Does this patient have toxic effects of poisoning?

The 2009 Annual Report of the American Association of Poison Control Centers (AAPCC) National Poison Data System (NPDS) included 2,849,086 human exposure cases reported by 62 poison centers during 2009. This was an increase of 20% compared with 1999, and an increase of 118% as compared to 1989. A total of 299.4 million people were served by the participating centers, with an average of 8 exposures per 1000 people. Of the exposures, 91% were acute and 92% involved a single poison. Sixty percent of the exposures were unintentional while suicidal intent comprised 8% and intentional misuse comprised 2% of the exposures.

Approximately 27% of recognizably poisoned patients are treated in a health care facility and 7% require admission to the hospital. Less than one tenth of 1% of all exposed patients die. The mortality rate is higher for intentional exposures where 0.3% die and these patients comprises 78% of all fatalities in exposure cases. Three percent of the recognizably poisoned patients require intensive medical care, including hemodynamic and ventilatory support with close monitoring in an special care unit. The remainder recover with general support and ward nursing supervision. Fewer than 5% of cases of recognizable poisoning are amenable to techniques that facilitate the elimination of the poison.

A number of exposures still carry a high morbidity and mortality secondary to the toxic effect of the poison or its metabolic byproducts. Lithium, for example, causes its toxic effect directly. On the other hand, the two most toxic alcohols, ethylene glycol and methanol, are converted to glycolate and formate, respectively. These metabolites cause the metabolic abnormalities and end organ damage seen with ingestion of ethylene glycol or methanol. Many of the toxic effects of poisonings are reversible, such as hypotension, acidosis, seizures and decreased mentation.

Continue Reading

This chapter reviews the strategies for limiting the toxic effects of various poisonings. It will start with the initial approach to the intoxicated patient, which includes techniques to limit further absorption, antidotes to specific toxins, strategies to enhance endogenous elimination, and evaluation to determine the nature and severity of the intoxication. General considerations to help with decisions regarding extracorporeal therapy initiation and prescription will then be detailed. A brief review of the types of therapies available to the nephrologist including hemodialysis, hemoperfusion and continuous modalities will follow. The intoxicants that are most effectively removed by these therapies will be discussed in greater detail in separate chapters.

Less commonly, the toxic effect can be permanent such as the neurologic effects of methanol or lithium, renal failure from ethylene glycol, or rarely subsequent death from the complications of any of the intoxications discussed in this chapter.

What tests to perform?

Initial laboratory evaluation

Anion gap

The calculation of the difference between the measured cations and the measured anions can be used to estimate the difference between the unmeasured anions and the unmeasured cations. The normal anion gap is 8 to 12 mEq/L and a value above 12 mEq/L can signify an increase in unmeasured anions. The most common intoxications to cause a high anion gap acidosis are ethylene glycol, methanol and salicylates. Also, an elevated anion gap from lactic acidosis can signify an intoxication with acetaminophen, carbon monoxide, metformin, cyanide and non-steroid anti-inflammatory agents (NSAIDs).

It is important to note that a normal anion gap does not rule out an intoxication since many toxins do not cause a gap or there may be a coexisting condition that lowers the gap. The most common condition to lower the gap is hypoalbuminemia: the anion gap falls 2.5 mEq/L for every 1 g/dL drop in serum albumin. A few toxins such as methanol and ethylene glycol need to be metabolized before they create an anion gap acidosis. In these cases, intoxication may not be associated with an anion gap early on, especially when there is ethanol co-ingestion.

Osmolar gap

Ingestion of low molecular weight toxins will increase the difference between the measured and the calculated plasma osmolarity or osmolar gap. The

calculated osmolarity = 2 x Na+ + BUN / 2.8 + glucose / 18 + ethanol / 4.6

Osmolar gap = measured Osm – calculated Osm

An osmolar gap greater than 10 mOsm indicates the presence of osmotically active substances such as ethanol, methanol, isopropyl alcohol, and ethylene glycol. Hospitalized patients may develop an osmolar gap from glycerol, IV immunoglobulin, propylene glycol, radiocontrast media and sorbitol. Propylene glycol is a common vehicle for intravenous medications and can cause an osmolar gap. Its metabolite, lactic acid, can contribute to a high anion gap acidosis.

