Digitalis Toxicity Workup

The plasma digoxin level can be used to monitor compliance and toxicity and can be used as a guide to the appropriate dosing of medication.

Therapeutic digoxin levels vary; the lower limit ranges from 0.6-1.3 ng/mL, while the upper limit generally is agreed to be 2.6 ng/mL. Serum concentrations associated with toxicity overlap between therapeutic and toxic ranges because of the myriad of factors potentiating digoxin toxicity.

Measure Na⁺, K⁺, Cl^-, CO^2-, Mg⁺⁺, Ca⁺⁺, blood urea nitrogen (BUN), and creatinine levels. Obtain cardiac markers such as creatine kinase MB or troponin I or T if myocardial infarction is a clinical concern.

False-negative assay results may occur in the setting of acute ingestion of nondigoxin cardiac glycosides, such as foxglove and oleander, even in the setting of profound clinical toxicity. This is caused by nonreactivity or minimal cross-reactivity with the digoxin radioimmunoassay.

Initial potassium levels are better correlated with the prognosis than either ECG changes or the initial serum digoxin level. In one series, all patients with an initial potassium level greater than 5.5 died, whereas 50% of patients with a serum digoxin level of 5-5.5 died.^[18]

The development of sensitive and accurate radioimmunoassays has improved the diagnosis and management of digitalis toxicity.

The therapeutic range is 0.5-2 ng/mL, but significant levels in patients with toxicity and levels in those without toxicity overlap significantly. Consequently, digoxin levels cannot be used as the sole indicator of toxicity.

Neonates and small infants rarely develop toxic symptoms or ECG abnormalities with serum levels of less than 4-5 ng/mL. Children without cardiovascular disease may tolerate levels as high as 10 ng/mL without serious toxicity, but they may have bradyarrhythmias or conduction delays on ECG. The general rule is that the smaller the infant, the higher the levels may be before toxic effects are observed.

levels determined less than 6-8 hours after an acute ingestion reflect the initial distribution of the drug but not the actual tissue levels, and they are not necessarily predictors of toxicity. The plasma half-life of digoxin is shortened to 10-25 hours with acute, massive ingestions, compared with a mean value of 36 hours in nontoxic ingestions.

In acute toxicity, repeat the digoxin level after 2-4 hours to guide therapy. Digoxin levels do not necessarily correlate with toxicity, especially after acute ingestion. Following acute ingestion, digoxin levels do not equilibrate quickly because of variable absorption and subsequent tissue distribution. Toxicity is related to intracellular levels, not serum levels.

To best way to guide therapy is to follow the digoxin level and correlate it with serum potassium concentrations and the patient's clinical and ECG findings.

DLIS

Endogenous digoxinlike immunoreactive substance (DLIS) can cause a false-positive result or an elevated digoxin level. DLIS is observed in neonates and in patients with renal insufficiency, liver disease or hyperbilirubinemia, subarachnoid hemorrhage, CHF, diabetes mellitus, or acromegaly; it may also be present in those who are pregnant or using spironolactone.

In some studies, premature infants had levels as high as 4 ng/mL, with peaks at age 6 days, and positive assay results until they were aged 3 months. Most authors agree that serum digoxin levels due to DLIS are usually less than 2 ng/mL and that the interference is assay dependent and may vary with the lot of the reagent. Some laboratories use ultrafiltration techniques to eliminate the contribution of DLIS.

Other variables

Other confounding variables include digoxin metabolites and drugs. While most patients metabolize less than 20% of digoxin, 10% of the population metabolizes as much as 55% of digoxin to initially active metabolites. Not all routinely used radioimmunoassays measure each of these metabolites. Additionally, the antibodies used in digoxin immunoassay can cross-react with numerous compounds, including steroids and spironolactone.

Because most digoxin assays measure total rather than free digoxin levels, serum digoxin levels are no longer useful after Fab fragment administration.

Initial potassium levels are better correlated with the prognosis than either ECG changes or the initial serum digoxin level. In one series, all patients with an initial potassium level greater than 5.5 died, whereas 50% of patients with a serum digoxin level of 5-5.5 died.^[18]

Hyperkalemia

In acute toxicity, hyperkalemia is common owing to inactivation of the Na⁺/K⁺ -ATPase pump. It is a predictor of morbidity and mortality and also reflects the degree of poisoning.

Hypokalemia

Long-term digoxin users very often develop hypokalemia because of concurrent diuretic use. The condition should be corrected promptly; treating hypokalemia may help to improve cardiac glycoside-related arrhythmia.

Hypomagnesemia

Long-term digoxin users often have hypomagnesemia secondary to diuretic usage. Intracellular magnesium depletion may occur in long-term diuretic use despite a normal serum magnesium level. Importantly, magnesium is a cofactor of the Na⁺/K⁺ -ATPase pump, and alterations of its concentration will affect the pump's actions.

Electrocardiography may be necessary to facilitate the diagnosis, the type of rhythm, and arrhythmia. Suspect digitalis toxicity when the evidence suggests increased automaticity and depressed conduction. Sinus bradycardia and AV conduction blocks are the most common ECG changes in the pediatric population, while ventricular ectopy is more common in adults.^[10] (Almost any dysrhythmia may occur, but rapid atrial fibrillation or flutter is rare.)

Nonparoxysmal atrial tachycardia with a block and bidirectional ventricular tachycardia are particularly characteristic of severe digitalis toxicity.

Digoxin effects on the baseline ECG include downward scooping of the ST segment and inverted T waves. These findings are not indicative of toxicity. New QRS prolongation, varying degrees of AV block, and arrhythmias may signify digoxin toxicity. Comparison with previous ECGs is helpful. Rhythm strips may be necessary to facilitate arrhythmia analysis

Nonspecific ECG findings include the following:

• Premature ventricular contractions, especially bigeminal and multiform
• First-, second- (Wenckebach), and third-degree AV block
• Sinus bradycardia
• Sinus tachycardia
• SA block or arrest
• Atrial fibrillation with slower ventricular response
• Atrial tachycardia
• Junctional (escape) rhythm
• AV dissociation
• Ventricular bigeminy and trigeminy
• Ventricular tachycardia
• Torsade de pointes
• Ventricular fibrillation

More specific, but not pathognomonic, ECG findings include the following:

• Atrial fibrillation with slow, regular ventricular rate (ie, AV dissociation)
• Nonparoxysmal junctional tachycardia (rate 70-130 beats per minutes [bpm])
• Atrial tachycardia with block (atrial rate usually 150-200 bpm)
• Bidirectional ventricular tachycardia