Control of rubella and congenital rubella syndrome (CRS) in developing countries, Part 2: Vaccination against rubella.

Control of rubella and congenital rubella syndrome (CRS) in developing countries, part 2: vaccination against rubella S.E. Robertson,1 F.T. Cutts,2 R. Samuel,3 & J.-L. Diaz-Ortega4 In 1995-96 we conducted a review of rubella immunization strategies. Worldwide, 78 countries (more than one-third) reported a national policy of using rubella vaccine. This was closely related to country economic status. Based on the United Nations country classification, rubella vaccine is used in 92% of industrialized countries, 36% of those with economies-in-transition, and 28% of developing countries. Cases of congenital rubella syndrome (CRS) may be prevented as follows: by providing direct protection to women and/or schoolgirls (a selective vaccination strategy); by vaccinating boys and girls to provide indirect protection by reducing the transmission of rubella virus (a childhood vaccination strategy); or by a combination of these approaches (a combined strategy). A combined strategy was most commonly reported (60% of countries); seven countries (9%) reported a selective strategy; and 24 countries (31%) reported only childhood immunization. Experience has shown that it is essential to include vaccination of women of childbearing age in any rubella control strategy. Childhood vaccination alone may pose a risk of an increase in CRS cases. Although many countries have introduced rubella vaccine, few report any data on the impact of vaccination. Countries using rubella vaccine need to establish surveillance for rubella and CRS and monitor coverage in each of the target groups. Introduction Live attenuated rubella vaccines were licensed in the USA in 1969 and introduced throughout much of the industrialized world soon afterwards. Rubella vaccine has, however, not been recommended for inclusion in the Expanded Programme on Immuni- zation (EPI) in developing countries (1) since when sustained high coverage cannot be guaranteed its introduction could increase the susceptibility of adult women by slowing, but not interrupting, ru- bella transmission (2). Even without a global recom- mendation, some countries have added rubella vaccine to their national immunization programmes, reflecting the high coverage levels (>80%. often >90%) with childhood vaccines in these countries, Medical Officer, Global Programme for Vaccines and Immuniza- tion, World Health Organization, 1211 Geneva 27, Switzerland. Requests for reprints should be sent to this author. 2 Senior Lecturer in Communicable Disease Epidemiology, Lon- don School of Hygiene and Tropical Medicine, London, England. 3Research Officer, Departments of Clinical Virology and Com- munity Health Christian Medical College and Hospital, Vellore, India; and DSc Fellow, Netherlands Institute for Health Sciences, Erasmus University Medical School, Rotterdam, Netherlands. 4Director, Research, Training, and Supervision, National Vaccina- tion Council, Ministry of Health, Mexico City, Mexico. Reprint No. 5755 as well as a national response to studies documenting the burden of congenital rubella syndrome (CRS) (3). There is considerable documentation of the bur- den of disease related to CRS in some developing countries (3). Approximately 50 developing coun- tries have already conducted substantial studies to assess their CRS disease burden. For countries that have not yet done so, part I of this review provides guidance on various methods suitable for surveil- lance of CRS. In part 2, we present information on the current use of rubella vaccine in different WHO regions, with emphasis on developing countries. Summarized are lessons learned about the effect of different vaccination policies on rubella and CRS control, and recommendations are made for devel- oping comprehensive rubella control programmes in those countries that have or are considering a rubella vaccination policy. Methods We conducted a literature review of congenital ru- bella syndrome. acquired rubella, and rubella vac- cine in developing countries as well as referring to key articles on rubella immunization in industrial- ized countries. In addition, we consulted UNICEF, which facilitates vaccine purchases for the poorest countries, about rubella vaccine prices. Bulletin of the World Health Organization. 1997, 75 (1): 69-80 © World Health Organization 1997 69 S.E. Robertson et al. In 1995, the WHO Global Programme for Vaccines and Immunization carried out a global sur- vey of rubella immunization policies. Information was gathered from national governments, and up- dates as reported to WHO by June 1996 are included in this review. We classified immunization strategies according to the primary target group. Cases of CRS may be prevented as follows: by directly protecting women and/or schoolgirls (a selective vaccination strategy); by vaccinating boys and girls in childhood to provide indirect protection by reducing the trans- mission of rubella virus (a childhood vaccination strategy); or by a combination of these approaches (a combined strategy). To examine the effect of economic status on use of rubella vaccine, we classified countries accord- ing to two schemes; first, using a United Nations scheme, which considers countries as industrialized, economies-in-transition (including Eastern European countries and most of the newly independent states of the former Soviet Union), or developing (4); second, according to categories for vaccine support developed by the WHO Global Programme for Vaccines and Immunization (5). The latter classifi- cation combines information on population and income to assess the capacity of a country to be self- sufficient in vaccine supply; band A consists of small low-income countries that should continue to receive support for vaccines from donors: and band B con- sists of countries that currently need donor support, but which will gradually assume financial responsi- bility for vaccine purchases. Bands C and D include countries that are already able to procure vaccines independently, while band E consists of high income countries. Results Rubella vaccine Between 1965 and 1967, several live, attenuated ru- bella strains were developed, and most industrialized countries began rubella vaccination soon afterwards. The live rubella vaccine currently distributed in most countries contains the RA 27/3 strain, which is pre- pared in human diploid cell culture (6). The vaccine is produced in monovalent form (rubella only), and in the following combinations: measles-rubella (MR) vaccine and measles-mumps-rubella (MMR) vaccine. The 1996 UNICEF discounted prices per dose for the neediest countries (those in bands A and B) were US$ 0.15 (monovalent rubella vaccine), US$ 0.55-0.59 (MR vaccine), US$ 0.72-0.95 (MMR vac- cine) (J. Gilmartin, personal communication, 1996). However, UNICEF reports that the total annual number of doses ordered was <500000 over the pe- riod 1992-95, suggesting that most rubella vaccine purchases are made using other mechanisms, such as national tenders. For developing countries in the Region of the Americas, PAHO offers rubella- containing vaccines at a discounted price through a revolving fund. Also, in the private sector of some countries, MMR is less expensive than MR, because of the greater demand (S. Migasena, personal com- munication, 1996). In clinical trials, ¢-95% of susceptible persons who received a single dose of rubella vaccine when they were at least 12 months of age developed anti- body (7), and some studies have shown 99-100% seroconversion (8, 9). In India, >95% of children seroconverted after being vaccinated at 9 months of age (10) and, as in South Africa (11), there was no difference in seroconversion rates when rubella vac- cine was given at 9 months or 15 months of age. Clinical efficacy and challenge studies indicate that >90% of vaccinees are protected against both clinical rubella and viraemia for at least 15 years (12- 15), and vaccine-induced protection is generally assumed to be lifelong. Reinfection of antibody- positive individuals has, however, been demon- strated by significant rises in antibody titre after exposure (16), and is more likely among vaccinees than among those with natural immunity (17, 18). Although the risk of infection of the fetus is much lower than after primary infection (19, 20), occa- sional cases of CRS have been reported in babies born to vaccinated women who had previously had at least one antibody-positive report (21). As with other live virus vaccines, there has been concern that rubella vaccine should not be ad- ministered during pregnancy, since it can infect the fetus although the isolation rate is low (3%) for RA27/3 vaccine (22), and the detection of vaccine virus in some aborted fetuses of vaccinated women does not mean that it can multiply to the level re- quired to produce congenital defects (23). England and Wales, Germany, Sweden, and the USA have kept registries of women who inadvertently received rubella vaccine within 3 months of conception and continued their pregnancy (24). None of the 515 infants born to known seronegative women have had congenital anomalies compatible with CRS; thus the observed risk of vaccine-associated CRS is zero. The maximum theoretical risk of vaccine-asso- ciated CRS, based on a binomial distribution of cases with 95% confidence limits, is <1 %, lower than the risk of major malformations among all preg- nancies (2-3%). In the USA, the registry was closed in 1988 (25). Most advisory bodies continue to con- sider pregnancy a contraindication to rubella vac- cination because of the theoretical risk of WHO Bulletin OMS. Vol 75 199770 Vaccination against rubella in developing countries vaccine-associated CRS. However, there is agree- ment that if a pregnant women is vaccinated inad- vertently, the risk of vaccine-associated defects is so small as to be negligible, and should not ordinarily be a reason to consider interruption of pregnancy (26). A review of adverse events associated with vac- cination carried out in 1991 by the U.S. Institute of Medicine concluded that available evidence was con- sistent with a causal relationship between RA27/3 vaccination and transient acute arthritis (27). Self- limiting joint symptoms that resolve spontaneously are estimated to occur among 13-15% of women following vaccination with a dose of RA27/3 rubella vaccine, with much lower frequencies among chil- dren, adolescents, and adult men. The risk of chronic arthritis among women vaccinated with RA27/3 vac- cine has been reported in three recent studies. A double-blind historical cohort study in Israel found no evidence for an association between postpartum rubella vaccination and subsequent development of arthritis (28), nor did a large retrospective cohort study in the USA (29). The preliminary results of a prospective study in Canada that provided postpartum rubella vaccine or placebo indicate no significant difference in the subsequent occurrence of arthritis/arthralgia in these two groups (30). Current use of rubella vaccine in national immunization programmes Worldwide, 78 countries (> one-third of all coun- tries) reported a national policy of rubella vaccina- tion (Fig. 1, Table 1). This does not include countries where rubella vaccine is only used in certain areas or the private sector. No countries in the African Re- gion include rubella vaccine in their national im- munization schedules. Rubella vaccine is used by approximately half the countries in the Region of the Americas and the Eastern Mediterranean Region, 64% in the European Region, and 31% in the West- ern Pacific Region. In the South-East Asia Region, Sri Lanka and Thailand introduced rubella vaccine in 1996. Of the 78 countries that use rubella vaccine, a combined strategy was reported by 47 (60%); 7 (9%) reported selective immunization of women and/or schoolgirls; and 24 (31%) reported only child- hood immunization. Fig. 1. Countries that include rubella vaccine in their national immunization programme (data reported to WHO up to June 1996). A-, * National programme 0 _. g0 _ 00 * 4 0~~~ &0 WH96880 WHO Bulletin OMS. Vol 75 1997 0 40 10 0 at 71 S.E. Robertson et al. Table 1: Countries/areas using rubella vaccine, based on information reported to WHO up to June 1996 No. of countries/areas using WHO Region rubella vaccine African Americas Eastern Mediterranean European South-East Asia Western Pacific Global total 0/48 (0)b 22/47 (47) 11/23 (48) 32/50 (64) 2/10 (20) 11/36 (31) 78/214 (36) % of population represented by countries/areas using rubella vaccinea 0 43 12 55 5 11C 20 a Ref (4). b Figures in parentheses are percentages. c Excluding China, 46% of the population is represented by coun- tries using rubella vaccine in the Western Pacific Region. Use of rubella vaccine is related to national eco- nomic status. Based on the United Nations country classification (4), rubella vaccine is used by 92% of industrialized countries, 36% of economies-in- transition, and 28% of developing countries. A simi- lar pattern emerges when countries are classified ac- cording to vaccine support criteria (5). Among high-income countries (band E), 92% use rubella vaccine. Among countries that are self-sufficient in vaccine financing (bands C and D), 48% use rubella vaccine. This contrasts with countries that need ex- ternal support for vaccine