Basic Analytical Toxicology
Qualitative tests for poisons
Many difficulties may be encountered when performing qualitative tests for poisons, especially if laboratory facilities are limited. The poisons may include gases, such as carbon monoxide, drugs, solvents, pesticides, metal salts, corrosive liquids (acids, alkalis) and natural toxins. Some poisons may be pure chemicals and others complex natural products. Not surprisingly, there is no comprehensive range of tests for all poisons in all samples.
When certain compounds are suggested by the history or clinical findings, simple tests may be performed using the procedures given in the monographs (section 6). However, in the absence of clinical or other evidence to indicate the poison(s) involved, a defined series of tests (a screen) is needed. It is usually advisable to perform this series of tests routinely, since circumstantial evidence of poisoning is often misleading. Similarly, the analysis should not end after the first positive result, since additional unsuspected compounds may be present.
The sequence of analyses outlined in section 5.2 will detect and identify a number of poisons in commonly available specimens (urine, stomach contents, and scene residues, i.e., material such as tablets or suspect solutions found with or near to the patient) using a minimum of apparatus and reagents. The compounds detected include many that give rise to nonspecific features, such as drowsiness, coma or convulsions, and which will not be indicated by clinical examination alone. Poisons for which specific therapy is available, such as acetylsalicylic acid and paracetamol, are also included. The analysis takes about 2 hours and may be modified to incorporate common local poisons if appropriate tests are available.
5.1 Collection, storage and use of specimens
5.1.1 Clinical liaison
Good liaison between the clinician and the analyst is of vital importance if the results of a toxicological analysis are to be useful (see section 2). Ideally, this liaison should commence before the specimens are collected, and any special sample requirements for particular analytes noted. At the very least, a request form should be completed to accompany the specimens to the laboratory. An example of such a form is given in Fig. 2.
Before starting an analysis it is important to obtain as much information about the patient as possible (medical, social and occupational history, treatment given, and the results of laboratory or other investigations), as discussed in sections 2 and 3. It is also important to be aware of the time that elapsed between ingestion or exposure and the collection of samples, since this may influence the interpretation of results. All relevant information about a patient gathered in conversation with the clinician, nurse, or poisons information service should be recorded in the laboratory using the external request form (Fig. 2) or a suitably modified version of this form.
5.1.2 Specimen transport and storage
Specimens sent for analysis must be clearly labelled with the patient's full name, the date and time of collection, and the nature of the specimen if this is not self-evident. This is especially important if large numbers of patients have been involved in a particular incident, or a number of specimens have been obtained from one patient. Confusion frequently arises when one or more blood samples are separated in a local laboratory and the original containers are discarded. When the plasma/serum samples are forwarded subsequently to the toxicology laboratory for analysis, it can be difficult, if not impossible, to ascertain which is which.
The date and time of receipt of all specimens by the laboratory should be recorded and a unique identifying number assigned to each specimen (see section 4.1.6). Containers of volatile materials, such as organic solvents, should be packaged separately from biological specimens to avoid the possibility of cross-contamination. All biological specimens should be stored at 4°C prior to analysis, if possible, and ideally any specimen remaining after the analysis should be kept at 4°C for 3-4 weeks in case further analyses are required. In view of the medicolegal implications of some poison cases (for example, if it is not clear how the poison was administered or if the patient dies) then any specimen remaining should be kept (preferably at -20°C) until investigation of the incident has been concluded.
Urine is useful for screening tests as it is often available in large volumes and usually contains higher concentrations of drugs or other poisons than blood. The presence of metabolites may sometimes assist identification if chromatographic techniques are used. A 50-ml specimen from an adult, collected in a sealed, sterile container, is sufficient for most purposes; no preservative should be added. The sample should be obtained as soon as possible, ideally before any drug therapy is initiated. However, drugs such as tricyclic antidepressants (amitriptyline, imipramine) cause urinary retention, and thus a very early specimen may contain insignificant amounts of poison.
Conversely, little poison may remain in specimens taken many hours or days later, even though the patient may be very ill, as in acute paracetamol poisoning. If the specimen is obtained by catheterization there is a possibility of contamination with lidocaine. If syrup of ipecacuanha has been given in an unsuccessful attempt to induce emesis there is a possibility of emetine being present in the urine.
