Buckley NA, Dawson AH, Reith DA. Controlled release drugs in overdose, clinical considerations. Drug Safety 1995; 12(1): 73-84 PMID 7741985
The main characteristic of poisoning with controlled release formulations is the delay in presentation and onset of clinical effects. There is a prolonged absorption phase, which leads to a delayed time to maximum plasma concentration (Tmax) and usually a sustained time close to the peak concentration (Cmax). Absorption may continue for more than 24 hours. Poisoning with controlled release formulations of toxic drugs requires a longer period of observation as the onset of symptoms may be as late as 16-20 hours. Treatment nomograms calculated for standard formulations are not appropriate for controlled release formulations. The optimal gastrointestinal decontamination method is controversial but in serious poisonings should include gastric lavage and activated charcoal followed by whole bowel lavage as a means of clearing whole tablets from the gastrointestinal tract. Pharmacobezoar formation should be suspected if, despite apparently effective gastrointestinal decontamination, there is evidence of continuing absorption. These are best diagnosed with endoscopy and the treatment options include endoscopic removal, whole bowel lavage, and surgery.
Controlled release formulations of many drugs are widely used. A number of potentially lethal drugs are available in these formulations and this may lead to additional problems in the management of overdose. Poisonings with these preparations have in common the pharmacokinetic characteristics produced by the controlled release vehicles. There is a prolonged absorption phase, which leads to a delayed time to maximum plasma concentration (Tmax) and usually a sustained time close to the peak concentration (Cmax). Poisoning with controlled release formulations of toxic drugs requires a longer period of clinical vigilance and a different approach to gastrointestinal decontamination. These poisonings may also present later and a follow a different clinical course to standard preparations. In addition, there are some unusual toxic syndromes due to the delivery mechanisms themselves.
We will use the term controlled release drugs to cover a number of different methods of modifying the release and hence absorption of medications. These include transdermal delivery of drugs and “slow release”, “extended release” and “delayed release” preparations. We will not discuss parenteral medications (e.g. insulin) in this article.
Controlled release formulations have been developed for the dual purposes of decreasing adverse effects and increasing efficacy of medications. Decreased adverse effects are achieved by delaying or slowing the release of gastric irritants until they have passed the pylorus, for example, enteric coated aspirin or sustained release potassium (Senel et al 1991), or reducing peaks in plasma drug concentration which may be associated with systemic toxicity (Todd & Faulds 1992). Increased efficacy may be achieved by improving compliance and by prolonging the action of a drug without increasing adverse effects (Powers-Cramer & Saks 1994). Such a rationale is appropriate for drugs with a relatively short half life which are intended for long term treatment and which are associated with acute concentration dependent toxicity (CPMP Working Party 1991, Urquart 1982).
The mechanisms of controlled release formulations include dissolution systems, diffusion systems, osmotic pump systems, ion exchange resins, and transdermal systems (Prisant et al 1992, Robinson & Gauger 1986, Ranade 1991). Combinations of different systems may also be used. These systems are briefly described in Table 1. These are not homogenous and even within groups there may be substantial differences in the rate of dissolution dependent on factors such as hardness and type of matrix material (Senel et al 1991, Robinson & Gauger 1986). These differences may lead to different absorption profiles (Sommers et al 1992).
The different controlled release vehicles can also be classified according to their absorption profile (CPMP Working Party 1991). This pharmacokinetic classification based on their idealised absorption in therapeutic doses is illustrated in Figure 1.
The half-life is normally defined as “the time it takes for the plasma concentration to be reduced by 50%” and is proportional to the volume of distribution and inversely proportional to drug clearance (“the rate of elimination by all routes relative to the concentration of drug in any biologic fluid”) (Benet & Massoud 1984). This elimination half-life is calculated after a drug is completely absorbed and distributed. The half-life may be calculated as:
* Half life = ln 2 * [T2-T1]/[ln Cp1) - ln Cp2)]
Where Cp1 is the plasma concentration at time T1 and Cp2 is the plasma concentration at time T2. However, if this equation is used when there is ongoing absorption, the half-life of the drug appears to be much longer and the mathematical relationship to clearance no longer holds. This “apparent half-life” in controlled release drug poisoning is therefore a clinically useful measure of both drug clearance and the success of gastrointestinal decontamination in preventing ongoing absorption.