purchase: only 210% of countries in band B and 4% of the poorest countries (band A) use rubella vaccine. Region of the Americas. In the Americas, 22 coun- tries (representing 43% of the total population in the region) reported national use of rubella vaccine (Table 1). The island countries of the Caribbean were the first developing countries in the Americas to introduce rubella vaccine on a national basis (31). Recently, regional strategies and activities for ru- bella control have interlinked with those of measles elimination. A plan of action for measles elimination by the year 2000 calls for achievement and mainte- nance of 95% measles vaccine coverage in all dis- tricts, with complementary periodic vaccination campaigns and careful surveillance of measles-like illnesses, including laboratory confirmation of the diagnosis (32). A regional measles laboratory net- work has been established, and in many instances sera that test negative for measles are screened for dengue and rubella. In the English-speaking Caribbean (CAREC), laboratory testing of specimens from the rash and fever surveillance system from 1990-95 indicated widespread circulation of rubella virus. A review of all rubella seroprevalence studies in the Caribbean showed that 30-50% of women of childbearing age remain susceptible. By 1996, all CAREC countries should use a rubella-containing vaccine as part of their routine infant immunization schedule, and all will try to include this in the follow-up campaigns of children under 5 years of age conducted for the measles elimination programme (33). The CAREC rubella control policies also include vaccinating all women of childbearing age, vaccinating as many 5-18-year-old schoolchildren as possible, and initia- ting surveillance for CRS and rash-in-pregnancy (33). In Cuba, rubella epidemics occurred every 5-7 years during the pre-vaccine era. In 1982. selective vaccination of 12-year-old girls was introduced, and in 1985-86 mass campaigns were conducted first for 18-30-year-old females (rubella vaccine), followed by those for 1-14-year-old males and females (MR vaccine). Since 1988, MMR vaccine has been given routinely to children at 12 months of age. With this combination of strategies, rubella infection has virtu- ally been eliminated in Cuba. Sao Paulo State, Brazil. has used a similar ap- proach to Cuba, but did not include vaccination of adult women. From the age-specific seroprevalence in a community-based survey in Caieiras city, Sao Paulo, the average age at infection was estimated to be 6 years (34). In 1992, the Brazilian Ministry of Health decided to conduct a nationwide vaccination campaign against measles for children between 9 months and 15 years of age. Sao Paulo State used MMR vaccine for children under 10 years of age in the campaign, and has subsequently implemented routine MMR vaccine for children. One year after the campaign, the average rubella seroprevalence among children aged 1-15 years had increased from 40% to 97% (35). The decrease in rubella transmis- sion in Sao Paulo State has reduced the chance of exposure among pregnant women who are still sus- ceptible, but it is important to continue monitoring rubella transmission among adults as well as the po- tential introduction of rubella from neighbouring states. Eastern Mediterranean Region. In the Eastern Medi- terranean Region 11 countries (12% of the regional population) reported national use of rubella vaccine (Table 1). Interest in vaccination increased during 1990-93 when rubella outbreaks were reported in Bahrain (36), Islamic Republic of Iran (37), Iraq (M.H. Wahdan, personal communication, 1993), Kuwait (36), and Oman (38). Subsequently, retro- spective reviews also identified increases in rubella incidence in Saudi Arabia and the United Arab Emirates (36). In Saudi Arabia, which implements WHO Bulletin OMS. Vol 75 199772 Vaccination against rubella in developing countries postpartum vaccination as well as childhood MMR vaccination, only 9% of 10824 women attending an- tenatal clinics in 1992-93 were seronegative (39). There was an increase in CRS cases, none the less, with 10 neonatal CRS cases being seen at one refer- ral hospital over a 12-month period in 1992-93 com- pared with a total of 27 CRS cases over the previous 11 years. In December 1995, the six Gulf states agreed to adopt standard case definitions for mea- sles, rubella, and CRS; to implement surveillance of rash illness (including serological confirmation of measles and rubella); and to focus efforts on pre- venting CRS through high coverage with a combined rubella immunization strategy (36). The Eastern Mediterranean Region plans to conduct work- shops to examine rubella epidemiology and con- trol strategies for other groups of countries in the region. European Region. In 1984. the European Region es- tablished a goal to eliminate CRS by the year 2000. The operational targets were as follows: all countries should have at least 90% rubella vaccine coverage and effective rubella and CRS surveillance by 1995; and by 1996, all countries should be investigating every suspected CRS case. However, many coun- tries have had difficulty in financing their national vaccination programmes. As of mid-1996, 32 coun- tries in the European Region (representing 56% of the regional population) reported national use of rubella vaccine (Table 1). but only 36% of the Eastern European countries have introduced rubella vaccine. A combined vaccination strategy has operated since 1976 in Croatia (at that time part of Yugoslavia) with 92-96% coverage of <3-year-olds and 96-98% coverage of 14-year-old girls; a large rubella out- break occurred in 1989, in which 9% of reported cases were aged -20 years (40). In 1992-93, rubella outbreaks were reported in several Eastern Euro- pean countries, including Poland (41). and rubella incidence also increased in England and Wales (42). In Israel, rubella vaccination began in 1973, af- ter a major rubella epidemic in 1972 that resulted in 542 confirmed and 739 suspected rubella cases dur- ing pregnancy (43). For the first 6 years. only school- girl vaccination was practised. Despite very high coverage levels, the biggest rubella epidemic ever recorded occurred in 1979, leading to 45 CRS cases (44) and over 480 rubella-associated abortions (45). The proportion of legal abortions due to rubella was 1()% during this epidemic, equal to that in 1972. but fewer CRS cases were documented than in 1972 (46). This epidemic showed the importance of protecting women of childbearing age, and in 1980 postpartum vaccination was introduced and soon extended to all women of childbearing age as well as to special risk groups. By 1983, 75% of seronegative