the week preceding delivery), or when the opportunity for preventing transmission has passed (e.g., long after rupture of membranes). In these situations, the potential benefits to the fetus may not outweigh the risk of potential maternal complications associated with cesarean section. Even when performed before rupture of membranes, cesarean section may not be completely effective in preventing neonatal infection. Randomized controlled trials to assess these issues may be difficult to perform for ethical reasons. Decision analysis models have shown that weekly cultures in pregnant women with recurrent herpes would, in a cohort of 3.6 million women, prevent 11.3 neonatal deaths and 3.7 cases of severe mental retardation, but 3.3 women would die as a result of operative deliveries necessitated by culture results.16 The costs of screening and cesarean sections for maternal HSV infections in this scenario were estimated at $61 million, or about $2 million for each case prevented.

References

1. Chuang TY, Su WPD, Perry HO, et al. Incidence and trend of herpes progenitals: a 15-year population study. Mayo Clinic Proc 1983; 58:436-41.

2. Guinan ME, Wolinsky SM, Reichman RC. Epidemiology of genital herpes simplex virus infection. Epidemiol Rev 1985; 7:127-46.

3. Corey L, Spear PG. Infections with herpes simplex viruses. N Engl J Med 1986; 314:749-57.

4. Corey L, Adams HG, Brown ZA, et al. Genital herpes simplex virus infections: clinical manifestations, course, and complications. Ann Intern Med 1983; 98:958-72.

5. Nahmias AJ, Keyserling HL, Kerrick GM. Herpes simplex. In: Remington JS, Klein JO, eds. Infectious diseases of the fetus and newborn infant. Philadelphia: WB Saunders, 1983:636-78.

6. Cates W Jr. The "other STDs": do they really matter? JAMA 1988; 259:3606-8.

7. Whitley RJ, Hutto C. Neonatal herpes simplex virus infections. Pediatr Rev 1985; 7:119-26.

8. Jenista JA. Perinatal herpes virus infections. Semin Perinatol 1983; 7:9-15.

9. Stagno S, Whitley RJ. Herpes virus infections of pregnancy. Part II. Herpes simplex virus and varicella-zoster virus infections. N Engl J Med 1985; 313:1327-30.

10. Brown ZA, Vontver LA, Benedetti J, et al. Genital herpes in pregnancy: risk factors associated with recurrences and asymptomatic viral shedding. Am J Obstet Gynecol 1985; 153:24-30.

11. Arvin AM, Hensleigh PA, Prober CG, et al. Failure of antepartum maternal cultures to predict the infant's risk of exposure to herpes simplex virus at delivery. N Engl J Med 1986; 315:796-800.

12. Prober SG, Sullender WM, Yasukawa LL, et al. Low risk of herpes simplex virus infections in neonates exposed to the virus at the time of vaginal delivery to mothers with recurrent genital herpes simplex virus infections. N Engl J Med 1987; 316:240-4.

13. Baker DA, Milch PO. Acyclovir for genital herpes simplex virus infections: a review. J Reprod Med 1986; 31:433-8.

14. Straus SE, Croen KD, Sawyer MH, et al. Acyclovir suppression of frequently recurring genital herpes: efficacy and diminishing need during successive years of treatment. JAMA 1988; 260:2227-30.

15. Grossman JH III, Wallen WC, Sever JL. Management of genital herpes simplex virus infection during pregnancy. Obstet Gynecol 1981; 58:1-4.

16. Binkin NJ, Kaplan JP, Cates W. Preventing neonatal herpes: the value of weekly viral cultures in pregnant women with recurrent genital herpes. JAMA 1984; 251:2816-21.

17. American College of Obstetricians and Gynecologists. Perinatal herpes simplex virus infections. ACOG Technical Bulletin No. 122. Washington, D.C.: American College of Obstetricians and Gynecologists, 1988.

18. American Academy of Pediatrics. Perinatal herpes simplex virus infection. Pediatrics 1980; 66:147-8.

Burden of Suffering: A number of disorders that cause bacteriuria, hematuria, and proteinuria are associated with significant morbidity and mortality. Asymptomatic bacteriuria often precedes symptomatic urinary tract infection, which accounts for over 6 million outpatient visits each year.' (There are over 300,000 hospitalizations each year for urinary tract infections,' but these generally involve patients with indwelling urethral catheters.) In adults, bacteriuria and urinary tract infection may be associated with renal insufficiency, hypertension, and increased mortality. In pregnant women, bacteriuria is a risk factor for prematurity and low birthweight.2 In children, about 13-17% of cases with recurrent bacteriuria develop chronic pyelonephritis, and 23-29% have evidence of vesicoureteral reflux.3 Children with significant structural abnormalities are at increased risk of renal scarring, obstructive atrophy, hypertension, and renal insufficiency.

