Comparison of Chemical Tests with the Pupillary Method for the Diagnosis of Narcotic Use


Experimental procedure
TABLE 1 Summary of studies on prison parolees comparing the pupillary test with chemical tests for detection of morphine or heroin usage


Author: E. Leong WAY, Henry W. ELLIOTT, Norman NOMOF, Thomas FIELDS
Pages: 29 to 33
Creation Date: 1963/01/01

Comparison of Chemical Tests with the Pupillary Method for the Diagnosis of Narcotic Use

E. Leong WAY
Norman NOMOF
Technical Assistance Thomas FIELDS
from the Department of Pharmacology, University of California Medical Center, San Francisco, California, and the Office of Norman NOMOF, M.D., and Joseph FISCHER, M.D.,48 East Santa Clara Street, San Jose, California.


In 1953 Wikler, Isbell & Fraser reported that in patients tolerant to heroin, morphine or methadone, a small single dose of nalorphine precipitated an abstinencelike syndrome including mydriasis. In 1956, Fraser et al. showed that nalorphine given alone decreased pupil diameter, but that when administered after single doses of morphine it partially antagonized the miosis due to morphine. These observations led to the use of the narcotic antagonists nalorphine and levallorphan in the detection of narcotic use and addiction.

Application particularly in the detection of narcotic use has become quite extensive since the report of Terry & Braumoeller in 1956. By May 1960, according to a bulletin by State of California Departments of Public Health and Justice in parole and probation work nalorphine had been used in 16,987 tests in about 2,695 subjects and levallorphan was utilized in 759 tests in 120 subjects by September 1960. The narcotic use test depends upon the antagonism between nalorphine or levallorphan and the morphine-like analgesics. Subjects under the influence of narcotics will show a pupillary dilation after the injection of an antagonist and others will show a constriction. An experimental study (Elliott & Way, 1961) indicated that 3 mg of nalorphine or 1-2 mg of levallorphan subcutaneously will usually produce a decrease in pupil diameter of 0.25 mm or more in normal subjects and field work has given the impression that subjects who have recently received an injection of a narcotic (usually heroin) or are addicted to a narcotic will show an increase of pupil diameter of 0.25 mm or more.

Since a positive test may be used as evidence of narcotic use or addiction, it is desirable to determine its validity and limitations under actual field conditions. It was felt that detection of narcotics in urine might serve this purpose so urine specimens were collected from subjects undergoing the pupillary test and sent to our laboratory for narcotic analysis. Later, the findings were correlated with the results of the pupil tests. This report is concerned with the methods used for detecting narcotics in urine as well as with the correlation of the findings with the pupillary tests.

Experimental procedure

Chemical tests for morphine and certain drugs with morphine-like effects were performed on the urine of subjects on the State of California Anti-Narcotic Testing Programme. The subjects were parolees from the state prison or probationers with records of previous narcotic usage. The urine was collected just before or at the time the subjects were undergoing the pupil test with nalorphine. The personnel involved in performing the pupillary and the chemical tests had no a priori knowledge of each other's results.

TABLE 1 Summary of studies on prison parolees comparing the pupillary test with chemical tests for detection of morphine or heroin usage

  Pupillary test

Chemical tests



339 81.0
13 3.0
18 4.3
10 2.4
16 3.8
12 2.9
11 2.6
419 100.0

Several chemical tests were used to establish the presence of morphine or drugs with morphine-like effects in the urine. Inasmuch as the vast majority of narcotic users are generally habituated to heroin or morphine, the methods used were directed mainly towards detecting and quantifying morphine in the urine. Both heroin (Oberst, 1943) and morphine (Oberst, 1941, 1942; Fujimoto, Way & Hine, 1954) are excreted largely in the urine of addicts as "bound" morphine, the latter being a conjugate of glucuronic acid (Fujimoto and Way, 1957). To increase the sensitivity of the test, therefore, "total" morphine concentration was determined; total morphine represents free unchanged morphine plus any morphine liberated from morphine glucuronide by acid hydrolysis of urine.