5.1.4 Stomach contents
Stomach contents may include vomit, gastric aspirate and stomach washings - it is important to obtain the first sample of washings, since later samples may be very dilute. A volume of at least 20 ml is required to carry out a wide range of tests; no preservative should be added. This can be a very variable sample and additional procedures such as homogenization followed by filtration and/or centrifugation may be required to produce a fluid amenable to analysis. However, it is the best sample on which to perform certain tests. If obtained soon after ingestion, large amounts of poison may be present while metabolites, which may complicate some tests, are usually absent. An immediate clue to certain compounds may be given by the smell; it may be possible to identify tablets or capsules simply by inspection. Note that emetine from syrup of ipecacuanha may be present, especially in children (section 2.2.1).
5.1.5 Scene residues
It is important that all bottles or other containers and other suspect materials found with or near the patient (scene residues) are retained for analysis if necessary since they may be related to the poisoning episode. There is always the possibility that the original contents of containers have been discarded and replaced either with innocuous material or with more noxious ingredients such as acid, bleach or pesticides. Note that it is always best to analyse biological specimens in the first instance if possible.
A few milligrams of scene residues are usually sufficient for the tests described here. Dissolve solid material in a few millilitres of water or other appropriate solvent. Use as small an amount as possible in each test, in order to conserve sufficient for possible further tests.
Blood (plasma or serum) is normally reserved for quantitative assays but for some poisons, such as carbon monoxide and cyanide, whole blood has to be used for qualitative tests. For adults, a 10-ml sample should be collected in a heparinized tube on admission. In addition, a 2-ml sample should be collected in a fluoride/oxalate tube, if ethanol poisoning is suspected. Note that tubes of this type available commercially contain the equivalent of about 1 g/l fluoride, whereas about 10 g/l fluoride (40 mg sodium fluoride per 2 ml of blood) is needed to inhibit fully microbial action in such specimens. The use of disinfectant swabs containing alcohols (ethanol, propan-2-ol) should be avoided. The sample should be dispensed with care: the vigorous discharge of blood though a syringe needle can cause sufficient haemolysis to invalidate a serum iron or potassium assay.
In general, there are no significant differences in the concentrations of poisons between plasma and serum. However, if a compound is not present to any extent within erythrocytes, the use of lysed whole blood will result in considerable dilution of the specimen. On the other hand, some poisons, such as carbon monoxide, cyanide and lead, are found primarily in erythrocytes and thus whole blood is needed for such measurements. A heparinized whole blood sample will give either whole blood or plasma as appropriate. The space above the blood in the tube (headspace) should be minimized if carbon monoxide poisoning is suspected.
5.2 Analysis of urine, stomach contents and scene residues
If any tests are to influence immediate clinical management, the results must be available within 2-3 hours of receipt of the specimen. Of course, a positive result does not in itself confirm poisoning, since such a result may arise from incidental or occupational exposure to the poison in question or the use of drugs in treatment. In some cases, the presence of more than one poison may complicate the analysis, and examination of further specimens from the patient may be required. A quantitative analysis carried out on whole blood or plasma may be needed to confirm poisoning, but this may not be possible if laboratory facilities are limited. It is important to discuss the scope and limitations of the tests performed with the clinician concerned, and to maintain high standards of laboratory practice (see section 4.1), especially when performing tests on an emergency basis. It may be better to offer no result rather than misleading data based on an unreliable test. In any event, it is valuable to have a worksheet to record the analytical results. An example of such a sheet is given in Fig. 3.
The qualitative scheme given below, possibly modified to suit local needs, should be followed in every case unless there are good reasons (such as insufficient sample) for omitting part of the screen, since this will provide a good chance of detecting any poisons present. The scheme has three parts: physical examination, colour tests and thin-layer chromatography, and is designed primarily for the analysis of urine samples. However, most of the tests and some additional ones are also applicable, with due precautions, to stomach contents and scene residues. Some compounds and groups of compounds not normally detected using this procedure are listed in Table 11. Simple tests for many of these compounds are given in the monographs (section 6).
