- Email: [email protected]
Hepatitis B: A case for prenatal screening of all patients
Poppe et al.
bilateral selective embolization of the uterine arteries was performed with polyvinyl alcohol (Ivalon) particles mixed with thrombin (Topostasine). The postembolization period was uncomplicated. Control intravenous subtraction angiography demonstrated total obliteration of the left uterine artery. A small part of the arteriovenous malformation was still filled through the right uterine artery. During a follow-up period of 6 months no abnormal bleeding was noticed. The patient had normal, regular menstrual cycles and conceived after the sixth postembolization month. During the first and second trimesters of pregnancy intermittent periods of slight vaginal bleeding were observed. There was no cervical infection; there was no placenta previa, and no signs of abruptio placentae were seen. The placenta was implanted on the left ventrolateral side. Clinical and ultrasonographic evaluation showed normal fetal growth. Doppler measurements showed normal flow at both the uterine and umbilical fetal sites. Except for rest, we did not advise any special treatment. The patient went into sudden premature labor at 35 weeks and was delivered of a normal male infant weighing 2600 gm. Macroscopic and microscopic ex-
May 1987 Am J Obstet Gynecol
amination of the placenta was normal, and the postpartum period was uncomplicated.
Comment In this case, although the left uterine artery was totally obliterated and the right uterine artery was partially obliterated, there was a normal adaptation of the uteroplacental vasculature resulting in optimal fetal growth and development, even with the placenta located at the site of the embolized arteriovenous malformation. The case report presented shows that conservative treatment of an arteriovenous malformation can stop the heavy vaginal bleeding and preserve the capability of reproduction. REFERENCES 1. Liggins GC. Uterine arteriovenous fistula. Obstet Gynecol 1964;23:214-7. 2. Wilms GE, Favril A, Baert AL, Poppe W, Van Assche FA. Transcatheter embolization of uterine arteriovenous malformations. Cardiovasc Intervent Radiol 1986;9:61-4.
Hepatitis B: A case for prenatal screening of all patients Amelia C. Cruz, M.D., Barbara H. Frentzen, M.S.N., and Marylou Behnke, M.D. Gainesville, Florida Evaluation of a mass screening program to detect hepatitis B surface antigen in the obstetric population of Shands Hospital, University of Florida in Gainesville, from January 1, 1983, through December 31, 1985, was undertaken. Prevalence of hepatitis B surface antigen seropositivity was 0.54%. Review of medical records revealed that 67% of patients with positive hepatitis B surface antigen screens had no risk factors identifiable by routine prenatal history. Although all patients were screened on admission to the hospital, the results of the immunoassay were not available in time for staff to institute isolation procedures for the laboring mother or unwashed neonate. In 82% of the cases, treatment of neonates occurred later than 12 hours after delivery (the current Centers for Disease Control recommendation). It is recommended that prenatal screening of all patients with a hepatitis B surface antigen immunoassay be done by 34 weeks' gestation where the patient population is of predominantly low socioeconomic status to ensure appropriate isolation and timely neonatal immunoprophylaxis. (AM J OSSTET GVNECOl 1987;156:1180-3.)
Key words: Hepatitis B, prenatal screening, neonatal immunoprophylaxis
From the Departments of Obstetrics and Gynecology and Pediatrics, University of Florida College of Medicine. Supported entirely, or in part, under an agreement with the Department of Health and Rehabilitative Services, State of Florida, Children's Medical Services. Receivedforpublication August 5, 1986; revised November 3,1986; accepted November 17, 1986. Reprint requests: Amelia C. Cruz, M.D., Department of Obstetrics and Gynecology, Box J-294 JHMHC, University of Florida, Gainesville, FL 32610.