women at fam- ily health clinics were vaccinated, and the overall susceptibility of women of childbearing age fell from >20% to <10%. In the next rubella epidemic, in 1983, only 3% of cases involved women of childbearing age, with the incidence among them being only 21 % of that in 1971-80, and no increase in CRS being seen (47). None the less, outbreaks of rubella continued to occur, especially among adoles- cent males, and in 1989 a combined strategy was adopted by the addition of MMR vaccine at 15 months of age. In 1987, 98% of 220 female military recruits and 88% of male recruits aged 18-19 years were seropositive for rubella (48). Western Pacific Region. In the Western Pacific Re- gion. 11 countries (representing 11 % of the regional population) reported national use of rubella vaccine (Table 1). Fiji, Japan, and Macao conduct selective vaccination of schoolgirls. Malaysia, one of the larger countries in the re- gion (1994 population, 19.7 million) used a different approach. In 1987, a mass campaign of females aged 12-44 years was conducted at numerous sites, includ- ing health clinics, factories, offices, markets and shopping malls. Altogether, 2.3 million doses of rubella vaccine were delivered (49). Rubella immu- nization requirements were introduced for govern- ment workers, female school and university personnel, and all women working in health care. The prevalence of rubella infections, diagnosed through TORCHES screening (Toxoplastna gondii. rubella virus, cytomegalovirus. herpes simplex virus, and syphilis) of newborns (mostly "Xweak and immature newborns"), decreased from 6.8% in 1987 to 20% in 1991 (50). In Singapore, vaccination of 11-year-old school- girls began in 1976, and 96% coverage was reached. Vaccination was extended to schoolboys and service- men in 1982. The immunization programme increased the immunity levels of women of child- bearing age from 56% seropositive in 1975-79, to 850% in 1987. However, the risk of CRS was still high, since 15% remained susceptible. Between 45 and 77 abortions per year were carried out for cases of clini- cal rubella during pregnancy, and an average of five CRS cases were reported per year over the period 1983-87 (51). In Hong Kong, a selective vaccination strategy, introduced in 1978, reduced the incidence of CRS. but recurrent rubella outbreaks showed that spread to pregnant women was still a possibility (Fig. 2). In 1990, Hong Kong introduced a combined rubella immunization strategy (52). WHO Bulletin OMS. Vol 75 1997 73 S.E. Robertson et al. Fig. 2. Annual reported number of cases of rubella and congenital rubella syndrome (CRS) in Hong Kong, 1977-94; MMR: measles-mumps-rubella (ref. 52). n u)0 C3 0 o z co Cuco 0 C.) 0 z Year Discussion A variety of factors may influence the choice of a strategy for rubella vaccination. According to Plotkin, vaccination of all infants will probably eradi- cate CRS in 30-40 years; vaccination of all school- girls will presumably eradicate CRS in 10-20 years; and vaccination of adult women will eradicate CRS immediately, but only if 100% are immunized (6). However, eradication of CRS based on immuniza- tion of a single target group remains to be demon- strated. Moreover, the achievement of high coverage may not be equally easy in each potential target group. Table 2 summarizes the advantages and dis- advantages encountered with the different strategies reviewed in this article. Irrespective of the strategy selected, it is important that high coverage of the selected target groups be attained and sustained. Experience in several countries shows that it is essential to include vaccination of women of childbearing age in any strategy (43, 51, 53, 54). In this respect, it is important to define this age group, taking into account the age-specific fertility pattern of the country, since this will influence the costs of strategies that include vaccination of all women of childbearing age. To avoid the theoretical risk of vaccinating pregnant women, postpartum vaccina- tion has been the usual strategy. In several industri- alized countries antenatal serological screening with postpartum vaccination of seronegative women has been difficult to implement (53,54). It may therefore be more practical to vaccinate all postpartum women without prior screening. The cost of the extra vaccine needs to be balanced against that of screening and the potentially higher coverage achievable without serological screening. However, this would still leave primigravidae unprotected, among whom 40% of reported CRS cases occur in the United Kingdom (55) and 50-60% in the USA (54). Vaccine should, therefore, be offered to all post-pubertal women at all opportunities; the usual procedure being to coun- sel women to avoid becoming pregnant for 3 months after being vaccinated (26). Vaccination of women of childbearing age through routine health services may be feasible in some countries, but in many develop- ing countries would be difficult since a lifetime im- munization record would be needed to determine whether they have been vaccinated previously. An alternative is a mass campaign for women of childbearing age, such as in Cuba and Malaysia. Selective vaccination of teenage girls directly protects the vaccinees, but has very little effect on overall transmission of rubella. In countries that can- not guarantee sustained high coverage of childhood vaccination, this strategy has the advantage of not shifting the average age at infection into child- bearing ages, but it requires high rates of female attendance at school and good liaisons between school and health authorities. Potentially, vaccina- tion against rubella can be conducted at the same time as school-based tetanus toxoid vaccination at a small marginal cost. However, coverage must be monitored since although high coverage of school- girls was attained in Bahrain (36), Hong Kong (52), Israel (47), Kuwait (36), Singapore (51) and Slovenia (A. Kraigher, personal communication, 1996), other industrialized (56,57) and developing countries have had difficulty in attaining this. In Jamaica, where schoolgirl vaccination was introduced in 1978, only 60-70% were immunized in many parishes in 1986 (58), and in the late 1980s, 31 % of 2448 females aged 15-45 years from all parishes were still seronegative by enzyme immunoassay (59). In Trinidad and Tobago, approximately 10 years after the start of schoolgirl vaccination, 410% of antenatal clinic attendees were still seronegative (60). The current rubella outbreaks in Guyana and Jamaica that began in 1995, coupled with evidence of widespread rubella virus circulation throughout the Caribbean, are thus of great concern (33). To eliminate CRS and