Download diagnostic testing for acute dysuria in adult women: the current state

VOLUME 22, NO. 3 FALL 1990
DIAGNOSTIC TESTING FOR ACUTE DYSURIA IN ADULT WOMEN: THE CURRENT STATE
Insurance Testing
DIAGNOSTIC TESTING FOR ACUTE DYSURIA IN ADULT
WOMEN: THE CURRENT STATE
BY
TERRY A. HURLBUT, III, MD
BENJAMIN LI~FENBERG, MD
INTRODUCTION BY
HUGO C. PRIBOR, MD, PHD
Introduction
Terry A. Hurlbut, III, MD, is a Research Associate in the
Medical Information Sciences program in the Department of
Pathology at Dartmouth Medical School, and Benjamin
Littenberg, M.D. is an Assistant Professor of Medicine at Dartmouth Medical School in Hanover, N.H.
The Medical Information Sciences program is a National Library of Medicine-sponsored training and research program
dedicated to the finding of new uses for computers in medical
research, education and patient care. Research fellows and
faculty members pursue computer-related research projects in
a variety of areas. For example, in clinical decision analysis,
they will extensively research a particular medical decision,
including all possible strategies and the consequences of those
strategies. The typical decision is between observing a patient
suspected of having a particular clinical entity, treating the
patient empirically, or testing the patient first to decide
whether to treat or not. They then draw a "decision tree"
listing all the possible consequences of all possible strategies,
including the probabilities associated with those consequences and their utilities, or costs, both economic and otherwise. By using an ordinary desktop computer, they can
calculate the maximum utility that one can realize from the
situation and determine the best strategy for achieving that
end. And since the computer relieves them of the onerous task
of hand-calculation of these utilities, they can make these
calculations multiple times, with multiple values for probabilities or outcome states about which they are uncertain. In this
manner, they can calculate the best decision in nearly every
conceivable medical setting.
Cost-benefit analysis and cost-effectiveness analysis are subsets of clinical decision analysis in which economic costs are
explicitly taken into account, and often balanced against a
recognizable clinical measure of utility, such as quality-adjusted life expectancy.
we practice medicine in the future. The combination of computer
skills plus medicine will allow them to design efficient, cost-effective techniques in diagnosis and management of patient care.
Acute dysuria in adult women is a serious problem in clinical
medicine. Cost-effective early diagnosis is important for therapy. In those instances when this is not done, many of these
women can have chronic recurring dysuria which has both
morbidity and mortality implications. This of course is important for the insurance industry for health care insurance and
possibly life insurance. Drs. Hurlbut and Littenberg are evaluating the most cost effective methods of diagnosis and management of this disorder. They are also evaluating the
significance of various different positive urine tests in the
differential diagnosis of acute dysuria versus acute vaginal
infection. I felt this column would be of interest as a hallmark
of things to come in directions of new research in computer
applications in clinical medicine and laboratory testing.
Acute Dysuria, Incidence and Importance
Acute dysuria in women is one of the most common health
problems in the United States today. Estimates of its incidence
range from 10% to 20% of women at least once in their lifetimes1 to as high as 25 per 100 woman-years.2 At least three
million U.S. doctors’ office visits are made by women complaining of dysuria.2
For nearly twenty years, the management of acute dysuria in
women was dominated by several traditional assumptions
about what disease these women usually have (bacterial urinary tract infection), the best test for that disease (bacterial
culture of mid-stream, clean-catch urine), the criterion for a
positive culture, and the optimal therapy for bacterial UTI.
Several events in more recent years have cast doubt upon most
or all of these assumptions.2 This discussion is intended to
summarize the current medical consensus regarding acute
dysuria in women, and to clarify the newer thinking that now
pertains to this common clinical problem.
Other projects of interest include interactive multimedia preWhat does acute dysuria actually mean?
sentations, computer-based instruction and examination of
medical students, patient data management, and expert sysAcute dysuria literally means pain on urination. Usually, it
tem design and implementation. I believe graduates of this refers to a burning sensation in the urethra that occurs whenprogram will be implementing broad-based changes in how ever a patient passes urine.
