Evaluation of house-to-house versus fixed-site oral poliovirus vaccine delivery strategies in a mass immunization campaign in Egypt.

Research I Recherche Evaluation of house-to-house versus fixed-site oral poliovirus vaccine delivery strategies in a mass immunization campaign in Egypt R.W. Linkins,1 E. Mansour,2 0. Wassif,3 M.H. Hassan,3 & P.A. Patriarca4 Among poliomyelitis eradication activities recommended by WHO are national immunization days. Most campaigns have delivered oral poliovirus vaccine (OPV) from fixed sites, reaching 80-90% of target populations. Although house-to-house vaccination provides nearly universal coverage, countries have been reluctant to use this approach because it is considered more costly and logistically difficult. To quantify the cost-effectiveness of both these strategies, we compared the vaccine coverage and vacci- nation costs per child for house-to-house and fixed-site delivery in a mass campaign in Egypt. While personnel and total costs were higher in house-to-house delivery (38% and 13% higher, respectively), the costs per child vaccinated were similar. This was due primarily to the high coverage levels achieved in house-to-house delivery (100% versus 86%) and the reduced vaccine wastage. Vaccinating children at highest risk of infection was only 25-50% as expensive on a per child basis using house-to-house delivery, since such children were less likely to visit fixed sites. These findings may not be generaliz- able to other countries where labour costs are higher and the population density lower; however, house- to-house delivery may prove to be the most cost-effective eradication strategy by ensuring universal access to immunization. Introduction In May 1988 the Forty-first World Health Assembly committed WHO to the global eradication of polio- myelitis by the year 2000 (1). Among immunization activities recommended in WHO's plan of actiona are national immunization days: administration of oral poliovirus vaccine (OPV) to all children in an appropriate age group (usually <5 years of age), regardless of their prior immunization status. Ideally. mass campaigns should be conducted over a limited I Epidemiologist, National Immunization Program, Centers for Dis- ease Control and Prevention, Mailstop E-05, Atlanta, GA 30 333, USA. Requests for reprints should be sent to Dr Linkins at this address. 2 Director, Child Survival Project, Cairo, Egypt. 3 Medical Epidemiologist, Child Survival Project, Cairo, Egypt. 4 Medical Epidemiologist, Center for Biologics Evaluation and Research, Food and Drug Administration, Bethesda, MD, USA. a Plan of action. Global poliomyelitis eradication by the year 2000. Unpublished document WHO/EPI/POLIO/92.2, 1992. Reprint No. 5637 period (a few days to weeks), and preferably during the low season of poliovirus transmission. A second OPV dose is recommended 4-6 weeks after the first dose. This strategy was based primarily on the suc- cess achieved in the Region of the Americas, where wild poliovirus infection was eliminated within a period of approximately 5 years (2). Mass immunization campaigns in most countries have delivered OPV at fixed immunization sites, and have reported covering 80-90% of target popula- tions. Occasionally, house-to-house campaigns have also been used, and by their very nature provide nearly universal vaccine coverage. However, because house-to-house vaccine delivery has been considered more costly and more logistically difficult than fixed-site delivery, many countries have been reluc- tant to implement it. In an effort to quantify better the relative cost- effectiveness of these two vaccine delivery strat- egies. we compared the vaccine coverage and per child vaccination costs for house-to-house and fixed- site OPV delivery in a mass immunization campaign conducted in Egypt in 1993. In addition, we evaluat- Bulletin of the World Health Organization, 1995, 73 (5): 589-595 589 R.W. Linkins et al. ed the characteristics of children vaccinated from fixed sites to identify children "at risk" of not parti- cipating in fixed-site mass campaigns. Methods All children aged 0-5 years inclusive were targeted to receive one dose of OPV in each of two rounds of vaccine delivery; no other vaccines were adminis- tered. To meet the objectives of the evaluation, two well-defined catchment areas were identified which were considered to be representative of other regions in Egypt; one of these was in an urban area and the other in a rural area. The OPV coverage in these catchment areas in 1992 was 86% and 88%, respec- tively, similar to the 89% coverage estimated for the whole of Egypt in 1992. In both areas, OPV