postnatal rubella, child- hood vaccination needs to be included in national programmes. The introduction of routine childhood vaccination alone, however, is not recommended, since CRS will continue subsequently for 20 years or more, until the cohorts that were vaccinated as chil- dren reach childbearing age. Mathematical models show how intermediate coverage levels for child- hood vaccination can increase the average age at infection and thus the incidence of CRS. For exam- ple, in countries such as the United Kingdom and USA, with an average age at infection pre- vaccination of 9 years, childhood vaccination would lead to an increase in cases of CRS at coverage levels WHO Bulletin OMS. Vol 75 199774 Vaccination against rubella in developing countries Table 2: Advantages and disadvantages of rubella vaccination strategies Strategy Selective vaccination of schoolgirls Selective vaccination of schoolgirls and postpartum women Selective vaccination of schoolgirls and all women of childbearing age Childhood vaccination alone Combined routine immunization of children, schoolgirls, and all women of childbearing age Mass campaign for 1-14-year- olds with routine MR or MMR vaccination of childrenb Mass campaign for women of childbearing age and 1-14-year- olds with routine MR or MMR vaccination of children Advantages Direct protection to future mothers before first pregnancy Relatively inexpensive if school health services already in place Ongoing rubella transmission in children adds to protection As above, plus: direct protection to women of childbearing age with no theoretical risk relating to vaccine in pregnancy. Performed well (e.g. Australia). As above, plus: immediate protection to women of all gravidities Performed well (e.g. Israel) In principle, high coverage may eventually eliminate rubella transmission Impact on CRS potentially immediate In longer term, potential to eliminate rubella Potentially interrupts rubella transmission, at least in short term (e.g. Sao Paulo) If implemented effectively, potential to eliminate rubella (e.g. Cuba) Disadvantages No effect on rubella transmission Delay of -10 years for impact on CRS No effect on rubella transmission Primigravidae not reached for - 10 years If antenatal serological screening conducted with postpartum vaccination of susceptibles, difficult to reach high coverage Vaccination costs higher Need to conduct pregnancy counsellinga Indirect protection to women of childbearing age NOT guaranteed Very long time until impact seen This strategy is not recommended May be costly to maintain Need to conduct pregnancy counsellinga May be difficult to achieve high coverage of each target group Leaves a pool oftbider susceptibles unprotected May get resurgence of rubella in teens/adults leading to CRS This strategy is not recommended Most costly Need to conduct pregnancy counsellinga Rubella transmission could still continue among adult males aWomen should be advised to avoid pregnancy for 3 months after vaccination; women who are pregnant should not receive vaccine. b MR: mumps-rubella; MMR: mumps-measles-rubella. <50%. In countries, such as Brazil, with an average age at infection of 6 years (3), an increase in CRS would occur unless at least 80% coverage were achieved (2). The experience of childhood rubella vaccination in the USA supports the predictions from these mathematical models. At the beginning of the pro- gramme in the USA, rubella vaccine was predomi- nantly targeted at 1-12-year-old boys and girls, and 65% coverage of this age group was reached by 1975. This greatly reduced the incidence of acquired ru- bella and interrupted the pattern of epidemics, but had proportionately less effect on CRS incidence. Although rubella vaccine was available for suscepti- ble women of childbearing age, coverage of this group was low (54), there were substantial missed opportunities (61), and the proportion of susceptible women of childbearing age remained unchanged at 10-20% (62). In 1978-79 and 1989-91 rubella epidemics led to substantial numbers of CRS cases (63). The move in many countries towards acceler- ated control and elimination of measles raises the question of the marginal cost of including rubella control strategies with this initiative. Through vacci- nation it should be easier, in principle, to reduce transmission of rubella than that of measles, in view of the lower basic reproductive rate of rubella (2); in practice, however, elimination of rubella may be more complex. Unlike the situation with measles in the pre-vaccination era, all adults cannot be assumed to be immune against rubella, and it is precisely the occurrence of rubella in adult (pregnant) women that must be avoided. The age group to include in mass campaigns for rubella control or elimination needs careful study. For measles, the age group <15 years was selected in the Americas, because cam- paigns were implemented approximately 15 years after the vaccination programme began, and older persons were assumed to be immune (64). For ru- bella, however, older individuals may need to be included, especially in countries where >25% of the WHO Bulletin OMS. Vol 75 1997 75 S.E. Robertson et al. adult population is susceptible to the disease. Mass vaccination of women of childbearing age may pro- vide direct protection to the target group for preven- tion of CRS. but it is plausible that rubella could continue to circulate for some years among adult males, and it will be important to maintain protec- tion of girls entering their childbearing years after the campaign ends. Additional vaccination activities targeted towards areas with high densities of adult males (26) might help reduce transmission further. but as recent experience in Scotland shows (65), transmission may be widespread in the community and not confined to institutions. Recommendations A total of 38% of all countries currently include rubella vaccine in their national immunization pro- grammes (Table 1): it is a cause for concern that several countries report only a childhood vaccina- tion strategy. For countries already using rubella vaccine and those considering its use, the experi- ence reported in this article leads to the following recommendations. * Ensure protection of women of childbearing age. It is important for countries with national ru- bella vaccination programmes to ensure that their strategies include protection of women of child- bearing age. Whether to achieve this using a one- time mass campaign or routine vaccination of women of childbearing age, with or without prior screening for susceptibility, will depend on the local situation and the capacity to attain high coverage. * Monitor vaccine coverage. Coverage of