Prevention of Infection in Pregnancy

We believe the prevention of infection-related adverse pregnancy outcome is the most important focus for obstetricians today. An emphasis upon immunization of susceptible women, prevention of transmissible disease by modification of patient behavior, and identification and treatment of silent infections should become standards of practice. This will require educational initiatives for physicians and their patients as well as continued clinical trials to determine costs and effectiveness.

It requires a change in practice priorities. As a specialty, we have focused upon operative skills plus the recognition and treatment of disease. Moving our priorities forward to prevention will be difficult because it represents an abrupt change in care. What are our standards today? In middle class America, the newly sexually active teenager switches her physician provider from the pediatrician to the obstetrician. Currently, the screening for this young woman is a Papanicolaou (Pap) smear and prevention is limited to a prescription for an oral contraceptive. This protects against an unwanted pregnancy, but offers no protection against sexually transmitted diseases. For urban poor women, no preventive care is available. The focus is on pathology. We treat urban poor teenagers with an unwanted pregnancy or symptomatic pelvic inflammatory disease. This is a narrow and unacceptable scope of practice.
An outline for infection-related adverse pregnancy outcomes has been the mnemonic TORCH, in which the significant pathogens for the fetus have been identified. We would like to expand this to STORCH s for the purposes of this discussion of prevention (see Table 1). The focus will be directed toward three strategies of prevention: vaccine administration, preventive counseling to achieve maternal behavior modification, and treatment of silent infections. All of these strategies require the use of screening laboratory studies to determine the risks in asymptomatic patients. This requires, on the part of the obstetrician, knowledge and understanding of laboratory studies, many of which are new. VACCINES An awareness of immunization standards is important for the obstetrician. In the United States, we are the primary care physician for most women during their childbearing years. We need to know the needs of our patients and the status of vaccines, i.e., are they live virus preparations or not. For the pregnant woman, live virus vaccines should be avoided.  (65.5C); meat cooked until no longer pink inside generally has an internal temperature of 165F (73.8C).
Patients should be advised to wash their hands after contact with raw meat and after gardening or contact with soil; in addition, they should wash fruit and vegetables well before eating them raw.
If the patient owns a cat, someone else should change the litter box daily; alternatively, the patient should wash her hands thoroughly after changing the litter box. Patients should be encouraged to keep their cats inside and not to adopt or handle stray cats. Frequent hand washing should be done, particularly after changing diapers. If a blood transfusion is needed, these patients should receive only CMV antibody-negative blood or leukocyte-reduced cellular blood products in non-emergency situations.
against toxoplasmosis screening are based upon cost effectiveness. The patient at risk is the susceptible woman who acquires the infection during pregnancy. Although acquisition of toxoplasmosis during pregnancy is uncommon, approximately 90% of those who get infected during pregnancy have no symptoms and thus are unaware of the infection, le This makes a strong argument for universal screening of pregnant patients. Suggestions that most mothers who become infected will seek medical care 9 fly in the face of reported reality, lz This is an uncommon infection. One study of pregnant women in New York City reported 6 toxoplasmosis infections among 4,048 pregnant women (1.5/1,000). 13 Not every newborn exposed to this pathogen will have persistent infection at birth. In a study in Massachusetts, only 52 of 635,000 infants delivered had congenital infection. 14 That is less than in 10,000 births. However, this is a serious infection for the newborn, it is treatable, and only 2 of the 52 infected infants were detected on clinical grounds. Without universal screening, 50 would have been denied early treatment. Despite the infrequency, we believe toxoplasmosis screening is indicated. For susceptible women, a series of preventive guidelines can be given (see Table 2). s A similar preventive program in Belgium has significantly reduced maternal acquisition of toxoplasmosis during pregnancy. 1