5.2.1 Physical examination of the specimen
High concentrations of some drugs or metabolites can impart characteristic colours to urine (Table 12). Deferoxamine or methylene blue given in treatment may colour urine red or blue, respectively. Strong-smelling poisons such as camphor, ethchlorvynol and methyl salicylate can sometimes be recognized in urine since they are excreted in part unchanged. Acetone may arise from metabolism of propan-2-ol. Turbidity may be due to underlying pathology (blood, microorganisms, casts, epithelial cells), or to carbonates, phosphates or urates in amorphous or microcrystalline forms. Such findings should not be ignored, even though they may not be related to the poisoning. Chronic therapy with sulfonamides may give rise to yellow or greenish brown crystals in neutral or alkaline urine. Phenytoin, primidone, and sultiame form crystals in urine following overdosage, while characteristic colourless crystals of calcium oxalate form at neutral pH after ingestion of ethylene glycol (Fig. 4).
Stomach contents and scene residues
Some characteristic smells associated with particular substances are listed in Table 13. Many other compounds (for example, ethchlorvynol, methyl salicylate, paraldehyde, phenelzine) also have distinctive smells. Very low or very high pH may indicate ingestion of acid or alkali, while a green/blue colour suggests the presence of iron or copper salts. Microscopic examination using a polarizing microscope may reveal the presence of tablet or capsule debris. Starch granules used as a filler in Some tablets and capsules are best identified using crossed polarizing filters, when they appear as bright grains marked with a dark Maltese cross.
aTake care: specimens containing cyanide may give off hydrogen cyanide, especially if acidified - not everyone can detect hydrogen cyanide by smell. Similarly sulfides evolve hydrogen sulfide - the ability to detect hydrogen sulfide (rotten egg smell) is lost at higher concentrations.
Undegraded tablets or capsules and any plant remains or specimens of plants thought to have been ingested should be examined separately. The local poisons information service will normally have access to publications or other aids to the identification of tablets or capsules by weight, markings, colour, shape and possibly other physical features.
5.2.2 Colour tests
The nine qualitative tests described here are based on simple colour reactions and cover a number of important drugs and other poisons. Full descriptions of these tests are given in the respective monographs (section 6), together with details of common sources of interference and detection limits. Other tests, such as the Reinsch test for antimony, arsenic, bismuth and mercury, are not discussed further here, but full details are given in the respective monographs.
Of the tests outlined (Table 14), that for salicylates such as acetylsalicylic acid (aspirin) (Trinder's test, also known as the ferric chloride test) is best performed on urine rather than stomach contents or scene residues, since acetylsalicylic acid itself does not react unless hydrolysed. Most of the others may be performed using either specimen, but the tests for chlorates and other oxidizing agents and for ferrous or ferric iron can only be carried out on stomach contents or scene residues. Examples of the colours obtained in these tests are given in Plates 1-8.
5.2.3 Thin-layer chromatography
The aim of the scheme outlined below is to obtain as much information as possible in a short time and with a minimum of sample. Drugs and, in some cases, metabolites present are extracted from the sample into an organic solvent under acidic and alkaline conditions. The extracts are analysed by thin-layer chromatography (TLC) on one plate using a single solvent system. The basic extract is acidified during the evaporation stage to minimize loss of volatile bases such as amfetamines. If the sample volume is limited, the acidic and basic extractions can be performed sequentially on the same portion of the specimen. However, it is important that the pH change between extractions is accomplished satisfactorily. Extracts of stomach contents may contain fatty material, which makes chromatographic analysis difficult, and purification by re-extraction into aqueous acid or alkali may be required.
Although simple examination of the developed chromatogram under ultraviolet light (254 nm and 366 nm) may reveal the presence of fluorescent compounds such as quinine, the use of a number of spray (visualization) reagents widens the scope of the analysis and increases the confidence of any identifications.
aTests for use on stomach contents or scene residues only.
The recommended TLC visualization reagents are as follows:
1. Mercurous nitrate reagent (acidic extract), which gives white spots with a grey centre on a darker background with barbiturates and related compounds such as glutethimide.
2. Acidified iodoplatinate reagent (basic extract), which gives mainly purple, blue or brown spots with a range of basic and neutral drugs and metabolites. Note that some authors recommend neutral iodoplatinate, which is more stable and which gives similar reactions with many basic drugs; this is oversprayed with sulfuric acid (500 ml/l) which facilitates reactions with neutral compounds such as caffeine and phenazone (from dichloralphenazone).
