OBJECTIVE: To estimate the cost and consequences of intrapartum polymerase chain reaction (PCR) screening on early-onset group B streptococcal (GBS) disease compared with the antenatal lower vagina culture screening recommended in France.
METHODS: This was a single-institution study comparing the intrapartum PCR screening strategy implemented in 2010 with antenatal culture strategy in place in 2009. Early-onset GBS disease in newborns was monitored exhaustively. We estimated direct costs, including screening test costs and hospital costs, for deliveries of healthy newborns compared with those infected with GBS. Costs in 2009 and 2010 were compared on an intention-to-treat basis.
RESULTS: Term deliveries were 2,761 and 2,814 in 2009 and 2010, respectively. Among the screened mothers, the vaginal GBS colonization rate was 11.7% based on antenatal GBS culture screening in 2009 compared with 16.7% in 2010 using the intrapartum PCR testing. The overall probabilities of neonatal GBS disease were 0.9% compared with 0.5%, and the average total cost per delivery was $1,759±1,209 in 2009 compared with $1,754±842 in 2010 (P=.9) in antenatal and intrapartum screening strategies, respectively. The number and severity of cases of early-onset GBS disease and the resulting hospital costs were higher in 2009.
CONCLUSION: Polymerase chain reaction intrapartum screening strategy was cost-neutral when compared with the 2009 antenatal lower vagina culture screening, with a significant decrease in early-onset GBS disease.
Intrapartum group B streptococcus polymerase chain reaction screening compared with antenatal lower vagina culture screening used in a French hospital, is cost-neutral with a significant decrease in neonatal group B streptococcal infections.
From the Groupe Hospitalier Paris-Saint-Joseph, Service de Microbiologie and Service de Gynécologie-Obstétrique, the AP-HP Groupe Hospitalier Henri Mondor Albert Chenevier, Service de Santé Publique, and the URC Eco Ile-de-France (AP-HP), Paris, France.
The authors thank Karen Brigham and Celine Quelen for the bootstrap analysis and Gilles Rejasse and Maryse Bobin for technical advice in collecting the data.
Corresponding author: Najoua El Helali, Clinical Microbiologist, Service de Microbiologie, Groupe Hospitalier Paris-Saint-Joseph, 185 rue Raymond Losserand, 75674 Paris Cedex 14, France; e-mail: firstname.lastname@example.org.
Early-onset group B streptococcal (GBS) disease in newborns is acquired from the colonized maternal birth canal during labor or after membrane rupture. Clinical trials showed that intrapartum antibiotic prophylaxis administered to GBS–colonized mothers was highly effective in preventing early-onset GBS disease in newborns.1 Since then, the main challenge has been to identify which women should receive intrapartum antibiotic phophylaxis. Until recently, a rapid, sensitive, and specific GBS detection test was not available and intrapartum screening was not feasible. Therefore, since 2002, the Center of Disease Control in the United States recommended a universal vagina–rectum GBS screening at 35–37 weeks of gestation.2,3 In France, since 2001, a universal lower vagina culture screening is recommended at 35–38 weeks.4
In Paris-Saint Joseph hospital, monitoring for early-onset GBS disease of 8,048 live newborns between April 2007 and December 2009 found that 65% of the newborns infected with GBS had mothers whose antenatal screening had been negative. During the same period, we conducted a prospective study comparing antenatal vagina culture screening with intrapartum vagina screening using a rapid real-time polymerase chain reaction (PCR) GBS assay on the GeneXpert system. The study revealed that the positive predictive value of our antenatal vagina screening for identifying colonization status at delivery was only 58.3%, whereas the negative predictive value was imperfect (92.1%),5 leading to inappropriate intrapartum antibiotic prophylaxis. Because term deliveries account for nearly 75% of early-onset GBS disease cases,6 these results led us to perform around-the-clock GBS intrapartum PCR screening for all term deliveries since January 2010.
Our next objective was to estimate the effectiveness and costs of systematic intrapartum vagina PCR screening for term deliveries by comparing the 2010 strategy with the 2009 antenatal vagina screening.