Cytomegalovirus (CMV)
CMV presents a different set of circumstances. It is a much more common infection and most authorities think it is the most frequent pathogen involved in newborn morbidity. A recent publication from the Centers for Disease Control (CDC) estimated that 30,000-40,000 infants are infected each year with CMV and 9,000 of these infants suffer permanent sequelae. 6 Other evaluations indicate that the most frequent and serious newborn infections occur in susceptible women who acquire their first infection during pregnancy. 17 The presence of antibody offers protection to the newborn. American obstetricians have not offered screening for CMV because no effective antiviral therapy has surfaced. This is a short-sighted approach. Antibody screening should be done in pregnant women to determine susceptibility. If susceptible, the pregnant women should have three options presented to her (see Table 3). s For this virus, there are a wide variety of risks. CMV is a sexually transmitted disease. A sexually active pregnant woman who is not in a monogamous relationship should require her male sexual partner to wear a condom. This requires attitudinal changes by both the obstetrician and the pregnant woman. The condom in this situation is not to prevent pregnancy, but to lower the risk of acquiring a sexually transmitted disease. Small children in a group setting such as a day care center are frequently infected with CMV and are asymptomatic. 18 They shed large amounts of virus their saliva and urine. Susceptible women should avoid "sloppy" kisses with these children and should wash their hands if they have to change a diaper. Finally, for the rare occasions when a susceptible pregnant woman needs a blood transfusion, the donated blood should be CMV antibody negative. If this is not possible, leukocyte depleted blood should be given to lower the risk of CMV acquisition. 19 None of these preventive measures should be difficult to institute. The starting point in this preventive strategy is antibody screening to determine susceptibility.

HIV
HIV is an increasingly important problem worldwide for obstetricians because of its frequency and serious long-term implications. In the United States, the number of women infected by heterosexual contact continues to increase and the result is a continuing increase in the number of women developing acquired immunodeficiency syndrome (AIDS). 2 Screening to determine susceptibility is important. If these susceptible women are intravenous drug users, they must be counseled to avoid needle sharing. If they are not in a monogamous sexual relationship or the HIV antibody status .of their partner is not known, they should be counseled to require condom use. Again, this is for disease prevention, not contraception.

Herpes
This is a developing new story for obstetricians. In the past decade, a number of studies have documented the fact that herpes simplex virus (HSV) acquisition can occur without symptomatology. 21 '22 These people can shed virus without symptoms and infect susceptible sexual partners. For the pregnant woman, the greatest risk of newborn infection with herpes occurs when the infection is first acquired shortly before the onset of labor. A recent study indicates that this acquisition can be best documented by repeated HSV antibody screening during pregnancy. 23 In this study, only 36% of the women who seroconverted had symptoms consistent with herpes infection. This is a higher incidence of symptoms with virus acquisition than is seen in the non-pregnant adult population, z There are interesting future preventive strategies for susceptible women. The male of the herpes susceptible pregnant patient can be tested. If the male is HSV-1 or HSV-2 antibody positive, the patient is at risk for herpes acquisition during pregnancy. Abstinence or the use of a condom during the last trimester may lower the risk for such a patient, z3 Future studies will determine the benefits of such a strategy. Listeria monocgtogenes L. monocytogenes is an uncommon cause of newborn infection, but maternal infection can result in pre-term labor and delivery with newborn sepsis and death, z4 Most outbreaks are associated with maternal ingestion of contaminated dairy products. To avoid this, we advise pregnant women in New York City to avoid the ingestion of all cheeses that have not been labeled pasteurized. This includes imported cheeses with a rind as well as feta cheese.

SILENT INFECTION
Screening and treating silent infections are relatively new concepts for the obstetrician. In the past, we have focused our efforts upon the recognition of the signs and symptoms of a clinical infection, confirming our clinical suspicions with specific laboratory testing. It is becoming increasingly clear that many maternal infections that can cause problems for the mother or fetus are either asymptomatic or cause such minimal symptoms that the mother to be will not seek medical care. This requires physicians to shift their concerns to screen asymptomatic women to detect these infections followed by a treatment strategy. There are a number of screens that should be regularly employed.

