Ethylene glycol is a colorless, odorless liquid commonly
used as automobile antifreeze. Ethylene glycol ingestion can cause severe central nervous system depression, as well as cardiac, pulmonary, and renal effects, and may be fatal.1 Treatment
may include fomepizole, ethanol, or dialysis and should be initiated as
soon as possible. An ethylene glycol serum level is the gold standard
for diagnosing this ingestion. However, because test results may be
unavailable for hours or days, several surrogate tests are used for the
detection of ethylene glycol ingestion. One of these tests is the
inspection of the patient's urine with a Wood's
lamp.1-3 Fluorescein is a fluorescent dye added to most
ethylene glycol antifreeze products so that mechanics can locate the
site of an antifreeze leak by inspecting a vehicle with a black light.
Ingestion of fluorescein at concentrations present in antifreeze causes fluorescent urine.3
We recently treated an adolescent who was suspected of drinking
antifreeze at home. Her urine was fluorescent, suggesting that she had
ingested the product; however, no ethylene glycol was present in her
serum. An informal analysis in the pediatric emergency department
showed multiple fluorescent urine samples.
We studied whether fluorescence is a common property of urine from
children.
 |
METHODS |
The study was performed at a tertiary care pediatrics hospital.
Institutional review board approval was obtained. Written informed
consent was obtained from all participants (or their parents) for the
second part; this requirement was waived by the institutional review
board for the first part.
In part one, a convenience sample of urine from children was collected
prospectively. To be included in this group, a patient had to be
admitted at our children's hospital for any reason unrelated to
poisoning and any urine test had to be performed. We obtained our
specimens from urine discarded after other tests.
For the second part, we used urine from a convenience group of healthy
children known to one of the investigators (M.J.C.).
All urine specimens were viewed with a Wood's lamp (black light) in a
dark room. Each sample was identified as fluorescent or not fluorescent
by independent observers (all emergency physicians). Three physicians
rated specimens in part 1; 2 rated specimens in part 2.
A positive control (fluorescein in tap water) and a negative control
(tap water) were viewed under the same conditions as the urine
specimens.
| |
RESULTS |
Of the 30 specimens in part 1, 21 were considered fluorescent by
all 3 investigators; 8 specimens were considered fluorescent by 2 of
the 3 investigators; and 1 specimen was considered fluorescent by only
1 investigator. The 3 investigators considered 90%, 93%, and 83% of
the specimens fluorescent, respectively (
< .2). No patient
sample was considered nonfluorescent by all 3 investigators.
Of the 16 specimens in part 2, 11 were considered fluorescent by both
investigators; 5 were judged fluorescent by only 1 investigator; and
none was considered nonfluorescent by both investigators. The
investigators, respectively, considered 93% and 75% ( < .2) of the specimens fluorescent.
| |
DISCUSSION |
Early treatment of children who have ingested ethylene glycol is
critical. Because serum ethylene glycol levels often are not
immediately available, surrogate markers are used to identify patients
who have ingested ethylene glycol.
One of these surrogates, the osmolal gap, has been challenged as an
appropriate marker for ethylene glycol overdose.4
Our data challenge fluorescent urine as a surrogate marker of ethylene
glycol overdose. We found that nearly all urine specimens from children
may be fluorescent. Thus, although some children who have ingested
ethylene glycol antifreeze products may demonstrate fluorescent urine,
the majority of fluorescent urine specimens obtained from a population
with a low incidence of ethylene glycol ingestion will represent
false-positives.
Our report contradicts a previous study of adults,3 which
included a control group, none of whom had fluorescent urine, and a
study group, in which patients had transient urine fluorescence after
fluorescein ingestion. Only glass containers were used in that study
because of high native fluorescence of plastic. Although we used
plastic tubes, our negative control specimens in plastic did not
fluoresce. In a companion study, 10 additional urine specimens were
studied first in plastic, then in glass. All appeared fluorescent in
both containers. However, when we repeated this study at another hospital, we found both glass and plastic tubes that were fluorescent.
Limitations and Future Questions
We did not measure intensity and wavelength of fluorescent
emission. Each specimen was interpreted by an emergency physician as
fluorescent or not fluorescent. Although this method allowed more
interrater variability than would be expected from a fluorometer, it
more closely matches the clinical use of this test, the accuracy of
which has been challenged even in adults.5
We do not know the source of the urine fluorescence. Fluorescein, as D
and C yellow No. 7, may be added to commercial drug and cosmetic
products. A number of drugs (indomethacin,7
triamterene,8 and others), food products (riboflavin and
its metabolite lumiflavin7 and niacin) and toxins
(dinitrophenol,7 rhodamine,7
aflatoxins,7 palmotoxins,7 and others) and
endogenous compounds (reduced nicotinamide adenine dinucleotide, reduced nicotinamide adenine dinucleotide
phosphate,7 porphyrins, and others) are
fluorescent. Additional research will be needed to characterize the
source(s) of urine fluorescence in healthy children.
Although we advocate abandoning the test of urine fluorescence as a
marker for antifreeze ingestion, we do not have another test to
recommend in its place. Development and approval of an inexpensive,
accurate, and rapid assay for ethylene glycol is needed.
No urine was obtained from children poisoned by ethylene glycol
antifreeze. Although comparing the fluorescence of urine from poisoned
and healthy children would be interesting, this luxury is not usually
available to the emergency physician and the absence of this comparison
does not weaken our conclusion: the urine of children is often
fluorescent.
| |
CONCLUSION |
Most urine specimens obtained from children are fluorescent.
Interrater variability exists. Some plastic and some glass tubes are
fluorescent. Fluorescent urine is not an indication of ethylene glycol
antifreeze ingestion by children. Because the management of patients
with ethylene glycol ingestion includes expensive, invasive, and
potentially toxic therapies, we recommend that the urine fluorescence
test be abandoned as a screen for ethylene glycol ingestion in
children.
| |
ACKNOWLEDGMENTS |
Dr Casavant is partially supported by National Institute of
Child Health and Human Development Pediatric Pharmacology Research Unit
HD31316 and the National Institute of Environmental Health Sciences F32
ESO5651-02.
We thank John R. Hayes, PhD, for statistical advice.
, and
Top
Abstract
Methods
System
Pharmacists; 1995:2528