MATERIALS AND METHODS
The effectiveness of early-onset GBS disease–prevention strategies (antenatal and intrapartum screenings) in term deliveries and the costs related to the diagnosis and treatment of newborns were compared in a before-and-after study. This study was conducted in a single institution without a control group. The approval of Paris-Saint Joseph Institutional Review Board was not required because of the routine use of intrapartum PCR screening.
In the 2009 antenatal vagina culture screening strategy, intrapartum antibiotic prophylaxis was administered if the antenatal screening was positive or in case of bacteriuria during the current pregnancy or a previous child with early-onset GBS disease. If GBS status is unknown at the time of delivery, a risk-factor assessment (eg, membrane rupture more than 12 hours, intrapartum fever higher than 38°C) is used to determine whether intrapartum antibiotic prophylaxis should be administered.4
In the 2010 intrapartum vagina PCR screening strategy, intrapartum antibiotic prophylaxis is administered to women who screened positive, to those presenting obstetrical risk factors when their vaginal samples did not give a PCR result (eg, PCR invalid or error), and to mothers with a history of neonatal GBS infection. Intrapartum antibiotic prophylaxis is not indicated in the following cases: negative intrapartum screening result, absence of obstetrical risk factors when the sample did not give a PCR result, and positive screening in a previous pregnancy.
For both strategies, penicillin G (5 million international units followed by 2.5 million international units every 4 hours until delivery) was used for intrapartum antibiotic prophylaxis. In case of high anaphylaxis risk, in 2009, clindamycin was used if GBS was susceptible and vancomycine if not; in 2010, vancomycine was the only possible alternative.
In 2009 and 2010, all term deliveries and early-onset GBS disease in newborns were monitored. Exclusion criteria were preterm delivery before 37 weeks of gestation. We used data from the medical information department and the microbiology laboratory. All medical records of the newborns with early-onset GBS disease and the medical records of the mothers were reviewed.
Diagnosis of early-onset GBS disease was based on criteria of the French Health Authority.7 Proven early-onset GBS disease was defined by positive results of blood or cerebrospinal fluid in the presence of clinical signs, biological abnormalities, or both clinical signs and biological abnormalities consistent with sepsis. Probable early-onset GBS disease was defined by positive results of GBS culture of gastric fluid aspiration, deep ear specimen, or both, in the presence of clinical signs, biological abnormalities, or clinical signs and biological abnormalities consistent with sepsis in which the blood, cerebrospinal fluid, or blood and cerebrospinal fluid cultures were negative.
Data collection of the characteristics of the mothers was performed using the database of the maternity ward, which included the following information: gestational age at delivery, parity, duration of labor, duration of membrane rupture before delivery, intrapartum fever, antibiotic prophylaxis or antibiotic therapy, number of doses of penicillin G, and previous child with early-onset GBS disease. Group B streptococcus screening results and incidence of bacteriuria during the current pregnancy were collected from laboratory software.
We estimated the cumulative costs for the mother and the newborn during the index admission from the perspective of a third-party payer. We did not include the costs of informal care supplied by family members or time lost from work or usual activities associated with early-onset GBS disease. All costs were adjusted to 2011 dollars. We estimated costs of inpatient care based on reimbursement rates of the diagnosis-related group (DRG). In France, as in other countries, this DRG price is calculated by multiplying standard amounts for operating and capital expenses found in yearly surveys by a national “weight” associated with the DRG for each hospitalization.8 Additional payments to public hospitals cover medical education or service to lower-income populations. Direct costs included screening test costs, costs for well newborn care, and hospital expenditures associated with treating neonates with early-onset GBS disease. We excluded the costs of treatment for the maternal outcomes of postpartum bacteremia and endometritis and focused on newborn-related costs. We obtained the DRG for each hospitalization from our institution's database and used the national standard euro amount.
