The prevention of early-onset group B streptococcal infections in the newborn.

GROUP B STREPTOCOCCI (GBS) ARE A MAJOR CAUSE OF morbidity and mortality among newborn infants. Various strategies to prevent infection in the newborn were reviewed by the Infectious Diseases and Immunization Committee and the Fetus and Newborn Committee of the Canadian Paediatric Society and by the Maternal/Fetal Medicine Committee of the Society of Obstetricians and Gynaecologists of Canada. Th e committees identified an u rgen t need for addit ional re-

BACKGROUND GBS (Streptococcus agalacLiae) continue to be a major cause of bacterial sepsis among newborns. The source of infection in the neonate is the colonized maternal birth canal; transmission occurs before or d u ring the birth process. Estimates of GBS colonization rates among pregnant wom en range from 15 to 40% (l-4) . GBS is transmitted to 40 to 70% of newbor ns of colonized moth ers; however, only l to 2% of s u ch infants develop disease (5). Two types of CBS infection occur in the newborn. Early-onset disease (less than seven days of age) i more common and has a higher rate of mortality. and it is the subject of lhis statement. Late-onset disea e (seven days to three months of age) is less common and has a lower associated mortality (6-8) . Factors that have been associated with increased risk of early-onset CBS disease include: maternal age less than 20 years: low socioeconomic status: gestation less than 37 weeks; prolonged rupture of membranes (greater lhan 12 h); maternal intrapartum pyrexia (temperature greater than 37.5°C); multiple pregnancy; CBS bacteriuria: and degree of maternal genital colonization (heavy versus light) (5.9.10). Colonization rates are relatively constant during pregnancy. Fewer than 10% of women who are culture-negative !ale in the second trimester are cullure-positive at delivery (1.2).