Accumulation of propylene glycol in patients receiving high doses of IV medications such as diazepam which have propylene glycol as their carrier may lead to severe acidosis with hemodynamic instability. Rarely, this may require treatment with hemodialysis. Figure 1 lists the contribution to the osmolar gap of various drugs and toxins and the expected concentration of a substance in mg/dL that would cause an osmolar gap of 10 mOsm/L.

A number of toxins such as ethylene glycol and methanol will no longer produce an osmolar gap as they are metabolized and in these cases a normal gap does not exclude intoxication, only a late presentation.

Figure 1.

Osmolar contribution of various toxins and drugs.

Evaluate for a specific toxin

In many cases, the specific toxin is known from the patient history. A pill bottle may be found near a patient in suicide attempt. In many cases the toxin is not known but initial labs may give a clue, such as a combination anion gap acidosis and osmolar gap in ethylene glycol and methanol intoxication or a combination anion gap acidosis and respiratory alkalosis in salicylate intoxication. In these cases, a confirmatory toxin level should be obtained. In many cases, treatment will need to be started before the confirmatory level is back to avoid delay that might lead to worsening complications of the intoxication.

Another factor that lowers the sensitivity of the osmolar gap is the considerable variation in the normal osmolar gap in the general population. Indeed, patients may have an increased gap that is still below 10 mOsm/kg. Thus, a high osmolar gap is supportive of intoxication but a normal gap does not rule it out.

On the other hand, the osmolar gap can also be falsely elevated. Patients who are critically ill may have an elevated gap because of the presence of endogenous substances such as amino acids. Patients with hyperlipidemia or hyperproteinemia will have spurious hyponatremia leading to an elevated gap. There is also an accumulation of osmotically active substances in chronic renal failure. For all these reasons, the osmolar gap should be used with caution as additional evidence of an alcohol intoxication but should not be used a the primary determinant of intoxication or as a screening test.

A few measurements that are commonly done in the emergency room can give a hint about the nature and amount of the toxin ingested. Three simple calculations are most helpful in determining the type of ingestion: anion gap, osmolar gap and oxygen saturation gap. The anion gap and osmolar gap are most relevant to our discussion. A review of the oxygen saturation gap can be found elsewhere.

How should patients with toxic effects of poisoning be managed?

Emergency management

After supplying supportive measures to maintain airway, breathing and circulation (“ABCs”), the management for a poisoned patient should be directed toward decreasing or limiting toxin accumulation.

Prevention of further absorption

The first therapeutic intervention should be directed at preventing further absorption of the compound in question. The four methods of gastrointestinal (GI) tract decontamination are emesis, gastric emptying, whole-bowel irrigation and adsorption using an oral sorbent with catharsis. Ipecac-induced emesis is rarely beneficial in toxic exposures for patients treated in the hospital. Gastric lavage has limited utility in the management of the poisoned patient; it has been associated with an increased risk of aspiration, arrhythmia and stomach perforation and no clinical studies have shown an improvement in outcome with the use of gastric lavage.

Whole bowel irrigation

Whole bowel irrigation with a solution of electrolytes and polyethylene glycol may be beneficial in the elimination of undissolved tablets or pills. It is most likely to be beneficial in the management of intoxication with sustained release or enteric-coated drugs or in toxins that are poorly adsorbed by activated charcoal, such as arsenic and lithium. The optimal regimen has not been well established but most of the studies used 1 – 2 L/h for 3 to 5 hours. It is time consuming and is contraindicated in patients with an ileus, hemodynamic instability or a compromised airway.

Oral sorbents

Oral sorbents (primarily activated charcoal) can bind unabsorbed drug in the gastrointestinal tract and therefore promote its elimination by decreasing its absorption. Activated charcoal is most helpful in the elimination of salicylates, phenobarbital and theophylline. It is administered as an aqueous suspension with a minimum of 8 mL of water to each gram of powder. Commercial premixed formulations are available that may contain activated charcoal with a lubricant (e.g., propylene glycol or carboxymethylcellulose) or a cathartic (e.g., sorbitol). The mean transit time of activated charcoal in fasting subjects is 25 hours; this can be reduced to 1.1 hours with sorbitol.

There are a number of associated risks with the use of cathartics including hypotension, dehydration and hyperglycemia. The administration of a cathartic alone has no role in the management of the poisoned patient. The American Academy of Clinical Toxicology recommends limiting cathartic use to a single dose to lower the risk of adverse effects.