The cost of antenatal culture screening was computed from the following resource utilization: Columbia nalidixic acid, colimycin sheep blood agar and commercial latex agglutination tests (bioMerieux) to identify GBS β-hemolytic and suspected nonhemolytic colonies and the blood Mueller Hinton agar and antibiotics discs used for susceptibility tests. Personnel costs included the time required for each step: 5 minutes for agar plating and looking for suspected GBS colonies; 5 minutes for identifying suspected colonies; and 7 minutes for susceptibility tests. The hourly rate of a senior technician in our laboratory was based on a gross salary of $63,291 per 1,820 hours per year.
Costs of intrapartum PCR screening included the acquisition cost of the Xpert GBS test cartridge (which incorporates sample processing control and internal control), costs of GBS culture and identification, for samples with either positive results or without PCR result (eg, invalid or error). Time required for each step was 5 minutes for PCR sample processing and communicating the result to midwives and 10 minutes for culture and identification of suspected colonies. The GeneXpert machine capital cost was allocated per test, assuming a 6-year depreciation and 10% salvage value based on approximately 3,000 deliveries per year. Table 1 summarizes the unit costs. Euros were converted into U.S. dollars using the OECD purchasing power parity index ($1 U.S.=0.79 euros).
The unit of analysis was the mother (not the newborn, because in case of twin births we assumed that prevention strategy would benefit both twins). We compared 2009 and 2010 costs between patients with and without intrapartum screening on an intention-to-treat basis. We compared health care utilization and costs using two-sided t tests for continuous variables and χ2 or Fisher exact tests for dichotomous and categorical variables. A probabilistic sensitivity analysis was performed to estimate an incremental cost-effectiveness ratio and to assess the uncertainty on costs and effectiveness (newborn infections averted).9,10 Statistical analyses performed on Excel 2007 or SAS statistical software for the bootstrap analysis.
During the study period, out of 2,961 and 3,017 deliveries in 2009 and 2010, respectively, 2,761 and 2,814 were term deliveries corresponding to 2,824 and 2,842 live births. Vaginal deliveries represented 82.5% of all term deliveries. The characteristics of women with term deliveries presented in Table 2 did not show differences in the population in terms of gestational age at delivery, parity, time from rupture of membranes to delivery, and presence of obstetrical risk factors at delivery. The GBS colonization rate found using intrapartum PCR screening was 16.7% (±4% 95% confidence interval [CI]), whereas antenatal GBS colonization rate based on cultures was 11.7% (±3% 95% CI). Screening tests were performed in 86% of term deliveries in 2009 and in 85% in 2010. The number of women receiving intrapartum antibiotic prophylaxis in term deliveries was 311 (11%±3% 95% CI) in 2009 and 436 (16%±4% 95% CI) in 2010 (P<10−4). Characteristics of GBS-infected newborns are presented in Table 3.
Patients with severe symptoms had presented rapid clinical deterioration with respiratory distress or cardiovascular instability leading to an intensive care survey in which the average duration of antibiotic therapy was 10 days (except the meningitis). Mildly ill patients had mild respiratory distress with biological abnormalities consistent with sepsis; they were hospitalized in the neonatal ward unit, where the average duration of antibiotic therapy was 7 days.
The flow diagram in Figure 1 provides the results of the antenatal screening strategy in 2009. Of the 101 women who did not undergo antenatal screening, 40 were treated. Of these, 20 had GBS bacteriuria during the current pregnancy, eight had a previous newborn infected with GBS, and 12 had obstetrical risk factors at delivery. Of the 61 women who did not receive intrapartum antibiotic prophylaxis, 21 had GBS bacteriuria and 10 had had a previous neonatal early-onset GBS disease, but no cases of early-onset GBS disease occurred among them. We recorded two severe probable early-onset GBS disease among the 30 remaining women without any obstetrical risk factors at delivery. Severe cases also occurred among women who had negative screening results antepartum: two GBS bacteremia, one GBS meningitis, and two severe probable early-onset GBS disease. The 16 remaining cases of GBS infections among newborns in 2009 were mildly ill probable early-onset GBS disease.
