Services on Demand
Print version ISSN 0042-9686
Bull World Health Organ vol.82 n.4 Genebra Apr. 2004
Enquêtes sérologiques sur la rubéole dans trois hôpitaux ophtalmologiques Aravind au Tamil Nadu (Inde)
Estudios serológicos sobre la rubéola en tres hospitales oftalmológicos Aravind de Tamil Nadu (India)
Perumalsamy VijayalakshmiI; Rajamanickam AnuradhaI; Karthik PrakashII; Kalpana NarendranIII; Meenakshi RavindranIV; Lalitha PrajnaI; David BrownV; Susan E. RobertsonVI, 1
IAravind Eye Hospital, Madurai, Tamil Nadu, India
IILions Aravind Institute of Community Ophthalmology, Madurai, Tamil Nadu, India
IIIAravind Eye Hospital, Coimbatore, Tamil Nadu, India
IVAravind Eye Hospital, Tirunelveli Junction, Tamil Nadu, India
VEnteric, Respiratory and Neurological Virus Laboratory, Health Protection Agency, London, England
VIDepartment of Immunization, Vaccines and Biologicals, World Health Organization, 1211 Geneva 27, Switzerland (email: firstname.lastname@example.org)
OBJECTIVE: To determine the susceptibility of female eye hospital staff to rubella infection and the potential risk for hospital-based rubella outbreaks.
METHODS: A prospective cohort study on the seroprevalence of rubella IgG antibodies was conducted at three large eye hospitals in Coimbatore, Madurai and Tirunelveli, Tamil Nadu, India, where young children with eye abnormalities attributable to congenital rubella are treated. A total of 1000 female hospital employees aged 18-40 years agreed to participate and gave written informed consent.
FINDINGS: The proportions of rubella-seronegative women were: 11.7% at Coimbatore, with a 95% confidence interval (CI) of 8.1-16.5; 15% at Madurai (95% CI = 12.3-18.1), and 20.8 at Tirunelveli (95% CI = 14.7-28.6). For the entire cohort the proportion seronegative was significantly higher among married women (21.5%) than among single women (14.0%) (P = 0.02). Rates of seronegativity were highest among physicians and lowest among housekeepers. All 150 seronegative women in the study sample accepted a dose of rubella vaccine.
CONCLUSION: These are the first rubella serosurveys to have been reported from eye hospitals in any country. The relatively high rate of susceptibility indicated a risk of a rubella outbreak, and this was reduced by vaccinating all seronegative women. A policy has been established at all three hospitals for the provision of rubella vaccine to new employees. Other hospitals, especially eye hospitals and hospitals in countries without routine rubella immunization, should consider the rubella susceptibility of staff and the risk of hospital-based rubella outbreaks.
Keywords: Rubella/epidemiology/immunology; Rubella syndrome, Congenital; Hospitals, Special; Personnel, Hospital; Cross infection; Rubella vaccine; Women; Seroepidemiologic studies; Prospective studies; Cohort studies; India (source: MeSH, NLM).
OBJECTIF: Déterminer la sensibilité du personnel féminin des hôpitaux ophtalmologiques à l'infection rubéoleuse et le risque potentiel de flambées hospitalières de rubéole.
MÉTHODES: Une étude de cohorte prospective sur la séroprévalence des anticorps antirubéoleux de la classe des IgG a été réalisée dans trois grands hôpitaux ophtalmologiques (Aravind Eye Hospitals) à Coimbatore, Madurai et Tirunelveli, dans le Tamil Nadu (Inde), où sont traités de jeunes enfants atteints d'anomalies oculaires dues à la rubéole congénitale. En tout, 1000 femmes faisant partie du personnel hospitalier et âgées de 18 à 40 ans ont accepté de participer à l'étude et ont donné par écrit leur consentement éclairé.
RÉSULTATS: La proportion de femmes séronégatives pour la rubéole était de 11,7 % à Coimbatore, avec un intervalle de confiance (IC) à 95 % de 8,1-16,5 %, 15,0 % à Madurai (IC 95 % : 12,3-18,1 %) et 20,8 % à Tirunelveli (IC 95 % : 14,7-28,6 %). Au niveau de la cohorte entière, la proportion de femmes séronégatives était significativement plus élevée chez les femmes mariées (21,5 %) que chez les femmes célibataires (14,0 %) (p = 0,02). Le taux de séronégativité le plus élevé a été observé chez les médecins et le plus faible chez les intendantes. Les 150 femmes séronégatives de l'échantillon étudié ont accepté de recevoir une dose de vaccin antirubéoleux.