Most standard preparations are absorbed from the upper small intestine. In contrast, many of the drugs in extended release formulations are released throughout the gastrointestinal tract including the colon. Many drugs available in controlled release preparations are significantly absorbed from the colon, though the absorption half-life is prolonged, as there is reduced surface area (Staib et al 1986). Despite unchanged drug clearance, the apparent half-life is longer because of continuing absorption. This has been termed absorption dependent or “flip flop” kinetics and is due to the absorption half-life being longer than the elimination half-life of the medication (Minocha & Spyker 1986)
Controlled release formulations may also lead to changes in bioavailability and drug or food interactions. Rapid gastrointestinal transit may allow insufficient time for drug absorption (Prisant et al 1992, Robinson & Gauger 1986). Drugs with saturable first pass metabolism will have reduced bioavailability, as with decreased portal blood concentrations there will be reduced saturation of drug metabolism (Welling 1986). There are also large intersubject and inter-formulation variations in the rate of absorption (Sommers et al 1992). Ingestion of food may lead to dose dumping decreasing Tmax (Hendeles et al 1985). Pharmacokinetic studies of these preparations are thus usually performed in both fasting and non-fasting subjects. Tampering with the delivery vehicle such as through crushing or chewing the vehicle or even in some cases breaking it in two will result in a more rapid absorption of the medication (CPMP Working Party 1991).
Dermal delivery systems operate by diffusion of the drug across a concentration gradient. The bioavailability of the preparation is much reduced as a large amount of the drug remains in the delivery system when it is discarded (Meyer-Harris 1990). Diffusion is affected by the integrity and thickness of the stratum corneum as a barrier and by the blood supply to the skin. Rapid absorption occurs if transdermal preparations are chewed (Harchelroad et al 1992), however, it is unclear whether they will behave as a standard or controlled release preparation if ingested (Raber 1993).
Clinical toxicity specifically related to controlled release formulations is uncommon in therapeutic use. There are a few combinations of specific drugs and formulations that have caused toxicity. Indomethacin in an osmotic pump vehicle was associated with intestinal perforation and withdrawn in the United Kingdom (Day 1983). The GITS system has been associated with gastric outlet obstruction when prescribed following surgery for morbid obesity (Prisant et al 1991). Enteric coated aspirin in the presence of gastric outlet obstruction has been associated with bezoar formation with unpredictable release of aspirin from the aggregation (Bogacz & Caldron 1987).
There may be loss of controlled release characteristics with ageing, increasing the rate of absorption and causing “dose dumping” (D'Arcy 1985, Robinson & Gauger 1986). Dose dumping has also been associated with some theophylline preparations when ingested with food (Hendeles 1985). Inadvertent substitution of a standard formulation at the same dose for the controlled release preparation will cause a very similar clinical picture. These effects will be most apparent for drugs with acute dose related toxicity. As this is frequently the reason for developing a controlled release preparation, a significant number of drugs available in controlled release formulations will cause toxicity with rapid absorption.p
The major clinical and kinetic differences between poisonings with controlled release and standard formulations are due to the duration and rate of absorption.
There have been no formal studies of the kinetics of controlled release drugs in overdose. It might be expected that with overdose the rate of absorption would be equivalent to that in therapeutic doses. In that case, the Tmax would be the same as in therapeutic dosing if bioavailability was unaltered and drug clearance was linear (1st order kinetics). However, Tmax is often greatly delayed due to a number of factors that may further prolong absorption. Such factors include drug-induced delays in gastric emptying, decreased intestinal motility, or the formation of pharmacobezoars. Many drugs have dose dependent (zero order) kinetics in overdose resulting in a markedly prolonged elimination half-life and a consequent increase in Tmax and Cmax . In contrast, the rate of absorption may be enhanced by crushing or chewing which destroy the delivery mechanism, gastrointestinal decontamination may reduce both the extent and duration of drug absorption and some interventions may enhance drug clearance; all leading to a reduced Tmax.
Thus it is not surprising that case reports vary widely, with Tmax occurring from as early as 2 hours to much later than 24 hours (Bosse & Arnold 1992, Kwong et al 1983, Todd et al 1981, Wortzman & Grunfeld 1987, Olsen et al 1985, Buckley et al 1983) and evidence for continued absorption even later (Connell et al 1982, Hendren et al 1989). Case series report Tmax occurring later than generally occurs in therapeutic use Amitai & Lovejoy 1987, Gaudrealt et al 1983). At our centre, the Tmax occurred between 10-12 hours in forty percent of the patients who had taken controlled release formulations despite the majority receiving gastrointestinal decontamination (unpublished data).
