ABSTRACT

Aftercare monitoring programs for health care professionals with documented substance abuse problems are managed differently by various states and their respective licensing boards. Many programs have reported surreptitious alcohol use as a significant concern, yet there is no clear indication of the best methods for detecting alcohol use since the detection of ethanol is difficult due to its limited half-life in saliva, urine and serum. Ethyl glucuronide (EtG), a minor metabolite of alcohol metabolism, is only present in urine when alcohol has been consumed. EtG testing may therefore improve the detection of alcohol use by health care professionals in monitored aftercare programs.

This study compared urine ethanol (Medpro B panel) and urine EtG in specimens from 126 clients enrolled in Arizona’s physician health aftercare monitoring program. Each client was tested twice per month for a two-month period in 2004. Of the 504 tests, there were no positive urine ethanol results using a standard, lab determined minimum cutoff level of 20mg/dL. There were four positive EtG results (one percent positive rate) using a minimum cutoff level of 100ng/mL. The four positive EtG results imply the presence of four false negative urine ethanol results. Therefore, the results suggest that the EtG test is more sensitive for alcohol detection and should be considered for future improvements in the testing and monitoring of health care providers enrolled in aftercare programs.

INTRODUCTION

Approximately 13 percent of all Americans, including the community of medical professionals, will develop an alcohol abuse or dependency problem.1  In 1972, the American Medical Association (AMA) began recognizing substance abuse as a problem among physicians.2  The AMA recommended state medical societies establish programs or committees to help identify and aid impaired physicians. The AMA also developed model legislation to amend state practice acts so treatment, rather than punitive disciplinary measures, could be made available.2  All states currently have physician health programs (PHPs) specifically designed to help impaired physicians. These programs share common goals for treatment, recovery, rehabilitation and the maintenance of sobriety.3 

Monitored aftercare is an important element of a recovery program. Strict monitoring of abstinence from drug and alcohol abuse influences the effectiveness of these programs.4  Noncompliance with the client’s contract may be the greatest contributing factor to relapse and, therefore, should be monitored strictly.1  While the urine drug and alcohol screen is useful to monitor abstinence, it is not without limitations. The detection of substances in urine depends upon the time elapsed since their last use. Because most drugs remain in the body long enough to be detected through weekly or even biweekly screenings, urinalysis may be sufficient. Alcohol is an exception to this. The liver quickly metabolizes alcohol such that detection in the urine is limited to approximately six to eight hours after use.5  Because alcohol is the most frequently abused substance,6  adequate monitoring for alcohol consumption is a major concern.

Ingesting any type of alcohol while in aftercare is a violation of the contractual agreement, potentially leading to relapse and abuse of other psychoactive substances, thus rendering the practitioner unsafe to practice his or her profession. During rehabilitation and recovery, the potential for alcohol abuse may increase due to the lack of efficient screening tests. Therefore, alcohol detection remains an important component of contract compliance, sobriety and public safety.

A recently available test for alcohol use identifies an ethanol metabolite in urine, Ethyl glucuronide (EtG). EtG can be detected in urine as long as three to five days after blood alcohol levels reach an undetectable level.7  The EtG test is reportedly sensitive enough to detect even small amounts (7g) of alcohol (one standard drink contains 15g of alcohol).8  The urine EtG test is also 100 percent specific since EtG is only present in urine after alcohol consumption or exposure. Testing urine for EtG may eliminate the need to test blood for less reliable markers (such as gamma glutamyltransferase, alanine aminotransferase, aspartate aminotransferase, mean-corpuscular volume or carbohydrate-deficient transferrin) and will not give false positive results for alcohol in diabetic clients. Therefore, EtG testing may provide more accurate results for surreptitious alcohol use. This study was conducted to compare the performance of the standard ethanol urinalysis and the Ethyl glucuronide test in detecting surreptitious alcohol use in a monitored aftercare program in Arizona.

METHODOLOGY

The subjects were all 126 clients from an aftercare program in Arizona that included allopathic physicians, osteopathic physicians, physician assistants, chiropractors and naturopathic physicians being monitored for documented substance abuse by their respective licensing boards. Subjects were sent a letter informing them of the new EtG test (including a description of its specificity and sensitivity) along with their chain of custody forms to be used during the 61-day study period in summer, 2004. Therefore, the sample included all clients in the aftercare program that were tested during the 61-day study period. Clients called their customary 800-phone number each day to determine if they were randomly selected to provide a urine specimen at one of several local collection sites. Each client was tested an average of twice per month for two months. The collected specimen was handled according to the required chain of custody and was sent for analysis to Northwest Toxicology Testing in Salt Lake City, Utah. Specimens were analyzed using the standard Medpro B panel that tests for ethanol and 14 categories of illicit and prescription drugs. The specimen was also analyzed using the EtG test. The results were sent to the director of the aftercare monitoring program who then reported the results to the Arizona Medical Board, Arizona Board of Osteopathic Examiners and other official licensing parties.