Activated charcoal can be administered orally or via a nasogastric tube. The recommended dose is 10 times the weight of the ingested chemical or as much as possible if the dose of poison is unknown up to 1 gm/kg patient weight. Single dose activated charcoal has been shown to be most effective if given within one hour of ingestion. Its use may be considered after 1 hour in ingestions where delayed GI absorption is more likely (e.g., sustained release and enteric coated preparations).It should be used only in patients with an intact or protected airway.

Enhancing endogenous elimination

The second step in minimizing toxin accumulation or promoting its removal is to facilitate endogenous excretion through forced diuresis, manipulation of urinary pH, or removal of toxin via the gut.

Forced diuresis

Forced diuresis is a technique using volume loading to decrease tubular reabsorption. The goal is to achieve urine flow rates of 6 cc/Kg/hr with the combination of isotonic fluids and diuretics. It has the potential to cause significant volume and electrolyte imbalance and has not been shown to be effective in enhancing toxin elimination. Forced diuresis is therefore not recommended as a technique to enhance endogenous elimination in the poisoned patient.

Urinary pH manipulation

Urinary pH manipulation can effectively decrease tubular reabsorption of weak non-polar acids and bases. Manipulation of the urine pH can enhance the excretion of acidic or basic chemicals through a mechanism known as ion trapping.

The membranes of the nephron are generally more permeable to non-ionized and nonpolar molecules. Compounds are filtered and secreted in the non-ionized form of weak acids or bases by non-ionic diffusion across cell membranes.

With manipulation of urinary pH, the change in the intraluminal pH promotes the formation of a higher intratubular fraction of the ionized drug, effectively trapping the ionized moiety in the urinary space since the ionized form can no longer cross the cell membrane. For weak acids, an alkaline urine increases the fraction that is ionized.

An acidic urine does the same for weak bases. In each case, an increase in the ionized form of the drug, decreases reabsorption, enhancing renal elimination. Urine alkalinization can be used to enhance the elimination of salicylates and phenobarbital.

Alkalinization of the urine can be achieved by adding 150 mEq sodium bicarbonate to 1 L of D5W to run at 100 – 250 cc/hr. The goal is to achieve a urinary pH > 7 which usually requires 0.25 – 0.5 mEq/Kg/hr (Table I).

Table I.
Sodium bicarbonate: 150 meq in 1 L D5W
Fluid to run at 100-250 cc/hr
Aim for 1-2 mEq/kg every 3-4 hr to achieve urine PH = 7.5-8.5
Treat hypokalemia by adding 20-40 mEq KCl per L
Avoid in patients with acute or chronic kidney disease

This can only be achieved if the patient has intact renal function; urinary alkalinization should be avoided in patients with severe acute kidney injury. Risks of urinary alkalinization include volume overload, alkalemia, hypernatremia, and hypokalemia. It is important to treat the hypokalemia since it will prevent the alkalinization of the urine by promoting distal hydrogen secretion in replace of potassium secretion.

Hypokalemia can be avoided by adding 20 – 40 mEq potassium chloride to each liter of D5W with sodium bicarbonate. Acetazolamide will enhance urinary alkalinization but should be avoided because of the risk of worsening systemic acidemia which can enhance toxicity of certain poisonings most notably salicylates.

Urinary acidification is rarely used because of the potential to worsen renal injury in many poisonings. Arginine hydrochloride or ammonium chloride have been shown to be effective urinary acidification agents. Although urinary acidification may enhance elimination of weak bases, it can not be recommended as a treatment for toxicity from these compounds. Complications of urinary acidification include myoglobinuria, acute renal failure, and hyperkalemia.

Multi-dose activated charcoal

A few drugs undergo enterohepatic or enteroenteric circulation. Multi-dose activated charcoal (MDAC) can enhance elimination of these drugs by interrupting this circulation. Drugs that will have enhanced elimination with MDAC include carbamazepine, dapsone, phenobarbital, quinine and theophylline. There have not been any studies, however, that show that MDAC improves mortality or morbidity from toxicity due to these drugs.

The American Academy of Clinical Toxicology recommends that its use be considered only in patients who have ingested a potentially lethal amount of carbamazepine, dapsone, phenobarbital, quinine or theophylline. The standard regimen is to administer activated charcoal 1 g/kg then 0.5 g/kg at 2- to 6-hour intervals.

