2 = 0.42). Commensurate physiologic changes were not found in the presence of elevated catecholamine levels in the cyclic antidepressant overdose group. One possible explanation for the blunted systemic response to the elevated catecholamine levels is adrenergic desensitization. Investigation of serial catecholamine levels during cyclic antidepressant overdose may lead to modification of our current theories of cardiotoxicity and therapy.

Demographic and electrocardiographic factors associated with severe tricyclic antidepressant toxicity

This study was designed to evaluate a historic cohort of pure tricyclic antidepressant overdose patients for factors associated with severe toxicity. Hospitalized tricyclic antidepressant overdose patients were identified by computerized discharge diagnosis (ICD-9 codes). Patients with a serum drug screen positive for tricyclic antidepressants and an emergency department 12-lead electrocardiogram were included in the study. Multiple drug overdoses were excluded. Patients were divided into two groups: minor toxicity (n = 41 and major toxicity (n = 65). Criteria for inclusion in the major toxicity group were the occurrence of seizures, endotracheal intubation, coma, arrhythmias requiring treatment, hypotension, or death. The following were found to be associated with increased likelihood of major toxicity (p less than 0.05): ingestion of amitriptyline (odds ratio (OR) 2.57), age greater than or equal to 30 years (OR 2.56), heart rate greater than or equal to 120 bpm (OR 2.86), serum tricyclic antidepressant level greater than or equal to 800 ng/mL (OR 5.20), terminal 40 ms QRS axis (T40-ms axis) greater than or equal to 135 degrees (OR 2.73), QRS interval greater than or equal to 100 ms (OR 2.74), QRS axis greater than 90 degrees (OR 3.68), and QTc interval greater than 480 ms (OR 3.89). The mean T40-ms axis on the initial ECG was more rightward in the major toxicity group (174 +/- 84 vs 125 +/- 91 degrees, p = 0.006). We conclude that patients with severe tricyclic antidepressant toxicity tended to have a more rightward T40-ms axis than those with minor toxicity and that the presence of the above parameters was associated with an increased likelihood of severe toxicity.

The electrocardiographic manifestations of cyclic antidepressant therapy and overdose: a review

Cyclic antidepressants may cause changes in the electrocardiogram at therapeutic or toxic serum levels. The most serious complications of cyclic antidepressant toxicity are dysrhythmias, hypotension, and seizures. It is predominantly the cardiotoxic effects that cause mortality. Once cardiotoxicity is evident, the treatment of choice is serum alkalinization, preferably by sodium bicarbonate therapy. In order to predict which overdose patients are at high risk for complications, electrocardiographic criteria have been identified as reliable screens. For "first generation" tricyclic antidepressants, QRS prolongation (particularly greater than 100 msec) and a terminal 40-ms frontal plane axis greater than 120 degrees are the most sensitive. This article reviews antidepressant pharmacology, electrocardiographic manifestations of antidepressant cardiotoxicity, and approaches to treatment of antidepressant-induced conduction disturbances and dysrhythmias.

Tricyclic antidepressant overdose: conservative management in a community hospital with cost-saving implications

Reports of late-onset cardiovascular complications following tricyclic antidepressant (TCA) overdose have led to a very conservative approach to these patients. Many patients have been hospitalized for continuous cardiac monitoring, regardless of the clinical presentation. Management algorithms based on clinical predictors of outcome have recently been proposed. We used the algorithm developed by Tokarski and Young to retrospectively evaluate the care of 33 TCA overdose patients admitted to our hospital over a 3-year period. We then identified 11 patients who could have been treated on an outpatient basis had the algorithm been employed. Ten were admitted to a monitored unit and spent a mean of 31.6 +/- 15.64 hours on the unit. None of the 11 patients developed complications during their hospital stay. Use of the algorithm would have resulted in an estimated cost savings of 13 hospital days and $14,000.

Toxicological findings after fatal amitriptyline self-poisoning

A case of fatal self-poisoning with amitriptyline is reported. Both the drug and its metabolite nortriptyline were quantified in several post-mortem tissues and fluids, including vitreous humor. Results are discussed in the light of the existing literature.