The results were also sent to the primary investigator of the current study for comparison. Each client’s urine ethanol results were compared to their urine EtG results to determine if the EtG test had a higher positive rate than the urine ethanol test. This two-month study was approved by Midwestern University’s Institutional Review Board prior to data collection. Client confidentiality was maintained throughout the study.

RESULTS

There were 506 collected specimens. Two specimens were discarded (one had insufficient quantity to test and one specimen was discarded due to the presence of iodate, an adulterant tested by the lab). The final sample of 504 specimens were tested and yielded the recorded results. However, six were analyzed more than 10 days after they were obtained. Also, one specimen was blue in color and one was black. The label date of three of the specimens did not match the date on the forms. Sixteen specimens had low specific gravity (≤1.002) and were considered dilute but testable since EtG tests on dilute specimens are reportedly accurate.9  The corresponding EtG and ethanol tests were negative on all of these 16 specimens.

Of 504 tests, there were no positive urine ethanol results using the standard laboratory minimum cutoff level of 20mg/dL. There were four positive EtG results using a minimum cutoff level of 100ng/mL (one percent positive rate). The positive EtG levels were recorded at 100ng/mL, 117ng/mL, 176ng/mL and 913ng/mL. Thus, there were four false negative results from the Medpro B urine ethanol tests.

DISCUSSION

The results of the current study support the notion that the detection of urine EtG is a more reliable measure for surreptitious alcohol use than the current ethanol urinalysis. In this study the four positive EtG test results coincided with negative urine ethanol results, yielding four false negatives in clients that had exposure to alcohol. The 100 percent specificity of the EtG metabolite to ethanol eliminates false positives and the possibility of fermentation in urine, which may occur with diabetic clients.8,11 

The EtG test may be sensitive enough to detect possible incidental exposure to ethanol, such as mouthwash, over-the-counter medication, cough syrup or communion wine. However, since EtG is a minor pathway of alcohol metabolism in the liver, a significant amount of alcohol must be ingested to yield a positive EtG result. The EtG pathway metabolizes .02 percent to .04 percent of alcohol, therefore it is unlikely that incidental exposure would yield a positive EtG test result.7 

In the study by Skipper et al.7 , a suggested cutoff level of 100–250ng/mL was used to eliminate incidental exposure to ethanol. Other authors have recommended an EtG level greater than 500ng/mL in urine, so that a positive result due to incidental exposure would be unlikely.10  The possibility exists that a positive EtG result may be due to surreptitious alcohol consumption several days prior to testing. Therefore, any positive test result over 100ng/mL more than likely results from surreptitious alcohol consumption rather than incidental exposure. State programs may consider testing clients with positive levels less than 500ng/mL more frequently. As a result, programs using EtG testing need to determine their own cut-off levels for EtG in order to eliminate incidental exposure to alcohol versus surreptitious consumption.

The fact that only a one percent positive rate was observed during the test period may be a reflection of the informed study population regarding the new EtG testing protocol. Skipper et al.10  reported 7 of 100 (7 percent) specimens resulted in positive rates for EtG in a study of uninformed clients tested in eight different state programs (See Table 1, 2004a).10  The results of the current study and of Skipper et al. indicate informing clients of an additional EtG test along with the standard urinalysis may be a deterrent which limits alcohol consumption in these populations.

Table 1

Research Studies Comparing Urine Ethanol Tests to EtG Analysis

Research Studies Comparing Urine Ethanol Tests to EtG Analysis
Research Studies Comparing Urine Ethanol Tests to EtG Analysis

Skipper et al.10  also tested a select population of 18 clients for three reasons: those clients considered “high risk” with a history of multiple relapses, those “for concern” due to reports of possible drinking, or those who had positive urine ethanol results. Four of the six positive urine ethanol results reported were false positives due to fermentation of the urine of the diabetic clients after collection. The corresponding EtG results were negative. Ten of the 18 tested positive for EtG (56 percent) and admitted consuming alcohol. These results indicate the EtG test is useful for testing clients who are considered at “high risk” by program directors. Because the EtG test remains somewhat cost prohibitive for regular use, state program directors will need to determine the best protocol for incorporating EtG testing into their programs.

A study by Wurst et al.11  followed 35 patients over the course of 12 months. The patients were tested upon return from weekday or weekend leaves using EtG and other alcohol markers. Wurst et al. collected a total of 146 urine specimens, 14 of which were positive for EtG. All 14 specimens were collected from patients who had been on leave (Table 1) and later reported drinking. The one positive ethanol came from a patient who reported consuming 100g of alcohol 12 hours prior to testing. The present study supports these results in that EtG is a more sensitive test for exposure to alcohol.

A recent study12  of state aftercare monitoring programs indicates surreptitious alcohol use is a significant concern for all monitoring aftercare programs, yet there is no clear indication of the best methods for detecting alcohol use. In the 2003 survey of 46 state physician health programs, directors for 45 programs believed clients might be using alcohol occasionally without detection. Thirty-three state program directors considered client alcohol use a ‘significant concern,’ while 25 state programs do not use a separate test for alcohol detection. However, 17 state programs reported using EtG testing on specific clients — primarily those who have demonstrated repeated positive urine ethanol screens, those under significant suspicion or concern, or those whose drug of choice was alcohol. One program tests all clients using EtG one time per month,12  while some state programs are using the EtG test to retest dilute specimens that the lab cannot test accurately using the Medpro B panel.9  When asked, 93 percent of state program directors recommended improvements in the methods for detection of alcohol use. Fifteen directors considered EtG testing to be an important means for this improvement.12  The results of the current study support this recommendation.