There is also some evidence for its efficacy in methotrexate toxicity and poisoning with the chlorophenoxy herbicides. Urinary acidification can be used to enhance the elimination of chloroquine, amphetamine, quinine, and phencyclidine.


After emergent treatment of the intoxication that includes stabilizing the patient, preventing further absorption and increasing endogenous elimination, the next strategy is to convey protection against the toxin by administering specific antidotes, antibodies, or substrate inhibitors. Antidotes and antibodies are available for a limited number of poisonings (Figure 2).

Figure 2.

Drugs and poisons treated with specific antidotes.

The timing of their administration can be crucial, and most antidotes are only adjunctive therapy to aggressive supportive care. The antidotes ethanol and fomepizole can be used for the toxins methanol and ethylene glycol and will be discussed below. There are a number of toxins for which administration of the antidote is the primary therapy and their therapy are well reviewed elsewhere.

Extracorporeal Therapy to Remove Toxin

To determine how well a specific extracorporeal technique will remove a specific drug or toxin, one should consider both dialysis-related factors and drug-related factors. The characteristics of a drug that determine whether it can be removed by a specific extracorporeal technique are molecular weight, protein binding, volume of distribution (Vd), lipid or water solubility, rate of equilibration with the vascular space, charge, and membrane binding.

The extracorporeal method (i.e., hemodialysis, peritoneal dialysis, or hemofiltration) also influences drug or toxin removal. Some of the important properties of the hemodialysis system that will be discussed are properties of the dialysis membrane, blood flow rate (Qb), dialysate flow rate (Qd), pH, and temperature.

Drug-related factors

Molecular weight

The molecular weight of a compound is the most reliable predictor of drug removal by a dialysis system. The molecular size, which comprises the molecular weight, shape, charge, and steric hindrance of a molecule, is also an important determinant of the molecule’s ability to permeate a dialysis membrane pore.

Low-molecular-weight compounds or small molecules are those classified as being less than 500 D. These molecules cross conventional low-flux (low porosity, low surface area) dialysis membranes readily, with the extent depending more on Qb, Qd, and effective membrane surface area. The clearance of these drugs is usually fairly close to the clearance of urea. For drugs that have low protein binding (to be discussed below), the clearance constant for these drugs can often be estimated as equal to the clearance constant of urea that is listed for the dialysis membrane being used.

High-molecular-weight compounds or “large solutes” are those greater than 5000 D; they diffuse very slowly or not at all across membranes.

Middle-molecular-weight compounds are those between 500 and 5000 D. Their removal is intermediate to the other two categories mentioned. Vancomycin, which has a molecular weight of 1500 D, and vitamin B12, with a molecular weight of 1355 D, are good examples of a middle-molecular weight compounds.

Drugs with molecular weights of more than 1000 D depend more upon convection for dialytic clearance and are substantially removed only with high-flux dialysis where there is a higher rate of water movement across the membranes. Common features of high-flux dialysis membranes include high urea clearance constants at high blood flows, high ultrafiltration coefficients ((Kuf) > 15 mL/mm Hg/hr) and higher B12 clearance. The B12 clearance for a dialysis membrane is often given as an estimate of its ability to clear middle-molecular weight compounds.

Removal of large solutes is enhanced by the use of a high flux filter with a porous membrane . Over the past 5 years, there has been a trend toward the use of higher flux dialysis membranes. Most membranes in use today have considerably higher ultrafiltration fractions, are more porous and have higher clearance of B12 as compared to filters used 5 years ago allowing for greater clearances of middle-molecular-weight compounds. Evidence for this increase in middle-molecular weight clearances can be seen with the change in the dosing of vancomycin. It is now routine to need to dose vancomycin after each dialysis session where 5 to 10 years ago it was dosed every 4 days or more for patients on dialysis.

Protein binding

The degree of protein binding of a drug or toxin will influence its clinical effect and its metabolism and excretion. Protein binding renders the drug or compound pharmacologically inactive; only the unbound fraction of the drug can be readily metabolized and excreted by the liver, kidney or filtered by a dialysis membrane. Only unbound drug is pharmacologically active because only free drug can cross the cell membrane and exert its pharmacologic effect.

The protein-drug complex is too large to cross the dialysis membrane and is therefore poorly cleared by conventional dialysis. Hemoperfusion and albumin-dialysis are more effective at removing drugs and toxins that are highly protein bound.

