Objective: To determine the incidence of possible neonatal herpes simplex virus (HSV) infections, HSV infection status of women with infected infants, and use of measures to reduce risk of HSV transmission to the neonate in a large US managed-care population.
Study Design: Retrospective analysis of administrative claims from the Integrated Health Care Information Services National Managed Care Benchmark database.
Results: Of 233,487 infants born to 252,474 mothers from January 1997 to June 2002, the numbers assigned an ICD-9 code reflecting possible neonatal HSV infection ≤30 and ≤90 days of birth were 178 (0.08%) and 338 (0.15%), respectively. Of the 338 mothers delivering infants with possible neonatal HSV ≤90 days postnatally, 12% had a prior HSV diagnosis, 5% were prescribed an antiviral medication during the study period, and 3% used antiviral medication and had a cesarean delivery.
Conclusion: These results support national surveillance of neonatal HSV to better define its incidence, strengthen health policies, and improve prevention and treatment.
The incidence of neonatal herpes appears to have been higher than previously estimated, and measures to reduce risk of viral transmission to the neonate were not taken in many at-risk pregnancies in a large US managed-care population.
From the *Department of Pediatrics, Microbiology, Medicine and Neurosurgery, University of Alabama at Birmingham, Birmingham, Alabama; †GlaxoSmithKline, Research Triangle Park, North Carolina; and ‡GlaxoSmithKline, Collegeville, Pennsylvania
GlaxoSmithKline funded the research described in this manuscript. The research does not involve a specific pharmaceutical product. The authors acknowledge Jane Saiers, PhD (Chapel Hill, NC), for assistance with writing this manuscript and Baiju Rasheed, MCA, and Reshma Chowdhry, BE (GlaxoSmithKline, Bangalore, India), for assistance with data analysis. Dr. Saiers' work was funded by GlaxoSmithKline.
Correspondence: Richard Whitley, MD, University of Alabama, 303 CHB, 1600-Seventh Avenue South, Birmingham, AL 35294. E-mail: firstname.lastname@example.org.
Received for publication October 24, 2006, and accepted December 18, 2006.
APPROXIMATELY 1 IN 5 PREGNANT WOMEN in the United States is infected with herpes simplex virus-Type 2 (HSV-2), the primary cause of neonatal herpes simplex virus (HSV) infections.1 Neonatal HSV infection is usually the consequence of intrapartum exposure of the fetus to HSV-2 or HSV-1 at birth.2 Neonatal HSV infection is defined as that occurring within the first 30 days of life, but it most often manifests during the first 2 to 3 weeks as encephalitis; disseminated multiorgan disease; or skin, eye, and mouth disease.3,4 Regardless of the manifestations of neonatal HSV infection, significant neurodevelopmental impairment or death occurs in at least one-third of infected babies.5
Prevention of neonatal infection is viewed as the cornerstone of management because treatment of the infected newborn does not avert severe sequelae in many survivors.1,6 Preventive strategies include educating women and their partners about the prevention of acquisition of HSV during pregnancy and reducing the risk of viral transmission from HSV-positive women to their infants through measures such as cesarean delivery for women with genital lesions during labor, avoidance of invasive perinatal procedures, and suppressive antiviral therapy during the last 4 weeks of pregnancy.1,6,7 As the majority of pregnant women with HSV are not diagnosed,1 serologic testing of pregnant women and their partners to identify infected women and, importantly, those at risk of infection during pregnancy can assist in preventing neonatal herpes.
Effective development and implementation of these and other preventative strategies depend on an understanding of the incidence of neonatal HSV infection.8 Accurate estimates of the incidence of neonatal herpes infections are not available in part because neonatal HSV infection, unlike many other communicable diseases affecting the neonate, is not required to be reported to US public health authorities with rare exceptions (e.g., NewYork and the states of Connecticut, Florida, Louisiana, Massachusetts, Nebraska, South Dakota, and Washington).9 Incidence estimates range widely from 1 in 3200 to 1 in 15,000 births in the United States.3,6,7,10,11 This study was undertaken to determine the incidence of possible neonatal HSV infections in a large managed-care population representing 7 US census regions. In addition, the HSV infection status of women with identified infants and the use of measures to reduce risk of viral transmission to the neonate were examined.