Limitations to the current study include the fact that all subjects were notified of the additional EtG alcohol test and the testing period was limited to two months. As a result, the positive EtG test rate may actually be an underestimation of client alcohol use prior to notification. In addition, due to the blinding of the test samples, it is not known whether the four positive tests came from four different individuals or if one person may have tested positive more than once. Nonetheless, the four positive EtG samples did correspond to four false negative Medpro B results. Areas for future studies include more rigorous evaluations of the influence of incidental alcohol exposure from common products or over-the-counter medications on EtG levels. These reports should determine the most appropriate cutoff EtG level and detection window to determine the most effective procedure for the detection of surreptitious alcohol use. Improving the sensitivity of screening tests for alcohol use may be a practical and effective alternative to (or in conjunction with) decreasing the specimen collection window. In addition, the use of more sensitive and specific toxicology tests, such as EtG, that demonstrate improved sensitivity and specificity may dramatically increase the effectiveness of aftercare programs to detect the use of alcohol. However, any changes to current testing protocols must always be balanced with consideration for their intrusiveness and invasiveness in the lives of the clients. Future studies may further address the need for uniform standards of monitoring and testing protocols by aftercare programs so that all state physician health programs may achieve their common goals.

ACKNOWLEDGEMENTS

This project was funded by the Biomedical Sciences Program in the College of Health Sciences at Midwestern University in Glendale, Arizona. The authors have no financial obligations or conflicts of interest related to the research study.

REFERENCES

REFERENCES
1.
Rivers
PA,
Bae
S.
Substance abuse and dependence in physicians: detection and treatment
.
Health Manpower Management
1998
;
24
(
5
):
183
87
.
2.
American Medical Association, Council on Mental Health.
The sick physician: impairment by psychiatric disorders, including alcoholism and drug dependence
.
JAMA
1973
;
233
:
684
687
.
3.
Nelson
HD,
Matthews
AM,
Girard
DE,
Bloom
JD.
Substance impaired physicians probationary and voluntary treatment programs compared
.
West J Med
July/Aug
1996
;
165
(
1–2
):
31
6
.
4.
Talbott
GD,
Gallegos
KV,
Angres
DH.
Impairment and recovery in physicians and other health professionals
.
In:
Graham
AW,
Schultz
TK,
eds.
Principles of Addiction Medicine
. 2nd ed.
Chevy Chase (MD)
:
American Society of Addiction Medicine
;
1998
. p.
1263
79
.
5.
Helander
A.
Biological markers of alcohol use and abuse in theory and practice
.
In:
Agarwal
DP,
Seitz
HK,
eds.
Alcohol in Health and Disease
.
New York
:
Marcel Dekker, Inc.
;
2001
:
177
205
.
6.
Boisaubin
EV,
Levine
RE.
Identifying and Assisting the Impaired Physician
.
Am J Med Sci
July
2001
;
322
(
1
):
31
6
.
7.
Skipper
G,
Weinmann
W,
Wurst
FM.
Ethylglucuronide (EtG): A new marker to detect alcohol use in recovering physicians
.
Journal of Medical Licensure and Discipline
2004
;
90
(
2
):
14
17
.
8.
Stephanson
N,
Dahl
H,
Helander
A,
Beck
O.
Direct quantification of ethylglucuronide in clinical urine samples by liquid chromatography-mass spectrometry
.
Therapeutic Drug Monitoring
2002
;
24
(
5
):
645
51
.
9.
Jansen
M,
Sucher
MA.
A study using ethylglucuronide testing in Arizona’s Physician Health aftercare program
.
Western Regional Meeting of the Federation of State Physician Health Programs
.
October 1, 2004
.
10.
Skipper
GE,
Weinmann
W,
Thierauf
A,
Schaefer
P,
Wiesbeck
G,
Allen
JP,
Miller
M,
Wurst
FM.
Ethyl glucuronide: A biomarker to identify alcohol use by health professionals recovering from substance use disorders
.
Alcohol Alcohol
Sept–Oct
2004
;
39
(
5
):
445
49
.
11.
Wurst
FM,
Vogel
R,
Jachau
K,
Varga
A,
Alling
C,
Alt
A,
Skipper
GE.
Ethyl glucuronide discloses recent covert alcohol use not detected by standard testing in forensic psychiatric inpatients
.
Alcohol Clin Exp Res
March
2003
;
27
(
3
):
471
476
.
12.
Jansen
M,
Bell
LB,
Sucher
MA,
Stoehr
JD.
Detection of Alcohol Use in Monitored Aftercare Programs: A National Survey of State Physician Health Programs
.
Journal of Medical Licensure and Discipline
2004
;
90
(
2
):
8
13
.