220
JOURNAL OF INSURANCE MEDICINE
VOLUME 22, No. 3 FALL 1990
When evaluating a man, this definition almost never produces clinics.3,4,1° Cell culture for chlamydiae is expensive1°,~L12 and
any confusion. But when evaluating a woman’s complaint of may not be cost-effective.
dysuria, this definition can be confusing, and this confusion is
What is the potential value of urinalysis in the management of acute
definitely not trivial. The confusion is this: Does the woman
feel this burning sen~. ation "inside" her pelvis, or "outside"? dysuria?
That is, does she feel the pain from the passage of the stream
Urinalysis traditionally has comprised testing of the urine
itself, or does she feel the pain in her vagina, when the stream with chemical reagents, and examination of urine under a
strikes the vaginal wall? This distinction may be, but is not microscope. Usually, the practitioner or laboratory technoloalways, helpful in determining whether the woman’s problem gist will centrifugate the urine, resuspend the sediment in a
is really a bladder or urethral infection at all, or whether her small amount of supernatant, and examine this concentrate
actual problem is a vaginal infection7, which requires a totally under the microscope. Sometimes, uncentrifugated urine is
different course of treatment. Bacterial or chlamydial infec- examined directly. Chemical tests, for the past twenty years or
tions of the urinary tract may both respond to the same antilonger, have been carried out with "dipsticks," which are
biotics, i.e., trimethoprim-sulfamethoxazole3’4, but mostplastic strips with chemically-treated pads mounted on them.
infections of the vagina are due to fungi and yeast, and these Dipsticks are, in general, easy to use and store; a dipstick
will not respond to trimethoprim-sulfamethoxazole.2
examination of urine can usually be completed within five
minutes or less.
What is the differential diagnosis of acute dysuria?
The differential diagnosis of acute dysuria includes acute and
subacute pyelonephritis (upper urinary tract infection) and
bacterial lower urinary tract infection. It also includes urethritis caused by Chlamydia trachomatis and occasionally by Neisseria gonorrheae, Trichomonas vaginalis, Candida albicans, and
herpes simplex virus. Dysuria may also be due to vaginitis.
Finally, a syndrome of dysuria with no known pathogen has
been described; this is usually due to noninfectious causes
such as trauma or post-menopausal desiccationY’2~
Stamm and colleagues6 in 1982 found Chlamydia trachomatis in
15 out of 63 women with bacteriologically sterile urine.
Komaroff and colleagues7 in 1978 found a 77% prevalence of
pure bacterial UTI and a 2% prevalence of vaginitis among
women who complained of dysuria but not of vaginal discharge. Among women who complained of dysuria with a
vaginal discharge, the prevalence of bacterial UTI was found
to be 9%, and of vaginitis, 61%.
How good are bacterial urine cultures?
Traditionally, the bacterial culture of mid-stream, clean-catch
urine has been considered the reference test for bacterial lower
urinary-tract infections. Kass, in 19628 first proposed 100,000
colony-forming units per milliliter (CFU/ml) as a criterion of
a positive culture, and this is currently the most common
criterion in use in modern microbiology laboratories. Kass was
attempting to promulgate a reference standard for the prevention or early detection of pyelonephritis, not lower-tract infection. Also, Kass set this criterion in an era when most catheters
were made of glass, not today’s flexible plastic, and thus
catheterized urine specimens were more difficult to obtain
safely than they are today. More recently, Stamm and colleagues6,9 have cast serious doubt on whether 100,000 CFU/ml
is a sensitive enough criterion for the detection of bacterial
urinary tract infection; Stamm now recommends 100 CFU/ml
in a catheter specimen or suprapubic aspirate as the new
bacterial culture criterion.
Another, more serious limitation of the bacterial culture is that
it cannot detect infections by Chlamydia trachomatis.
Chlamydiosis has become more prevalent in recent years in
men and women presenting to sexually-transmitted disease
In the diagnosis of acute dysuria, attention has focussed on
two recent additions to the multiple-reagent dipstick in common use today. One of these is based on the Greiss test for
nitrite in urine; nitrite is a metabolic by-product of Gram-negative bacteria, still the most common (but not the sole) agents
of bacterial urinary tract infection. The other is designed to
detect leukocyte esterase, which is found only in neutrophils.
The leukocyte esterase test pad has been shown to detect white
blood cells in urine with equal reliability whether the white
ceils have been lysed or not.