Materials and Methods
Study Design and Data Source
This retrospective cohort analysis of medical and pharmacy claims was conducted to determine the incidence of possible neonatal HSV infections, the HSV status of women with infected infants, and the use of measures to reduce risk of HSV transmission to the neonate in a large managed-care population. The Institutional Review Board at the University of Alabama at Birmingham approved the study protocol.
Data were obtained from the Integrated Health Care Information Services (IHCIS) National Managed Care Benchmark database12 (Waltham, MA). The IHCIS database, which is fully deidentified and compliant with the 1996 Health Insurance Portability and Accountability Act (HIPAA), includes data from more than 30 health plans covering 7 US census regions. For the study period from January 1997 to June 2002, the database contains medical and pharmacy claims for more than 17 million members covered by a mix of health maintenance organization (HMO), preferred provider organization (PPO), and point-of-care service (POS) plans.
Women who had a current procedural terminology (CPT) code for pregnancy delivery were identified from database records from January 1997 to June 2002, a time period when records for neonates could be linked to their mothers' records via a family identification number. For infants delivered by women with a CPT code for delivery during the study period, medical claims and presence of an acyclovir prescription claim were used to identify infants with possible neonatal HSV infection as reflected in International Classification of Diseases, 9th edition (ICD-9) diagnostic codes. Because a definitive ICD-9 code for neonatal herpes does not exist, the following codes for the newborns charts were considered to be possible indicators of neonatal herpes: 054.10 (genital herpes), 054.2 (herpetic gingivostomatitis), 054.3 (herpetic meningoencephalitis), 054.40 (herpes simplex with unspecified ophthalmic complication), 054.41 (herpes simplex dermatitis of eyelid), 054.42 (dendritic keratitis), 054.43 (herpes simplex iridocyclitis), 054.49 (herpes simplex with ophthalmic complications, other), 054.5 (herpetic septicemia), 054.6 (herpetic whitlow), 054.72 (herpes simplex meningitis), 054.73 (herpes simplex otitis externa), 054.79 (herpes simplex other), 054.8 (herpes simplex with unspecified complication), 054.9 (herpes simplex without mention of complication), 790.8 (viremia, unspecified), 771.2 (other congenital infections: herpes simplex, listeriosis, malaria, toxoplasmosis, tuberculosis), 695.1 (erythema multiforme: erythema iris, herpes iris, Lyell's syndrome, scalded skin syndrome, Stevens–Johnson syndrome, toxic epidermal necrolysis), and 286.6 (defibrination syndrome). The numbers of neonates assigned one or more of these diagnostic codes were summarized for neonates assigned the code at any time during the study period, ≤30 days of birth, and ≤90 days of birth. As neonatal HSV infection typically manifests during the first 1 to 3 weeks of life,4 diagnoses assigned ≤30 days were considered the most accurate for the diagnosis of this disease. In addition, babies with diagnoses assigned ≤90 days of age were identified to develop a broader understanding of the contribution of HSV to disease of very young infants.
For mothers of neonates assigned one of the aforementioned diagnostic codes and identified in the windows of ≤30 days or ≤90 days of birth, medical and prescription claims were used to identify mothers who had a record of a diagnosis of HSV, who had prescription claims for antiviral medications any time during the study period during the pregnancy, and during the last 4 weeks of pregnancy, and who had a prescription for antiviral medications and also delivered via cesarean section. These parameters were summarized (as numbers of mothers) for the maternal subgroups defined by time of diagnosis of their neonates' HSV infection (i.e., ≤30 days of birth and ≤90 days of birth).
From January 1997 through June 2002, 252,474 mothers with a CPT code for delivery of a pregnancy were identified from the database. These mothers delivered 233,487 infants. The discrepancy between the number of mothers and the number of infants could be attributed to several factors including coverage of the infant under a different medical plan than the mother, withdrawal of the mother from the medical plan before delivery of the infant, miscarriage, or incorrect assignment of the CPT code for delivery.