Malnutrition and proteinuria lower serum protein levels and therefore lead to a higher fraction of free drug owing to a reduced number of protein binding sites. Also, acute kidney injury, chronic kidney disease and critical illness can change the degree of protein binding.

The effect on protein binding by these clinical states depends on the illness and the drug in question. Medications that are acidic, such as penicillins, cephalosporins, phenytoin, furosemide and salicylates, are most severely affected by the reduced protein binding in chronic kidney disease. Acidic drugs are bound to albumin, plasma concentrations of which are often decreased in uremic patients.

Conversely, alkaline drugs (e.g. propranolol, morphine, oxazepam, vancomycin) bind primarily to non-albumin plasma proteins, such as alpha1-acid glycoprotein (AAG). AAG is an acute-phase protein whose plasma concentrations are often elevated in renal dysfunction and acute illness. Finally, as the dose of a drug increases, the level of protein binding may decrease such as in salicylate toxicity.

In some cases, protein binding may change because of competition for binding sites by other drugs, metabolites and accumulating endogenous substances. These other substances can displace medications from plasma protein binding sites. One such example is CKD-induced accumulation of hippuric acid with a resultant inhibition of theophylline protein binding.

Another example is the increase in free fatty acids in critical illness or heparin use. Free fatty acids can compete with drugs such as tryptophan, sulfonamides, salicylates, furosemide, phenytoin, benzodiazepine, and valproic acid for protein binding sites. Free fatty acid levels can change in a number of disease states such as critical illness, shock and with the use of heparin during dialysis.

Heparin use during dialysis stimulates the activity of lipoprotein lipases, subsequently increasing free fatty acid levels by triglyceride breakdown. This increases the free fraction of the previously mentioned drugs during the time that heparin is in use.

Changes in protein binding of a drug can have a significant clinical effect in the setting of highly protein bound drugs with a narrow therapeutic index such as theophylline. Table II displays the protein binding of the drugs and toxins discussed in this chapter.

Table II.

Characteristics of drugs and toxins that are amenable to removal with extracorporeal therapy.

Volume of distribution

A drug or toxin’s Vd is derived by dividing the total amount of drug in the body by its plasma concentration. It should be noted that this ratio often does not refer to a specific anatomic compartment in the body especially when it is large (i.e., > 1 L/kg) .

V(L) = dose (mg) /Cp (mg/L)

Vd may be affected by a number of physiologic determinants, including plasma protein and tissue binding, lipid partitioning, active transport systems and overall body composition. A large volume of distribution implies a high degree of tissue binding. Drugs and toxins that have high lipid solubility will have a high volume of distribution as well. They are likely to be able to diffuse more rapidly into the brain and are usually cleared poorly by the kidney or hemodialysis.

The drugs and toxins that have a Vd < 1 L/Kg are usually water soluble and have low tissue binding and lipid solubility. Drugs that meet these criteria include the alcohols (ethanol, ethylene glycol, methanol, and isopropyl alcohol), salicylates, lithium, theophylline, aminoglycosides and most cephalosporins. These drugs are more likely to be amenable to removal with extracorporeal techniques. Table II displays the volume of distribution (Vd) of toxins and drugs discussed in this chapter.

Compounds with a high volume of distribution (Vd) with a high degree of tissue binding are not substantially removed by hemoperfusion or hemodialysis. Some drugs and toxins that are known to cause severe toxic syndromes in overdose but are not well removed by extracorporeal therapy because of their high Vd.

A few compounds have a large volume of distribution but are also cleared by the kidney and therefore in the situation with a toxic level and kidney failure, extracorporeal therapy may be of some benefit. In these cases, the half life of elimination may be significantly prolonged by the kidney failure and drug removal by an extracorporeal therapy may still benefit the patient even when it is very slow. Metformin intoxication with renal failure may fit this model and hemodialysis may be beneficial in some cases.

The Vd of a compound can change in a number of disease states and other circumstances. In cases such as acute kidney injury, chronic kidney disease and critical illness where the protein binding of the drug changes, drugs with a high degree of protein binding will have a change in their volume of distribution. As the protein binding goes down, the volume of distribution will usually increase. Most water soluble agents will have a further increase in Vd in acute kidney injury as the total body water increases.