As a consequence of the slow rate of absorption, there may be time for adaptive responses to poisonings with some drugs for which homeostatic compensation or tolerance develops rapidly. Amelioration of potentially toxic effects has been described for many drugs in therapeutic use (Castaneda-Henandez et al 1994) and in chronic (v acute) poisonings (Shannon 1993). There is less evidence that this is a significant factor in acute poisoning with controlled release drugs as any tolerance is likely to be offset by the larger ingested dose and delayed presentation.
The specific clinical effects will be determined by the drug ingested. All will demonstrate a relatively delayed and prolonged time course.
The delayed and prolonged absorption may result in later presentation to medical attention due to the delay in onset of symptoms. In our own experience, the median time to presentation for all poisonings is 2-3 hours but for controlled release preparations is 4-5 hours (Buckley et al 1994 & unpublished data). This extra delay may make decontamination procedures used for standard formulations less effective as significant quantities of drug may be past the pylorus. The lack of acute toxicity may lead to the continued ingestion of drugs compounding the effect of late presentation.
The subsequent time course is variable and dependent on pharmacokinetic factors outlined above. The onset of symptoms may be as late as 16-20 hours (Buckley et al 1993, Spiller et al 1991, Colledge et al 1988). Death and serious complications may also occur late. Deaths as late as 44-58 hours after sustained release verapamil poisoning has been reported (Mc Donald et al 1992, Rankin & Edwards 1990). Similar late effects from poisoning with drugs such as aspirin, theophylline, and lithium have probably been averted with enhancement of clearance with haemodialysis and haemoperfusion.
The release of the drug from the formulation may be further prolonged due to the formation of a concretion of tablets in the stomach or intestine. These pharmacobezoars have been reported with diffusion and dissolution formulations (Coupe 1986, Cereda et al 1986, Rankin & Edwards 1990, Sporer et al 1993, Buckley et al 1993, Whyte KF & Addis GJ 1983, Bernstein et al 1992) but have not specifically been reported with the osmotic pump systems. Some drugs, such as theophylline and verapamil, are frequently reported and may be particularly prone to form bezoars (irrespective of the vehicle) and others, such as carbamazepine and aspirin, may even form bezoars with standard formulations (Coutselinis & Poulis 1980, Vree et al 1986, Ellenhorn & Barceloux 1987). Increased propensity to form bezoars may be attributable to the drugs solubility, drug effects on gastrointestinal motility and the agglutinating effect of the vehicle. Bezoar formation should be suspected if despite apparently effective gastrointestinal decontamination there is evidence of continuing absorption either by an increase in clinical toxicity or in the apparent half-life.
Pharmacobezoars of radio-opaque medications may be diagnosed (but not excluded) by a plain abdominal X-ray (Jaeger et al 1981 & Anon 1994). For other medications, gastro-endoscopy is the method of choice and may also aid subsequent management (Cereda et al 1986,Coupe et al 1986). Barium studies may precipitate release of the medication from the bezoar and are not recommended (Bogacz & Caldron 1987).
Definite diagnosis of intestinal bezoars of non radio-opaque medication is usually only made post mortem.
Toxicity due to the controlled release vehicle itself (causing obstruction - Prisant et al 1991) or the effect of prolonging contact of irritant drugs with the mucosa (Day 1983) has been reported in therapeutic use. Presumably, these effects can also occur in overdose however they do not appear to present additional significant clinical problems. Perforation of the stomach has been reported with both enteric and standard formulations of aspirin in overdose though this is very uncommon (Robins et al 1985, Farrand et al 1975).
Transdermal formulations can cause clinical toxicity in unusual circumstances. Absorption is limited by the thickness of the stratum corneum and if preparations are either accidentally or therapeutically used in children at adult doses they may cause toxicity (Hamblin & Martin 1987, Sennhauser & Schwarz 1986). The few reports of transdermal overdose in adults with nitrates and nicotine patches have not caused significant clinical toxicity (Ehrenpreis et al 1989, Engel & Parmentier 1993), however ingestion may lead to significant toxicity (Raber 1993, Harchelroad et al 1992).