The mean age of the 252,474 mothers who delivered during the study period was 31.15 years. Mothers were predominantly from the Mid-Atlantic and Northeast census regions and had 1 of 3 plan types: PPO, HMO, or POS.
Incidence of Possible Neonatal HSV Infection
Of the 233,487 infants delivered during the study period, the number of infants assigned at least one of the ICD-9 codes considered to reflect the possible presence of neonatal HSV ≤30 days after birth was 178 (Fig. 1) for an incidence of 0.08% (178/233,487). The number of infants with such a diagnosis ≤90 days of birth was 338 (Fig. 1) for an incidence of 0.15% (338/233,487). The most common ICD-9 codes for infants diagnosed with presumed neonatal HSV ≤30 days or ≤90 days after birth were 771.2 (other congenital infections), 054.9 (herpes simplex without mention of complication), 790.8 (viremia, unspecified), and 695.1 (erythema multiforme) (Table 1).
ICD-9 code 771.2 can be used to define several congenital infections including cytomegalovirus. To explore the potential effect of the lack of specificity of this code on the results, the incidence of possible neonatal herpes was assessed in a post hoc sensitivity analysis that excluded this code. The number of infants with a diagnosis (excluding 771.2) reflecting the possible presence of possible neonatal HSV ≤30 days of birth was 129 for an incidence of 0.06% (129/233,487) (Table 2). The number of infants with such a diagnosis (excluding 771.2) ≤90 days of birth was 283 for an incidence of 0.12% (283/233,487) (Table 2).
HSV Status of Mothers of Infants with Possible Neonatal HSV Infection
The numbers of mothers delivering infants with possible neonatal HSV infection diagnosed ≤30 days and ≤90 days of birth were the same as the numbers of infants, namely 178 and 338, respectively. Of the 338 mothers delivering infants diagnosed with possible neonatal HSV ≤90 days of birth, only 39 had a prior diagnosis of HSV infection. The rate of prior maternal diagnosis of HSV infection was similarly low between the maternal subgroups defined by time of diagnosis of possible neonatal HSV infection (15% for ≤30 days of birth and 12% for ≤90 days of birth) (Fig. 2).
Measures for Reducing Risk of Viral Transmission
Of the 338 mothers delivering infants diagnosed with possible neonatal HSV infection ≤90 days of birth, only 17 were prescribed antiviral medication during the study period. The rate of maternal use of antiviral medication was similarly low between the maternal subgroups defined by time of diagnosis of possible neonatal HSV infection (7% for ≤30 days of birth and 5% for ≤90 days of birth) (Fig. 2). Of the 17 mothers with a prescription claim for an antiviral medication during the study period, only 11 had a claim during their pregnancy, and only 8 had a claim within the last 4 weeks of pregnancy. Furthermore, of the 338 mothers delivering infants diagnosed with possible neonatal HSV infection ≤90 days of birth, the number who both had a prescription claim for an antiviral medication during the study period and had a medical claim for cesarean delivery was 9.
During the study period in this managed-care population, the incidence of possible neonatal HSV infection as defined by the presence of various ICD-9 codes and occurrence within the first 30 days of life was 1 in 1312 live births. When the postnatal window of time for inclusion in the incidence calculation was increased from the first 30 days to the first 90 days of life, the incidence of HSV infection in the young infant increased to 1 in 691 liveborns. The ≤30-day cutoff for diagnosis is sufficient to capture neonatal HSV infection.4 However, the finding that potential HSV infections with a ≤90-day cutoff was approximately double than that with the ≤30-day cutoff suggests a vastly greater-than-anticipated burden of disease secondary to HSV infection early in life. However, those cases identified between 30 and 90 days do not qualify as neonatal HSV infection. Importantly, the incidence of possible neonatal HSV infections in this large managed-care population in the United States was higher than previously reported in other cohorts in this country.7,10 In the most extensive previous study to assess the incidence of neonatal HSV infections in the United States, 1 in 3200 liveborns were determined to have neonatal HSV infection.7
The current study was conducted in an era of high genital herpes prevalence in the US population at large as indicated by recent population-based and community family-practice-based studies seroprevalence studies.13–15 In contemporary studies, the seroprevalence of HSV-2 ranges from17% to 25%.13–15 Furthermore, an increase in the incidence of primary HSV-1 infection of the genital tract has been detected.16,17 Like HSV-2, HSV-1 poses a significant threat for the newborn.7 Thus, enhanced attention to disease in the newborn is essential. As an increasing number of previously diagnosed women with HSV-2 infection approach term, the use of antiviral drugs will likely increase and may ultimately influence these incidence data.