was delivered from fixed sites in the first round of the mass campaign, and by house-to-house visits in the second round. During both rounds, the age, sex, poliomyelitis vaccination histories, method of vaccine history con- firmation (card/history), location (inside/outside the catchment area), and residence area (urban/rural) were recorded for each child on standardized forms. The costs associated with each round were calculated based on the number and type of health care person- nel needed, cold chain requirements, and the number and size of OPV vials used. In the first round, the locations of fixed vaccine delivery sites (four sites in the urban area and two sites in the rural area) were temporarily established so that the campaign would be evenly spaced throughout the catchment areas, providing the maxi- mum number of children with easy access to vacci- nation. At mosques within both catchment areas announcements were made throughout each day for adults to bring their children to the sites for vaccina- tion. OPV was administered each day from 10:00 until no additional children arrived for vaccination. In the second round, house-to-house vaccine delivery was conducted in the catchment areas (non- catchment areas delivered vaccine from fixed sites). Street maps were obtained for both urban and rural catchment areas and each vaccination team was assigned a specific area to cover. These areas were then subdivided into six sub-areas to ensure that the entire area would be covered during the 6-day campaign. Teams then made door-to-door visits and vaccinated all eligible children. Households were re- visited repeatedly if eligible children were identified but absent when the teams called; thus, all eligible children were probably vaccinated during the 6-day vaccination period. Statistical analysis The vaccine coverage achieved in the first round was estimated in both the urban and rural catchment areas by calculating the proportion of eligible chil- dren enumerated in the house-to-house visits who reported having been vaccinated during this round, after excluding children who were younger than 1 month of age (i.e., those children not yet bom during the first round). The costs per child vaccinated and the costs per high-risk child vaccinatedb were calcu- lated for each of the vaccine delivery strategies by summing the actual costs for health care personnel, cold chain equipment, and the number and size of vaccine vials; these costs were then divided by the number of children vaccinated to estimate the costs associated with each child vaccinated. Children were categorized into the number of doses they had receiv- ed prior to the campaign, and the costs associated with vaccinating high-risk children were calculated. In an effort to identify variables that may have been associated with not being vaccinated from fixed sites, vaccine coverage estimates were calculated for each of the variables collected on the standardized questionnaire in the second round, and x2 tests were then used to assess statistically significant differences. Results Vaccine coverage by type of vaccine delivery Approximately 86% of children enumerated in both the urban and rural catchment areas in the second round reported having been vaccinated in the first round (Table 1). Because all eligible children who resided in the catchment areas in the second round were believed to have been enumerated, owing to the repeat household visits made by vaccination teams, vaccine coverage in the second round was assumed to be 100%. Differences between these vaccine cover- age estimates were statistically significant (P <0.01, X2 test). Vaccine delivery costs Overall costs were about 25% higher for house-to- house vaccination than for fixed-site vaccination in the urban area; costs in the rural area were similar (Table 2). More health care personnel and cold boxes were needed for house-to-house vaccination, particu- larly in the urban areas. In both vaccination rounds the cost of vaccine accounted for approximately 75% of the total cost of the campaign. bHigh risk was defined as having received either zero or one dose of OPV prior to the mass campaign. 