In preliminary studies, when urine specimens were collected before nalorphine administration, a screening test for alkaloids and phenols was performed by spotting solvent extracts of acid hydrolysates of urine on filter paper and spraying respectively with iodoplatinate and the Folin-Ciocalteu reagent. However, in the present series the quantitative test for total morphine of Fujimoto, Way & Hine (1954) was most frequently used. A concentration of morphine of 2 micrograms/ml or greater was considered to be a positive test. In such instances confirmatory tests for morphine were performed by obtaining a differential ultraviolet absorption curve as described by Goldbaum & Kazyak (1952) and by the paper chromatographic method of Mannering et al. (1954). However, all these chemical tests may become invalidated if the urine is collected shortly after injection of nalorphine. Nalorphine extracts and reacts like morphine under the conditions of both the Fujimoto, Way & Hine and the Goldbaum & Kazyak procedures. Although the Mannering chromatographic procedure, using a solvent system consisting of butanol, acetic acid and water on filter paper buffered to pH 7 effectively separates nalorphine from morphine, normorphine shows virtually the same mobility as morphine in this system. When traces of normorphine were noted in some commercial preparations of nalorphine, it became necessary to find a solvent system that would differentiate normorphine from morphine. The solvent system described by Brossi et al. (1955) was adapted and used in the general procedure described below. A more detailed report will be published later (Lin & Way).

To 15 ml of urine, 1.5 ml of concentrated hydrochloric acid is added, and the acidified urine is heated in an autoclave for 40 minutes at 20 lb pressure. After cooling; 10 ml of 16 N potassium hydroxide is added and the urine is extracted with 20 ml of n-butanol. A 15 ml aliquot of the butanol layer is equilibrated with 6 ml of N/1 sulfuric acid. After separation of the two layers the aqueous acid phase is neutralized to approximately pH 9 by addition of concentrated potassium hydroxide (approximately 1.2 ml). After addition of approximately 0.5 g of sodium bicarbonate the alkalinized solution is extracted with 10 ml of ethyl acetate. The organic solvent phase is removed, concentrated nearly to dryness (about 0.1 ml) and spotted on Whatman No. 3 filter paper buffered to pH 7. This is prepared by soaking the paper in phosphate buffer of pH 7 and then allowing it to dry at room temperature. Development of the chromatogram is effected by the descending technique in a system consisting of tertiary amyl alcohol (10 parts), butyl ether (1 part) and distilled water (5 parts).


The results of the pupillary test and the chemical tests for morphine were compared on a total of 419 cases and are summarized in table 1. The two tests showed accord in 352 or 84 per cent of the cases, 339 cases being reported as negative and 13 cases as positive by both procedures, with morphine levels in the urine of the positive cases ranging from 4 to 30 micrograms/ml.

An analysis of the 67 cases where the chemical tests and the pupil test were not in agreement yielded the following statistics. In 18 cases reported to be negative by the pupillary test, morphine was detected in concentrations ranging from 2 micrograms/ml, to as high as 16 micrograms/ml. In 10 cases where the pupillary test was reported to be equivocal, morphine was noted to be present in concentrations ranging from 3 micrograms/ml. to 8 micrograms/ml. Thus, in 28 cases the chemical tests clearly proved to be more sensitive than the pupillary tests for establishing morphine usage. Furthermore, in 16 cases where the pupillary test was reported to be negative, the chemical test by the Fujimoto, Way & Hine (1954) procedure was equivocal that is, morphine was apparently detected, but the concentrations (> 1 < 2 micrograms/ml) were too low to make an unequivocal identification. These cases were reported as "negative" to officials involved in the antinarcotic testing programme.

There were 12 cases, on the other hand, where the pupillary test was reported to be positive and no morphine was detected. However, additional study provided an explanation for the discrepancy in 7 instances. The presence of codeine was established in 5 instances. In one case, although the compound was not detected in the urine, the subject admitted having taken methadone. In another case the chemical test might have been invalidated because the urine appeared to have been

Comparison of chemical tests with the pupillary method for the diagnosis of narcotic use 31diluted with water. However, in 5 cases no objective evidence for a difference in the pupillary and chemical tests could be established. Moreover, in 11 cases where the pupillary test was reported to be equivocal, the chemical tests were negative. If the 5 positive codeine cases are included in the number of cases wherein the chemical and pupillary tests are in accord, then the total would be 357 and the percentage of the tests in agreement would be 85.


Identification of morphine in urine by chemical tests in general should be quite reliable. While no single test may be considered to be absolutely specific for morphine, a positive test for morphine by two or three methods that rigidly define its different chemical and physical attributes constitutes very convincing evidence. Although the Fujimoto, Way & Hine (1954) procedure is based on a phenol colour test with silicomolybdic acid, measurement of morphine can be made with considerable reliability because, in the initial extraction step, the compound is partitioned selectively from a highly concentrated alkalinized urine by n-butanol. Subsequent extractions of morphine from acid and basic solution in the conventional manner further enhance its purification before the phenol test is applied. Thus, this procedure in combination with the ultraviolet and paper chromatographic methods yields extremely reliable evidence for morphine when the compound is present in urine at concentrations in excess of 2 micrograms/ml. While the administration of nalorphine may complicate urinary chemical tests for morphine, the paper chromatographic procedure described in this paper using tertiary amyl alcohol, butyl ether and distilled water, easily circumvents this objection. Thus, the results obtained with the chemical tests may be used as a base line for making comparisons with respect to the accuracy, sensitivity and usefulness of the pupillary test.