3. Mandelin's reagent (basic extract) gives colours ranging from blue and green to orange and red with a variety of basic compounds. Some, especially tricyclic antidepressants such as amitriptyline and nortriptyline, give fluorescent spots if viewed under ultraviolet light (366 nm) after spraying with this reagent.
4. Sulfuric acid (500 ml/l) (basic extract) alone gives red, purple or blue spots with many phenothiazines and their metabolites. This is especially valuable since some phenothiazines (chlorpromazine, for example) are given therapeutically in relatively high doses and have many metabolites, which can give a very confusing picture if unrecognized.
Of course, many additional mobile phase and spray reagent combinations could be used as well as, or in place of, those suggested here, and details of some are given in the references listed in the Bibliography and in the monographs (section 6). For example, methanol:concentrated ammonium hydroxide (99:1.5) (methanol:ammonia, MA) is widely used in the analysis of basic drugs, and is especially useful in the detection of morphine and related opioids (see section 6.73). Of the spray reagents, Marquis' reagent gives a variety of colours with different basic drugs, and is again especially valuable for the detection of morphine and other opioids which give blue/violet colours.
In all cases, the colour obtained from a particular compound may vary depending on concentration, the presence of co-eluting compounds, the duration and intensity of spraying, and the type of silica used in the manufacture of the plate, among other factors. Some compounds may show a gradation or even a change in colour from the edge of the spot towards the centre (usually a concentration effect), while the intensity or even the nature of the colour obtained may vary with time. A further problem is that the interpretation and recording of colour reactions are very subjective. It is thus important to analyse authentic compounds, ideally on the same plate as the sample extracts. Even so, compounds present in sample extracts sometimes show slightly different chromatograms from the pure compounds owing to the presence of co-extracted material. Interfering neutral compounds, especially fatty acids from stomach contents, can be removed by back-extraction of acidic or basic compounds into dilute base or acid, respectively (the neutral compounds stay in the organic extract), followed by re-extraction into organic solvent.
Detailed records of each analysis should be kept, not only for medicolegal purposes, but also to establish a reference data bank to aid in the interpretation of results. This can be used to supplement the data given in Table 15 and elsewhere, and has the advantage of being generated in the laboratory actually involved in analysing the specimens.
The recommended TLC screening system is as follows.
Applicable to urine, stomach contents or scene residues.
a With certain compounds, there may be a colour difference between the centre and the outer part of the spot; this is indicated by the mention of a second colour in parentheses,
b Diamorphine itself is not found in urine, but is detected as monoacetyl morphine and morphine conjugates
Reagents and equipment
1. Aqueous hydrochloric acid (1 mol/l).
2. Aqueous sodium hydroxide (0.5 mol/l).
3. Ammonium chloride buffer. Saturated aqueous ammonium chloride adjusted to pH 9 with concentrated ammonium hydroxide (relative density 0.88).
4. Hydrochloric acid (2 ml/l in methanol).
5. Ethyl acetate:methanol:concentrated ammonium hydroxide (EMA) (85:10:5) (relative density 0.88).
6. Mercurous nitrate reagent. Place 1 g of mercurous nitrate in 100 ml of purified water and add concentrated nitric acid (relative density 1.42) until the solution is clear.
7. Acidified iodoplatinate reagent. Mix 0.25 g of platinic chloride, 5 g of potassium iodide and 5 ml of concentrated hydrochloric acid (relative density 1.18) in 100 ml of purified water.
8. Mandelin's reagent. Suspend 1 g of finely powdered ammonium vanadate in 100 ml of concentrated sulfuric acid (relative density 1.86). Shake well before use.
9. Aqueous sulfuric acid (500 ml/l).
10. Silica gel thin-layer chromatography plate (20 × 20 cm, 20 µm average particle size; see section 4.4.1).
All 1 g/l in chloroform:
1. Acidic drugs mixture (amobarbital, mefenamic acid, phenobarbital, theophylline).
2. Basic drugs mixture (amitriptyline, codeine, nicotine, nortriptyline).
3. Phenothiazine mixture (perphenazine, trifluoperazine, thioridazine).
1. Acidic extract (extract A)
(a) To 10 ml of urine in a 30-ml glass centrifuge tube add 1 ml of dilute hydrochloric acid and 10 ml of chloroform.