590 WHO Bulletin OMS. Vol 73 1995 House-to-house versus fixed-site OPV delivery in Egypt Table 1: Coverage with OPV in the study population, by vaccination round and by catchment area, mass cam- paign, Egypt, 1993 No. Target Vaccination vaccinated populationa coverage (%) First round: fixed site delivery Urban 4 095b 4 747 86.3 Rural 1 931b 2 256 85.6 Second round: house-to-house delivery Urban 4 747 4 747 100 Rural 2 256 2 256 100 a Based on the number of eligible children enumerated in the catchment areas in the second round, after excluding children who were aged <1 month in the second round (i.e., children not yet born in the first round). b Based on the proportion of eligible children enumerated in the catchment areas in the second round who reported having received a vaccine in the first round, after excluding children who were aged <1 month in the second round (i.e., not yet born in the first round). Estimated cost of vaccine delivery per child Although vaccination costs on a per child basis could be precisely determined for the second round, owing to the house-to-house enumeration of eligible chil- dren, those for the first round could not, because 24% (1570/6422) of the children vaccinated at the fixed sites reported living outside the catchment areas. Consequently, information collected in the house-to-house enumeration on first round participa- tion was used to estimate costs. Although this infor- mation does not differentiate children vaccinated at fixed sites inside versus those outside the catchment areas, it none the less provides cost estimates consis- tent with the 86% observed coverage. In the urban areas, the costs per child vaccinated at fixed sites were equivalent to the house-to-house vaccination costs (£E 0.37 (US$ 0.11) per dose), but were LE 0.10 (US$ 0.03) per dose higher in the rural areas (Table 3). In both the urban and rural areas, the costs of vaccinating children from fixed sites who reported having received zero or one dose of OPV prior to the campaign were approximately 2-4 times higher than the costs of house-to-house vaccination of these high-risk children. Although vaccine wast- age in the urban areas was approximately 25% for both vaccine delivery strategies, in the rural areas wastage was nearly twice as high for fixed-site vaccine delivery (41.5%) than for house-to-house delivery (23.5%). Table 2: Total costs for OPV delivery, by catchment area, mass campaign, Egypt, 1983a Fixed-site delivery: House-to-house delivery: Urban Rural Urban Rural Expenditure No. Cost (US$)b No. Cost (US$)b No. Cost (US$)b No. Cost (US$)b Personnel: Physicians 12 21.62 6 10.81 12 21.62 6 10.82 Nurses 12 16.22 12 16.22 24 32.43 18 24.32 Hygienists 12 18.02 0 0 12 18.02 0 0 Clerks 12 14.41 12 14.41 24 28.83 18 21.62 Community workers 18 16.22 18 16.22 24 21.62 18 16.22 Drivers 6 9.01 6 9.01 12 18.02 6 9.01 Subtotal 95.50 66.67 140.54 81.99 Vaccine vials c 20-dose 45 54.05 0 0 0 0 0 0 50-dose 93 279.28 66 198.20 128 384.38 59 177.18 Subtotal 333.33 198.20 384.38 177.18 Cold boxesd 0 0 0 0 24 0.30 18 0.30 Total costs (US$) 428.83 264.87 525.22 259.47 a Because the costs of one refrigerator, transportation, and publicity were the same in both the first and second rounds, they were not included in the cost estimates. Additionally, although 252 extra ice-packs were used in the house-to-house campaign (144 in the urban area and 108 in the rural area) the costs of ice-packs over the 6-day vaccination period were considered too small to estimate.b US$1 = LE 3.33. c Based on US$ 0.06 per dose of OPV. d Costs were based on the price of one cold box (US$ 19.17) and on its estimated life (5-years' daily use by the immunization pro- gramme): the cost of each cold box was estimated to be US$ 0.01 per day. WHO Bulletin OMS. Vol 73 1995 591 R.W. Linkins et al. Table 3: Costs per child vaccinated, by type of delivery and by catchment area, mass campaign, Egypt, 1993 No. of prior OPV doses r 0 .2 Unknown Cost per child wirrinntpri (MR$M5 Variables associated with nonvaccination from fixed sites Fixed-site House-to-house Although fixed-site vaccine coverage estimates delivery: delivery: (Table 4) were similar in urban and rural areas and Urban Rural Urban Rural for males and females, there were statistically signi- eceiveda ficant differences (P <0.01) for age at vaccination 24 2 190 74 (youngest and oldest children were less likely to be 110 40 169 98 vaccinated), the number of previous OPV doses 3 961 1 889 4 364 2 084 reported (children with <3 prior OPV doses were961 0 24 2 less likely to be vaccinated), and the method of veri-0 0 24 2 fication of vaccine history (children without vaccina- 0.11 0.14 0.11 0.11 tion cards were less likely to be vaccinated). va"l lamcu kuQPJ- Cost per high-risk childc 3.20 6.31 vaccinated (US$) OPV wastage (%) 26.2 41.5 Total number vaccinated 4 095 1 931 1.46 1.51 25.8 23.5 4 747 2 256 a Received prior to either round. b US$ 1.00 = EE 3.33. c 0-1 OPV doses received prior to the vaccination campaign. Table 4: Variables associated with vaccine 1993 Characteristic Location Urban Rural Age range (months) 1-11 12-23 24-35 36-47 