In view of the reliability of the chemical test, in the cases where the chemical test for morphine was positive and the pupillary test was either negative or equivocal, the chemical test must be giving a more sensitive indication that the subject under test had been taking heroin or morphine. The logical explanation is that the effects of heroin or morphine had dissipated in these subjects by the time the pupil test was performed, but sufficient amounts were present or still being excreted in the urine. It may be that these subjects had been taking narcotics only occasionally or for a short period and would have developed little or no physical dependence on heroin or morphine.

The finding of codeine in 5 cases where a positive pupillary test was obtained is of considerable interest, since former heroin addicts are known to have used it as a partial substitute for their former habit and addiction to cough syrups containing dihydrocodeinone has been reported (Rosenwald & Russel, 1961). Codeine would not be detected by the Fujimoto, Way & Hine (1954) and the Goldbaum & Kazyak (1952) methods for morphine, but would be easily detected by both the Mannering et al. (1954) paper chromatographic procedure and the one described in this communication. The latter two methods can also be applied for detecting other morphine surrogates in urine, but their general usefulness in this respect needs to be defined. The applicability of general alkaloidal tests for detecting the various morphine-like substances needs to be studied in terms of the sensitivity of the test reagent for each substance recovered from biologic fluids and correlated with what is known about its fate.

In instances where the pupillary test did not correlate with the chemical tests, several explanations come to mind:

  1. An error may have been made in measuring the pupil. This does not seem unlikely, particularly if only one measurement was made with the card pupillometer.

  2. A narcotic substance other than the ones we could detect had been taken.

  3. The subject may have been strongly addicted to heroin or morphine but recently withdrawn from the drug. In such circumstances the pupillary test may still be responsive after the drug has been eliminated from the body. A more definitive explanation can be obtained only by further studies.

Despite the need for this additional information, the present studies enable some conclusions to be made concerning the relative merits of the pupillary and the chemical tests in terms of reliability, sensitivity, general applicability, and simplicity.

With respect to reliability, as was pointed out earlier, there can be no question that the chemical tests, properly performed, provide very convincing evidence for establishing that narcotics are used.

With respect to the relative sensitivity of the pupil and the chemical tests, it is difficult to generalize, because the criteria for establishing usage of narcotics by the two tests are so dissimilar. In the pupillary test, drug potency is being assessed, while in the chemical test drug concentration is important. As a rule, one might expect that a narcotic with low activity would be easily detected by chemical means because the high dose of the compound needed to elicit a pharmacologic effect would usually be reflected by appreciable amounts of the compound in the urine. On the other hand, with a highly potent narcotic, its low dosage would not favour its detection in urine, especially if it had undergone extensive biotransformation; however, there should be no difficulty in establishing its usage by the pupillary test.

In terms of applicability the pupillary method may be considered to be a useful general test for narcotics. Most drugs with morphine-like effects exhibit mydriasis following nalorphine, but meperidine appears to be an exception. Wikler (1958) reported that nalorphine precipitated little or no abstinence phenomena in patients tolerant to meperidine. Moreover, unlike morphine, meperidine, in therapeutic doses, usually leaves the pupillary size unaltered (Batterman, 1943); in fact addicts may show mydriasis (Woods, 1958).

The chemical tests are deficient in that a given test may have only limited applicability. A universal test for narcotics in urine that would be simple, sensitive and reliable for application in the field is not available and, indeed, would be extremely difficult to devise. Inasmuch as a "negative" test for narcotic substances merely defines the limits of sensitivity of the test procedure under application, and since different narcotic compounds have varying degrees of sensitivity to the same alkaloidal reagent, the development of a general narcotic test for use on the same urine sample can be accomplished only by sacrificing sensitivity. Thus, enhancing the sensitivity of tests for heroin or morphine by subjecting the urine to acid hydrolysis virtually invalidates any alkaloidal tests for ester-type narcotics such as meperidine and anileridine. On the other hand, omitting the acid hydrolysis step for bound morphine reduces the sensitivity of the test for morphine by several times. This dilemma is compounded further by the fact that the narcotic compounds are largely metabolized in the body. In general less than 10 per cent of the administered dose appears in the urine unchanged and the chief biotransformation products, unlike their parent alkaloids, do not necessarily share common chemical and physical attributes (Way & Adler, 1960-1961). Since heroin is by far the agent most subject to abuse, it does not appear feasible to sacrifice sensitivity of tests for its detection for wider applicability to other agents. The paper chromatographic procedure we have described does have fairly wide applicability and sensitivity, and we consider it most useful for screening purposes.