The risk of acquiring bacteriuria varies with age and sex. Asymptomatic bacteriuria during infancy is more common in boys (prevalence of 2-4%), but pediatric bacteriuria is considerably more common in girls after age 1. Approximately 5-6% of girls have at least one episode of bacteriuria between first grade and their graduation from high school, and as many as 80% of these children experience recurrent infections. Bacteriuria occurs in 4-7% of pregnant women and 10-20% of diabetic women. The incidence of asymptomatic bacteriuria increases with age, and thus it is a common finding in the elderly, especially the very old and the institutionalized elderly.*

5,6

Hematuria and proteinuria are often the first detectable signs of urologic cancer and end-stage renal disease due to hypertension, diabetes, or glomerulonephritis. These diseases carry a significant burden of suffering. Kidney, bladder, and other urologic cancers will account for over 70,000 new cases and over 20,000 deaths in the United States in 1989.7 Renal dysfunction of any form can contribute to the morbidity and mortality of diseases such as hypertension, heart failure, and liver disease, and it can

limit the use of therapeutic drugs and contrast agents. About 92,000 Americans receive chronic dialysis for end-stage renal disease, while an additional 175,000 acute dialyses are performed yearly on hospitalized patients.' Kidney transplantation is necessary in about 8400 persons each year.

Efficacy of Screening Tests: Dipstick urinalysis is the most common test for detecting urinary tract disorders in asymptomatic persons. Multi-pad "dipstick" reagent strips can detect a variety of disorders, including bacteriuria (nitrite test), pyuria (leukocyte esterase test), hematuria (heme test), and proteinuria (tetrabromphenol test). Screening for bacteriuria may be impractical in infants, however, because positive dipstick tests are often contaminated (false positives) and require the collection of confirmatory sterile culture specimens by suprapubic aspiration. This procedure is too invasive and costly to be considered in a routine screening protocol for asymptomatic infants.

9-14

The most accurate test for bacteriuria is urine culture, but laboratory charges make this test too expensive for routine screening. Moreover, results are usually not available for at least 24 hours. The dipstick leukocyte esterase (LE) test, by detecting pyuria, is an indirect test for bacteriuria. When compared with culture (at least 100,000 organisms/mL), it has a sensitivity of 72-97% and a specificity of 64-82%. The nitrite reduction test, which detects bacteriuria, has variable sensitivity 9-13,15-23 (35-85%), but good specificity (92-100%). The sensitivity of this test can be improved by obtaining first-morning specimens, preferably on consecutive days, rather than by performing random collection. In detecting hematuria (more than two to five red blood cells per high-powered field in the sediment of centrifuged urine), dipstick urinalysis has a sensitivity of 91-100% and a specificity of 6599%;9 9,24-35 the sensitivity and specificity are 95-99% in detecting proteinuria." False-positive and falsenegative urinalysis results are due to a variety of factors, including specimen contamination, certain organisms, the timing of specimen collection, inter

15

fering substances (urobilinogen, glucose, ascorbic acid, drugs, urine cells and bacteria), other urine properties (specific gravity, pH, concentration), and biological factors (exercise, cold exposure, prolonged recumbency, medical illness). Examination of the sediment by microscopic urinalysis has limited value as a screening test for asymptomatic persons. Studies have reported a diagnostic yield of less than 3% in detecting clinically significant disorders by routine microscopic examination of urine that is grossly normal. 34,35

For

In populations with a low prevalence of urinary tract disorders, most positive urinalyses are falsely positive. Thus, in asymptomatic men and in asymptomatic women under age 60, a dipstick test for bacteriuria has a positive predictive value of less than 10% (assuming a sensitivity of 85% and a specificity of 70%).17,37-39 In groups at increased risk for urinary tract infection, the positive predictive value is higher: 13% in pregnant women, 18% in women over age 60, 33% in diabetic women, and 44% in the institutionalized elderly.5,6,17,37,39-44 similar reasons, the positive predictive value of a dipstick test for hematuria or proteinuria in the general population is 6-45% for disorders of possible clinical significance (e.g., asymptomatic bacteriuria, nonstaghorn calculi, mild glomerulonephritis) and less than 2% for serious urologic diseases (e.g., urogenital tumors, tuberculosis, staghorn calculi, vasculitis, nephritides, obstructive lesions). In older men, a population at increased risk, studies have found that only 4-5% of men develop cancer or other urologic diseases in the first three years after the test. A more recent study of men over age 50, however, reported a higher positive predictive value (26%), possibly due to the use of more sophisticated diagnostic follow-up studies to detect cancer. Due to the frequently intermittent nature of hematuria and proteinuria in persons with urologic cancer, a single screening test for these abnormalities may have less sensitivity in detecting cases than periodic testing, but further comparative data are needed.