The flow diagram in Figure 2 provides the results of the intrapartum screening strategy in 2010. Polymerase chain reaction test results were not available for 212 (8.9%) women because of invalid PCR or errors in PCR. Of these, although 46 had obstetrical risk factors, only 41 were treated and no GBS infection was recorded. Of the 171 women who did not receive intrapartum antibiotic prophylaxis, 166 did not present any obstetrical risk factor and two probable early-onset GBS disease cases occurred among them. Only four probable early-onset GBS disease cases were found among mothers who had negative screening results, three of whom delivered more than 22 hours after screening and were administered antibiotic therapy because of fever higher than 38°C and prolonged membrane rupture. All early-onset GBS disease in the intrapartum screening strategy were mildly ill cases.
Probabilities of GBS infection presented in Table 4 showed an overall probability of 0.9% and 0.5% for antenatal screening and intrapartum screening strategies, respectively (P<10−4). The probability of newborn infection given negative GBS screening decreased from 0.6% in 2009 to 0.17% in 2010 (P<10−4).
Prolonged hospital admissions for neonates associated with early-onset GBS disease occurred in 23 in 2009 and in 12 in 2010, among the 2,824 and 2,842 live births, respectively. Of the 23 early-onset GBS disease cases in 2009, one was GBS meningitis, six were severe neonatal GBS infections including two bacteremia cases, and 16 were mildly ill cases. In 2010, all 12 neonatal GBS infections were mildly ill cases.
The corresponding total days of hospitalization for infected newborns was 177 days in 2009 compared with 94 days in 2010, or average day per newborn of 7.7 days in 2009 compared with 7.8 days in 2010. The total cost of delivery and treatment of infected newborns was reduced from $146,057 in 2009 to $25,433 in 2010, reflecting the reduction both in caseload and severity.
When costs were estimated over the entire population, the average total costs per delivery were identical for patients in 2009 with antenatal vagina screening and in 2010 with intrapartum vagina screening ($1,759 compared with $1,754). All results are presented in Table 5. The cost-effectiveness ratio was estimated using a probabilistic sensitivity analysis with 2,000 bootstrap replications. The results are presented in Figure 3. More than half of the replications were in the bottom right quadrant, indicating that intrapartum screening was a dominant (both cost-saving and infection reducing) strategy.
This study demonstrated the feasibility of intrapartum PCR screening, which resulted in a significant decrease in probability of GBS infection in newborns with fewer severe cases and the same average costs compared with the antenatal lower vagina culture screening strategy recommended in France.
The key element for success was that the probability of newborn infection considering negative GBS screening decreased dramatically from 0.6% to 0.17%. This may be attributable to false-negative results in the antenatal culture screening, as described in previous studies,5,11–14 which found that more than 60% of GBS-infected newborn were born to mothers with negative antenatal screening results,6,15–17 even when the vagina–rectum swabbing was used.
The effectiveness of any screening strategy depends on timely administration of antibiotic prophylaxis. In our population, 75% of the women delivered their newborns in more than 3 hours. When comparing the two approaches, 50% in 2010 and 55% in 2009 of women who received prophylaxis were treated with at least two penicillin doses. Although duration of 4 hours of intrapartum antibiotic prophylaxis before delivery was considered the optimal time to interrupt GBS transmission, several studies showed that a duration of 2 hours or less is still effective.18,19
Our survey showed a significant increase in the number of women receiving intrapartum antibiotic prophylaxis with the intrapartum PCR screening approach. Even though the wider use of prophylaxis could raise concerns about the potential increase in antibiotic resistance among Gram-negative organisms, it was shown that ampicillin-resistant Escherichia coli infections increased in preterm newborns and not in term newborns, possibly as a result of prolonged maternal ampicillin treatment before delivery.20 In our maternity department, a continued survey did not show any increase in Gram-negative-resistant neonatal sepsis.
Probable early-onset GBS diseases were taken into account in our analysis because, in the era of intrapartum antibiotic prophylaxis, a portion of newborns have probable GBS sepsis diagnosed, possibly as a result of intrapartum antibiotic prophylaxis treatment that inhibits growth in blood and cerebrospinal fluid but does not alleviate clinical symptoms or death.21 Furthermore, there can be false-negative blood cultures related to the small volume of blood inoculated. Thus, the real burden of disease is underestimated if only proven early-onset GBS diseases are evaluated, as suggested by some studies22,23; in our study, four severe GBS infections were recorded among probable early-onset GBS disease cases.