Perinatal exposure of neonates to the hepatitis B surface antigen (HBsAg) and e antigen (HBeAg) by mothers who are seropositive, but clinically symptom free, is known to be a major source of hepatitis B virus (HB V) infection. I·' Treatment at birth with hepatitis B immunoglobulin (HBIG) and vaccine has been shown to dramatically reduce the incidence of acute and chronic infection in neonates. I. 4. 5 Therefore identification of all pregnant women who are HBsAg carriers is nec-
Volume 156 Number 5
essary to identify and prophylactically treat exposed neonates after delivery. Serum screening for HBsAg is not a routine procedure during prenatal care for all patients because HBV infection is of low endemicity in the general population of the United States. However, it is highly endemic among certain population subgroups. Because the lifetime risk of contracting HBV varies greatly (from almost 100% for the high-risk groups to approximately 5% for the population as a whole), the Centers for Disease Control (CDC) have recommended routine screening for pregnant women who belong to high-risk groups.! It is unclear whether screening only members of high-risk groups effectively identifies the majority of seropositive women. In 1973 Szmuness et al. 6 interviewed blood donors who were HBsAg positive and found that "only a minority gave a past history that could have been related to the mode of acquisition of hepatitis B infection." Other researchers, including a group from Shands Hospital in 1973, found that a majority of hospital patients who were HBsAg positive did not have clinical hepatitis and were unaware of their infectious condition.7.8 In 1986 Malecki et al. 9 reported the results of screening all maternity patients attending four county health department clinics in south Florida. They found that 67% of women with positive HBsAg screens had no risk factors identifiable by chart review or interview. Their sample was small; of 741 women tested, eight had positive test results. However, their results are consistent with the finding that only one third of persons with hepatitis B antibodies can recall a history of clinical symptoms, suggesting that the infection frequently may be nonspecific or asymptomatic.!O Low rates of detection because of the lack of reported historic risk factors in pregnant women are of concern. Although chronic infection of HBV is relatively uncommon among immunologically competent adults,!! infection in neonates usually leads to the chronic carrier state and increases the risk of viral transmission to others. Those infected as neonates and infants are also at risk for chronic liver disease and hepatocellular carcinoma? Cases of fatal hepatitis B infection have been reported in infants born to women with no known risk factors. s, !2 Results from studies indicate that perinatal transmission occurs most often during delivery, although there is a 75% chance of the fetus becoming HBsAg positive following an acute infection in the mother during the third trimester.!! Therefore the majority of infants born to carrier mothers should be HBsAg seronegative at birth," and prophylactic treatment should effectively prevent infection. The CDC recommends treatment as soon as physiologic stabili-
Prenatal screening for hepatitis 8
1181
zation has occurred, preferably within 12 hours after delivery. Knowledge of each woman's HBsAg status before she is admitted for delivery is important. Safety of staff and other patients is contingent on isolation of seropositive women and their unwashed neonates. Careful handling of all objects that may have come into contact with infected blood is important because infectivity is retained by blood that has dried on environmental surfaces.!4 Among physicians, surgeons are at highest risk for infection.!S In the delivery room, exposure to blood during both vaginal and surgical deliveries is high, not only for the obstetrician, but for nursing and pediatric staffs as well. The purpose of this study was to conduct a retrospective evaluation of the effectiveness of a screening program for hepatitis B in the obstetric population served by Shands Hospital of the University of Florida. Screening for the HBsAg has been part of the hospital's admission policy for all patients admitted to the adult inpatient service since 1981, with laboratory data stored in a computerized data base. Charts of obstetric patients were reviewed to collect data on prevalence of hepatitis B infection, to correlate positive serologic findings with known risk factors, and to determine the time relationship between screening and prophylactic treatment of exposed neonates. Subjects and methods
This retrospective study covered the period January 1, 1983, through December 31, 1985, during which time 21,517 patients between the ages of 14 and 44 years were admitted to Shands Hospital and screened for HBsAg. Of these, 7962 (37%) were obstetric patients. All patients with positive HBsAg findings were identified for this period. The medical record for each obstetric patient with a positive HBsAg finding was then reviewed to determine whether risk factors as defined by the CDC were identifiable by routine prenatal history. Risk factors considered were (1) Asian, Pacific Island, Alaskan Eskimo, Haitian, or Sub-Saharan African descent; (2) birth or travel in countries outside the United States, where the incidence of hepatitis is higher; (3) acute or chronic liver disease; (4) work or residence in an institution for the mentally retarded; (5) frequent blood transfusions; (6) frequent occupational exposure to blood; (7) household contact with an HBV carrier; (8) multiple episodes of venereal disease; (9) illicit intravenous drug use; (10) rejection as a blood donor; and (11) work or treatment in a hemodialysis unit. The immunology laboratory used an enzyme-linked immunosorbent assay to determine the presence of HBsAg. All test results were confirmed with controls.