Value of initial ECG findings and plasma drug levels in cyclic antidepressant overdose

Electrocardiographic changes and quantitative plasma drug levels associated with cyclic antidepressant (TCA) overdose have been the subject of many reports in the medical literature. Heart rate, QRS duration, QT interval, and most recently, the terminal 40 ms QRS (T40) axis have been reported to be valuable indicators in TCA overdose. The value of plasma drug levels and their ability to predict complications has been debated. To evaluate the discriminant and predictive abilities of ECG parameters and quantitative drug levels in a large series of TCA overdoses, we retrospectively reviewed the charts of all overdose patients admitted to ICU in our facility during a 30-month study period. The TCA-positive group (187) had statistically significant differences (P less than .001) from the TCA-negative group (171) in heart rate, QRS duration, QTc, and T40 axis. However, none of these parameters could correctly classify more than 66% of cases. The sensitivity and specificity of the T40 axis were found to be only 29% and 83%, respectively. Correlation coefficients for plasma quantitative TCA levels by ECG parameter yielded an R of less than .33 in all cases. QRS duration of 100 ms or more and heart rate of 100 or more, or TCA level of 1,000 ng/mL or more, were present in all of a few patients with complications. We conclude that ECG parameters cannot be relied on to include or exclude the diagnosis of TCA overdose and that TCA levels do not correlate with ECG parameters.

Tricyclic antidepressant toxicosis

Accidental ingestions of TCAs by companion animals often occur. During the past 4 years, over 450 cases have been reported to the IAPIC. At least 7% of the animals that displayed clinical signs of toxicosis eventually died. Overdoses of TCAs adversely affect the cardiovascular, parasympathetic, and central nervous systems. The cardiovascular system is involved most seriously and ventricular arrhythmias with severe hypotension are believed to be the primary cause of death. Animals that ingest a potentially lethal dose (over 15 mg/kg) may die within 1 to 2 hours if appropriate treatment is not administered. Treatment involves the use of initial life-supportive measures (control of seizures, maintenance of an airway, ventilation, and so on), detoxification of the animal (enterogastric lavage, activated charcoal, etc.), and the intravenous use of sodium bicarbonate (2-3 mEq/kg) to control signs of acidosis, hypotension, tachycardia, bradycardia, and other cardiac conduction abnormalities. The animal must then be monitored closely for the return of the clinical signs and sodium bicarbonate therapy should be repeated as needed. In addition, to enhance removal of the TCAs from the gastrointestinal tract and, ultimately, from the body, activated charcoal should be repeated at 3-hour intervals until the animal is asymptomatic.

The toxicity of antidepressants

Considerable attention has been paid, both by regulatory authorities and clinicians, to the side-effects of antidepressants, but despite the known high suicide risk in depression, toxicity in overdosage has been largely ignored. On the basis of the evidence currently available, 'older' antidepressants appear to be more toxic than 'newer' agents. A case can be made for limiting the use of some of the older tricyclic antidepressants, especially in patients who remain in the community, when continuous supervision is lacking, in favour of newer, safer drugs.

Terminal 40-ms frontal plane QRS axis as a marker for tricyclic antidepressant overdose

Tricyclic antidepressant (TCA) poisoning has been reported to cause a right-axis deviation of 130 degrees to 270 degrees in the terminal 40-ms frontal plane QRS axis (T40-ms axis) of the ECG. This retrospective cohort study was designed to determine if the T40-ms axis could discriminate TCA-toxic patients from other overdose patients and whether a correlation exists between TCA plasma concentration and T40-ms axis rotation. Only symptomatic overdose patients with plasma and urine drug screens and an ECG obtained within two hours of each other were included in the study. Patients were divided into two groups: TCA overdose patients (TCA OD, n = 48) and nonTCA overdose patients (nonTCA OD, n = 30). The mean T40-ms axis was significantly more rightward in the TCA OD group compared with the nonTCA OD group (179 +/- 74 vs 86 +/- 87, P less than .001). A TCA OD patient was 8.6 times more likely to have a T40-ms axis of more than 120 degrees than was a nonTCA OD patient (odds ratio, 8.6; 95% confidence interval, 2.7 to 29.1). Eight of the TCA poisoned patients (17%) did not demonstrate a T40-ms axis between 120 degrees and 270 degrees. Receiver operating characteristics demonstrated that the T40-ms axis was a better indicator of TCA toxicity than the QRS interval (P less than .05). A T40-ms axis of 120 degrees or more was 83% sensitive and 63% specific for TCA overdose. A correlation between plasma TCA concentration and T40-ms axis deviation was not found (r = .04).(ABSTRACT TRUNCATED AT 250 WORDS)

Effects of magnesium citrate and clidinium bromide on the excretion of activated charcoal in normal subjects