rubella vaccine should be monitored in each of the target groups. This should be straightforward for children under 2 years of age, by incorporating information on rubella vaccine in the national EPI coverage sys- tem. For schoolgirls, the cooperation of school health authorities and the Ministry of Education will be needed. Routine monitoring of coverage will probably be most difficult for women of childbearing age, and a lifetime immunization record will prob- ably be needed, as for tetanus toxoid vaccine. Missed opportunitv surveys should be conducted in settings where rubella vaccine is indicated (61). * Conduct surveillance of CRS and acquired ru- bella. Proposed methods for carrying out such sur- veillance have been described in part I of this review (3). Strengthening of laboratory and epidemiological capacity for rubella investigation, as part of EPI sur- veillance of rash illnesses in all countries, is strongly recommended. This should be integrated with activi- ties to strengthen measles and dengue surveillance, with extension of surveillance to adults with rash illnesses. In countries where therapeutic abortions are available, abortion conducted because of a rash illness in pregnancy may be a much more sensitive indicator of programme impact than the reported incidence of CRS (46, 51, 55, 66). * Establish serological surveillance of suscep- tibility if resources permit. Ongoing (longitudinal) serological surveillance may be a useful adjunct to clinical surveillance for monitoring the effect of the programme on susceptibility in different age groups, particularly among women of childbearing age. The simplest method is to monitor susceptibility among women attending antenatal clinics (53): ideally, this should be combined with postpartum vaccination of seronegative women. In countries aiming for elimi- nation, monitoring changes in age- and sex-specific seroprevalence provides data on which to base addi- tional vaccination strategies (67). * Investigate the effect of partial childhood vac- cination coverage. Because of concerns about the effect of partial coverage with childhood rubella vaccination, there is a need to investigate the effect of childhood MMR vaccination in countries with low coverage, or in those where vaccination is conducted in the private sector only. If a substantial proportion of children is vaccinated via the private sector, for example in urban areas, the average age at infection could increase, potentially raising the risk of expo- sure of pregnant women. Information on differences in rubella and CRS epidemiology between areas with and without MMR vaccination is needed. * Update models of the effectiveness of differ- ent vaccination strategies. In the past, mathema- tical models have been helpful for illustrating the potential effect of childhood versus selective vacci- nation strategies at different coverage levels (2). However, such models deal with industrialized country settings, and do not include alternative vac- cination approaches such as initial mass campaigns of different potential age groups. The effectiveness of the various approaches outlined here, in devel- oping countries with a range of demographic and epidemiological situations, should be susceptible to modelling. Ideally, empirical data should be ob- tained on the costs of different strategies, so that models can include this important element. * Countries considering introduction of rubella vaccine must be able to sustain a control pro- gramme. National governments should ensure that they have the economic and logistical capacity to sustain a rubella and CRS control programme in WHO Bulletin OMS. Vol 75 199776 Vaccination against rubella in developing countries the long-term, before introducing vaccination. For countries considering whether to introduce vaccine, data on the burden of disease should be obtained using the methods described in part 1 of this review (3). In the past, rubella has not been considered a public health priority in developing countries, which is probably appropriate for the poorest countries where infant, child, and maternal mortality rates are high and the health and education services infra- structure is lacking. However, developing countries represent a wide range of economies and demo- graphic structures, with some having taken the initia- tive to include rubella vaccine in their national immunization programmes, while others are consid- ering doing so. It is therefore important to evaluate programmes that are in progress, including the cost- effectiveness of different control strategies. Acknowledgements We are grateful to the staff in the WHO Regional Offices for sharing their information and experience with us. C. Broome, R. Chen, S. Cochi, C. de Quadros, A. Galazka, R.H. Henderson, B. Hull, F.M. LaForce, B. Melgaard, J. Nokes, G. Nossal, and J.-M. Olive provided helpful com- ments. A. Batson and C. Torel are thanked for their tech- nical assistance. This project was supported by the Steering Commit- tee on Epidemiology and Field Research of the WHO Global Programme for Vaccines and Immunization with funds donated by the Rockefeller Foundation and the United Nations Development Programme. Resume Lutte contre la rubeole et la rubeole congenitale dans les pays en developpement, deuxieme partie: vaccination contre la rubeole En 1995-96, nous avons mene une enquete mondiale sur les strat6gies de vaccination contre la rub6ole. Au total, 78 pays (plus du tiers) ont signale qu'ils avaient une politique nationale de vaccination contre cette maladie; dans ce nombre ne figurent pas les pays ou la vaccination antirubeolique n'est pratiqu6e que dans quelques r6gions ou par le sec- teur priv6. Aucun pays de la R6gion africaine n'a inscrit la rubeole dans son calendrier national de vaccination. Environ la moitie des pays de la Region des Ameriques et de la Region de la Mediterranee orientale, 64 % des pays de la R6- gion europ6enne et 31 % de ceux de la Region du Pacifique occidental utilisent le vaccin anti- rub6olique. Dans la Region de l'Asie du Sud-Est, Sri Lanka et la Thaflande ont commenc6 a l'utiliser en 1996. Les cas de rubeole cong6nitale peuvent etre prevenus en prot6geant directement les femmes et/ ou les 6colieres (strat6gie de vaccination selective), en vaccinant les enfants (garcons et filles) pour reduire la transmission du virus et obtenir ainsi une protection indirecte (strat6gie de vaccination des enfants), ou en associant ces deux methodes (strat6gie combinee). La majorite des pays (60 %) ont indique qu'ils avaient adopte la strat6gie com- binee. Seulement sept pays (9 %) ont d6clare avoir adopte une strat6gie s6lective de vaccination des femmes etlou des ecolieres, tandis