Personal View Ann C/in Biochem 1997; 34: 579-581

Effects of drugs 00 clinical laboratory tests

D S Young
From the Department of Pathology and Lahoratory Medicine, University of Pennsylvania, 3400 Spruce
Street, Philadelphia, PA 19104-4283, USA

Additional key phrases: drug interferences; labora- growing. One continuing concern is the point at
tory testing; clinical databases which data should be considered obsolete and
not reported. Generally, data on analytical in ter-
In 1962, Caraway' highlighted the potential for ferences of drugs have been deleted from the
therapeutic drugs and endogenous compounds to master file when references are more than 15
modify the concentrations of various anal ytes in years old. However, for procedures that con-
body ñuids, A few years later, others, including tinue to be widely used, e.g. urine dipsticks, the
Elking and Kabat! and Sunderman;' amplified reported effects have been republished. In the
Caraway's earlier reporto These early papers early editions of the publication there was con-
prompted rnany clinical chemists to begin to siderable emphasis on the effects of drugs on the
question their test results, something that clini- analytical methods. In the later editions there
cians often did when the results did not appear to has been les s emphasis on analytical interfer-
fit the clinical picture of their patients. At the time en ces as analytical specificity has improved.
these papers were written there was legitima te In all the editions I have adopted the
concem about the analytical specificity of many philosophy that it is as important to document
laboratory methods, and clinicians did not have that a drug has been demonstrated not to have
confidence in the test results. The clinician's view an effect on a particular method as it is to
was that any test result that did not conform to describe the effects that it does have. The
his or her clinical impression was erroneous. clinician is often faced with an unexpected
My personal interest in the effects of drugs on abnormal result. He or she needs to know all
clinical laboratory tests was triggered by the possible causes. Invariably, clinicians are
physicians criticizing certain test results that unaware of the analytical methods used to
my laboratory produced. They felt that the measure the analytes that they require to be
results were not in keeping with their clinical measured. They are also often unaware of sorne
impressions. They concluded the test results of the side-effects of drugs that they use
were wrong. Yet I knew that quality assurance relatively infrequently. Thus, to pro vide all the
data were within acceptable limits. Obtaining information that clinicians and laboratory staff
drug histories often identified likely causes of need to properly interpret the test results, the
abnormal results. Sometimes, the clinicians were database of the effects of drugs must be fully
not aware of all the in vivo effects of the drugs comprehensive.
they were using. It was possible to show that the A repeatedly asked question is whether the
abnormal test result was related to a side-effect, availability of databases of effects of drugs on
Sometimes we were able to demonstrate an clinical laboratory tests can influence medical
interference of a drug with an analytical method. practice. Most users of the database can provide
It seemed worth capturing the effort put into anecdotal accounts of instances when it was able
tracking the causes of spurious results so that to pro vide an explanation for unexpected
other laboratory staff might find it easier to abnormal results. Shortly after the publication
explain unanticipated test results. The informa- of the second compilation of the database," we
tion I accumulated was first published in a embarked on a study at the University of
regular issue of Clinical Chemistry.t updated in a Wisconsin Hospitals? to determine whether
special issue of the journal, and then, periodi- automatic reporting to physicians of potential
cally, in books published by the American drug-induced causes of abnormal test results
Association for Clinical Chemistry (AACC), could modify medical practice. More than
the most recent in 1995.5 13 000 patient-days in patients in general
The amount of information related to the surgical, minimal or modera te medical care, or
effect of drugs on clinical laboratory tests is intensive care units, were followed.