INTERVENTION STRATEGIES
The high rates of morbidity and mortality associated with CBS infections and the rapidity of onset of postnatal disease have led to numerous strategies to prevent infection in the newborn . They include antenatal. intrapartum and postnatal ch emoprophylaxis. and active and passive immunoprophylaxis. Controversy persi ts about the optimal strategy because of insufficient data or differing interpretations of existing data. Immunoprophylaxis -Active: Serum antibody to CBS capsular polysaccharide correlate with protection against infection. Antibody levels lo CBS in infected infants and their mothers are lower than in uninfected maternal-newborn pairs (11). This has led to efforts to develop CBS capsular polysaccharide vaccines for the active immunization of pregnant women to prevent early-onset and late-onset disease in the newborn ( 12). As with other bacterial vaccines. a substantial number of individuals do not respond to polysaccharide antigens either because of genetic inOuences such as immunoglobulin allotype or becau e of the T cell-independent nature of the antigen (13). Lack of significant placental transfer of immunoglobulin (Ig) G before 32 weeks of gestation also diminishes lhe usefulness oflhis intervention for protection of the very premature infant. However. efforts to develop an immunogenic vaccine that can be administered to all women. or during pregnancy. continue to be made. Although active immunization may be the ultimate solution . no vaccine is likely to be available in the foreseeable future. Passive: The use of intravenous immunoglobulin (lg-IV) ha been proposed to increase antibody titres lo CBS in newborns. Clinical studies using passive immunoprophylaxis have been disappointing (14). Variability in CBS-specific antibody in Ig-IV preparations may account for these poor results (15.16). Although recent efforts to produce a CBS hyperimmune lg-IV preparation may make passive immunization a viable intervention in the future (17). at present. routine use of lg-IV cannot be 252 recommended for prevention or therapy of CBS infection in newborns (18) . Chemoprophylaxis: Several trategies have been employed using antibiotics for the prevention of CBS infections in th e newborn. These can be categorized as neonatal. antepartu m and intrapartum. However. no trategies preven t a ll ca es of early-onset CBS disease. Fu lminant CBS sepsis and neonatal demise can occur in pile of any planned intervention. Neonatal: Although penicillin 01-ampicillin U1erapy of lhe neonate may prevent some cases of CBS in term infants ( 19). more than 60% of infants are already symptomatic at. or shortly after, birth and are at hi<rh risk of cBs complications and poor outcome (7.9.20,21).
Chemoprophylaxis of neonates appears to be an inferior strategy for prevention of CBS infections.
Antep artum: The efficacy of antibiotic therapy of CBScolonized women during pregnancy for preventing neonatal infection has not been established. Studies using newborn infection as an outcome measure have been of insufficient size to demonstrate efficacy (22)(23)(24). Studies in which maternal CBS colonization was fo llowed after a course of antibiotic therapy have demonstrated high rates of CBS recurrence (67%) by lhe time of delivery. This was true even when antibiotics were given to both pregnant women and lheir sex.rual partners (7,25). An exception to the lack of benefit of antepartum antibiotics is women with CBS bacteriuria, which increases the risk for premature rupture of membranes and preterm labour. Oral antibiotic therapy resolves CBS bacteriuria and decreases lhe risk of preterm labour. although genital colonization typically persists (2.10). Intrapartum: Most experts agree that intrapartum maternal chemoprophylaxis is lhe best strategy available to prevent early-onset neonatal CBS disease (7,8,21,26). Several studies have assessed the benefit of intrapartum chemoprophylaxis (3.27-31): a meta-analysis of selected randomized controlled trials (5, [29][30][31] showed benefit of intrapar tum prophylaxis (32). The strategies used by the studies differ primarily in the methods used to select women for prophylaxis (including whether and when to screen pregnant women for colonization with CBS). laboratory methods lo detect CBS colonization. and selective or nonselective chemoprophylaxis. Timing of GBS screening: Anorectal and vaginal swabbing for CBS culture early in the lhird trimester (26 to 28 weeks) permits adequate lime to identifY CBS colonization. Problems with this approach inclu de late acquisition of colonization and lack of prenatal care. A single cullure has a predictive value of CBS colonization at delivery of only 67% (27). Screening late in the third trimester (36 weeks) decreases lhc likelihood of missing late acquisition of CBS but misses women with premature labour, and therefore, neonates who are at highest risk of CBS complications. A strategy that screens women presenting in premature labour (less than 37 weeks) and that treats CBS carriers or those women in whom Lh e lest result is unavaila ble would result in the prophylaxis of as many as 7.5% of pregn a nt women (33). b u t it does not address term infants , w ho account for over two-thirds of early-onset GBS disease. Intrapartum screening s uffers from the lack of ra pid , sensitive la b o ratory tests, which wou ld provide th e opportunity to make decis ions about chemoprophylaxis with sufficient li me before delivery. Labor atory tests: Ch emoprophylaxis strategies th at involve identificatio n of CBS-colonized women requi re tests with h igh sen s itivity and specificity. Th e ra te of isolation of GBS is improved through culture ofth e lower vagina and anorectum and through use of elective media for transpor t and cu lture ( 1. 7). Bacteri al i ola li on is time-consuming, and resul ts a re nol rapid ly available for the wom an pres enting in la b ou r . Op tima l u sc of ch em oproph ylaxis requires a s ufficient interval (a t leas t 4 h) between beginning antibio ti cs and d elivery (34) . Antigen detection tests provide res ults m ore quickly (5 h or less) but wi th loss of sen s itivi ty rela tive to cu lture (29,(35)(36)(37)(38)(39). The shortcomin<>"s of th e la b oratory tests have led to strategies fo r selec ti ve ch emoprophylaxis based only on epidemiological risk factors withou t la bora tory sampling (2 7 ,40) . Selective versus n onselective prophylaxis : Identification of ri s k fac tors th at increase th e likelihood of GBS infection in th e n ewborn of a colon ized woman h as led to strategies fo r u se of prophylaxis in h igh ti sk s itua tions such as pre term la b our (less tha n 37 weeks ). prolonged rupture of mem branes (greater than 12 h) and intrapar tu m m ate rnal fever. This strategy has been s hown to be efficaciou s (21,4 1,42) b u t is estimated to fail to prevent 25 to 30% of cases of early-on et GBS infection  crease lh e num ber of undetected cases (except for th e late pren atal care) but would dra matically increas e the u se of intrap a rt um antibiotics. with the associa ted risks of adverse reactions a nd , dep end ing on the coloniza tion rate of th e pop u la tion , poss ibly th e loss of cost-effectiven ess. Alth ough identification of women wi th h eavy GBS colon ization predic ted a high er ris k of n ewborn GBS infection . 16% of infection s occurred in newborns of lightly colonized wom en (28). s u ggesting th a t this s tra tifica tion method was of limi ted benefit. More s imply. inc reased sen s itivity might be achieved by expandin g th e list of ris k factors th at wo uld be indications for chemoprophylaxis (ie. GBS bacteri u ria. multiple births in prelerm la b our) . Cost-effectiveness: Seve r a l s t rategi es h ave b ee n s h own to be cost-effective in the United S lates. including screening a t 26 to 28 weeks and selective intra partum c h emoprophylaxis (43). in tra pa r tu m screening a nd ch e moprophylaxis of all CBS-colon ized wo men regardless of ris k factors (44) and in tra pa rtum prophylaxis based on ris k factors only (43). Th e clinical outcom es and costs of 19 potential strategies wer e recently compared u s ing d ecis ion ana lysis (45). Universal intra partu m ch em oprophylaxis. intrapartum ch em oprophylaxis based on ris k factors. and unive rsal screening a t 36 wee ks ' gesta ti on plus intra pa rtum c hem oprop hylaxis of a ll GBS culture-positive women and women in pretenn lab our were identified as tl1e optin1al s tra tegies. However, most of the s tra tegies modelled are theoretical and n eed to be studied in practice before in1plemen tation . None of the cost analyses addressed the s ubstantial implications or costs associa ted with the investigation and managemen t of n eon a tes wh ose mothers received intra partum ch emoprophylaxis, especially newborns without GBS disease. Managemen t of in fant s of mot her s who received intrapartum chemoprophylaxis: No studies h ave been performed ass essing th e optimal managem en t of newb orn s whose m other s received intra partum chem oprophylaxis; therefore, recommenda tions are n ecessarily empirical (4 6) . Management s hould tak e into account gestational age and clinical evalu ation.