The efficacy of cathartics in shortening the gastrointestinal transit time of activate charcoal (AC) in the presence of drugs that alter gastrointestinal motility has not been determined. We evaluated the effects of magnesium citrate (MC) on the excretion of activated charcoal in healthy volunteers alone and with concurrent administration of the anticholinergic drug clidinium bromide. Forty subjects were randomized to clidinium bromide 5 mg or placebo capsule (PC), followed by activated charcoal 15 g and magnesium citrate or a placebo liquid (PL). The onset and duration of excretion of activated charcoal were noted. Mean onset times for activated charcoal were: group I (CB, MC) 4.5 +/- 2.1 h; group II (CB, PL) 17.0 +/- 10.0 h; group III (PC, MC) 6.3 +/- 5.8 h; and group IV (PC, PL) 20.6 +/- 8.4 h. The onset of excretion of activated charcoal was statistically different in both magnesium citrate groups as compared with the placebo liquid groups. The duration of activated charcoal in the stool was similar among the groups. The addition of clidinium bromide did not appear to affect gastrointestinal transit time. These results support previous studies of the effects of cathartics on the excretion of activated charcoal, and suggest that cathartic efficacy is not inhibited by anticholinergic drugs when used in therapeutic doses.

The cardiovascular effects of antidepressants

This monograph comprises a review of the cardiovascular effects of the various types of antidepressant drugs in clinical use. The frequency, severity and clinical importance of these effects are placed in perspective. Most antidepressants can cause changes in blood pressure. Both the tricyclic type (TCA) and the monoamine oxidase inhibitors (MAOIs) can produce postural hypotension which may be dose-limiting. In addition, the MAOIs may be associated with severe hypertension when amine-containing foods or medicines are ingested. It is unlikely that therapeutic doses of any available antidepressant drug could impair cardiac contractility. Typical TCAs can cause abnormalities of cardiac conduction and arrhythmias, but this affects less than 5% of patients, mostly to a clinically insignificant extent. Newer compounds such as lofepramine, mianserin, trazodone and viloxazine seem safer in this respect. Reports of an association between therapeutic use of TCAs and sudden death are far from convincing. Overdosage with the MAOIs, lithium and carbamazepine is dangerous but not common; overdose with a TCA is a major source of morbidity and mortality. Lofepramine, mianserin and trazodone are relatively safe in overdose. The use of various antidepressants in patients with hypertension, cardiac failure, angina pectoris, myocardial infarction, or cardiac arrhythmias is discussed and guidelines suggested for the selection and use of antidepressant medication.

Impact of unusual gastrointestinal problems on the treatment of tricyclic antidepressant overdose

We report the cases of two patients with tricyclic antidepressant overdose in which the use of charcoal was hampered by gastrointestinal abnormalities. In the first patient, a previous gastric bypass procedure impeded the placement of an orogastric tube and subsequent charcoal administration, while potentially facilitating rapid absorption of the drug--factors that may have contributed to her death. In the second patient, severe esophageal spasm delayed therapy until IV nitroglycerin relieved the functional obstruction. Both circumstances are previously unreported complications associated with tricyclic antidepressant overdose. Potentially corrective measures are proposed.

Hypotension in severe tricyclic antidepressant overdose

Sixty four patients who presented to the emergency department following severe acute tricyclic antidepressant (TCA) overdose (defined as an antidepressant ingestion associated with a QRS interval greater than or equal to 0.10 seconds, TCA level greater than or equal to 500 ng/mL, or grade IV coma) were prospectively evaluated to determine the incidence of hypotension and the factors associated with its development. Among these patients, the mean antidepressant level was 1,094 ng/mL. The overall frequency of admission hypotension (systolic BP less than 95 mmHg) was 34% (22 of 64 patients). Using regression analysis, systolic BP showed poor correlation with TCA level (r = -.37) and maximal QRS interval (r = -.17) following severe TCA overdose. Using multivariate analysis with a logistic regression model, the influence of BP (as well as TCA level, QRS interval, and coingestion of another drug) was evaluated on four clinical outcomes: seizures, arrhythmias, aspiration pneumonia, and pulmonary edema. The occurrence of arrhythmias and pulmonary edema was significantly associated (inversely) with hypotension (P less than .01). Seizures and aspiration pneumonia were unrelated to admission BP. These results suggest that hypotension is common after severe TCA overdose and occurs independently of TCA level and prolongation of the QRS interval. Hypotension is strongly associated with the development of arrhythmias and pulmonary edema. Seizures and aspiration pneumonia may occur regardless of initial BP.

Criteria for admitting patients with tricyclic antidepressant overdose

Several investigators have recently developed guidelines for determining which patients with tricyclic antidepressant overdose should be hospitalized. The width of the QRS complex on the ECG and several clinical parameters have been proposed to identify patients at risk for major complications. To validate these, we developed an algorithm and then applied it to 45 patients who had overdosed on tricyclic antidepressants. This algorithm correctly predicted which patients required admission, whether due to present or impending complications, and which patients could have been discharged without morbidity or mortality. We conclude that use of the modified algorithm can identify patients with tricyclic antidepressant overdose who can be safely discharged from the emergency department.