que 24 pays (31 %) ont signal6 qu'ils ne pratiquaient que la vaccination des enfants. L'utilisation du vaccin antirubeolique est 6troitement liWe au niveau econ- omique du pays. Si l'on reprend la classification des pays par les Nations Unies, on constate que le vaccin est utilis6 dans 92 % des pays industrialis6s, 36 % des pays a economie de transition et 28 % des pays en developpement. Si beaucoup de pays ont introduit le vaccin antirub6olique dans leur programme, peu d'entre eux fournissent des donn6es sur l'impact de la vaccination. Dans la Region des Am6riques, on a constat6 une circulation g6n6ralis6e du virus de la rubeole dans les Caraibes anglophones au cours de la p6riode 1990-95. Malgre les programmes nationaux de vaccination, 30 a 50 % des femmes des Caraibes en age de procr6er restent sensibles a la maladie; on s'efforce donc actuellement d'ameliorer la couverture vaccinale des femmes. Dans la Region de la Mediterranee orientale, des flamb6es de rub6ole se sont produites entre 1990 et 1993 a BahreTn, en Iraq, en R6publique is- lamique d'Iran, au KoweTt, a Oman, en Arabie saoudite et dans les Emirats arabes unis. En 1996, les six pays du Golfe ont 6tabli un systeme de surveillance des maladies eruptives (avec confirma- tion serologique pour la rougeole et la rubeole) et ont d6cid6 de signaler regulierement les cas de rub6ole et de rub6ole cong6nitale repondant a la d6finition officielle. Dans la Region europeenne, une flamb6e de rub6ole s'est produite en 1989 en Croatie (qui faisait alors partie de la Yougoslavie) et une importante 6pid6mie a eu lieu en Pologne en 1992-93. Avant d'introduire le vaccin antirub6olique, les gouvernements devraient s'assurer qu'ils disposent des moyens economiques et logistiques n6ces- saires pour appliquer un programme a long terme de lutte contre la rubeole et la rubeole cong6nitale. Ils devraient 6galement reunir des donn6es afin WHO Bulletin OMS. Vol 75 1997 77 S.E. Robertson et al. d'6valuer le fardeau que repr6sente pour eux la rub6ole cong6nitale. Les pays qui utilisent deja le vaccin anti- rub6olique devraient s'assurer que les femmes en age de procr6er sont prises en compte dans leur strat6gie de lutte contre la rub6ole. Une solution pratique pourrait etre la vaccination des femmes apres l'accouchement. Les autres possibilites sont la vaccination de masse des femmes en age de procr6er etlou la vaccination des 6colieres. Les pays devraient 6galement controler la couverture vaccinale de chacun des groupes cibles. Le ren- forcement des moyens de laboratoire et d'etude epidemiologique de la rub6ole, dans le cadre de la surveillance de la rougeole et des autres maladies 6ruptives, est fortement recommand6. References 1. Global Programme for Vaccines and Immuniza- tion, Expanded Programme on Immunization: im- munization policy. Unpublished document WHO/EPI/ GEN/95.03/Rev.1, 1995 (available upon request from Global Programme for Vaccines and Immuniza- tion, World Health Organization, 1211 Geneva 27, Switzerland). 2. Anderson RM, May RM. Vaccination against rubella and measles: quantitative investigations of different policies. Journal of hygiene, 1983, 90: 259-325. 3. Cutts FT et al. Control of rubella and congenital ru- bella syndrome (CRS) in developing countries, part 1: burden of disease from CRS. Bulletin of the World Health Organization, 1997, 75: 55-68. 4. The sex and age distribution of the world populations, the 1994 revision. New York, United Nations, 1994. 5. Global Programme for Vaccines and Immunization: programme report 1994. Unpublished document WHO/GPV/95.1, 1995 (available upon request from Global Programme for Vaccines and Immuniza- tion, World Health Organization, 1211 Geneva 27, Switzerland). 6. Plotkin SA. Rubella vaccine. In: Plotkin SA, Mortimer EA, eds. Vaccines, 2nd edition. London, W.B. Saunders, 1994. 7. Weibel RE et al. Clinical and laboratory studies of live attenuated RA 27/3 and HPV 77-DE rubella virus vaccines (40931). Proceedings of the Society for Ex- perimental Biology and Medicine, 1980, 165: 44-49. 8. Lerman SJ, Bollinger M, Brunken JM. Clinical and serologic evaluation of measles, mumps, and rubella (HPV-77:DE5 and RA27/3) virus vaccines, singly and in combination. Pediatrics, 1981, 68: 18-22. 9. Robertson CM et al. Serological evaluation of a mea- sles, mumps, and rubella vaccine. Archives of disease in childhood, 1988, 63: 612-616. 10. Singh R et al. Immune response to measles, mumps and rubella vaccine at 9, 12 and 15 months of age. Indian journal of medical research, 1994, 100: 155- 159. 11. Schoub BD et al. Measles, mumps and rubella immu- nization at nine months in a developing country. Pediatric infectious disease journal, 1990, 9: 263- 267. 12. Baba K et al. Rubella epidemic in an institution: pro- tective value of live rubella vaccine and serological behavior of vaccinated, revaccinated and naturally immune groups. Biken journal, 1978, 21: 25-31. 13. Balfour HH Jr et al. Rubella viraemia and antibody responses after rubella vaccination and reim- munization. Lancet, 1981, 1:1078-1080. 14. Greaves WL et al. Clinical efficacy of rubella vaccine. Pediatric infectious disease journal, 1983, 2: 284-286. 15. O'Shea S et al. Persistence of rubella antibody 8-18 years after vaccination. British medical journal, 1984, 288: 1043. 16. Boue A, Nicolas A, Montagnon B. Reinfection with rubella in pregnant women. Lancet, 1971, 1: 1251- 1253. 17. Chang T-W, DesRosiers S, Weinstein L. Clinical and serologic studies of an outbreak of rubella in a vaccinated population. New England journal of medi- cine, 1970, 283: 246-248. 18. Horstmann DM et al. Rubella: reinfection of vacci- nated and naturally immune persons exposed in an epidemic. New England journal of medicine, 1970, 283: 771 -778. 19. Cradock-Watson JE. Laboratory diagnosis of ru- bella: past, present and future. Epidemiology and in- fection, 1991, 107: 1-15. 20. Morgan-Capner P et al. Detection of rubella-specific IgM in subclinical rubella reinfection in pregnancy. Lancet, 1985, 1: 244-246. 21. Public Health Laboratory Service and National Congenital Rubella Surveillance Programme. Ru- bella surveillance to December 1990, United King- dom. Weekly epidemiological record, 1991, 66(30): 217-220. 22. Preblud SR, Alford CA Jr. Rubella. In: Remington JS, Klein JO, eds. Infectious diseases of the fetus and newborn infant, 3rd edition. London, W.B. Saunders, 1990. 23. Sabin AB. Perspectives on current use of some virus vaccines in the USA. Pediatric research, 1979, 13: 674-683. 24. Best JM. Rubella vaccines: past, present and future. Epidemiology and infection, 1991, 107: 17-30. 25. Rubella vaccination during pregnancy - United States, 1971-1988. Morbidity and mortality weekly report, 1989, 38(17): 289-293. 