1182 Cruz, Frentzen, and Behnke
Results of anti-HBs, HBeAg, and anti-HBe were reviewed when available. The medical records of infants born to HBsAg-positive mothers were then reviewed to determine whether the time of detection of seropositivity was early enough to follow CDC guidelines for neonatal prophylaxis. Results
Of the 21,517 patients who were screened, 192 (0.89%) had positive results of antigen tests. Of those with positive screening data, 43 (22%) were in the obstetric group. Thus the prevalence rate was 0.54% for obstetric patients and 1.09% for nonobstetric patients. When all 43 obstetric charts were reviewed, only 14 of these women (33%) had risk factors that were identifiable by patient history: two were Asian; three were from countries with higher hepatitis rates (Iran, Jamaica, and Mexico); three reported prior cases of hepatitis B; two worked at an institution for the mentally retarded and another was mentally retarded but prior institutionalization was unclear; one reported multiple episodes of sexually transmitted diseases; and two stated they had used intravenous drugs. Eight additional women had no identifiable risk factors but were found to have HBsAg-positive results during one pregnancy early in the study and thus were known to be carriers during subsequent pregnancies occurring within the study period. Therefore slightly less than one third of these patients could have been identified by patient history and the CDC criteria. Of the 43 women with positive findings for HBsAg, only 13 were identified as potential carriers and screened during the prenatal period. The remaining 30 were screened for HBsAg for the first time on hospital admission. Blood was drawn for assay as soon as possible following the decision to admit the patient to labor and delivery. However, because the laboratory runs its immunoassay procedures only once daily, results of the tests were frequently received after the patients had delivered. Isolation procedures were not instituted in these cases and staff exposure to infected blood was high. Of the 43 women with HBsAg-positive findings, 33 were also screened for HBeAg, which indicates greater infectivity potential.' Findings were positive in 12 of these women; four (33%) of them were in the subgroup that could be identified by routine prenatal history. Neonatal treatment protocols at birth followed CDC recommendations. This included injection of HBIG within 12 hours after delivery. The protocol was revised twice to conform with the CDC recommended timing of the first hepatitis B vaccine injection: from 3 months of age (protocol No.1) to within 7 days of birth (protocol No.2) to within 12 hours of delivery (protocol No.3). There were 45 singleton deliveries from 40 women
May 1987 Am J Obstet Gynecol
who had positive screens. Two of the remaining women spontaneously aborted; the third delivered at another facility and was lost to follow-up. Twelve infants were treated by protocol No.1, 16 by protocol No.2, and 13 by protocol No.3. Treatment was not documented in four infants. One infant (2%) tested HBsAg positive at birth. There was no documentation of acute maternal disease during the pregnancy; however, the mother was Asian and HBeAg positive. One infant (2%), treated only with HBIG and then lost to follow-up for 8 months, was found to be a chronic carrier with HBeAg positivity. All other treated infants were HBsAg negative after completion of treatment. As a result of the lag time between blood draw and immunoassay result, only eight infants (18%) were treated within 12 hours after birth. The time from birth to initial injection ranged from 4 to 156 hours. Comment
To prevent the development of acute hepatitis B infection and the carrier state in neonates whose mothers carry the virus, HBIG and hepatitis B vaccine should be administered as soon as possible after delivery. This can be accomplished only by identifying mothers with positive antigen before delivery. In our study population, 37 of the 45 exposed infants did not receive treatment within the 12-hour period recommended by the CDC because positive serologic results from screening of the mothers on admission to the hospital were not available in time. In our population of women, if only those for whom risk factors were apparent had been tested, 67% of positive HBsAg carrier mothers would have been missed. This percentage of unidentified patients correlates well with the findings of Szmuness et al. 6 and Malecki et al. 9 This is particularly significant when one considers that more than one third of those tested for HBeAg had positive findings and had a high probability of infecting their neonates. It can therefore be concluded that two thirds of patients who carry the HBsAg can not be identified as high risk for HBV infection by history alone. Detection of maternal HBsAg seropositivity before delivery is necessary for appropriate isolation and timely prophylactic treatment of exposed neonates. Therefore prenatal screening before hospital admission for all patients is recommended where the patient population is predominantly of low socioeconomic status as in our patient group. Since repeat testing for gonorrhea and syphilis is done at 34 weeks' gestation, this would be a convenient time to include hepatitis screening as part of the prenatal evaluation. Cost-effectiveness must be considered in every screening recommendation. The average cost of an HBsAg screening in our community is $27.60. For the individual patient this amount is small when the total
Prenatal screening for hepatitis B
Volume 156 Number 5
cost of pregnancy care is considered. However, for the general population the cost is significant. The cost of identifying each seropositive patient not otherwise identified by known risk factors would be $7562.40. The cost is high; however, screening is critical in the effort to prevent both neonatal infection and the carrier state with its long-term morbidity. REFERENCES 1. Centers for Disease Control. Recommendation for protection against viral hepatitis. MMWR 1985;34:313-35. 2. Stevens CD, Toy PT, Tong MJ, et al. Perinatal hepatitis B transmission in the United States. JAMA 1985; 253:1740-5. 3. Lee AK, Ip HM, Wong VC. Mechanisms of maternal-fetal transmission of hepatitis B virus. J Infect Dis 1978; 138:668-71. 4. Reesink HW, Reerink-Brongers EE, Lafeber-Schut BJ, et al. Prevention of chronic HBsAg carrier state in infants of HBsAg-positive mothers by hepatitis B immunoglobulin. Lancet 1979;2:436-8. 5. Delaplane D, Yogev R, Crussi F, et al. Fatal hepatitis B in early infancy: the importance of identifying HBsAgpositive pregnant women and providing immunoprophylaxis to their newborns. Pediatrics 1983;72: 176-80. 6. Szmuness W, Prince AM, Brotman B, et al. Hepatitis B antigen and antibody jn blood donors: an epidemiological study. J Infect Dis 1973;127:17-25.