RECOMMENDATIONS
Althou gh intra partum chemoprophylaxis in h igh ri s k s itua tions reduces the morbidity and mortality due to early onse t of GBS infection. no method prevents all cas d eaths. De tailed review of this informa tion indica tes lha t s pecific Cana dian da ta are required to determine the b es t stra tegy for che moprophylaxis in the prevention of early-onset n eona ta l GBS infection (Table  1) and to test various me thods of id entifying women whose n eonates m ay b e at risk of GBS disease . Two widely recommended strategies a re universal screening co upled with selective intrapa rtum che moprophylaxis of colonized women at high risk or intrapartum ch emoprophylaxis of all women at high risk. Currently a vaila ble da ta fa vour the first approach; however , the cost. implications of universal screening are considera ble.
Uniil the resul ts of Canadian studies are a vaila ble, th e Infectious Diseas es and Immunization Committee a nd the Fetus and Newborn Committee of the Canadian Paedia tric S ocie ty, and lhe Maternal/Fetal Medicine Committee of the Society of Obste tricians and Gynaecologists of Canada recommend th e following: requisition. transport. testin~ and rcporlln~ of the~e ::.pecimcns.

4.
Antepartum oral antibiotic therapy s hould be used for women with GI3S bacteriuria. These women s hould still be considered to be cus-colonizecl at the time of going into la bour. They do nol need to be rescreened once they have been ide nllfied.

5.
U~l' Of intrapartum Chem oprophylaxis Of WOmen who previously have given birth to a n infant with GBS disease regardless q.f ems colonization status.
6 . The a ntibiotic· regimen of c h oice for intrapartum c-hemoprophyta. ..... is is intravenous a mpicillin (2 g initia lly followed by I lo 2 g every 4 lo 6 h) or penicillin G (5 million U every 6 h) until delivery or until labour is ~lopped. Women who arc a llergic to penicillin may be t?;ivcn intravenous clinclamycm (300 to 600 mg every 8 h until delivery). 7. Women with a culture positive for cus at 26 to 28 weel<s· gestation do not reqUire treatment unless th ey become high risk and are in labour.

8.
Mamt~emenf of asymptomatic inl~tnls of mothers who received intrapartum c-h emoprophylaxis s hould bl' based on the infant'~ ~estationa l age. results of investigations for sepsis and the adeq uacy of the ma ternal intrapartum ehemo prophylaxis (Fi~ure I). Newbo rns with clinical si~ns of sepsis should be investi~ated a nd treated with antibiotics to c-over GllS and other organisms regardlc~s of maternal GllS colonization or chemo prophyi:.Lxis. l:kC'ause of the conc·em U1at mtrapartum antibiotic::. could interfere with the ability to obtain a positivt• blood culture. symptomatic-infants in whom blood c ultures a rc negative should receive a full course or antibioti('S. the duration of whiC'h will depend on lhe C'linical pic·ture.