Asymptomatic Bacteriuria
This is a familiar screen for asymptomatic disease by the obstetrician. A clean voided urine is obtained at the first prenatal visit and those women with a colony count of 100,000 or greater are treated. There is good evidence that this strategy reduces the subsequent incidence of pyelonephritis. zs This has been an effective preventive strategy for obstetricians.

Bacterial Vaginosis (BV)
This is the current infectious disease "hot item" for obstetricians. BV is manifested by a tremendous increase in the number of bacteria in the vagina and many women so detected are asymptomatic.
More important, the Vaginal Infections in Pregnancy (VIP) study showed an association with BV and the delivery of a low birth weight infant, z6 An optimistic sign is that treatment studies with both systemic clindamycin 27 and metronidazole 28 have been effective in reducing the rates of preterm births. Vaginal antibiotic preparations which are effective in the treatment of BV have not resulted in a lowering of the rate of prematurity, z9, Armed with this information, a system of screening and treatment of those who test positive has been devised.
The diagnosis of BV can be made within minutes of the obstetrician-patient contact. Every time a vaginal examination is done on a pregnant woman, the following screening tests should be done. The vaginal pH should be tested to see if it is alkaline and a drop of vaginal secretion placed on a drop of dilute potassium hydroxide on a slide and checked immediately to see if a "fishy" odor emanates. If either of these tests is positive, another vaginal secretion sample should be placed on a drop of normal saline on a slide and examined under a microscope. If over 20% of the epithelial cells are "clue" cells, the diagnosis is confirmed. The treatment should be with oral metronidazole z8 or oral clindamycin z7 for at least 7 days. There is one study from Germany suggesting that va[ginal clindamycin treatment given before 14 weeks gestation lowers the incidence of preterm births in the treated population (Dennemark, personal c.ommunication). Further studies on this early pregnancy population will be needed to confirm this observation. If so, it would simplify the care of women with BV in the first trimester.

Trichomonas vaginalis
There is increasing interest in the asymptomatic pregnant patient with T. vaginalis vaginitis. In the VIP study, the finding of T. vaginalis by a culture technique was associated with a higher incidence of preterm labor than was seen with BV. 31 Another recent study using the polymerase chain reaction (PCR), a more sensitive test than the culture technique used in the VIP study, found a high incidence of vaginal Trichomonas infection in an urban poor population with a high incidence of prematurity. 3z Further information about the risk of Trichomonas for the pregnant patient requires a dual strategy of physician education and prospective study. Physicians must learn that asymptomatic vaginal Trichomonas exists and the microscopic examination of vaginal secretions in saline is not a sensitive screening test. 33 Universal screening needs to be done on pregnant populations, comparing the pregnancy outcomes in PCR and culture-positive patients. Finally, this infected population needs to be treated with metronidazole to determine if the preterm delivery rate is decreased.

Neisseria gonorrhoege
Pregnant patients can be culture positive for N. gonorrhoeae without enough symptomatology to cause them to seek medical attention. If untreated, these patients have a higher incidence of preterm labor and delivery. 34 Part of pregnancy care should involve screening for N. gonorrhoeae. This can be done with a DNA probe for the incidence if penicillin resistance is high enough that cephalosporins are now recommended. Treatment with a single dose of ceftriaxone 250 mg 1M should be adequate. s Antibiotic treatment of the pregnant woman is also important, because an infected mother can infect her newborn's eyes through the process of labor and delivery.