579
580 Young
 When an abnormal test result was obtained in products on common laboratory tests. Such
a patient receiving a drug whose presence had information should be available from the clinical
been reported to afTect that test, a computer tri al s prior to market release.
generated a report stating that the drug might Both laboratory staff and clinicians must
account for the abnormal result, These reports, recognize that any database of effects of drugs
including the explanation of the effect in the file, on clinical laboratory tests can only provide
were sent with the test results to the ordering a possible cause of a change in a test value, not a
physician. The drugs prompting most alerts were definitive explanation of an abnormal value in
frusemide (10·0%), acetaminophen (9·8%), a specific patient. Furthermore, it should be
penicillin (6·5%) and hydrochlorothiazide noted that in vivo efTects are usually determined
(4·7%). The laboratory tests most afTected were in healthy individuals, and not in patients whose
total leukocyte count (11·3%), haemoglobin disease might influence the metabolism of the
(9·0%), potassium (8·7%) and glucose (7·3%). drugs to a difTerent extent.
After the study was concluded we interviewed Laboratory staff are now well aware of the
the medical staff to determine the usefulness of potential of drugs to affect the concentration of
the programo They felt that approximately 30% certain analytes. However, many c1inicians still
of the reports were useful and 4% actually trivialize the importance of these efTects, focus-
caused them to change the management of their ing only on the therapeutic intent of the drug.
patient. Our analysis of the reports suggested To enhance their role in the management of
that the interaction contained in the file patients, laboratory staff should think of
provided the most probable explanation for the themselves more as an information provider
abnorma! test result: (a) in the general surgical than an analyst. They shou!d assume a consult-
(11 %); (b) rninimal medical (21 %); (e) moderate ing role, providing an interpretation of the test
medical (23%); and (d) intensive care units results, The three databases published by the
(6%). Even though the file was not designed for AACC now contain more than lOO 000 docu-
this application, it is reassuring that many of the mented influences on test values.s.lo.11 Medical
explanations were pertinent; suggesting the progress is such that about 10 000 new effects
potential for an automatic application of a are likely to accrue each year. 1 use an electronic
database to enhance medical careo screen of the published literature, review the
Discussions of the possible mechanisms by abstracts of pertinent papers, and then enter
which drugs afTect laboratory tests have been those effects (clinical, drug and pre-analytical)
thorough1y documented.! It is now expected that that appear to be significant into a single data-
manufacturers will pro vide purchasers of their base that has been sorted, in the past, to produce
equipment with documentation of the effects of separate publications. The AACC intends to
common drugs on the analytical methods used disseminate the information electronically In
in their analysers. The National Committee for several different ways in the future.
Clinica! Laboratory Standards (NCCLS) has
been instrumental in developing a protocol for
REFERENCES
manufacturers to study the effects of drugs at
physiological concentrations on the methods 1 Caraway WT. Chemical and diagnostic specificity of
included in their ínstruments.? Such information laboratory tests. Am J Clin Pathol 1962; 37: 445--{í4
2 Elking MP, Kabat HF. Drug-induced modifications
is invaluable. Most pharmaceutical manufac- of laboratory test values. Am J Hosp Pharm 1968;
turers routinely pro vide information about the 25: 484-519
in vivo effects of their drugs on organ function 3 Sunderman FW ir. Drug interference in clinical
and often document their effects on analytical biochemistry. Crit Rev Clin Lab Sci 1970; 1: 427-49
methods. 4 Young DS, Thomas DW, Friedman RB, Pestaner
LC. Effects of drugs on clinicallaboratory tests. C/in
Often overlooked when assessing the effects of
Chem 1972; 18: 1041-301
drugs is the effect of metabolites of a drug that
5 Young DS, Pestaner LC, Gibberman V. Effects of
may afTect a test result to a greater extent than drugs on clinical laboratory tests. Clin Chem 1975;
the parent compound. The excipient or vehicle 21: 10-432
in which the drug is adrninistered, may also have 6 y oung DS. Effects of Drugs on Clinica/ Laboratory
an effect unrelated to the drug itself. It would be Tests. Washington, DC: American Association for
use fui if pharmaceutical manufacturers provided Clinical Chemistry, 1995
7 Friedman RB, Young DS, Beatty ES. Automated
readily available documentation of both the
monitoring of drug-test interactions. Clin Pharmaco/
effects and the absence of effects of their Ther 1978; 24: 16-21

Ann Clin Biochem 1997: 34

Effects of drugs on clinical laboratory tests 581
 8 Forman OT, Young OS. Orug interferences In oc: American Association for Clinical Chemistry,
laboratory testing. Ann Clin Lab Sci 1976; 6: 1997
263-71 II Young OS. Effects of Preanalytical Variables 0'1
9 Powers OM, Boyd lC, Glick MR, Kotschi ML, Clinical Laboratory Tests, 2nd edn. Washington,
Letellier G, Miller WG, et al. Interference Testing OC: American Association for Clinical Chemistry,
in Clinical Chemistry, NCCLS Document EP7-P, 6, 1997
No. 13, 1986
10 Friedman RB, Young OS. Effects of Disease 0'1
Clinical Laboratory Tests, 3rd edn. Washington, Accepted for publication l July 1997

Ann Clin Biochem 1997: 34