The total drug taken as well as the chronicity of the ingestion can also influence the Vd in overdoses. For example, the Vd of salicylates increases in a toxic ingestion and with chronic ingestion. The former is due to a decrease in protein binding and the latter is due to an increased tissue binding. The fraction of unbound drug in the blood and tissue can influence the Vd. In patients with impaired plasma protein binding, there is an increase in the apparent Vd of the drug. This is seen in patients with renal failure, owing to decreased albumin and impaired binding capacity of albumin. Renal failure decreases the Vd for digoxin but increases the Vd for phenytoin.

Vascular equilibrium

Even for substances with a small Vd (i.e., Vd ≤ 0.6 L/Kg), most of the compound is outside the vascular space. The proportion of a compound that is extra-vascular increases with increasing Vd. After an ingestion, the substance’s concentration in the vascular space will often decline following first order kinetics such that the proportional change in concentration over time remains the same. This first order elimination suggests that the compound is in pseudoequilibrium between the movement into the vascular space and elimination out of the vascular space.

This movement takes time and with rapid removal of the drug from the vascular space, the pseudoequilibrium will be disturbed as the substance is removed from the vascular space faster than it can be replaced from extracellular and cellular stores (i.e., no longer following first order kinetics). With discontinuation of the extracorporeal elimination, there will be a rebound in the concentration as the movement into the vascular space catches up. The degree of rebound increases with Vd, the degree of tissue binding, and inversely with the rate of clearance of the drug from the vascular space. It is not a concern in continuous therapies.

It is important to keep in mind the degree of rebound that is likely when reviewing literature on the effectiveness of an extracorporeal therapy to treat a drug intoxication. A report of a significant drop in concentration of a drug with therapy may just represent a disruption in the pseudoequilibrium between the intravascular space and the tissues. The important data is the fraction of the total drug burden removed and is best determined by measuring the drug in the effluent.

There are a number of reports of success with removing toxins with high Vd and slow tissue equilibration such as tricyclic antidepressants or metformin but in many cases, the authors report concentrations drawn after termination of the technique and not fraction of the drug removed.

Lithium has an intermediate Vd (≈0.8 l/Kg) and slow equilibration into the vascular space. Although hemodialysis is effective at removing lithium in intoxication, the rebound can be significant following termination of the therapy and repeated treatments with hemodialysis or the use of a continuous therapy may be necessary.

Device-related factors

In addition to drug or toxin characteristics, the effectiveness of the extracorporeal removal is also determined by the properties of the extracorporeal device. How effectively does the device remove the toxin from the plasma delivered to the device? The extraction ratio and the clearance rate are measures of the efficiency of the device.

Extraction ratio and clearance rate

The extraction ratio (ER) is determined by measuring the concentration of the drug (plasma or blood levels) before it enters the hemoperfusion cartridge or hemodialyzer filter (A) and just after it exits (V). The ER can refer to the removal or the extraction of a drug from whole blood or plasma. It is calculated by the following formula:

ER = A – V / A

A value of 1.0 indicates that the drug was completely removed (extracted) in one pass through the extracorporeal system. The clearance rate (Cl) is a measure of the rate at which blood or plasma is cleared of the substance by the device in a given time. The clearance rate can thus be calculated by knowing the flow rate (blood or plasma) through the system:

Cl = flow rate x ER

It is important to differentiate between a high extraction ratio and effective total drug removal. A high clearance rate is necessary but not sufficient for effective removal of a drug. This will be discussed below.

Effective extracorporeal elimination

As stated above, efficient removal of a toxin from the plasma or blood is necessary to achieve effective elimination of the drug from the patient but it is not sufficient. The pharmacokinetics of digoxin will help illustrate this concept. The Vd of digoxin is approximately 10 L/kg. If a 70Kg man ingests a toxic dose of one hundred 250 mcg tablets of digoxin, his plasma level would be ≈ 36 mcg/L after distribution to tissues. Hemodialysis has a fairly high ER for digoxin but even if we assume an ER of 1 with a plasma flow of 200 ml/min the total removal of digoxin would be small after 4 hours.

In this case, the Cl = 200 ml/min or 48 L per 4-hour dialysis session. At most, this would remove 48 L x 36 mcg/L = 1.7 mg or 6 % of the total dose. The actual removal of digoxin is even less because of the slow equilibration of digoxin from tissue stores. In the case of digoxin, the ER is sufficiently high, but because of characteristics of the drug (i.e., high Vd and slow tissue equilibration) hemodialysis is not an effective therapy for digoxin toxicity and the antidote digoxin immune Fab needs to be used to bind the drug.