Fatal toxicity of antidepressant drugs in overdose

A fatal toxicity index (deaths per million National Health Service prescriptions) was calculated for antidepressant drugs on sale during the years 1975-84 in England, Wales, and Scotland. The tricyclic drugs introduced before 1970 had a higher index than the mean for all the drugs studied (p less than 0.001). In this group the toxicity of amitriptyline, dibenzepin, desipramine, and dothiepin was significantly higher, while that of clomipramine, imipramine, iprindole, protriptyline, and trimipramine was lower. The monoamine oxidase inhibitors had intermediate toxicity, and the antidepressants introduced since 1973, considered as a group, had significantly lower toxicity than the mean (p less than 0.001). Of these newer drugs, maprotiline had a fatal toxicity index similar to that of the older tricyclic antidepressants, while the other newly introduced drugs had lower toxicity indices, with those for mianserin, nomifensine, trazodone, and viloxazine reaching significance. Provided that these drugs are equally effective clinically, serious consideration should be given to prescribing antidepressants with a lower fatal toxicity.

Level of consciousness as a predictor of complications following tricyclic overdose

In order to determine whether patients at low risk for complications following tricyclic antidepressant (TCA) overdose can be identified, we analyzed the records of all admissions for TCA overdose from 1975 to 1985. Of the 92 patients reviewed, 37 had serious complications including hypoventilation, loss of protective airway reflexes, seizures, hypotension, hemodynamically significant arrhythmias, or death. Patients with complications had increased heart rate, depressed level of consciousness, depressed Glasgow Coma Scale, and prolonged QRS interval, as compared to patients without complications. All 37 patients with serious complications required intubation in the emergency department. A Glasgow Coma Scale of less than 8 was the most sensitive predictor of serious complications (sensitivity = .86; specificity = .89). This was significantly better than the QRS interval (QRS greater than or equal to 100 msec; sensitivity = 59; specificity = .76). We conclude that level of consciousness is a better predictor of risk of complications and need for ICU admission than the QRS interval. Patients who are responsive to verbal stimuli on leaving the ED are at very low risk of developing complications.

QRS interval in tricyclic antidepressant overdosage: inaccuracy as a toxicity indicator in emergency settings

Emergency department treatment and disposition of tricyclic antidepressant (TCA)-overdose patients remains a common and difficult problem. Various clinical findings have been proposed as toxicity indicators. To study the performance of QRS duration as a predictor of toxicity in our patient population, we retrospectively reviewed the cases of all patients presenting to our ED with TCA overdosage. The charts of 102 patients with quantitative or qualitative laboratory confirmation of TCA ingestion were reviewed for ED findings and hospital course with specific attention to the occurrence of ventricular arrhythmias (VAs) or seizures. The ED ECG revealed that 57 patients had a maximal 12-lead ECG QRS interval duration (QRS) of less than .10 seconds (Group 1). The remaining 45 patients had QRS greater than or equal to .10 seconds (Group 2). Three patients (5%) in Group 1 and three (6%) in Group 2 experienced VAs. Four patients (7%) in Group 1 and five (11%) in Group 2 experienced seizures. There was no significant difference in the rate of occurrence of VAs or seizures between the two groups. Of note was the fact that five of eight VAs and nine of 11 seizures occurred in the ED setting. We conclude that determination of QRS duration is not an accurate indicator of VA or seizure risk for all TCA-overdose populations. In particular, risk of toxic events during the emergency phase of TCA overdose does not appear to be indicated by evaluation of the QRS duration in the ED.

Physical assessment and differential diagnosis of the poisoned patient

The rapid diagnosis and immediate intervention required in patients with serious drug overdose or poisoning makes toxicological screening of limited value to the emergency department physician. Instead, a careful clinical evaluation using the history, physical examination, and the more readily available laboratory tests may allow a tentative diagnosis and the initiation of life-saving treatment. Laboratory tests should include serum osmolality, electrolytes, glucose, BUN and an estimation of the anion and osmolar gaps. The ECG can also provide useful information. Clinical findings of important include altered blood pressure, pulse, respiration and body temperature, the presence of coma, agitation, delirium or psychosis, and muscular weakness. An ophthalmological examination is also of importance in the acutely poisoned patient. Oral burns or dysphagia may occur following ingestion of any strongly reactive substance, but the absence of oral burns does not preclude the possibility of oesophageal or stomach injury. Odours and skin colour may also contribute to the diagnosis. Comprehensive toxicology screening may not be immediately available, or may be inaccurate, thus adding little to the information obtained during the initial evaluation of the poisoned patient.