7. Linnemann CC, Hegg ME, Remundo N, et al. Screening hospital patients for hepatitis B surface antigen. Am J Clin Pathol 1977;67:257-9. 8. Mahoney J, Richman A, Teague P. Admission screening for hepatitis B surface antigen in a university hospit~1. South Med J 1978;71 :624-8. 9. Malecki MJ, Guarin O~ Hulbert A, et al. Prevalence of hepatitis Bsurface antigen among women receiving prenatal care at the Palm Beach County Health Department. AMJ OBSTET GYNECOL 1986;154:625-6. 10. Hollinger FB, Melnick JL. Epi4emiology. In: Hollinger FB, Melnick JL, Robinson WS, eds. Viral hepatitis. New York: Raven Press, 1985:101-46. 11. Tong MJ, Thursby MW, Lin J-H, et a\. Studies on the maternal-infant transmission of the hepatitis B virus and HBV infection within families. Prog Med Virol 1980; 27:137-47. 12. Ewing cr, Davidson DC. Fatal hepatitis B in infant born to a HBsAg carrier with HBeAb. Arch Dis Child 198::>; 27:137-47. 13. Okada K, Yamada T, Miyakawa Y, et al. Hepatitis B surface antigen in the serum of infants after delivery from asymptomatic carrier mothers. J Pediatr 1975;87:360-3. 14. Bond WW, Favero MS, Peterson NJ, et al. Survival of hepatitis B virus after drying and storage for one week. Lancet 1981; 1:550-1. 15. Denes AE, Smith JL, Maynard JE, et al. Hepatitis B infection in physicians. JAMA 1978;239:210-2.
The reliability of perinatal mortality statistics in The Netherlands Johannes P. R. Doornbos, M.D., Hendrik J. Nordbeck, M.D., and Pieter E. Treffers, M.D. Amsterdam, The Netherlands All 13 hospitals in the municipality of Amsterdam cooperated in this study; labor ward records for the years 1981 and 1982 were personally searched for cases of perinatal deCith that corresponded with the World Health Organization definitions of perinatal mortality, presently applied in The Netherlands. Th~ 360 cases that had apparently occurred in the Amsterdam hospitals were individually linked to the cases that had been entered into the national records at the Central Bureau of Statistics. The study established that the magnitude of underregistration of perinatal mortality is considerable (14.3%). Underregistration at the national level is dl!e to underreporting by physicians and not due to errors in statistical bookkeeping. Underreporting is related to birth weight, that is, viability of the infant, and immigrant status of the mother. Recommendations are made to improve uniformity of notification criteria and reliability of perinatal mortality statistics. (AM J OSSTET GVNECOL 1987;156:1183:7.)
Key words: Perinatal mortality, records, vital statistics
From the Department of Obstetrics and Gynecology, Academic Medical Center of the University of Amsterdam, and the Department of Tropical Hygiene, Royal Tropical Institute. Received for publication July 29, 1986; accepted December 17,
1986. Reprint requests: Prof P. E. Treffers, Department of Obstetrics and Gynecology, Academic Medical Center of the University of Amsterdam, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands.