48-59 >60 Sex Male Female No. of prior OPV dosesd 0 1 2 .3 Vaccine verification Card History Total No. vaccinateda 4 095 1 931 854 934 1 065 1 170 1 239 764 3 110 2 916 26 150 222 5 628 1 035 4 991 6 026 Discussion Although the level of population immunity required to achieve elimination of wild poliovirus infection has not been determined, the emergence of large- scale outbreaks of poliomyelitis in highly vaccinated receipt from fixed sites, mass campaign, Egypt, No. in target populationb 4 747 2 256 1 062 1 061 1 184 1 335 1 420 941 3 589 3 414 140 195 266 6 402 1 131 5 872 7 003 Vaccination coverage (%) 86.3 85.6 80.4 88.0 90.0 87.6 87.3 81.2 86.7 85.4 P-valuec 0.45 0.00 0.13 18.6 76.9 83.5 87.9 91.5 85.0 86.1 0.00 a Based on the proportion of eligible children enumerated in the catchment areas in the second round who reported having been vaccinated in the first round, after excluding children who were aged <1 month in the second round (i.e., children not yet born in the first round). b Based on the number of eligible children enumerated in the catchment areas in the second round, after excluding children who were aged <1 month in the second round (i.e., children not yet born in the first round). c By x2 test. d Excluding doses received in the first round. e By x2 test for linear trend. WHO Bulletin OMS. Vol 73 1995592 House-to-house versus fixed-site OPV delivery in Egypt populations in the Gambia (3), Oman (4), Jordan (5), and South Africa (6) suggests that only relatively small numbers of susceptible children are needed to sustain transmission. These observations, along with the results of serological surveys which indicate that up to 64% of children will fail to develop neutral- izing antibody to poliovirus types 1 or 3 following routine administration of three doses of OPV (7), have reinforced the need to develop supplementary immunization strategies to minimize the number of susceptible children. Mass vaccination campaigns can achieve this goal in a number of ways, including the following: more aggressive efforts to vaccinate children who may have been missed in the routine programme; higher seroconversion rates than those achieved by routine delivery (8); repeated boosting of secretory antibody (which is generally short lived) in seropositive children; higher levels of immunity in the general population owing to intensive spread of vaccine virus to household and community contacts (9-11); and potential "displacement" of wild polio- virus following vaccination of large numbers of children simultaneously (12). These considerations, combined with the remarkable success of national vaccination days in Latin America (2), provide sub- stantial support for the use of this approach in polio- endemic countries. Because the principal goal of mass vaccination campaigns is to reduce the number of children who are susceptible to poliovirus infection, the effective- ness of campaigns depends primarily on the level of coverage achieved. Although house-to-house admin- istration of OPV ensures that virtually 100% of chil- dren are vaccinated, many countries have been reluc- tant to implement this approach because it is perceived to be more labour intensive and costly than vaccination from fixed sites. Although these perceptions were borne out in the present study, where personnel and total costs were approximately 38% higher and 13% higher for house-to-house than for fixed-site delivery, respectively, the costs on a per child basis were similar for both approaches. This finding was due primarily to the higher levels of coverage achieved in house-to-house delivery (100% versus 86%), as well as the reduced vaccine wastage, particularly in rural areas. Of perhaps great- er relevance was the observation that the cost of vac- cinating children at highest risk of wild poliovirus infection (those with one or zero prior doses of OPV), was only one-quarter to one-half as expensive on a per child basis in the house-to-house campaign. The finding that such children are less likely to par- ticipate in a fixed-site campaign was also reported in a study in Cameroon, where most campaign partici- pants had already been fully vaccinated in a routine programme (13). Although in Egypt the costs involved in a house- to-house campaign were essentially offset by higher coverage rates, particularly for high-risk children, our findings may not be generalizable to other coun- tries. First, the apparent cost advantages of house-to- house delivery in Egypt were due largely to very low expenses for labour; if such costs were only 2-4 times higher, as is the case in many other countries (14), the cost advantages