Despite the large number of medications available as controlled release preparation the list of drugs associated with fatality or significant morbidity is relatively small (Table 2). The history of the ingested dose may predict patients at significant risk of toxicity. Management, in particular the need for gastrointestinal decontamination and monitoring, should be based on this as the onset of clinical symptoms and signs may be greatly delayed (Bernstein et al 1992, Buckley et al 1993). Patients should be monitored for signs of clinical toxicity until the expected time to peak concentration has passed. Even in the presence of apparently effective gastrointestinal decontamination, this may take 12-24 hours (Connell et al 1982, Todd et al 1981).
Monitoring of drug concentrations
Treatment nomograms based on drug concentrations, (e.g. aspirin, iron, and paracetamol), calculated for standard release formulations are not appropriate for controlled release formulations and their application may lead to inappropriate management (Kwong et al 1983, Todd et al 1981). Treatment decisions for these patients may need to be based on clinical toxicity and calculations of the total ingested dose. For drugs where the absolute concentration may indicate the need for further treatment (e.g. lithium, theophylline, procainamide), the concentrations should be measured until there is a sustained decline to non-toxic concentrations. Poisonings with these preparations have demonstrated multiple peak concentrations indicating continuing and variable absorption for greater than 24 hours (Connell et al 1982, Hendren et al 1989, Colledge et al 1988).
At the time of presentation, many patients may have tablets throughout the gastrointestinal tract. There are numerous reported cases where routine gastrointestinal decontamination for standard preparations, with gastric lavage and one or two doses of activated charcoal, has been followed by delayed but significant toxicity and even death (Buckley et al 1993, Hendren et al 1989, Spiller et al 1991, Wortzman & Grunfeld 1987 ). Decontamination should therefore aim to remove tablets and prevent absorption from the entire gastrointestinal tract.
The need for decontamination should be clearly explained, as the potential toxicity may not be apparent while the decontamination techniques may appear extreme or punitive. Possible methods of achieving decontamination include induced emesis, gastric lavage, activated charcoal, and whole bowel irrigation. Other than induction of emesis, all of these may have a role in some poisonings.
Induced emesis, with ipecac, is inferior to activated charcoal for standard formulations in both real and simulated poisonings (Albertson et al 1989, Curtis et al 1984, Underhill et al 1990). Any delay in presentation or the presence of tablets beyond the pylorus will even further reduce the effectiveness of emesis. In addition, protracted emesis can also lead to significant practical difficulties and delays in administering activated charcoal or whole bowel irrigation (Kulig et al 1985) and increase the risk of aspiration (Albertson et al 1989).
The utility of gastric lavage is controversial for standard preparations as it has a minimal effect on tablet absorption and there is no evidence for a clinical benefit compared with the use of activated charcoal without lavage in patients presenting later than one hour post ingestion (Merigan et al 1990, Kulig et al 1985). There are a number of potential reasons for gastric lavage having a more useful role in controlled release poisonings. Gastric lavage has been demonstrated to produce significant tablet return at 12 hours following salicylate poisoning (Ellenhorn & Barceloux 1987). The formation of a bezoar in the stomach may also increase late tablet return (however, the absence of any tablet return does not exclude the presence of a bezoar). It has been suggested that massage of the left upper quadrant of the abdomen while the knees are flexed may increase tablet return from lavage (Bartecchi 1977).
We would recommend, where there is a significant risk of serious toxicity (Table 2), gastric lavage be performed on admission with an orogastric tube large enough to remove whole tablets. Patients presenting with calcium antagonists or propranolol may require pretreatment with atropine to avoid vagal stimulation, which may precipitate complete heart block or asystole (Soni et al 1983).
The role of activated charcoal in the treatment of poisonings is well established and has been reviewed previously (Palatnick & Tenenbein 1992, Pond 1986). Its exact place in the treatment of poisonings of controlled release preparations is still being defined.
Its efficacy in reducing absorption of sustained release theophylline is critically dependent on the time of administration. Administration of activated charcoal at 1 and 6 hours and at 6 hours with sorbitol reduced the bioavailability of theophylline to 8.8%, 42.7% & 36.7% respectively (Minton et al 1990). The addition of sorbitol as an osmotic cathartic did not reduce absorption in that study but did have a modest benefit in a model of salicylate poisoning (Keller et al 1990).