Perinatal mortality rates are widely used as indicators of the outcome of pregnancy and childbirth and of the quality of maternal and neonatal care. At the same time their usefulness and validity are being questioned for a number of reasons. Obviously, an unequivocal definition of perinatal mortality is a prerequisite for any application. A number of World Health 1183
bilateral selective embolization of the uterine arteries was performed with polyvinyl alcohol (Ivalon) particles mixed with thrombin (Topostasine). The postembolization period was uncomplicated. Control intravenous subtraction angiography demonstrated total obliteration of the left uterine artery. A small part of the arteriovenous malformation was still filled through the right uterine artery. During a follow-up period of 6 months no abnormal bleeding was noticed. The patient had normal, regular menstrual cycles and conceived after the sixth postembolization month. During the first and second trimesters of pregnancy intermittent periods of slight vaginal bleeding were observed. There was no cervical infection; there was no placenta previa, and no signs of abruptio placentae were seen. The placenta was implanted on the left ventrolateral side. Clinical and ultrasonographic evaluation showed normal fetal growth. Doppler measurements showed normal flow at both the uterine and umbilical fetal sites. Except for rest, we did not advise any special treatment. The patient went into sudden premature labor at 35 weeks and was delivered of a normal male infant weighing 2600 gm. Macroscopic and microscopic ex-
May 1987 Am J Obstet Gynecol
amination of the placenta was normal, and the postpartum period was uncomplicated.
Comment In this case, although the left uterine artery was totally obliterated and the right uterine artery was partially obliterated, there was a normal adaptation of the uteroplacental vasculature resulting in optimal fetal growth and development, even with the placenta located at the site of the embolized arteriovenous malformation. The case report presented shows that conservative treatment of an arteriovenous malformation can stop the heavy vaginal bleeding and preserve the capability of reproduction. REFERENCES 1. Liggins GC. Uterine arteriovenous fistula. Obstet Gynecol 1964;23:214-7. 2. Wilms GE, Favril A, Baert AL, Poppe W, Van Assche FA. Transcatheter embolization of uterine arteriovenous malformations. Cardiovasc Intervent Radiol 1986;9:61-4.
Hepatitis B: A case for prenatal screening of all patients Amelia C. Cruz, M.D., Barbara H. Frentzen, M.S.N., and Marylou Behnke, M.D. Gainesville, Florida Evaluation of a mass screening program to detect hepatitis B surface antigen in the obstetric population of Shands Hospital, University of Florida in Gainesville, from January 1, 1983, through December 31, 1985, was undertaken. Prevalence of hepatitis B surface antigen seropositivity was 0.54%. Review of medical records revealed that 67% of patients with positive hepatitis B surface antigen screens had no risk factors identifiable by routine prenatal history. Although all patients were screened on admission to the hospital, the results of the immunoassay were not available in time for staff to institute isolation procedures for the laboring mother or unwashed neonate. In 82% of the cases, treatment of neonates occurred later than 12 hours after delivery (the current Centers for Disease Control recommendation). It is recommended that prenatal screening of all patients with a hepatitis B surface antigen immunoassay be done by 34 weeks' gestation where the patient population is of predominantly low socioeconomic status to ensure appropriate isolation and timely neonatal immunoprophylaxis. (AM J OSSTET GVNECOl 1987;156:1180-3.)
Key words: Hepatitis B, prenatal screening, neonatal immunoprophylaxis
From the Departments of Obstetrics and Gynecology and Pediatrics, University of Florida College of Medicine. Supported entirely, or in part, under an agreement with the Department of Health and Rehabilitative Services, State of Florida, Children's Medical Services. Receivedforpublication August 5, 1986; revised November 3,1986; accepted November 17, 1986. Reprint requests: Amelia C. Cruz, M.D., Department of Obstetrics and Gynecology, Box J-294 JHMHC, University of Florida, Gainesville, FL 32610.