CONCLUSION: Ces enquêtes sérologiques sur la rubéole sont les premières rapportées au niveau mondial pour des hôpitaux ophtalmologiques. Le taux de sensibilité relativement élevé observé indiquait un risque de flambée de rubéole, que l'on a réduit en vaccinant toutes les femmes séronégatives. Les trois hôpitaux ont établi une politique consistant à vacciner les nouvelles employées. D'autres hôpitaux, en particulier les ôpitaux ophtalmologiques et les hôpitaux des pays qui ne pratiquent pas la vaccination systématique contre la rubéole, devraient prendre en compte la sensibilité de leur personnel à la rubéole et le risque de flambées de rubéole en milieu hospitalier.
Mots clés: Rubéole/épidémiologie/immunologie; Syndrome de rubéole congénitale; Hòpital spécialisé; Personnel hôpital; Infection hospitalière; Vaccin antirubéoleux; Femmes; Etude séroépidémiologique; Etude prospective; Etude cohorte; Inde (source: MeSH, INSERM).
OBJETIVO: Determinar la vulnerabilidad del personal femenino de hospitales oftalmológicos a la infección rubeólica, así como el riesgo potencial de brotes de rubéola en los hospitales.
MÉTODOS: Se llevó a cabo un estudio de cohortes prospectivo sobre la seroprevalencia de anticuerpos IgG contra la rubéola en tres grandes hospitales oftalmológicos de Coimbatore, Madurai y Tirunelveli, en el Estado de Tamil Nadu, India, donde reciben tratamiento los niños pequeños que sufren anomalías oculares causadas por la rubéola congénita. Un total de 1000 empleadas de hospital de 18-40 años respondieron positivamente a la invitación de participar y dieron por escrito su consentimiento informado.
RESULTADOS: Las proporciones de mujeres seronegativas para la rubéola fueron las siguientes: 11,7% en Coimbatore, con un intervalo de confianza (IC) del 95% de 8,1%-16,5%; 15% en Madurai (IC95% = 12,3%-18,1%), y 20,8% en Tirunelveli (IC95% = 14,7%-28,6%). En el conjunto de la cohorte la proporción seronegativa fue significativamente mayor entre las mujeres casadas (21,5%) que entre las solteras (14,0%) (P = 0,02). Las tasas de seronegatividad fueron máximas entre las médicas y mínimas entre las trabajadoras de los servicios de limpieza. Las 150 mujeres seronegativas aceptaron una dosis de la vacuna antirrubeólica.
CONCLUSIÓN: Estas son las primeras encuestas serológicas de que se informa sobre la rubéola en los hospitales oftalmológicos en país alguno. La tasa relativamente alta de vulnerabilidad mostró que existía el riesgo de que se produjera un brote de rubéola, riesgo que se redujo vacunando a todas las mujeres seronegativas. Se ha instaurado en los tres hospitales una política destinada a suministrar la vacuna antirrubéolica a las nuevas empleadas. Otros hospitales, especialmente los hospitales oftalmológicos y los hospitales de los países sin inmunización sistemática contra la rubéola, deberían estudiar la vulnerabilidad del personal a esta enfermedad y el riesgo de brotes de rubéola en los hospitales.
Palabras clave: Rubéola/epidemiología/inmunología; Síndrome de rubéola congénita; Hospitales especializados; Personal de hospital; Infección hospitalaria; Vacuna contra la rubéola; Mujeres; Estudios seroepidemiológicos; Estudios prospectivos; Estudios de cohortes; India (fuente: DeCS, BIREME).
If health-care workers are not immune to rubella they are at risk of contracting it, especially from their patients. This is particularly important in hospitals treating paediatric patients with congenital rubella syndrome (CRS) and in countries that do not include rubella vaccine in their national immunization programmes. Rubella has been a special concern of ophthalmologists for more than 60 years. In 1941, Gregg reported cataracts in 78 infants, many of whom were also affected by congenital heart disease and failure to thrive (1). Most of the mothers of these children had contracted rubella during the early months of pregnancy, and Gregg therefore postulated that rubella virus was the cause of the infant malformations, later called CRS. Subsequent studies confirmed that the risk of rubella defects was high in infants whose mothers were infected by rubella virus in the first 16 weeks of pregnancy (2). WHO estimates that, worldwide, more than 100 000 children are born with CRS each year, most of them in developing countries (3). A study in southern India during 1993-94 found that CRS was the cause of 26% of cases of children born blind with congenital cataracts (4).