of house-to-house delivery would largely disappear. Second, the population density in Egypt is high (1450 per km2) (15), even in rural areas; this may also have resulted in lower costs for labour and supplies than those in other, less densely populated areas. Thus, in considering wheth- er house-to-house campaigns should be implement- ed, the following factors, inter alia, must be evaluat- ed: the local availability and cost of labour, transportation, and supplies; population density; the estimated number of high-risk children and the degree of difficulty in reaching them using routine immunization services or fixed-site campaigns; the desirability of providing vaccines other than OPV (often difficult to carry out in house-to-house cam- paigns); and the concern that parents may become more complacent in seeking immunization and other primary health care services for their children (16). Nevertheless, in view of the large and continuing burden of poliomyelitis among unvaccinated children in Egypt (Centers for Disease Control, Atlanta, GA, USA, unpublished data, 1993) and elsewhere (17), house-to-house delivery of OPV could very well prove to be the most cost-effective approach. Acknowledgements We thank the following team members for their assistance with data collection: F. Abdeen and A. Hassan (Ministry of Health, Cairo); and M. Salah, M. Gendy, A. Younis, A. Abdo, and M. Wally (Benha University, Cairo). Resume Comparaison des strategies d'adminis- tration du vaccin antipoliomyelitique oral a domicile et dans des centres fixes lors d'une campagne de vaccination de masse en Egypte Une des activit6s recommandees par l'OMS en vue de l'radication de la poliomyelite est l'organi- sation de journ6es nationales de vaccination: tous les enfants de la tranche d'age appropri6e, quelle que soit leur situation vaccinale, recoivent une dose de vaccin oral (VPO), suivie quatre a six WHO Bulletin OMS. Vol 73 1995 593 R.W. Linkins et al. semaines plus tard d'une deuxieme dose. II est souhaitable que cette campagne de masse ne dure que quelques jours ou quelques semaines, de pr6ference pendant la saison ou la transmis- sion du poliovirus est ralentie. Le plus souvent, le VPO est administr6 dans des centres fixes qui touchent 80 a 90% de la population cible. Meme si la vaccination a domicile assure une couverture presque totale, les pays hesitent a adopter cette m6thode car ils estiment qu'elle est plus coOteuse et pr6sente des difficult6s logistiques. Pour quantifier le rapport cout/efficacite de ces deux strategies de vaccination, nous avons compar6 la couverture et le coOt par enfant de I'administration du VPO a domicile et dans des centres fixes lors d'une campagne de vaccination de masse en Egypte en 1993. En outre, nous avons evalu6 les caracteristiques des enfants vac- cin6s dans les centres fixes afin de determiner ceux qui risquent d'etre laisses a l'ecart lorsque cette strat6gie est retenue. Deux secteurs bien d6finis et repr6sentatifs (un secteur urbain et un secteur rural) ont ete 6valu6s. Dans ces deux secteurs, le VPO a et6 administr6 dans des centres fixes lors de la pre- miere phase de la campagne et a domicile au cours de la deuxieme phase. Durant les deux phases, les informations pertinentes ont ete recueillies pour chaque enfant vaccine. Pour cal- culer les coOts associes a chaque phase, on a pris en compte le nombre et la qualification des personnels de sant6 necessaires, les besoins de la chaine du froid, ainsi que le nombre et la contenance des flacons de VPO utilises. Malgr6 un coOt en personnel et un coOt total plus 6leves pour la vaccination a domicile (res- pectivement 38% et 13% de plus), le coOt par enfant vaccine a et6 comparable. Ce r6sultat est dO principalement a la couverture 6lev6e de la vaccination a domicile (100% contre 86%) et a un moindre gaspillage de vaccin. La vaccination des enfants les plus exposes au risque d'infection (ceux qui n'avaient jamais 6te vaccin6s ou qui n'avaient recu qu'une dose de VPO avant la cam- pagne) s'est r6v6l6e deux a quatre fois moins coOteuse par tete dans le cas de la vaccination a domicile, car ces enfants avaient moins de chan- ce d'tre vaccin6s dans les centres fixes. Bien que l'on ne sache pas exactement quel niveau doit atteindre le taux de vaccination dans la population pour 6viter l'infection par le virus sauvage, les flamb6es importantes survenues dans des populations bien vaccin6es, en Gambie, en Oman, en Jordanie et en Afrique du Sud, per- mettent de supposer qu'il suffit d'un nombre relati- vement faible d'enfants sensibles pour que la transmission du virus se poursuive. Ces observa- tions, ainsi que les r6sultats