Perinatal exposure of neonates to the hepatitis B surface antigen (HBsAg) and e antigen (HBeAg) by mothers who are seropositive, but clinically symptom free, is known to be a major source of hepatitis B virus (HB V) infection. I·' Treatment at birth with hepatitis B immunoglobulin (HBIG) and vaccine has been shown to dramatically reduce the incidence of acute and chronic infection in neonates. I. 4. 5 Therefore identification of all pregnant women who are HBsAg carriers is nec-
Volume 156 Number 5
essary to identify and prophylactically treat exposed neonates after delivery. Serum screening for HBsAg is not a routine procedure during prenatal care for all patients because HBV infection is of low endemicity in the general population of the United States. However, it is highly endemic among certain population subgroups. Because the lifetime risk of contracting HBV varies greatly (from almost 100% for the high-risk groups to approximately 5% for the population as a whole), the Centers for Disease Control (CDC) have recommended routine screening for pregnant women who belong to high-risk groups.! It is unclear whether screening only members of high-risk groups effectively identifies the majority of seropositive women. In 1973 Szmuness et al. 6 interviewed blood donors who were HBsAg positive and found that "only a minority gave a past history that could have been related to the mode of acquisition of hepatitis B infection." Other researchers, including a group from Shands Hospital in 1973, found that a majority of hospital patients who were HBsAg positive did not have clinical hepatitis and were unaware of their infectious condition.7.8 In 1986 Malecki et al. 9 reported the results of screening all maternity patients attending four county health department clinics in south Florida. They found that 67% of women with positive HBsAg screens had no risk factors identifiable by chart review or interview. Their sample was small; of 741 women tested, eight had positive test results. However, their results are consistent with the finding that only one third of persons with hepatitis B antibodies can recall a history of clinical symptoms, suggesting that the infection frequently may be nonspecific or asymptomatic.!O Low rates of detection because of the lack of reported historic risk factors in pregnant women are of concern. Although chronic infection of HBV is relatively uncommon among immunologically competent adults,!! infection in neonates usually leads to the chronic carrier state and increases the risk of viral transmission to others. Those infected as neonates and infants are also at risk for chronic liver disease and hepatocellular carcinoma? Cases of fatal hepatitis B infection have been reported in infants born to women with no known risk factors. s, !2 Results from studies indicate that perinatal transmission occurs most often during delivery, although there is a 75% chance of the fetus becoming HBsAg positive following an acute infection in the mother during the third trimester.!! Therefore the majority of infants born to carrier mothers should be HBsAg seronegative at birth," and prophylactic treatment should effectively prevent infection. The CDC recommends treatment as soon as physiologic stabili-
Prenatal screening for hepatitis 8
1181
zation has occurred, preferably within 12 hours after delivery. Knowledge of each woman's HBsAg status before she is admitted for delivery is important. Safety of staff and other patients is contingent on isolation of seropositive women and their unwashed neonates. Careful handling of all objects that may have come into contact with infected blood is important because infectivity is retained by blood that has dried on environmental surfaces.!4 Among physicians, surgeons are at highest risk for infection.!S In the delivery room, exposure to blood during both vaginal and surgical deliveries is high, not only for the obstetrician, but for nursing and pediatric staffs as well. The purpose of this study was to conduct a retrospective evaluation of the effectiveness of a screening program for hepatitis B in the obstetric population served by Shands Hospital of the University of Florida. Screening for the HBsAg has been part of the hospital's admission policy for all patients admitted to the adult inpatient service since 1981, with laboratory data stored in a computerized data base. Charts of obstetric patients were reviewed to collect data on prevalence of hepatitis B infection, to correlate positive serologic findings with known risk factors, and to determine the time relationship between screening and prophylactic treatment of exposed neonates. Subjects and methods
This retrospective study covered the period January 1, 1983, through December 31, 1985, during which time 21,517 patients between the ages of 14 and 44 years were admitted to Shands Hospital and screened for HBsAg. Of these, 7962 (37%) were obstetric patients. All patients with positive HBsAg findings were identified for this period. The medical record for each obstetric patient with a positive HBsAg finding was then reviewed to determine whether risk factors as defined by the CDC were identifiable by routine prenatal history. Risk factors considered were (1) Asian, Pacific Island, Alaskan Eskimo, Haitian, or Sub-Saharan African descent; (2) birth or travel in countries outside the United States, where the incidence of hepatitis is higher; (3) acute or chronic liver disease; (4) work or residence in an institution for the mentally retarded; (5) frequent blood transfusions; (6) frequent occupational exposure to blood; (7) household contact with an HBV carrier; (8) multiple episodes of venereal disease; (9) illicit intravenous drug use; (10) rejection as a blood donor; and (11) work or treatment in a hemodialysis unit. The immunology laboratory used an enzyme-linked immunosorbent assay to determine the presence of HBsAg. All test results were confirmed with controls.