In summary: extracorporeal techniques are directed at the compound available in the plasma or blood. For the device to work effectively, the following conditions must be met:

1. There most be a high clearance rate of the compound from the blood.

2. A large proportion of the compound must be intravascular (i.e., small Vd).

3. The compound must equilibrate quickly from the tissue stores to the vascular compartment.

As discussed above, the clearance rate is determined by a combination of drug and device characteristics. The drug characteristics include low protein binding and small size. Below, we will discuss the device characteristics.

Device characteristics

Hemodialyzer properties

For hemodialysis, the factors that determine the clearance rate are membrane permeability, surface area of the membrane, blood flow and dialysate flow. Solute removal during hemodialysis is accomplished primarily by diffusion, with a smaller contribution coming from convection.

Convection becomes a little more important with high flux dialysis membranes. To maximize clearance of a compound, one would pick a large permeable membrane with high blood and dialysate flows. A high efficiency membrane is able to support higher blood and dialysis flows and achieves higher urea clearance. A high flux membrane is more permeable to middle molecules and allows for clearance of toxins with higher molecular weights.

Most dialysis membranes in use today are high efficiency and high flux and therefore maximize the clearance rate of small and middle molecular weight compounds. Clearances for small, unbound toxins can approach 300 cc/min with a high efficiency membrane given high blood and dialysate flows.

Peritoneal dialysis properties

Clearance rates for peritoneal dialysis are significantly less than for hemodialysis and are rarely adequate to obtain significant toxin removal. The factors in peritoneal dialysis that contribute to drug or toxin removal are (1) transport characteristics of the peritoneal membrane, (2) composition of the dialysate solution, (3) frequency of exchanges, and (4) dwell times.

Hemofiltration properties

For hemofiltration, the solutes, drug, or toxin is removed primarily by convective mass transfer. As such, solutes dissolved in plasma water are removed in the filtrate. Most small molecules will cross the membrane close to their concentration in the serum. The sieving coefficient (S) is the ratio of the concentration in the ultrafiltrate to that in the serum.

S = C(f) / C(p)

Where C(f) is concentration in ultrafiltrate and C(p) is concentration in plasma. The sieving coefficient is usually close to 1 for small, non protein bound molecules. The clearance rate (Cl) is proportional to the sieving coefficient and the ultrafiltration rate:

Clearance rate = ultrafiltration rate x sieving coefficient.

Thus, toxin removal depends on high rates of ultrafiltration.

Hemoperfusion properties

Hemoperfusion allows for the removal of compounds by direct contact with a material that adsorbs the compound. The material that acts as the sorbent is activated charcoal. Unlike hemodialysis, hemoperfusion is able to remove highly protein bound and lipophilic compounds. The ER ratio for most toxins approaches 1 and the clearance rate is therefore mostly determined by the blood flow. To maintain this high ER, the activated charcoal cartridge needs to be changed after a few hours of therapy.

Clearance rates of over 200 cc/hr have been described for many toxins. Because of superior blood flows with hemodialysis, however, compounds with good ER in hemodialysis will have better clearance rates for hemodialysis as compared to hemoperfusion.

Indications for extracorporeal therapy

Indications for extracorporeal elimination of drugs or toxins depend most strongly on the clinical severity and potential complications of the poisoning. The following issues must be considered:

1. Characteristics of the individual patient: does the patient have impaired endogenous clearance of the toxin(e.g., older age, decreased renal function, congestive heart failure, liver failure) and are they more likely to have clinical toxic effects from the compound (e.g., older age, chronic ingestion, critical illness)?

2. Characteristics of the compound: What are the toxic effects of the substance ingested, are there antidotes available, are the adverse effects likely to be severe, permanent or life threatening?

3. Characteristic of the ingestionss: Was it a toxic dose, what is the plasma concentration, how is the level changing with time, is it likely to go up over time or fail to fall?

Appropriate interpretation of the drug concentration must take into account hepatic or renal elimination, delayed gastrointestinal absorption, active metabolites, altered distributional characteristics, and saturable elimination pathways. Extracorporeal elimination that increases the total body clearance by 30% or more is believed to be a worthwhile intervention in the proper clinical setting.