The WHO-recommended case definition for probable rubella is a patient with fever, maculopapular rash and cervical, suboccipital or postauricular lymphadenopathy or arthralgia/arthritis (5). Because of the difficulty of clinical diagnosis of rubella, laboratory confirmation is required. This involves a rubella-specific IgM test on a serum specimen obtained within 28 days after the onset of rash.
The WHO-recommended case definition for clinically confirmed CRS is an infant with two of the complications described in (a) below or with one of those in (a) and one in (b) (5).
(a) Cataract(s), congenital glaucoma, congenital heart disease, loss of hearing, pigmentary retinopathy.
(b) Purpura, splenomegaly, microcephaly, mental retardation, meningoencephalitis, radiolucent bone disease, onset of jaundice within 24 hours after birth.
Laboratory confirmation of CRS involves a rubella-specific IgM test on a blood specimen obtained within the first year of life, preferably within the first six months. Where appropriate laboratory expertise is available, the detection of rubella virus in specimens from the nasopharynx or urine of an infant with suspected CRS may also provide laboratory confirmation. Infants with CRS shed rubella virus for long periods in nasopharyngeal secretions and urine, and close contact with such infants can therefore lead to rubella infection. In a large case series in the USA, rubella virus was isolated from nasopharyngeal secretions of 84% of infants with CRS during the first month of life and from 62% aged 1-4 months, 33% aged 5-8 months and 11% aged 9-12 months (6). Eye disease is a common sign of CRS: in a case series of 46 children with CRS at the Aravind Eye Hospital, Madurai, India, cataract was present in 81 eyes (7). A special risk occurs during cataract surgery for a child with CRS because the lens aspirate may contain live rubella virus. Rubella virus has been identified in lens aspirates from children aged up to 3 years (8).
Nearly 50% of rubella infections are subclinical, and an infected health care worker may therefore unknowingly transmit the virus to patients or other staff. This poses a risk of CRS if a woman becomes infected with rubella in the early months of pregnancy. Hospital-based rubella outbreaks have been reported from industrialized countries (9, 10). In India, such an outbreak was reported from the St John's Medical College and Hospital, Bangalore, in 1990 (11), and another occurred in 1996 among medical and nursing students and staff at the Christian Medical College, Vellore, Tamil Nadu (12).
This paper reports the results of a serological study to assess the rubella susceptibility of female employees at three Aravind Eye Hospitals.
From May to December 2002 a rubella serosurvey was conducted among female personnel at three Aravind Eye Hospitals in Coimbatore, Madurai and Tirunelveli, Tamil Nadu, India, with a total of 2675 beds. In 2002 these hospitals conducted approximately 190 000 operations and more than 1.3 million outpatient visits. All female personnel aged 18-40 years were eligible and were given background information about the study; those who provided voluntary and written informed consent to participate were enrolled. The study was approved by: the Aravind Eye Hospital Ethics Committee, Madurai; the Indian Council for Medical Research, New Delhi; and the WHO Secretariat Committee for Research Involving Human Subjects. A standard questionnaire was administered on demographic characteristics, marital and vaccination history, and type and duration of employment. A 5-ml blood specimen was obtained from each participant. The serum was separated and stored at the study site at 4-8 ºC before being transferred to the central laboratory at Madurai Hospital, where it was stored at -20 ºC. Rubella-specific IgG antibodies were detected using a commercial IgG enzyme-linked immunosorbent assay (ELISA) kit (catalogue number 51208, Human Gesellschaft für Biochemica und Diagnostica mbH, Wiesbaden, Germany) in accordance with the manufacturer's instructions. The Enteric, Respiratory and Neurological Virus Laboratory of the Health Protection Agency, London, served as the reference laboratory for the study and provided on-site training of the laboratory technician. The London laboratory conducted quality control tests on the human rubella IgG ELISA kit every six months and performed a blinded proficiency test on the Aravind Hospital laboratory.