1182 Cruz, Frentzen, and Behnke
Results of anti-HBs, HBeAg, and anti-HBe were reviewed when available. The medical records of infants born to HBsAg-positive mothers were then reviewed to determine whether the time of detection of seropositivity was early enough to follow CDC guidelines for neonatal prophylaxis. Results
Of the 21,517 patients who were screened, 192 (0.89%) had positive results of antigen tests. Of those with positive screening data, 43 (22%) were in the obstetric group. Thus the prevalence rate was 0.54% for obstetric patients and 1.09% for nonobstetric patients. When all 43 obstetric charts were reviewed, only 14 of these women (33%) had risk factors that were identifiable by patient history: two were Asian; three were from countries with higher hepatitis rates (Iran, Jamaica, and Mexico); three reported prior cases of hepatitis B; two worked at an institution for the mentally retarded and another was mentally retarded but prior institutionalization was unclear; one reported multiple episodes of sexually transmitted diseases; and two stated they had used intravenous drugs. Eight additional women had no identifiable risk factors but were found to have HBsAg-positive results during one pregnancy early in the study and thus were known to be carriers during subsequent pregnancies occurring within the study period. Therefore slightly less than one third of these patients could have been identified by patient history and the CDC criteria. Of the 43 women with positive findings for HBsAg, only 13 were identified as potential carriers and screened during the prenatal period. The remaining 30 were screened for HBsAg for the first time on hospital admission. Blood was drawn for assay as soon as possible following the decision to admit the patient to labor and delivery. However, because the laboratory runs its immunoassay procedures only once daily, results of the tests were frequently received after the patients had delivered. Isolation procedures were not instituted in these cases and staff exposure to infected blood was high. Of the 43 women with HBsAg-positive findings, 33 were also screened for HBeAg, which indicates greater infectivity potential.' Findings were positive in 12 of these women; four (33%) of them were in the subgroup that could be identified by routine prenatal history. Neonatal treatment protocols at birth followed CDC recommendations. This included injection of HBIG within 12 hours after delivery. The protocol was revised twice to conform with the CDC recommended timing of the first hepatitis B vaccine injection: from 3 months of age (protocol No.1) to within 7 days of birth (protocol No.2) to within 12 hours of delivery (protocol No.3). There were 45 singleton deliveries from 40 women
May 1987 Am J Obstet Gynecol
who had positive screens. Two of the remaining women spontaneously aborted; the third delivered at another facility and was lost to follow-up. Twelve infants were treated by protocol No.1, 16 by protocol No.2, and 13 by protocol No.3. Treatment was not documented in four infants. One infant (2%) tested HBsAg positive at birth. There was no documentation of acute maternal disease during the pregnancy; however, the mother was Asian and HBeAg positive. One infant (2%), treated only with HBIG and then lost to follow-up for 8 months, was found to be a chronic carrier with HBeAg positivity. All other treated infants were HBsAg negative after completion of treatment. As a result of the lag time between blood draw and immunoassay result, only eight infants (18%) were treated within 12 hours after birth. The time from birth to initial injection ranged from 4 to 156 hours. Comment
To prevent the development of acute hepatitis B infection and the carrier state in neonates whose mothers carry the virus, HBIG and hepatitis B vaccine should be administered as soon as possible after delivery. This can be accomplished only by identifying mothers with positive antigen before delivery. In our study population, 37 of the 45 exposed infants did not receive treatment within the 12-hour period recommended by the CDC because positive serologic results from screening of the mothers on admission to the hospital were not available in time. In our population of women, if only those for whom risk factors were apparent had been tested, 67% of positive HBsAg carrier mothers would have been missed. This percentage of unidentified patients correlates well with the findings of Szmuness et al. 6 and Malecki et al. 9 This is particularly significant when one considers that more than one third of those tested for HBeAg had positive findings and had a high probability of infecting their neonates. It can therefore be concluded that two thirds of patients who carry the HBsAg can not be identified as high risk for HBV infection by history alone. Detection of maternal HBsAg seropositivity before delivery is necessary for appropriate isolation and timely prophylactic treatment of exposed neonates. Therefore prenatal screening before hospital admission for all patients is recommended where the patient population is predominantly of low socioeconomic status as in our patient group. Since repeat testing for gonorrhea and syphilis is done at 34 weeks' gestation, this would be a convenient time to include hepatitis screening as part of the prenatal evaluation. Cost-effectiveness must be considered in every screening recommendation. The average cost of an HBsAg screening in our community is $27.60. For the individual patient this amount is small when the total
Prenatal screening for hepatitis B
Volume 156 Number 5
cost of pregnancy care is considered. However, for the general population the cost is significant. The cost of identifying each seropositive patient not otherwise identified by known risk factors would be $7562.40. The cost is high; however, screening is critical in the effort to prevent both neonatal infection and the carrier state with its long-term morbidity. REFERENCES 1. Centers for Disease Control. Recommendation for protection against viral hepatitis. MMWR 1985;34:313-35. 2. Stevens CD, Toy PT, Tong MJ, et al. Perinatal hepatitis B transmission in the United States. JAMA 1985; 253:1740-5. 3. Lee AK, Ip HM, Wong VC. Mechanisms of maternal-fetal transmission of hepatitis B virus. J Infect Dis 1978; 138:668-71. 4. Reesink HW, Reerink-Brongers EE, Lafeber-Schut BJ, et al. Prevention of chronic HBsAg carrier state in infants of HBsAg-positive mothers by hepatitis B immunoglobulin. Lancet 1979;2:436-8. 5. Delaplane D, Yogev R, Crussi F, et al. Fatal hepatitis B in early infancy: the importance of identifying HBsAgpositive pregnant women and providing immunoprophylaxis to their newborns. Pediatrics 1983;72: 176-80. 6. Szmuness W, Prince AM, Brotman B, et al. Hepatitis B antigen and antibody jn blood donors: an epidemiological study. J Infect Dis 1973;127:17-25.