Extracorporeal therapy may be considered when all of the following conditions are met:

1. The ingestion is likely to cause severe morbidity or mortality and the removal of the drug from the serum will lessen this risk. In some intoxications the effect is too rapid and irreversible for extracorporeal removal to help (e.g., cyanide) or the removal from the serum does not remove it from the tissues where it has its toxic effect (e.g., paraquat). Keep in mind the patient characteristics that may make the risk of severe toxic effects more likely, such as older age or chronic ingestion in salicylate and theophylline ingestions.

2. The extracorporeal therapy will add significantly to the total body elimination of the drug (>30%). In this case, the device must have a high clearance rate for the compound and the compound must be mostly in the vascular space (i.e., Vd ≤ 0.6 L/Kg) or equilibrate into the vascular space quickly. In some cases, this condition might be met partly because of a decreased endogenous clearance in the patient in question.

A patient with lithium toxicity may be more likely to benefit from hemodialysis when there is impaired renal clearance because of heart failure or liver or kidney disease. A patient with lactic acidosis from metformin intoxication may benefit from hemodialysis if she also has severe acute kidney injury.

There is some controversy about which poisons are likely to respond to extracorporeal therapy. Those for which there is some consensus regarding effectiveness of extracorporeal therapy are listed in Figure 3 with their important characteristics. Please see the related topics in intoxications that deal with the specific toxins and drugs that are frequent causes of poisonings and whose elimination is significantly enhanced with either hemodialysis or hemoperfusion.

Toxins in which hemodialysis is likely more effective include ethanol, methanol, ethylene glycol, isopropyl alcohol, salicylates, metformin and lithium. Other drugs such as theophylline, phenytoin, carbamazepine, valproic acid and phenobarbital have a higher degree of protein binding and may benefit from hemoperfusion as compared to hemodialysis. Table III shows the number of reported exposures and the mortality rates for these toxins as reported to NPDS in 2009. For specifics on each of these intoxications and their management, please see those subject headings.

Table III.

Exposures and fatalities from toxins that are substantially removed by extracorporeal technique

What is the evidence?

Winchester, JF, Harbord, NB, Rosen, H. “Management of poisonings: core curriculum 2010”. Am J Kidney Dis. vol. 56. 2010. pp. 788-800.

Bayliss, G. “Dialysis in the poisoned patient”. Hemodialysis Int. vol. 14. 2010. pp. 158-167.

Lynd, LD, Richardson, KJ, Purssell, RA, Abu-Laban, RB, Brubacher, JR, Lepik, KJ, Sivilotti, ML. “An evaluation of the osmole gap as a screening test for toxic alcohol poisoning”. . vol. 8. 2008. pp. 5

Kraut, JA, Madias, NE. “Serum anion gap: its uses and limitations in clinical medicine”. Clin J Am Soc Neph. vol. 2. 2007. pp. 162-174.

Mokhlesi, B, Leikin, JB, Murray, P, Corbridge, TC. “Adult toxicology in critical care”. Chest. vol. 123. 2003. pp. 897-922.

Bronstein, AC, Spyker, DA, Cantilena, LR, Green, JL, Rumack, BH, Giffin, SL. “2009 Annual Report of the American Association of Poison Control Centers' National Poison Data System (NPDS): 27th Annual Report”. . vol. 48. 2010. pp. 979-1178.

Levine, M, Ruha, AM, Graeme, K, Brooks, DE, Canning, J, Curry, SC. “Toxicology in the ICU: part 1: General overview and approach to treatment”. Chest. vol. 140. 2011. pp. 795-806.

Levine, M, Ruha, AM, Graeme, K, Brooks, DE, Canning, J, Curry, SC. “Toxicology in the ICU: part 2: Specific toxins”. Chest. vol. 140. 2011. pp. 1072-85.

Levine, M, Ruha, AM, Graeme, K, Brooks, DE, Canning, J, Curry, SC. “Toxicology in the ICU: part 3: natural toxins”. Chest. vol. 140. 2011. pp. 1357-70. These three articles are an excellent overview of intoxications and are an update on the 2003 article from the same journal.

Mowry, JB, Spyker, DA, Cantilena, LR, Bailey, JE, Ford, M. “2012 Annual Report of the American Association of Poison Control Centers' National Poison Data System (NPDS): 30th Annual Report”. Clinical Toxicology: The Official Journal of the American Academy of Clinical Toxicology & European Association of Poisons Centres & Clinical Toxicologists. vol. 51. 2013. pp. 949-1229. An excellent compilation of the data in the trends in types of intoxications including those requiring treatment with extracorporeal therapies.