In conclusion, we may state that although certain discrepancies were noted between the pupillary test and the chemical tests, we still consider the biologic test to be highly useful for establishing the diagnosis of narcotic usage. While the pupillary test alone should not be used as the absolute criterion of narcotic usage, it certainly can furnish strong suggestive evidence. The application of the test under conditions recommended by State of California Department of Public Health (1961) appears to have a relatively small risk with respect to serious untoward complications. A controlled study of the dose and time relationship of the pupil's response to nalorphine and morphine surrogates is needed to provide more definitive information since it is impossible to evaluate completely the pupillary test until known amounts of morphine-like drugs can be administered under controlled conditions. Permission to use confined parolees for this purpose has been obtained, and this will permit more complete studies in the future.


The validity of the nalorphine or pupillary test for determining use of narcotics was evaluated by chemical analyses for narcotic substances in urine collected from narcotic parolees or probationers undergoing the pupillary test. The findings of the two tests were correlated on a total of 419 tests. There was accord in 357 or 85 per cent of the cases, 339 cases being reported as negative and 18 as positive (morphine, 13; codeine, 5) by both procedures. In 28 tests the chemical method proved to be more sensitive; a positive chemical test was obtained when the pupillary test was reported as negative in 18 tests and equivocal in 10 others. There were 7 cases in which in the pupillary test was found to be positive and the chemical test negative; a satisfactory explanation for the discrepancy could be explained in two of the tests. Some possible reasons for a difference in the chemical and pupillary tests in the other 5 tests were considered. It is concluded that the pupillary test with nalorphine is useful for determining whether an individual has taken narcotic agents, but change in pupillary size alone is insufficient evidence to establish this for certain.



BATTERMAN, R. C.: Archiv. Int. Med., 71: 345, 1943.


BROSSI, A. O., HÖFLIGER, O. and SCHNIDER O.: Arzneimittel-Forschung, 5: 62, 1955.


ELLIOTT, H. W. & WAY, E. Leong: Clin. Pharm. & Therap., 2: 1961.


FRASER, H. F., VAN HORN, G. D. & ISBELL. H.: Am. J. Med. Sci, 231: 1, 1956.


FUJIMOTO, J. M. & WAY, E. Leong: J. Pharmacol, 121: 340, 1957.


GOLDBAUM, L. R. & KAZYAK, L.: J. Pharmacol., 106: 388, 1952.


GOODMAN, L. & GILMAN, A.: The Pharmacological Basis of Therapeutics , 2nd edition, MacMillan, 1955.


MANNERING, G. L., DIXON, A. C., CARROL, N. & COPE, O.: J. Lab. Clin. Med., 44 : 292, 1954.


OBERST, F. W.: J. Pharmacol., 73 : 401, 1941.


OBERST, F. W.: J. Pharmacol., 74 : 37, 1942.


OBERST, F. W.: J. Pharmacol., 79 : 266, 1943.


Recommended Procedure for Narcotic Use Testing of Probationers and Parolees . Bulletin issued by State of Calif. Dept. of Public Health and Dept. of Justice, Bureau of Narcotic Enforcement. May 1961.


ROSENWALD, R. J. & RUSSEL, D. H.: New Eng. J. Med. 264: 927, 1961.


TERRY, J. G. & BRAUMOELLER, F. L.: Cal. Med. 85: 299-301, 1956.


WAY, E. Leong & ADLER, T. K.: Pharmacol. Rev. 12: 383, 1960.


WIKLER, A.: Public Health Monograph No. 52, 1958.


WIKLER, A., FRASER, H. F., & ISBELL, H.: J. of Pharmacol., 109: 8, 1953.


WOODS, L. A. in Pharmacology in Medicine edit. by V. A. Drill, McGraw Hill, 1958.


During this study the authors served as consultants to the Department of Justice, State of California. The study is a part of the Department's Anti-Narcotic Programme and was supported in part by Contract No. 1137 with the State of California and in part by a United States Public Health Service Grant, RG 1839