Seronegative women were offered rubella vaccine and, as recommended by WHO, were advised to avoid pregnancy for one month following vaccination (13). The RA 27/3 rubella vaccine (Serum Institute of India, Pune) was supplied in single-dose vials that were stored at 4-8 ºC during the study. The vaccine was reconstituted with diluent supplied by the manufacturer and administered intramuscularly.
Data were double-entered in Epi Info, version 6.04 (Centers for Disease Control and Prevention, Atlanta, GA, USA and WHO, Geneva, Switzerland), and cleaned. Statistical analyses were performed using Stata, version 7.0 (Stata Corporation, College Station, TX, USA). Seroprevalence levels were compared by c2 tests.
Of the 1038 female employees aged 18-40 years, 1000 (96%) consented to participate in the study, 248 of them at Coimbatore Hospital, 608 at Madurai Hospital and 144 at Tirunelveli Hospital. The age range of the participants was 18-40 years (mean, 21.7 years). Overall, 150 were negative for rubella IgG antibody (15%; 95% confidence interval (CI) = 13.3-16.7%). All the participants were informed of their test results and the seronegative women all accepted a dose of rubella vaccine. The seronegativity rates at the Coimbatore, Madurai and Tirunelveli hospitals were 11.7 % (95% CI = 8.1-16.5), 15.0% (95% CI = 12.3-18.1) and 20.8 (95% CI = 14.7-28.6), respectively (Fig. 1). The differences between the hospitals were non-significant, nor were there significant differences in seronegativity rates between age groups (Table 1). However, the proportion of seronegative women was significantly higher (P = 0.02) among the 135 married women (21.5%) than among the 865 single women (14.0%).
The categories of employees participating in the study included doctors, nurses, opticians/refractionists, laboratory staff, medical equipment production/repair staff, and housekeepers/caterers. The percentage of seronegativity was highest among doctors (26.7%) and lowest among housekeepers/caterers (11.1%) (Table 2).
Of the 1000 participants, only two physicians and one nurse reported previous receipt of rubella vaccine.
These are the first rubella serosurveys reported from eye hospitals anywhere in the world. At the three hospitals, 12-21% of female health workers and 27% of female physicians were rubella seronegative. These are moderate-to-high rates. A WHO-sponsored global review identified more than 100 serosurveys conducted among women of childbearing age in 45 developing countries before the introduction of rubella vaccine (14). The proportion who were rubella seronegative was less than 10% in 13 countries, 10-24% in 20 countries and 25% or higher in 12 countries. A number of previous studies have demonstrated the potential risk of rubella to hospital employees, most commonly those working in obstetrics or paediatrics departments (15-18).
The high proportion of rubella-seronegative physicians is unexplained. However, only 30 physicians participated in the study, of whom 19 (63%) came to the Aravind Eye Hospitals from different states in India where their previous exposure to rubella virus may have differed from that of other hospital employees. The women employed at the Aravind Eye Hospitals who were not physicians came from the local areas and were possibly representative of working women in their respective cities, although they may have had additional exposure to rubella virus in the hospital work environment. Unmarried women employees at the three Aravind Eye Hospitals lived in dormitory residences. No outbreaks of rubella or measles had been reported among dormitory residents during the past eight years, although there were varicella outbreaks. We did not demonstrate an inverse relationship between age and the proportion of seronegative women, perhaps because of the small number of women aged >30 years. Previous studies indicated great variability in the proportions of women in given age groups who were susceptible to rubella, and some studies showed little decrease in susceptibility over the age range studied (19). A clearer picture of the community-based risk of rubella in the female population would be provided by an antenatal serosurvey with stratified sampling, ensuring a more even distribution of age groups. Such a study is planned.
Rubella serosurveys in India
A review of the literature identified 15 rubella serosurveys in women of childbearing age in India (20-28, Table 3). However, the data should be viewed with caution because of the diversity of the laboratories and assays in question. Almost all the studies carried out in 1990 and subsequently involved the use of rubella IgG ELISA assays, whereas earlier studies employed haemagglutination inhibition assays. Nevertheless, a review of rubella test methods indicated general agreement between these assays (29). The 15 serosurveys from India showed that susceptibility ranged from 5% to 45%, reflecting the large size of the country and the pattern of rubella virus circulation. A larger proportion of women can be expected to be susceptible during periods when rubella virus circulates at an endemic level than during periods following rubella outbreaks. There is evidence that the proportion of susceptible women in Tamil Nadu has increased gradually from 4% in the 1980s to the 15% found in the present study, suggesting endemic levels of virus circulation with no major outbreaks.