(b) Shake on a mechanical shaker for 5 minutes, centrifuge in a bench centrifuge for 10 minutes and transfer the lower, organic layer to a 15-ml tapered glass tube.
(c) Evaporate the extract to dryness on a water-bath at 60°C under a stream of compressed air.
2. Basic extract (extract B)
(a) To a further 10 ml of urine in a 30-ml glass tube, add 2 ml of ammonium chloride buffer and 10 ml of chloroform:propan- 2-ol (9:1).
(b) Shake on a mechanical shaker for 5 minutes, centrifuge in a bench centrifuge for 10 minutes and transfer the lower, organic layer to a 15-ml tapered glass tube.
(c) Add 0.5 ml of methanolic hydrochloric acid (to minimize losses of volatile bases; see section 6.1) and evaporate the extract to dryness in a water-bath at 60°C under a stream of compressed air.
3. Purification of extracts of stomach contents
(a) Prior to the solvent evaporation stage, add 5 ml of aqueous sodium hydroxide solution to extract A, and 5 ml of aqueous hydrochloric acid to extract B.
(b) Shake on a mechanical shaker for 5 minutes, centrifuge in a bench centrifuge for 10 minutes and discard both organic layers.
(c) Add 5 ml of aqueous hydrochloric acid to the aqueous residue from extract A, and 5 ml of ammonium chloride buffer to the aqueous residue from extract B, and re-extract into chloroform or chloroform:propan-2-ol as in methods 1 and 2 above.
1. Divide the plate into eight columns by scoring with a pencil (see section 4.4.2). Lightly draw a pencil line about 1 cm from the bottom of the plate to indicate the origin, and score a horizontal line 10 cm from the origin to indicate the limit of development, as shown in Fig. 5.
2. Reconstitute each extract in 100 µl of chloroform:propan-2-ol and apply 25 of extract A and 3 portions of 25 µl of extract B at the origin of the plate, as shown in Fig. 5.
3. Apply 10 µl of the respective standard mixtures to the plate, as shown in Fig. 5.
4. Ensure that the plate is dry, and then develop the chromatogram using ethyl acetate:methanol:concentrated ammonium hydroxide (EMA) (saturated tank, see section 4.4.3).
5. Remove the plate and dry under a stream of air in a fume cupboard or under a fume hood.
6. View the plate under ultraviolet light (254 nm and 366 nm) and note any fluorescent spots.
7. Invert the plate and spray each portion with the visualization reagents as shown in Fig. 6. Take care to mask with a clean glass plate those portions of the plate not being sprayed.
Take care - all the spray reagents used are very toxic. Spraying must be performed in a fume cupboard or under an efficient fume hood.
8. View the plate again under ultraviolet light (254 nm and 366 nm) and note any fluorescent spots, especially in the portion sprayed with Mandelin's reagent.
9. If necessary, the colours obtained with Mandelin's reagent can be enhanced by heating the plate in an oven at 100°C for 10 minutes.
hRf values and reactions with the spray reagents of some of the compounds of interest are given in Table 15 and illustrated in Plates 9-12.
Plate 9. An example of a chromatogram obtained from the analysis of an acidic and a basic urine extract from a patient following the ingestion of codeine and methadone. The plate is divided into four sections for the application of separate visualization reagents: A - mercurous nitrate; B - acidified iodoplatinate; C - Mandelin's reagent; D - sulfuric acid. The acidic sample extract (TA) has been applied at the origin in section A and run with test mixture S1 containing: amobarbital (1), phenobarbital (2), and theophylline (3). The "basic" extract (TB) has been applied at the origin in sections B, C and D and run with test mixtures: S2 containing amitriptyline (4), nicotine (5), nortriptyline (6), codeine (7), and mefenamic acid (8); and S3 containing thioridazine (9), trifluoperazine (10), and perphenazine (11).
Note that a complex pattern of drug and metabolites is obtained for the basic sample extract (TB) visualized with iodoplatinate (B) and Mandelin's reagent (C). No response is seen using sulfuric acid (D). No response is seen for the acidic sample extract (TA) visualized using mercurous nitrate in section A.