26. Rubella prevention: recommendations of the Immuni- zation Practices Advisory Committee (ACIP). Morbid- ity and mortality weekly report, 1990, 39(RR-15): 1- 18. 27. Howson CP, Fineberg HV. The ricochet of magic bullets: summary of the Institute of Medicine Report, "Adverse effects of pertussis and rubella vaccines". Pediatrics, 1992, 89(2): 318-324. 28. Slater PE et al. Absence of an association between rubella vaccination and arthritis in underimmune postpartum women. Vaccine, 1995, 13(16): 1529- 1532. 78 WHO Bulletin OMS. Vol 75 1997 Vaccination against rubella in developing countries 29. Ray P et al. Retrospective cohort evaluation of chronic rubella vaccine arthropathy. In: Abstracts of the 34th Interscience Conference on Antimicrobial Agents and Chemotherapy, Orlando, FL, 4-7 October 1994. Abstract No. 110, 240. 30. Frey T. Report of an international meeting on rubella vaccines and vaccination, 9 August 1993, Glasgow, United Kingdom. Journal of infectious diseases, 1994, 170: 507-509. 31. Expanded Program on Immunization in the Ameri- cas. English-speaking Caribbean rubella policy. EPI newsletter, 1993, 15(4): 6. 32. Expanded Program on Immunization in the Ameri- cas. PAHO measles laboratory network workshop. EPI newsletter, 1995, 17(4): 1-2. 33. Expanded Programme on Immunization. Immuni- zation in the English-speaking Caribbean. Weekly epidemiological record, 1996, 71: 217-223. 34. Azevedo-Neto RS de et al. Rubella sero- epidemiology in a non-immunized population of Sao Paulo State, Brazil. Epidemiology and infection, 1994, 113: 161-173. 35. Massad E et al. Assessing the efficacy of a mixed vaccination strategy against rubella in Sao Paulo, Brazil. International journal of epidemiology, 1995, 24(4): 842-850. 36. Report of a measles/rubella epidemiology workshop for the Arabian Peninsula, Kuwait, 9-10 December 1995. Alexandria, WHO Office for the Eastern Medi- terranean Region, 1995. 37. Kabiri M, Moattari A. The rubella immuno- surveillance of Iranian females: an indication of the emergence of rubella outbreak in Shiraz, Iran. Iranian journal of medical sciences, 1993, 18: 134- 137. 38. Expanded Programme on Immunization. Rubella outbreak, Oman. Weekly epidemiological record, 1994, 69(45): 333-336. 39. Saeed AA, Abu-Shagra S, Al-Rasheed R. Congeni- tal rubella syndrome revisited. Saudi medical journal, 1993, 14:256. 40. Bakasun V, Suzanic-Karnincic J. A rubella outbreak in the Region of Rijeka, Croatia. International journal of epidemiology, 1995, 24: 453-456. 41. Rudnicka H. [Epidemiology of rubella in 1992]. Przeglad epidemiologiczny (Warszawa), 1994, 48: 75-78 (in Polish). 42. Rubella surveillance, United Kingdom. Weekly epide- miological record, 1993, 68(17): 119-120. 43. Fogel A et al. The impact of rubella epidemics on pregnant women, proof for the necessity of pre-marital screening and vaccination. Developments in biologi- cal standardization, 1979, 43: 349-353. 44. Brand N et al. Congenital rubella in Israel following the 1978-79 rubella epidemic. Israel journal of medi- cal sciences, 1983, 19: 925-928. 45. Berger SA, Ginsberg GM, Slater PE. Cost-benefit analysis of routine mumps and rubella vaccination for Israeli infants. Israel journal of medical sciences, 1990, 26: 74-80. 46. Isacsohn M et al. Surveillance of congenital rubella and the immunity status to rubella in the Jerusalem area. Israel journal of medical sciences, 1983, 19: 929-933. 47. Swartz TA, Hornstein L, Epstein I. Epidemiology of rubella and congenital rubella infection in Israel, a country with a selective immunization programme. Reviews of infectious diseases, 1985, 7(suppl. 1): S42-S46. 48. Danon YL et al. Prevalence of antibodies against measles, mumps and rubella in Israeli young adults. Israeljournal of medical sciences, 1992, 28: 733-736. 49. Robertson SE. Report of a visit to Manila and Kuala Lumpur, 30 June-21 July 1990. Upublished docu- ment, Geneva, World Health Organization, 1990. 50. Proceedings of the 90th Anniversary Scientific Semi- nar, Institute for Medical Research, Kuala Lumpur, Malaysia, 23-25 June 1992. Kuala Lumpur, Institute for Medical Research: 85-86. 51. Rubella immunization programme, Singapore. Weekly epidemiological record, 1989, 64(19): 144- 145. 52. Chan CM, Tsang HF. Rubella in Hong Kong. Public health and epidemiology bulletin, Department of Health, Hong Kong, 1995, 4(3): 17-20. 53. Miller CL, Miller E, Waight PA. Rubella susceptibility and the continuing risk of infection in pregnancy. Brit- ish medical journal, 1987, 294: 1277-1278. 54. Orenstein WA et al. The opportunity and obligation to eliminate rubella from the United States. Journal of the American Medical Association, 1984, 251(15): 1988-1994. 55. Smithells R et al. Congenital rubella in Great Britain 1971-1988. Health trends, 1990, 22(2): 73-76. 56. de Spiegelaere M, Dramaix M, Hennart P. ln6galit6s sociales et pr6vention: le statut vaccinal des adoles- cents. Revue d'epid6miologie et de sant6 publique, 1996, 44: 228-236. 57. Peckham CS, Marshall WC, Dudgeon JA. Rubella vaccination of schoolgirls: factors affecting vac- cine uptake. British medical journal, 1977, 1: 760- 761. 58. Baxter DN. Control of the congenital rubella syn- drome in Jamaica. West Indian medicaljournal, 1986, 35: 50-54. 59. Prabhakar P et al. Seroprevalence of Toxoplasma gondii, rubella virus, cytomegalovirus, herpes simplex virus (TORCH) and syphilis in Jamaican pregnant women. West Indian medical journal, 1991, 40: 166- 169. 60. Expanded Program on Immunization in the Ameri- cas. Measles campaign improves anti-rubella cover- age. EPI newsletter, 1994, 16(1): 5-6. 61. Robertson SE et al. Preventing rubella: assessing missed opportunities for immunization. American journal of public health, 1987, 77(10): 1347- 1349. 62. Cochi SL et al. Congenital rubella syndrome in the United States, 1970-1985: on the verge of elimina- tion. American joumal of epidemiology, 1989, 129: 349-361. 63. Lindegren ML et al. Update: rubella and congenital rubella syndrome, 1980-1990. Epidemiologic re- views, 1991, 13: 341-348. WHO Bulletin OMS. Vol 75 1997 79 S.E. Robertson et al. 64. de Quadros CA et al. Measles elimination in the Americas: evolving strategies. Journal of the Ameri- can Medical Association, 1996, 275(3): 224-229. 65. Stevenson J et al. Resurgence of rubella. Lancet, 1996, 347: 980-981. 66. Isacsohn M, Nishmi M, Swartz TA. Rubella in Jeru- salem. 2. Clinical and serologic findings in children with congenital rubella. Israel journal of medical sci- ences, 1979, 15(1): 17-22. 67. Morgan-Capner P et al. Surveillance of antibody to measles, mumps, and rubella by age. British medical journal, 1988, 297: 770-772. 80 WHO Bulletin OMS. Vol 75 1997