Chlamydia trachomatis
C. trachomatis is a good example of a bacterial pathogen that can cause serious problems for both the mother and the fetus. Women infected with C. trachomatis have no symptoms or minimal symptomatology that is not severe enough for them to seek medical care. These symptom-free women, if infected, have the potential for a series of problems for the pregnant patient and the fetus. Pregnant women with C. trachomatis antibodies, evidence of a past infection, have a higher than expected incidence of spontaneous abortion. 6 Among premature newborns delivered of women colonized with C. trachomatis, there is a risk of developing a C. trachomatis pneumonia. 7 The newborns of culturepositive women, both premature and at term, have the potential for developing ophthalmia neonatorum. These serious problems and lack of patient symptomatology demand universal screening of pregnant women. PCR should be used because it is so much more sensitive and specific than the DNA probes. s PCR-positive women and their sexual partners should be treated. The CDC recommends 7 days of oral erythromycin therapy for the pregnant woman, for it is safe during pregnancy. s The problem is that erythromycin is poorly tolerated by most pregnant women. They have enough gastrointestinal distress that many of these asymptomatic women do not complete their course of treatment.
For this reason, we favor the use of amoxicillin in those pregnant women who are not allergic to penicillin. This antibiotic is better tolerated by the pregnant recipient and it is effective in eliminating the organism. 39 In addition to screening at the time of the first prenatal visit, we favor another PCR screen at 35-37 weeks gestation. Term infants born of infected mothers can develop ophthalmia neonatorum from the C. trachomatis and the antibiotic ophthalmic ointments used in the newborn are not very effective in curing the infection. 4 These culture-positive women should be treated prior to delivery.

Group B Streptococcus
The group B streptococcus is a feared organism by obstetricians. Vaginal colonization with this organism which causes no maternal symptomatology can result in vertical transmission to the fetus with newborn sepsis and death despite newborn antibiotic treatment. 41 There is evidence that treatment of the mother with antibiotics prior to delivery markedly reduces newborn infectious morbidity and mortality. 42 The problem with any strategy of antibiotic prophylaxis is the selection of the mother at risk for newborn group B streptococcal infection.
Much is known about maternal colonization and newborn infections with the group B streptococcus. Vaginal colonization with the group B streptococcus is not a constant. Women colonized in the first, second, or third trimester may not be culture positive at the time of delivery. Alternatively, women culture negative at the first, second, or third trimester vaginal culture screening may be culture positive at the time of delivery. In one study, only 51% of women culture positive in the first trimester remained culture positive at the time of delivery. 43 This raises natural concerns about preventive strategies using antepartum cultures. Some women who are culture positive with an antepartum screen will be negative at the time of birth and some women who are culture positive at the time of delivery will not be detected at the time of the antepartum screen. Rapid screening for the group B streptococcus would obviate this problem, but the systems studied to date have not been sensitive or specific. 44 This is not an acceptable option at present. Another problem is the discrepancy between the incidence of maternal colonization and newborn infection. This varies between 1:204s and 1:102, 46 with the highest incidence of newborn problems seen in those women who deliver premature infants. This means many mothers will be treated to prevent one case of newborn sepsis. No one rec-LEDGER AND SKUPSKI ommends treating antepartum culture-positive patients with antibiotics, even though there is an association with first trimester group B streptococcus colonization and preterm premature rupture of the membranes and preterm delivery. 47 There are two major reasons for this. This antibiotic intervention has not proved successful in lowering the premature delivery rate 48 and systemic antibiotic treatment has not been effective in the persistent eradication of the group B streptococcus from the vagina of these women. 49 Finally, there is the question of the selection of pregnant women to receive antibiotic prophylaxis in labor. The CDC has suggested two strategies for the patient who is not penicillin allergic; they recommend penicillin G with ampicillin as an alternative, s There is controversy surrounding both of these choices. Penicillin G has a narrow range of antibacterial activity. For the obstetrician concerned about other organisms which cause newborn infection with some frequency such as Escherichia coli, Hemophilus influenzae, Listeria monocytogenes, Ureaplasma urealyticum, and C. trachomatis are not covered. On the positive side, widespread use of this narrow spectrum antibiotic should not have a major impact on the bacterial flora found on a labor and delivery floor. Many obstetricians favor ampicillin, because its wider spectrum of activity means that some strains of E. co/i, H. influenzae, and L. monocytogenes will be covered as well as the group B streptococcus. This widespread use of a broader spectrum antibiotic has a cost. There will be a greater impact upon the antibiotic flora of an obstetrical service and there already has been one report of an increase in the number of gram-negative aerobic infections in the newborn when this ampicillin prophylaxis was employed, sl There are important considerations to be kept in mind when a preventive regimen is proposed. All antibiotic intervention regimens have the capability of lowering the incidence of group B streptococcal disease, s2 All are different in their technique of selecting mothers at risk (see Table 4) and none is 100% successful in preventing group B streptococcal newborn disease, s2 Intrapartum antibiotic administration to the mother will not prevent all cases. There have been instances in which mothers given antibiotics when they were admitted in labor still have delivered infants with group B sepsis and death due to this infection, s's2 All of this is prologue to the current CDC rec-   We have two major reservations about the CDC recommendations. They enforce a practice pattern that may not pass the test of time. I have no doubt these recommendations will increase the use of intrapartum antibiotics in the United States and decrease the number of group B streptococcal newborn infections in the short run. Over the long term, this strategy increases the risk of altering bacterial flora causing newborn disease. Surveillance of the organisms causing newborn sepsis will be necessary to detect this. These guidelines will also limit the effectiveness of prospective studies trying to pinpoint mothers at risk for having babies with newborn disease and it will impede the study of alternative approaches. For example, some obstetrical services in Europe use an antiseptic vaginal INFECTIOUS  If membrane ruptured at < 37 weeks' gestation, and the mother has not begun labor, collect group B streptococcal culture and either a) administer antibiotics until cultures are completed and the results are negative or b) begin antibiotics only when positive cultures are available. No prophylaxis is needed if culture obtained at 35-37 weeks' gestation was negative.
t Broader spectrum antibiotics may be considered at the physician's discretion, based on clinical indications. E. coli E. coli is another pathogen of concern for obstetricians in their quest to control newborn infections.
It is a frequent cause of newborn infection 5s and on most services ranks next to the group B streptococcus in frequency. There are some aspects of E. coli newborn infection that lend themselves to strate-gies of control. Newborn infection results from vertical transmission from the mother. Unlike the group B streptococcus, maternal vaginal colonization with E. coli seems to be a constant, s6 This suggests that antepartum vaginal screening for this organism could identify those women at risk. In addition, specific strains of E. coli seem to have more invasive properties and to be more dangerous for newborns, s7 Again, identification screening techniques could be employed to delineate women at risk for infection. Intrapartum treatment should be effective, but many E. coli strains are resistant to Are any of the following risk factors present? Previously delivered infant who had invasive GBS disease GBS bacteria during this pregnancy Delivery at <37 weeks' gestation Duration of ruptured membranes > 18 hours Intrapartum temperature > 100.4 F (> 38.0 C)