Based on multiple reported comparisons of the nitrite and
leukocyte esterase dipstick tests to the traditional bacterial
culture, a disjunctive combination of the two tests (calling the
dipstick positive if one or both tests are positive) is about 75%
sensitive and 80% specific, on average, in detecting bacteriuria
defined as 100,000 CFU/ml on culture. No data currently exist
on the ability of the dipstick to detect bacteriuria defined as
100 CFU/ml, and it is entirely possible that the dipstick could
be less sensitive (but more specific) in detecting bacteriuria
defined by this criterion.
What do the Experts Currently Recommend?
A thorough clinical history remains the comerstone of effective diagnosis of acute dysuria in women.7,~3 If a patient’s
symptoms are unambiguous for urinary tract infection (no
vaginal discharge, dysuria described as strictly "internal"),
then a pelvic examination may not be required7, and indeed
the patient may best be treated empirically with trimethoprimsulfamethoxazole, which is equally effective against coliform
bacteria and chlamydiae.3,4 If her symptoms are unambiguous
for vaginitis (vaginal discharge present; dysuria described as
strictly "external"), then a urinalysis need not be done, but a
pelvic examination may be required.7 Women with ambiguous symptoms and signs should probably have a pelvic examination and a urinalysis, and probably a urine culture.7 If a
culture is obtained, the best criterion for a positive bacterial
culture is probably 100 CFU/ml as recommended by Stamm,
rather than Kass’s 100,000 criterion.6’9 Whether dipstick tests
can replace the more expensive and time-consuming urinalysis, or the urine culture, is open to question.
221
VOLUME 22, NO. 3 FALL 1990
DIAGNOSTIC TESTING FOR ACUTE DYSURIA IN ADULT WOMEN: THE CURRENT STATE
REFERENCES
1. Johnson, J. R., and Stamm, W. E., "Urinary Tract Infections in Women:
Diagnosis and Treatment," Annals of Internal Medicine, 111:906-917,
1989.
2. Komaroff, A. L., "Acute Dysuria in Women," The New England Journal
of Medicine, 310(6):368-375,1984.
8. Kass, E. H., "Pyelonephritis and Bacteriuria: A Major Problem in
Preventive Medicine," Annals of Internal Medicine, 56(1):46-53, 1962.
9. Stamm, W. E., "Protocol for Diagnosis of Urinary Tract Infection:
Reconsidering the Criterion for Significant Bacteriuria," Urology, 32(2
Suppl)., 1988.
3. Brunham, R. C., Kuo, C-C, Stevens, C. E., and Holmes, K, K., "Treatment of Concomitant Neisseria gonorrheae and Chlamydia trachomatis
Infections in Women: Comparison of Trimethoprim-Sulfamethoxazole with Ampicillin-Probenecid," Reviews of Infectious Diseases, 4(2):491-499, 1982.
10. Fraiz, J., and Jones, R. B., "Chlamydial Infections," Annual Review of
Medicine, 39:357-70, 1988.
4. Hammerschlag, M. R., "Activity of Trimethoprim-Sulfamethoxazole
Against Chlamydia trachomatis in Vitro," Reviews of Infectious Diseases,
4(2): 500-505, 1982.
12. Nettleman, M. D., and Jones, R. B., "Cost-Effectiveness of Screening
Women at Moderate Risk for Genital Infections Caused by Chlamydia
trachomatis, Journal of the American Medical Association, 260(2):207-213,
5. Komaroff, A. L., ’q_lrinalysis and Urine Culture in Women with Dysuria," Annals of Internal Medicine, 104:212-218, 1986.
6. Stamm, W. E., Counts, G. W., Running, IC R., et al., "Diagnosis of
Coliform Infection in Acutely Dysuric Women," The New England
Journal of Medicine, 307(8):463-468, 1982.
7. Komaroff, A. L., Pass, T. M., McCue, J. D., et aL, "Management Strategies for Urinary and Vaginal Infections," Archives of InternaIMedicine,
138:1069-1073, 1978.
222
11. Nettleman, M. D., Jones, R. B., et al., "Cost-Effectiveness of Culturing
for Chlamydia trachomatis," Annals of Internal Medicine, 105:189-196,
1986.
1988.
13. Berg, A. O., Heidrich, E E., Fihn, S. E., et al., "Establishing the Cause
of Genitourinary Symptoms in Women in a Family Practice: Comparison of Clinical Examination and Comprehensive Microbiology," Journal of the American Medical Association, 251(5):620-625, 1984.