Plate 10. An example of a chromatogram obtained from the analysis of an acidic and a basic urine extract from a patient following the ingestion of phenobarbital and methadone. The plate is divided into four sections for the application of separate visualization reagents: A - mercurous nitrate; B acidified iodoplatinate; C - Mandelin's reagent; D - sulfuric acid. The acidic sample extract (TA) has been applied at the origin in section A and run with test mixture S1 containing: amobarbital (1), phenobarbital (2), and theophylline (3). The basic extract (TB) has been applied at the origin in sections B, C and D and run with test mixtures: S2 containing amitriptyline (4), nicotine (5), nortriptyline (6), codeine (7), and mefenamic acid (8); and S3 containing thioridazine (9), trifluoperazine (10), and perphenazine (11).
Note that the phenobarbital is clearly visualized in the acidic urine extract (TA) using the mercurous nitrate spray (A), whereas methadone and its metabolites are best visualized in the basic urine extract (TB) with iodoplatinate (B). No distinct response is seen with the remaining two sprays, Mandelin's reagent (C) and sulfuric acid (D).
Plate 11. An example of a chromatogram obtained from the analysis of an acidic and a basic urine extract from a patient following the ingestion of a phenothiazine (thioridazine) overdose. The plate is divided into four sections for the application of separate visualization reagents: A -mercurous nitrate; B - acidified iodoplatinate; C - Mandelin's reagent; D - sulfuric acid. The acidic sample extract (TA) has been applied at the origin in section A and run with test mixture S1 containing: amobarbital (1), phenobarbital (2), and theophylline (3). The basic extract (TB) has been applied at the origin in sections B, C and D and run with test mixtures: S2 containing amitriptyline (4), nicotine (5), nortriptyline (6), codeine (7), and mefenamic acid (8); and S3 containing thioridazine (9), trifluoperazine (10), and perphenazine (11).
Note that a complex pattern of drug and metabolites is obtained for the basic sample extract (TB) visualized with iodoplatinate (B), Mandelin's reagent (C) and sulfuric acid (D). No response is seen with mercurous nitrate (A) for the acidic sample extract (TA). The pattern of many spots and distinct colours with sulfuric acid is typical of phenothiazine-type compounds, but the pattern is quite dissimilar to that seen for the pure compound (thioridazine).
Plate 12. An example of a chromatogram obtained from the analysis of an acidic and a basic urine extract from a patient following the ingestion of the tricyclic antidepressant, dosulepin. The plate is divided into four sections for the application of separate visualization reagents: A -mercurous nitrate; B - acidified iodoplatinate; C - Mandelin's reagent; D - sulfuric acid. The acidic sample extract (TA) has been applied at the origin in section A and run with test mixture S1 containing: amobarbital (1), phenobarbital (2), and theophylline (3). The basic extract (TB) has been applied at the origin in sections B, C and D and run with test mixtures: S2 containing amitriptyline (4), nicotine (5), nortriptyline (6), codeine (7), and mefenamic acid (8); and S3 containing thioridazine (9), trifluoperazine (10), and perphenazine (11).
Note that dosulepin and its metabolites are visualized in the basic urine extract (TB) most clearly using the iodoplatinate spray (B). No response is seen with the other sprays (C and D), or with mercurous nitrate spray (A).
Benzodiazepines and their metabolites may appear as light green or yellow spots under ultraviolet light (366 nm) before the mercurous nitrate column (acidic extract) is sprayed, but these compounds are not considered further here (see section 6.11). Quinine (often from bitter drinks), carbamazepine and their metabolites give fluorescent spots (254 nm and 366 nm) before the columns containing the basic extracts are sprayed, and also undergo characteristic reactions with some of the spray reagents. These compounds are relatively easy to identify, as are tricyclic antidepressants (amitriptyline, and imipramine). Others such as nicotine and its metabolites (normally from tobacco), caffeine (from caffeinated beverages) and lidocaine (from catheter lubricant) occur frequently and should be recognizable with practice.
In difficult cases it may be useful to calculate the hRf value for unknown compounds (section 4.4.5) and to compare the findings with reference values (see Bibliography).