Difficulties in confirming cases of neonatal HSV infection in this managed-care population could have resulted in over- or underreporting of cases. No ICD-9 codes specific to neonatal herpes exist. Therefore, the authors used diagnostic codes that they considered, based on clinical experience, to be likely to reflect possible neonatal HSV infection. Some of the codes could have been assigned for diagnoses other than neonatal HSV infection. For example, defibrination (code 286.6), which equates with disseminated intravascular coagulopathy in the absence of a bacterial organism or enteroviral infection, is very common in disseminated neonatal HSV infection. Code 695.1 (erythema multiforme) contains a herpes diagnostic category but is also used for many skin ailments unrelated to HSV infection. Erythema multiforme of other etiologies is very rare in the first 30 days of life. Code 771.2 (other congenital infections) contains 5 congenital conditions, only one of which is HSV infection. The finding that a post hoc sensitivity analysis excluding code 771.2 yielded results similar to analyses that included this code suggests that the lack of specificity of the diagnostic code did not significantly affect reporting of cases in this instance. Future investigations of this type could be improved by identifying, through systematic examination, diagnostic codes previously used in known cases of neonatal herpes in a given population and by applying those codes to the sample being studied. The current authors recommend the creation of ICD-9 and/or CPT codes that indicate a diagnosis of neonatal HSV infection and include the extent of disease (i.e., skin, eye, and mouth; central nervous system; or disseminated multiorgan disease).
The inability to identify individual medications prescribed to infants while in the hospital (because individual medications were not coded in patients' hospital records) also affected the ability to identify cases of possible neonatal HSV. Postnatal medication use could not be used to identify infants as potentially having neonatal HSV. For example, whether infants received intravenous acyclovir in the hospital after birth—a common practice for infants at risk of neonatal HSV—was not discoverable.
Although the incidence data from this study cannot be considered definitive, they do suggest the need for further research and support recent calls to make neonatal herpes a reportable disease in the United States.8,18 Neonatal herpes is reportable in 7 states and in New York City.9 Serious communicable diseases are usually made reportable to public health authorities so that epidemiologic trends can be monitored and effective disease-control and -prevention efforts can be implemented. For example, congenital syphilis, which like neonatal herpes is an often fatal but preventable disease resulting from maternal infection, is reportable in the majority of the states.19 Other generally reportable conditions in newborns and children include congenital rubella syndrome, meningococcal infections, rheumatic fever, mumps, and polio. Among adults, several sexually transmitted diseases including syphilis, gonorrhea, chlamydia, acquired immune deficiency syndrome (AIDS), and hepatitis are almost universally reportable.20 Neonatal herpes is as severe, costly, and preventable as many of these currently reportable conditions. Moreover, as suggested by the relatively high incidence of possible neonatal HSV infection in the managed-care population in the current study, neonatal herpes is also as common as many currently reportable conditions. However, the absence of national surveillance for neonatal herpes might hinder efforts at prevention by promulgating the misperceptions that the disease is insufficiently severe and common to warrant such efforts.