Early-onset GBS disease cases in 2009 were found to be more severe (7 out of 23, including 1 meningitis and 2 bacteremia cases), whereas only mildly ill cases and no proven infections were recorded in 2010. As a result, the overall hospital care cost of early-onset GBS disease was nearly six times higher in 2009 than in 2010. When the costs of screening and antibiotic prophylaxis were included, we found that intrapartum PCR was cost-neutral compared with antenatal vagina culture screening. The probabilistic sensitivity analysis showed a high probability that intrapartum screening was cost-saving.
Our results do not entirely match results from previous analyses based on decision models. Daniels et al24 showed that antenatal culture strategy was cost-effective but not cost-neutral compared with intrapartum PCR screening. In their analysis, the unit costs of delivery were close to our figures, but the diagnostic accuracy of the PCR test differed dramatically. The sensitivity and specificity of the Xpert GBS test in our previous study were found to be 98.5% and 99.6%,5 whereas in the analysis by Daniels et al the sensitivity and specificity of the Smart GBS test (Smartcycler system) on vaginal samples were only 58% and 92%, respectively. This could be explained by the fact that Smartcycler system requires manual DNA extraction by technically skilled operators, whereas in their study the swabs were also tested by trained midwifery assistants. By contrast, the only technical step in the completely automated GeneXpert system is to insert the swab into the test cartridge.
The decision model by Haberland et al25 resulted in a net benefit of $6 per birth when using intrapartum molecular screening. These authors included in their model the long-term health care costs of newborns who were healthy and who were disabled, as well as the expected benefits for a healthy newborn. Schroeder et al26 estimated that the health and social care costs of newborns with GBS disease is, on average, two times higher during the first 2 years of life than for newborns without GBS disease. The intrapartum screening strategy would result in a benefit in terms of early-onset GBS disease and its associated long-term medical costs.
Group B streptococcus intrapartum PCR screening using the GeneXpert system has been implemented successfully since March 2011 at the point of care around the clock in the delivery room. A key success factor was the empowerment of the midwives in the management of PCR processing without losing time either in bringing the sample to the laboratory or in waiting for the communication of the screening result.
1. Boyer KM, Gottoff SP. Prevention of early-onset neonatal group B streptococcal disease with selective intrapartum chemoprophylaxis. N Engl J Med 1986;314:1665–9.
2. Schrag SJ, Zell ER, Lynfield R, Roome A, Arnold KE, Craig AS, et al.. A population-based comparison of strategies to prevent early-onset group B streptococcal disease in neonates. N Engl J Med 2002;347:233–9.
3. Centers for Disease Control and Prevention. Prevention of perinatal group B Streptococcal disease: Revised guidelines from CDC. MMWR 2010;59(RR-10):1–31.
5. El Helali N, Nguyen JC, Ly A, Giovangrandi Y, Trinquart L. Diagnostic accuracy of a rapid real-time polymerase chain reaction assay for universal intrapartum group B streptococcus screening. Clin Infect Dis 2009;49:417–23.
6. Van Dyke MK, Phares CR, Lynfield R, Thomas AR, Arnold KE, Craig AS, et al.. Evaluation of universal antenatal screening for group B Streptococcus. N Engl J Med 2009;360:2626–36.
8. Agence Technique de l'Information sur l'Hospitalisation. Echelle nationale de coûts par GHM (en euros), Référentiel national de coûts. Available at: www.atih.sante.fr
. Retrieved August 29, 2011.
9. Briggs AH, O'Brien BJ. The death of cost-minimization analysis. Health Economics 2001;10:179–84.
10. Briggs AH, Wonderling DE, Mooney CZ. Pulling cost-effectiveness analysis up by its bootstraps: a non-parametric approach to confidence interval estimation. Health Economics 1997:6:327–40.