45-50

32

48,49

Effectiveness of Early Detection: The early detection of urinary tract malignancy may improve survival, and the detection of hematuria or proteinuria may be the first indication of disease in asymptomatic persons.51 Survival from bladder cancer, for example, is directly related to the stage of the disease at diagnosis; the five-year survival rate is 72% for persons with localized disease but only 3% for those with advanced bladder cancer." However, lead-time and length biases may contribute to these

with untreated asymptomatic bacteriuria are at increased risk of developing symptomatic urinary tract infection 52,54 and other complications (e.g., structural damage, renal insufficiency, hypertension, mortality).38,40-44,55-57 There is little conclusive evidence, however, that these clinical outcomes are caused by bacteriuria (especially in the absence of a structural abnormality) or that they can be prevented by antibiotic therapy. The treatment of asymptomatic bacteriuria in the elderly, although associated with high recurrence rates in institutionalized patients, may be of benefit in the ambulatory setting. Two randomized controlled trials in elderly women have shown that treatment can reduce the incidence of subsequent bacteriuria (and possibly of symptomatic urinary tract infection). 53,54 It is not clear, however, whether this effect is of sufficient benefit to justify routine screening or the potential adverse effects of antibiotic therapy in the elderly, including drug toxicity and the development of resistant organisms while treating recurrent infections.

Recommendations against routine screening urinalyses in asymptomatic adults have been issued by the Canadian Task Force and other reviewers.67-70 Urinalysis screening in adults is generally considered more appropriate for persons at risk for bladder cancer (e.g., persons with heavy exposure to cigarette smoke and other bladder carcinogens); the Canadian Task Force recommends urine cytologic screening in such persons." The Canadian Task Force also recommends urine cultures during pregnancy. The American College of Obstetricians and Gynecologists recommends a urinalysis at each prenatal visit and urine culture for women with abnormal urinalysis or risk factors for urinary tract infection."

66

Discussion: Screening urinalysis appears to be especially important during pregnancy, where there is strong evidence that treatment is efficacious. There are, however, inadequate data to determine the optimal frequency of urine testing during pregnancy or whether prenatal testing should be carried out by urine culture (rather than by urinalysis) to

reduce the risk of false negatives. Screening for asymptomatic bacteriuria may also be beneficial in preschool children to help prevent permanent renal damage, but further studies are needed to establish its effectiveness. Screening urinalysis may be appropriate in high-risk groups, such as the elderly and persons with diabetes mellitus, but again firm evidence of benefit is limited. Screening is not justified in the general population because serious urinary tract disorders are relatively uncommon, the positive predictive value of screening urinalysis is low, and the effectiveness of early detection and treatment is unproved.

Clinical Intervention: Periodic testing for asymptomatic bacteriuria is recommended for persons with diabetes and for pregnant women, and it may also be clinically prudent in preschool children. The optimal frequency for urine testing in these groups has not been determined and is left to clinical discretion. The urine specimen should be obtained in a manner that minimizes contamination. Persons with abnormal results should receive further evaluation. In general, dipsticks combining the leukocyte esterase and nitrite tests should be used to detect asymptomatic bacteriuria. However, urine culture is a more accurate screening test than is dipstick urinalysis, and it is recommended for detecting asymptomatic bacteriuria during pregnancy.

Dipstick urinalysis for asymptomatic bacteriuria, hematuria, and proteinuria may also be clinically prudent in persons over age 60. Urinalysis is not recommended as a screening test to detect diabetes mellitus (see Chapter 16) or preeclampsia (Chapter 35) in asymptomatic persons.

Note: See Appendix A for the U.S. Preventive Services Task Force Table of Ratings for this topic. See also the relevant Task Force background papers: Woolhandler S, Pels RJ, Bor DH, et al. Screening asymptomatic adults for hematuria and proteinuria: dipstick urinalysis; and Pels RJ, Bor DH, Woolhandler S, et al. Screening asymptomatic adults for bacteriuria. In: Goldbloom RB, Lawrence RS, eds. Preventing disease: beyond the rhetoric. New York: Springer-Verlag (in press).

References

1. National Center for Health Statistics. Detailed diagnoses and procedures for patients discharged from short-stay hospitals: United States, 1985. Vital and Health Statistics, series 13, no. 90. Washington, D.C.: Government Printing Office, 1987. (Publication no. DHHS (PHS) 871751.)

2. Institute of Medicine, Division of Health Promotion and Disease Prevention. Preventing low birth weight. Washington, D.C.: National Academy Press, 1985.