Plates 9-12 give some examples of the spot shapes and colours that should be expected for the standards and for some other commonly occurring compounds. Even if the interpretation of the chromatography plate is relatively straightforward, it is important to record systematically the data generated. This can be done by photographing or photocopying the plate (taking great care to clean the photocopier and any other surfaces very carefully afterwards), but it is as easy to record spot positions and shapes on a standard form such as that illustrated in Fig. 7.
Note the position of the yellow-brown streak near the top of the plate observed with Mandelin's reagent (visible light) on analysis of blank urine. Colours, including those observed under ultraviolet light, and any temporal changes, can be noted either in writing or by using coloured pencils.
To ensure reproducible chromatography, attention should be given to the factors discussed in section 4.4.3, especially the use of saturated tanks and the need to ensure that the concentrated ammonium hydroxide is of adequate strength (relative density 0.88, 330 g/l), both in the tank and in the reagent bottle. It is good practice to buy either small (500-ml) bottles or to transfer the contents of large (2.5-litre) bottles of ammonium hydroxide to 500-ml bottles ( with care), which can be kept tightly stoppered until needed. Never use a batch of ethyl acetate:methanol:concentrated ammonium hydroxide mobile phase more than five times at room temperatures of 20-25°C (fewer times at higher ambient temperatures).
Even when all due precautions are taken, it is invariably found that the chromatographic characteristics of sample extracts are different from those of pure compounds. In extreme cases, broad streaks may be obtained rather than discrete spots. This is often attributable to the presence of polyethylene glycol used as a vehicle in, for example, temazepam capsules and can be minimized by back- extraction of acidic or basic compounds into dilute base or acid, respectively, as described above.
It is not possible to give detection limits for all the compounds under study. The experience of the analyst, the extraction efficiency, the spot density, the intensity of the chromogenic reaction with the spray reagents and even the type of silica gel used can all affect sensitivity. Nevertheless, a general limit of sensitivity of 1 mg/l is reasonable.
5.2.4 Reporting the results
The results of emergency analyses must be communicated direct to the clinician without delay, and should be followed by a written report as soon as possible. An example of an analytical toxicology report form is given in Fig. 8. Ideally, confirmation using a second, independent method, or failing this an independent duplicate, should be obtained before positive findings are reported. However, this may not always be practicable, especially if only simple methods are available. In such cases it is vital that the appropriate positive and negative controls have been analysed together with the specimen (see section 4.1.5).
When reporting quantitative results it is important to state clearly the units of measurement used (SI mass units are preferable; see section 4.1.6). In addition, any information necessary to ensure that the clinical implications of the result are fully understood should be noted on the written report. The clinical features associated with poisoning by a number of compounds are given in the appropriate monographs (section 6). Information on additional compounds will usually be found in one of the clinical toxicology textbooks listed in the Bibliography.
Although it is often easy for the analyst to interpret the results of analyses in which no compounds are detected, such results are sometimes difficult to convey to clinicians, especially in writing. It is important to give information as to the poisons excluded by the tests performed with all the attendant complications of the scope, sensitivity and selectivity of the analyses, and other factors such as sampling variations. Because of the potential medicolegal and other implications of any toxicological analysis, it is important not to use laboratory jargon such as "negative" or sweeping statements such as "absent" or "not present".
The phrase "not detected" should convey precisely the laboratory result, especially when accompanied by a description of the specimen analysed and the limit of sensitivity of the test (detection limit). However, it can still be difficult to convey the scope of analyses, such as the thin-layer chromatography screen for acidic and basic drugs discussed above. Even with Trinder's test for example, - a relatively simple test normally used to detect acetylsalicylic acid ingestion (Table 14) - a number of other salicylates including, of course, salicylic acid itself also react. One way of giving at least some of this information in a written report is to list the compounds or groups of compounds normally detected by commonly used procedures. If these groups are listed on the back of the report then it is relatively simple to refer to the qualitative tests performed by number and thus to convey at least some of the information required. An example of a grouping system based on that used in the analytical toxicology laboratory worksheet (see Fig. 3) is given in Table 16.
It should be clear that performing a qualitative poisons screen involves more than simply analysing specimens as they arrive in the laboratory and reporting the bare facts of the analysis. The suggested scheme of operation is summarized in Table 17.