A major limitation of our study was the lack of background rubella surveillance data that would provide an understanding of the local endemic-epidemic cycles of rubella virus. In the future such data should be increasingly available from countries in the South-East Asia Region of WHO. During 2002-03, staff from one or more laboratories in Bangladesh, Bhutan, India, Indonesia, Maldives, Myanmar, Nepal, and Sri Lanka attended a one-week workshop in Bangkok, Thailand or Pune, India. Participants received training in the IgM ELISA tests for measles and rubella, and their laboratories were subsequently enrolled in a proficiency-testing scheme. To complement the laboratory training a surveillance and data management workshop was held during September 2003 in New Delhi for countries of the South-East Asia Region. The surveillance and reporting of measles and rubella are expected to benefit from the experience gained in the highly successful surveillance of acute flaccid paralysis in this Region.
Risk groups and rubella vaccination
Our study identified physicians as the employee group at highest risk. The vaccination of seronegative individuals can help to lower the institutional risk of hospital-based rubella outbreaks. In addition, a new policy is planned for vaccinating hospital physicians and nurses against rubella at the start of their employment, without serological screening. This is consistent with the WHO recommendation that the health of employees, including their immunization history, be reviewed at recruitment (30). Rubella vaccination is important for doctors and nurses working at eye hospitals anywhere in the world, and for obstetricians, midwives, paediatricians, neurologists, cardiac surgeons, ear, nose and throat surgeons and other specialists who see children with congenital rubella. Moreover, the provision of rubella vaccine to both medical and nursing students before they entered the hospital environment would help to prevent hospital-based outbreaks and would protect female health personnel before their first pregnancies.
We are grateful for helpful comments made by Dr Jennifer Best and members of the Steering Committee on Epidemiology and Field Research of the WHO Department of Immunization, Vaccines and Biologicals. Financial support for the study was provided by WHO.
Conflicts of interest: none declared.
1. Gregg NM.Congenital cataract following German measles in the mother. Transactions of the Ophthalmological Society of Australia 1941;3:35-46. [ Links ]
2. Miller E, Cradock-Watson JE, Pollock TM.Consequences of confirmed maternal rubella at successive stages of pregnancy. Lancet 1982:2;782-4. [ Links ]
3. Robertson SE, Featherstone DA, Gacic-Dobo M, Hersh BS.Rubella and congenital rubella syndrome: global update. Pan American Journal of Public Health 2003;14:306-15. [ Links ]
4. Eckstein M, Brown DWG, Foster A, Richards AF, Gilbert CE, Vijayalakshmi P. Congenital rubella in south India: diagnosis using saliva from infants with cataract. BMJ 1996;312:161. [ Links ]
5. Cutts FT, Best J, Siqueira MM, Engstrom K, Robertson SE. Guidelines for surveillance of congenital rubella syndrome and rubella, field test version, May 1999. Geneva: World Health Organization; 1999. WHO document WHO/V&B/99.22. [ Links ]
6. Cooper LZ, Ziring PR, Ockerse AB, Fedun BA, Kiely BK, Krugman S.Rubella: clinical manifestations and management. American Journal of Diseases of Children 1969;11:818-29. [ Links ]
7. Vijayalakshmi P, Kakkar G, Samprathi A, Banushree R. Ocular manifestations of congenital rubella syndrome in a developing country. Indian Journal of Ophthalmology 2002;50:307-11. [ Links ]
8. Menser MA, Harley JD, Hertzberg R, Dorman DC, Murphy AM. Persistence of virus in lens for three years after prenatal rubella. Lancet 1967;2:387-8. [ Links ]
9. Carne S, Dewhurst CJ, Hurley R. Rubella epidemic in a maternity unit. BMJ 1973;1:444-6. [ Links ]
10. Polk BF, White JA, DeGirolami PC, Modlin JF. An outbreak of rubella among hospital personnel. New England Journal of Medicine 1980;303:541-5. [ Links ]