The finding that a prior diagnosis of HSV infection had been made in only 12% of mothers having infants with possible neonatal HSV in this study (i.e., in 39 of the 338 mothers delivering infants diagnosed with possible neonatal HSV ≤90 days after birth) is consistent with the previous finding that the majority of individuals with HSV, including pregnant women, are not diagnosed.1 The high rate of unrecognized maternal HSV infection in this cohort is alarming given the pivotal role of awareness of maternal HSV status in efforts to prevent viral transmission to the neonate. Likewise, the low frequency of use of measures that can reduce the risk of transmitting HSV to the neonate was disturbing. The risk of neonatal infection can be mitigated by antiviral prophylaxis during pregnancy, delivering via cesarean section instead of vaginally, and limiting the use of invasive procedures in women shedding HSV during labor.5,8,21 These measures were employed in a small minority of women in this study.
In conclusion, neonatal herpes is a potentially preventable disease that often results in fatalities or severe, lifelong neurologic sequelae. Data from this study in a large managed-care population suggest that the incidence of possible neonatal herpes might be higher than previously estimated and that measures to reduce risk of viral transmission to the neonate are not taken in many at-risk pregnancies. These results support recent calls to implement national surveillance of neonatal HSV in order to better define its incidence, strengthen health policies, and improve efforts at prevention and treatment.
1. Brown ZA, Gardella C, Wald A, et al. Genital herpes complicating pregnancy. Obstet Gynecol 2005; 106:845–856.
2. Kimberlin DW. Herpes simplex virus infections in neonates and early childhood. Semin Pediatr Infect Dis 2005; 16:271–281.
3. Marques AR, Straus SE. Herpes simplex type 2 infections—An update. Dis Mon 2000; 46:327–359.
4. Jenkins M, Kohn S. New aspects of neonatal herpes. Infect Dis Clin North Am 1992; 6:57–74.
5. Baker DA. Antiviral therapy for genital herpes infections in pregnancy. Expert Rev Anti Infect Ther 2005; 3:385–392.
6. Whitley R. Neonatal herpes simplex virus infection. Curr Opin Infect Dis 2004; 17:243–246.
7. Brown ZA, Wald A, Morrow RA, et al. Effect of serologic status and cesarean delivery on transmission rates of herpes simplex virus from mother to infant. JAMA 2003; 289:203–209.
8. Handsfeld HH, Waldo AB, Brown ZA, et al. Neonatal herpes should be a reportable disease. Sex Transm Dis 2005; 32:521–525.
10. Brown Z. Preventing herpes simplex virus transmission to the neonate. Herpes 2004; 11(Suppl 3):175A–186A.
11. Whitley RJ. Neonatal herpes simplex virus infections: Pathogenesis and therapy. Pathol Biol 1992; 40:729–734.
13. Fleming T, McQuillan G, Johnson R, et al. Herpes simplex virus type 2 in the United States, 1976 to 1994. N Engl J Med 1997; 337:1105–1111.
14. Leone P, Fleming D, Gilsenan A, et al. Seroprevalence of herpes simplex virus-2 in suburban primary care offices in the United States. Sex Transm Dis 2004; 31:311–316.
15. Xu F, Sternberg MR, Kottiri BJ, et al. Trends in herpes simplex virus type 1 and type 2 seroprevalence in the United States. JAMA 2006; 296:964–973.
16. Kimberlin DW, Rouse DJ. Genital herpes. N Engl J Med 2004; 350:1970–1977.
17. Solomon L, Cannon M, Reyes M, et al. Epidemiology of recurrent genital herpes simplex virus types 1 and 2. Sex Transm Dis 2003; 79:456–459.
18. Kimberlin DW. Neonatal HSV infections: The global picture. Herpes 2004; 11:31–32.
19. Roush S, Birkhead G, Koo D, et al. Mandatory reporting of diseases and conditions by health care professionals and laboratories. JAMA 1999; 282:164–170.
© Copyright 2007 American Sexually Transmitted Diseases Association
21. American College of Obstetricians and Gynecologists. ACOG practice bulletin. Management of herpes in pregnancy. Number 8 October 1999. Clinical management guidelines for obstetrician–gynecologists. Int J Gynaecol Obstet 2000; 68:165–173.