11. Davies HD, Miller MA, Faro S, Gregson D, Kehl SC, Jordan JA. Multicenter study of a rapid molecular-based assay for the diagnosis of group B Streptococcus colonization in pregnant women. Clin Infect Dis 2004;39:1129–35.
12. Alfa MJ, Sepehri S, De Gagne P, Helawa M, Sandhu G, Harding GK. Real-time PCR assay provides reliable assessment of intrapartum carriage of group B Streptococcus. J Clin Microbiol 2010;48:3095–9.
13. Whitney C, Daly S, Limpongsanurak S, Festin M, Thinn KK, Chipato T, et al.. The international infections in pregnancy study: group B streptococcal colonization in pregnant women. J Matern Fetal Neonatal Med 2004;4:267–74.
14. Lin FY, Weisman LE, Azimi P, Young AE, Chang K, Cielo M, et al.. Assessment of intrapartum antibiotic prophylaxis for the prevention of early-onset group B Streptococcal disease. Pediatr Infect Dis J 2011;30:759–63.
15. Puopolo KM, Madoff LC, Eichenwald EC. Early-onset group B streptococcal disease in the era of maternal screening. Pediatrics 2005;115:1240–6.
16. Pulver LS, Hopfenbeck MM, Young PC, Stoddart GJ, Korgenski K, Daly J, et al.. Continued early onset group B streptococcal infections in the era of intrapartum prophylaxis. J Perinatol 2009;29:20–5.
17. Poyart C, Réglier-Poupet H, Tazi A, Billoët A, Dmytruk N, Bidet P, et al.. Invasive group B streptococcal infections in infants in France. Emerg Infect Dis 2008;14:1647–9.
18. de Cueto M, Sanchez MJ, Sampedro A, Miranda JA, Herruzo AJ, Rosa-Fraile M. Timing of intrapartum ampicillin and prevention of vertical transmission of group B streptococcus. Obstet Gynecol 1998;91:112–4.
19. Berardi A, Rossi C, Biasini A, Minniti S, Venturelli C, Ferrari F, et al.. Efficacy of intrapartum chemoprophylaxis less than 4 hours duration. J Matern Fetal Neonatal Med 2011;24:619–25.
20. Hyde TB, Hilger TM, Reingold A, Farley MM, O'Brien KL, Schuchat A, et al.. Trends in incidence and antimicrobial resistance of early-onset sepsis: population-based surveillance in San Francisco and Atlanta. Pediatrics 2002;110:690–5.
21. Carbonell-Estrany X, Figueras-Aloy J, Salcedo-Abizanda S, de la Rosa-Fraile M, Castrillo Study Group. Probable early-onset group B streptococcal neonatal sepsis: a serious clinical condition related to intrauterine infection. Arch Dis Child Fetal Neonatal Ed 2008;93:F85–9.
22. Heath PT, Schuchat A. Perinatal group B streptococcal disease. Best Pract Res Clin Obstet Gynaecol 2007;411–24.
23. Bedford Russel AR, Breathnach A, Sender P. Confirmed group streptococcus infection: the tip of the iceberg. Arch Dis Child Fetal Neonatal Ed 2001;45:84.
24. Daniels J, Gray J, Pattison H, Roberts T, Edwards E, Milner P, et al.. Rapid testing for Group B streptococcus during labor: a test accuracy study with evaluation of acceptability and cost-effectiveness. Health Technol Assess 2009;13:1–154.
25. Haberland CA, Benitz WE, Sanders GD, Pietzsch JB, Yamada S, Nguyen L, et al.. Perinatal screening for group B streptococci: cost benefit analysis of rapid polymerase chain reaction. Pediatrics 2002;110(3):471–80.
© 2012 The American College of Obstetricians and Gynecologists
26. Schroeder EA, Petrou S, Balfour G, Edamma O, Heath PT, Health Protection Agency Group B Streptococcus Working Group. The economic costs of group B Streptococcus (group B Streptococcus) disease: prospective cohort study of infants with group B Streptococcus disease in England. Eur J Health Econ 2009;10:275–85.