11. Srinivasa H, Raj IS, Elizabeth J. Rubella in medical and nursing students, Bangalore, India. Virus Information Exchange Newsletter for South-East Asia and the Western Pacific 1991;8:26-7. [ Links ]
12. World Health Organization. Report of a meeting on preventing congenital rubella syndrome: immunization strategies, surveillance needs, Geneva, 12-14 January 2000. Geneva: World Health Organization; 2000. WHO document WHO/V&B/00.10. [ Links ]
13. World Health Organization. Rubella vaccines: WHO position paper. Weekly Epidemiological Record 2000;75;161-9. [ Links ]
14. Cutts FT, Robertson SE, Diaz-Ortega JL, Samuel R. Control of rubella and congenital rubella syndrome (CRS) in developing countries, part 1: burden of disease from CRS. Bulletin of the World Health Organization 1997;75:55-68. [ Links ]
15. Chan PKS, Li CY, Tam JS, Cheng AF. Rubella immune status among healthcare workers in the department of obstetrics and gynaecology of a regional hospital in Hong Kong: the need for a vaccination policy. Journal of Hospital Infection 1999;42:239-42. [ Links ]
16. Villasis-Keever MA, Pena LA, Miranda-Novales G, Munoz TA, Damasio- Santana L, Lopez-Fuentes G, et al. Prevalence of serological markers against measles, rubella, varicella, hepatitis B, hepatitis C, and human immunodeficiency virus among medical residents in Mexico. Preventive Medicine 2001;32:424-8. [ Links ]
17. Olusanya O, Blomberg J. Antibody prevalence against rubella among hospital personnel in Nigeria: implications for health care system and immunisation policy. Acta Tropica 1991;48:101-7. [ Links ]
18. Pruksananonda P, Bumrungtrakul P. Serosurvey of rubella antibody among health personnel of Songklanagarind University Hospital, Thailand. Southeast Asian Journal of Tropical Medicine and Public Health 1983;14:380-4. [ Links ]
19. Cutts FT, Vynnycky E. Modelling the incidence of congenital rubella syndrome in developing countries. International Journal of Epidemiology 1999;28:1176-84. [ Links ]
20. Bhaskaram P, Ramalakshmi BA, Rama Raju LA, Raman L. Need for protection against rubella in India. Indian Journal of Pediatrics 1991;58:811-4. [ Links ]
21. Pal SR, Chitkara NL, Broor S, Murthy JG, Choudhury S, Devi PK. Serological investigation of rubella virus infection in and around Chandrigarh - a preliminary communication. Indian Journal of Medical Research 1974;62:240-5. [ Links ]
22. Seth P, Balaya S, Mohapatra LN. Sero-epidemiological study of rubella infection in female subjects of Delhi and its surrounding villages. Indian Journal of Medical Research 1971;59:190-4. [ Links ]
23. Satpathy G. Seroepidemiology of rubella in Indian women and in children with congenital malformations. Virus Information Exchange Newsletter for South-East Asia and the Western Pacific 1989;6:126. [ Links ]
24. Yadav S, Gupta S, Kumari S. Seroprevalence of rubella in women of reproductive age. Indian Journal of Pathology and Microbiology 1995;38:139-42. [ Links ]
25. Shanmugam J, Raveendranath M, Nair VR. Seroprevalence of rubella and cytomegalovirus (CMV) infection in pregnant women from Kerala State. Journal of the Indian Association of Communicable Diseases 1982;5:58-63. [ Links ]
26. Black FL, Berman LL, Borgono JM, Capper RA, Carvalho AA, Collins C, et al. Geographic variation in infant loss of maternal measles antibody and in prevalence of rubella antibody. American Journal of Epidemiology 1986;124:442-52. [ Links ]
27. Mathur A, Chaturvedi UC, Mehrotra RML. Serological study for the prevalence of rubella at Lucknow. Indian Journal of Medical Research 1974;62;307-10. [ Links ]
28. Chakraborty MS, Mukherjee MK, Sarkar JK. Rubella antibody profile in women of child-bearing age in Calcutta area. Indian Journal of Medical Research 1973;61:340-3. [ Links ]
29. Cradock-Watson J. Laboratory diagnosis of rubella: past, present and future. Epidemiology and Infection 1991;107:1-15. [ Links ]
30. World Health Organization. Prevention of hospital-acquired infections: a practical guide. 2nd ed. Geneva: World Health Organization; 2002. [ Links ]WHO document WHO/CDS/CSR/EPH/2002.12.
Submitted: 17 October 03
Final revised version received: 06 December 03
Accepted: 09 December 03
1 Correspondence should be sent to this author.