d'enquetes s6rolo- giques selon lesquelles la proportion d'enfants ne produisant pas d'anticorps diriges contre les virus de type 1 ou 3 apres administration de trois doses de VPO peut atteindre 64%, confirment la n6ces- sit6 de mettre au point des strat6gies de vaccina- tion complementaires pour reduire au minimum le nombre d'enfants sensibles a l'infection. Etant donne que c'est la pr6cisement le premier objectif des campagnes de vaccination de masse, leur efficacite d6pend principalement du niveau de couverture atteint. Les resultats de cette evaluation ne sont pas necessairement applicables a d'autres pays ou le coOt de la main-d'ceuvre est plus eleve et la den- sit6 de population plus faible; n6anmoins, I'admi- nistration du VPO a domicile peut se reveler la strategie la plus rentable pour l'6radication de la poliomyelite en assurant un acces universel aux services de vaccination. References 1. Resolution WHA41.28: Global eradication of polio- myelitis by the year 2000. In: Handbook of resolu- tions and decisions of the World Health Assembly and Executive Board, vol. III, 3rd edit. (1985-1992). Geneva, World Health Organization, 1992: 100-101. 2. de Quadros C et al. Eradication of poliomyelitis: progress in the Americas. Pediatric infectious dis- ease journal, 1991, 10: 222-229. 3. Otten M et al. Epidemic poliomyelitis in The Gam- bia following the control of poliomyelitis as an endemic disease. I. Descriptive findings. American journal of epidemiology, 1992, 135: 381-392. 4. Sutter R et al. Outbreak of paralytic poliomyelitis in Oman: evidence for widespread transmission among fully vaccinated children. Lancet, 1991, 338: 71 5-720. 5. Reichler M et al. Outbreak of poliomyelitis in a highly immunized population in Jordan. In: Program and Abstracts of the 32nd Interscience Conference on Antimicrobial Agents and Chemotherapy, Ana- heim, 11-14 October 1992. Washington, DC, Ameri- can Society for Microbiology, 1992: 398. 6. Schoub B et al. Poliomyelitis outbreak in Natal/ KwaZulu, South Africa, 1987-1988. 2. Immunity aspects. Transactions of the Royal Society of Tropi- cal Medicine and Hygiene, 1992, 86: 83-85. 7. Patriarca PA et al. Factors affecting the immuno- genicity of oral poliovirus vaccine in developing countries: review. Reviews of infectious diseases, 1991, 13: 926-939. 8. Richardson G et al. Immunogenicity of oral polio- virus vaccine (OPV) given in mass campaigns versus routine immunization programs. In: Program and Abstracts of the 33rd Interscience Conference on Antimicrobial Agents and Chemotherapy, New 594 WHO Bulletin OMS. Vol 731995 House-to-house versus fixed-site OPV delivery in Egypt Orleans, 17-20 October 1993. Washington, DC, American Society for Microbiology, 1993: 346. 9. Gelfand H et al. Revised preliminary report on the Louisiana observations of the natural spread within families of living vaccine strains of poliovirus. In: Live poliovirus vaccines: papers presented and discussions held at the First International Confer- ence on Live Poliovirus Vaccines, Washington, DC, 22-26 June 1959. Washington, DC, Pan Ameri- can Sanitary Bureau, 1959 (Scientific Publication No. 44). 10. Benyesh-Melnick M. Studies of the immunogen- icity, communicability and genetic stability of oral poliovaccine administered during the winter. Ameri- can journal of epidemiology, 1967, 86: 112-136. 11. Chen R et al. Extent of secondary spread of oral polio vaccine virus in inner-city preschool children. In: Program and Abstracts of the 32nd Interscience Conference on Antimicrobial Agents and Chemother- apy, Anaheim, 11-14 October 1992. Washington, DC, American Society for Microbiology, 1992: 344. 12. Sabin A et al. Live, orally given poliovirus vaccine. Effects of rapid mass immunization on population under conditions of massive enteric infection with other viruses. Journal of the American Medical Association, 1960, 173: 1521-1526. 13. McBean A et al. Evaluation of a mass measles immunization campaign in Yaounde, Cameroun. Transactions of the Royal Society of Tropical Medi- cine and Hygiene, 1976, 70: 206-212. 14. Shepard D. Cost-effectiveness of routine and cam- paign vaccination strategies in Ecuador. Bulletin of the World Health Organization, 1989, 67: 649-662. 15. Egypt. In: Bair F, ed. Countries of the world and their leaders yearbook 1992, vol. 1. Detroit, Ml, Gale Research International Limited, 1992: 511. 16. Bryce J et al. Mass immunization campaigns and quality of services. Lancet, 1990, 33: 739-740. 17. World Health Organization. Expanded Pro- gramme on Immunization. Poliomyelitis in 1992. Weekly epidemiological record, 1993, 68(31): 225-230. WHO Bulletin OMS. Vol 731995 595