Repeated doses of activated charcoal can increase the clearance of a number of drugs by either interrupting enterohepatic circulation or by direct “dialysis” from capillaries in the gastrointestinal mucosa (Berlinger et al 1983, Pond et al 1981, Levy 1982, Berg et al 1982). The technique may also have advantages for some drugs (e.g. salicylates) which exhibit pH dependent binding to charcoal resulting in desorption after single doses of charcoal (Filippone et al 1987). The addition of sorbitol to repeated doses of activated charcoal was associated with a significant reduction in AUC in a model of SR theophylline poisoning probably by enhancing clearance (Goldberg 1987). The addition of sorbitol was noted to hasten the appearance of charcoal in the stool although the mean time was 10.3 hours. The addition of sorbitol is associated with significant fluid shift into the intestinal lumen and patients often require intravenous fluids. Although the problem of prolonged dissolution of drug from controlled release preparations and bezoars can be addressed with repeated doses of activated charcoal, in one series 25% of patients had a continued rise in theophylline concentrations despite such treatment (Henderson et al 1992).
Activated charcoal should be administered at a dose of 25-50 g for children and 50-100 g for adults (Palatnick & Tenenbein 1992). This dose can be repeated every 4 hours or can be fractioned to give a quarter of the dose every hour, which is equally effective and may be better tolerated (Ilkhanipour et al 1992).
Whole bowel irrigation
Whole bowel irrigation with polyethylene glycol electrolyte lavage solution (PEG ELS) is the treatment of choice for those drugs (e.g. lithium, iron) that are not absorbed onto charcoal (Tenenbein et al 1987, Tenenbein 1987 & 1988). In one study using an ampicillin overdose model PEG ELS was marginally better than activated charcoal in decreasing absorption. It has also been effective in controlled release poisonings (Tenenbein 1989, Laggner et al 1984, Buckley et al 1993). As PEG ELS is iso-osmolar, faecuresis is not dependent on fluid shifts into the gut lumen. There is no clinically significant volume loss by the patient. Whole bowel lavage should be performed using PEG ELS at a dosage of 2 L/hour in an adult and 500 ml/hour in a small child (Tenenbein 1988, Tenenbein 1985). It should be continued until the rectal effluent is clear, the mean time for this to occur is four hours (Tenenbein 1987).
Enhanced drug elimination, by interruption of enterohepatic circulation or “dialysis”, has also been described with PEG ELS (Lenz et al 1983) and the combination of activated charcoal and PEG ELS for some drugs (Mayer 1992, Laine et al 1994).
Charcoal or whole bowel irrigation
PEG ELS is the treatment of choice for those drugs that do not bind to charcoal. The optimal regime for other controlled release poisoning is controversial with repeated doses of activated charcoal, whole bowel irrigation and the combination of the two all having advocates (Tenenbein 1988, Burkhart et al 1992). The evidence for the benefit of one regime over the others is based on volunteer and animal studies and the results are conflicting (Kirschenbaum et al 1989, Burkhart et al 1992, Rosenberg et al 1988). These studies have also been criticised as differing in many important respects (time from ingestion, number of tablets, gastrointestinal decontamination regime used, Esc) from genuine poisonings (Buckley & Dawson 1993, Tenenbein 1989).
PEG ELS and activated charcoal may be given together. However PEG ELS decreases binding of drugs to activated charcoal in vitro (Kirschenbaum et al 1990) and it has been suggested the dose of activated charcoal should be increased by 30%. In human studies PEG ELS is better tolerated than activated charcoal and has a shorter gastrointestinal transit time.
Our practice is to use whole bowel irrigation alone for those medications not adsorbed on to charcoal. Other poisonings receive a single dose of activated charcoal (to adsorb drug that is not held within the vehicles) followed by whole bowel irrigation. Repeated doses of activated charcoal are only added for those drugs where it has clearly been shown to enhance clearance (Table 2). This decision tree is shown in the Figure.
It should be assumed that bezoars may occur with any slow release formulation although not reported with osmotic pump delivery systems. The presence of a pharmacobezoar necessitates prolonged gastrointestinal decontamination to deal with dissociating toxins or definitive treatment.
Gastric bezoars have been successfully fragmented or removed via gastro-endoscopy (Coupe et al 1986, Cereda et al 1986). This should be followed by continued gastrointestinal decontamination with activated charcoal and whole bowel lavage. Whole bowel lavage alone may cause the passage of a pharmacobezoar which presumably have either been small enough to pass through the pylorus or have formed in the small intestine (Buckley et al 1993).