No intrapartum prophylaxis needed
Give intrapartum penicillint If membrane ruptured at <37 weeks' gestation, and the mother has not begun labor, collect group B streptococcal culture and either a) administer antibiotics until cultures are completed and the results are negative or b) begin antibiotics only when positive cultures are available.
t Broader spectrum antibiotics may be considered at the physician's discretion, based on clinical indications. ampicillin. Prospective studies need to be done to determine the effectiveness of cephalosporins or ampicillin with a beta lactamase inhibitor added.
H. influenzae H. inf/uenzae is another important cause of newborn infection, ss A current therapeutic problem is that many strains of H. influenzae are beta lactamase producers and as such are resistant to ampicillin. Alternative non-beta lactam antibiotics would be better first-line agents in hospitals where ampicillin resistant H. influenzae is a frequent finding. U. urealyticum U. urea/yticum remains an enigma. It is associated with chronic lung disease in the premature newborn. s9 The difficulty is that the majority of pregnant women are colonized with this organism 6 so the problem of determining which newborn of these colonized women will be at risk remains unsolved. This organism has the potential to cause problems. It can occasionally be isolated from the amniotic fluid of women in premature labor. 61 In evaluating the pregnancy history of newborns who become infected, the route of delivery did not alter the risk of infection, but chorioamnionitis did. 6z The antibiotic of choice in the past has been erythromycin, but administration of this drug to pregnant women was ineffective in reducing the number of women who remained vaginal culture positive for this organism. 6 Clearly, more studies will be needed to delineate risk factors.