7. Linnemann CC, Hegg ME, Remundo N, et al. Screening hospital patients for hepatitis B surface antigen. Am J Clin Pathol 1977;67:257-9. 8. Mahoney J, Richman A, Teague P. Admission screening for hepatitis B surface antigen in a university hospit~1. South Med J 1978;71 :624-8. 9. Malecki MJ, Guarin O~ Hulbert A, et al. Prevalence of hepatitis Bsurface antigen among women receiving prenatal care at the Palm Beach County Health Department. AMJ OBSTET GYNECOL 1986;154:625-6. 10. Hollinger FB, Melnick JL. Epi4emiology. In: Hollinger FB, Melnick JL, Robinson WS, eds. Viral hepatitis. New York: Raven Press, 1985:101-46. 11. Tong MJ, Thursby MW, Lin J-H, et a\. Studies on the maternal-infant transmission of the hepatitis B virus and HBV infection within families. Prog Med Virol 1980; 27:137-47. 12. Ewing cr, Davidson DC. Fatal hepatitis B in infant born to a HBsAg carrier with HBeAb. Arch Dis Child 198::>; 27:137-47. 13. Okada K, Yamada T, Miyakawa Y, et al. Hepatitis B surface antigen in the serum of infants after delivery from asymptomatic carrier mothers. J Pediatr 1975;87:360-3. 14. Bond WW, Favero MS, Peterson NJ, et al. Survival of hepatitis B virus after drying and storage for one week. Lancet 1981; 1:550-1. 15. Denes AE, Smith JL, Maynard JE, et al. Hepatitis B infection in physicians. JAMA 1978;239:210-2.
The reliability of perinatal mortality statistics in The Netherlands Johannes P. R. Doornbos, M.D., Hendrik J. Nordbeck, M.D., and Pieter E. Treffers, M.D. Amsterdam, The Netherlands All 13 hospitals in the municipality of Amsterdam cooperated in this study; labor ward records for the years 1981 and 1982 were personally searched for cases of perinatal deCith that corresponded with the World Health Organization definitions of perinatal mortality, presently applied in The Netherlands. Th~ 360 cases that had apparently occurred in the Amsterdam hospitals were individually linked to the cases that had been entered into the national records at the Central Bureau of Statistics. The study established that the magnitude of underregistration of perinatal mortality is considerable (14.3%). Underregistration at the national level is dl!e to underreporting by physicians and not due to errors in statistical bookkeeping. Underreporting is related to birth weight, that is, viability of the infant, and immigrant status of the mother. Recommendations are made to improve uniformity of notification criteria and reliability of perinatal mortality statistics. (AM J OSSTET GVNECOL 1987;156:1183:7.)
Key words: Perinatal mortality, records, vital statistics
From the Department of Obstetrics and Gynecology, Academic Medical Center of the University of Amsterdam, and the Department of Tropical Hygiene, Royal Tropical Institute. Received for publication July 29, 1986; accepted December 17,
1986. Reprint requests: Prof P. E. Treffers, Department of Obstetrics and Gynecology, Academic Medical Center of the University of Amsterdam, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands.
Perinatal mortality rates are widely used as indicators of the outcome of pregnancy and childbirth and of the quality of maternal and neonatal care. At the same time their usefulness and validity are being questioned for a number of reasons. Obviously, an unequivocal definition of perinatal mortality is a prerequisite for any application. A number of World Health 1183