If gastric bezoars that cannot be fragmented there are two options. Decontamination can be continued indefinitely to deal with dissociating toxins or bezoars can be surgically removed (Landsman et al 1987, Swartz H 1976). Surgical removal has most commonly been reported for iron bezoars. It should be considered in patients with bowel obstruction and in those with other contraindications for gastrointestinal decontamination.
Flow chart: Decontamination algorithm for poisonings
The main characteristic of poisoning with controlled release formulations is the delay in presentation, the onset of clinical effects and in the time to reach maximum plasma concentrations. The delivery vehicles themselves are inherently safe and there is little toxicity at therapeutic doses.
Bezoar formation is a unique feature of poisoning with controlled release preparations and further prolongs the time course. Further study is required to define the optimal method of gastrointestinal decontamination and non-invasive means of preventing, diagnosing and managing bezoar formation.
Table 1. Classification of controlled release drugs based on the differences in formulation & design
|Dissolution||Encapsulated||The drug is held in multiple beads, multiple layers, or a gel, which gradually dissolves.|
|Matrix||Diffusion Membrane||The drug diffuses slowly across an intact non-soluble membrane or out of an insoluble matrix.|
|Matrix||Osmotic Mini-osmotic||A membrane separates a drug from an osmotic agent. As the compartment containing the agent swells, the drug is pushed through a pore in the dispensing system|
|Elementary||An osmotic core containing the drug swells and is forced through a pore|
|Ion Exchange||The drug is weakly bound within the preparation and is exchanged for elemental ions in the gut|
|Resins||Transdermal||The drug diffuses across a semipermeable membrane into the stratum corneum creating a second reservoir|
Table 2. Controlled release drugs associated with fatality or significant morbidity
|Drug||Toxic Dose & Comment|
|Amphetamine & Dexamphetamine||>1 mg/kg|
|Carbamazepine||>20 mg/kg Bezoars (Coutselinis 1980) Repeat Dose Charcoal Enhances Clearance Charcoal Haemoperfusion Enhances Clearance|
|Chlorpromazine||* 10 mg/kg Radio-opaque¹|
|Iron||>20 mg/kg Radio-opaque¹ Bezoars (Landsman et al 1987) Does not bind to charcoal4|
|Lithium||>10 mg/kg Radio-opaque¹ Does not bind to charcoal4 Haemodialysis Enhances Clearance|
|Meprobamate||Bezoars (Swartz H 1976)|
|Paracetamol||>150 mg/kg Nomogram not useful treat ingestions >150 mg/kg|
|Potassium Chloride||* 2 mEq/kg if renal function is normal Radio-opaque¹ Does not bind to charcoal4|
|Procainamide||>100 mg/kg Haemoperfusion Enhances Clearance|
|Quinidine||* 50 mg/kg Bezoars (McGuigan 1986) Radio-opaque¹ Haemodialysis Enhances Clearance|
|Theophylline||>20 mg/kg Bezoars (Cerada et al 1986, Whyte et al 1983, Bernstein et al 1992) Repeat Dose Charcoal Enhances Clearance³ Charcoal Haemoperfusion Enhances Clearance|
|Verapamil||Extreme Pharmacodynamic Variation² Bezoars (Buckley et al 1993, Sporer 1993) Repeat Dose Charcoal Enhances Clearance³|
¹ Radio-opaque medications may lose this property with tablet dissolution. The absence of tablets on plain X-ray does not exclude ingestion.
² Pharmacodynamic Variation: Toxicity from these drugs is significantly determined by patient factors such as pre-existing conditions and other medications. It is therefore difficult to define a safe lower limit for ingestion. All such patients should receive gastrointestinal decontamination.
³ Repeat Dose Charcoal Enhances Clearance: The increased clearance by using repeat dose charcoal is clinically significant. The charcoal dose may need to be increased in the presence of PEG-ELS
4 Drugs which do not bind to charcoal should be treated with whole bowel lavage with PEG-ELS
Nicholas A Buckley B Med FRACP Lecturer
Andrew H Dawson MB BS Dip Epid MRCP FRACP Director *
David A Reith MB BS Registrar *
DISCIPLINE OF CLINICAL PHARMACOLOGY, UNIVERSITY OF NEWCASTLE
* DEPARTMENT OF CLINICAL TOXICOLOGY & PHARMACOLOGY,
MATER MISERICORDIAE HOSPITAL, NEWCASTLE NSW 2298.
Acknowledgment: Dr Ian Whyte for helpful comments
Copyright Drug Safety, reproduced and adapted with permission