Immunogenicity of oral poliovirus vaccine administered in mass campaigns versus routine immunization programmes.

Immunogenicity of oral poliovirus vaccine administered in mass campaigns versus routine immunization programmes* G. Richardson,1 R.W. Linkins,2 M.A. Eames,3 D.J. Wood,4 P.J. Campbell,5 E. Ankers,6 M. Deniel,7 A. Kabbaj,8 D.l. Magrath,4 P.D. Minor,4 & P.A. Patriarca2 Reported are the results of a study to investigate the immunogenicity of oral poliovirus vaccine (OPV) when administered in mass campaigns compared with that following routine immunization programmes. For this purpose, paired sera were collected from a cohort of children before and after a mass vaccina- tion with OPV in Morocco in 1987. Serum samples and information on vaccination status and other con- founding factors that could influence antibody responses to OPV were collected. Neutralizing antibody titres to poliovirus types 1, 2 and 3 were determined using a standardized assay. OPV doses adminis- tered exclusively during the mass campaign were consistently associated with higher type-specific sero- prevalence rates than the same number of doses administered in the routine programme. These find- ings could not be attributed to differences in confounding factors. Enhanced secondary spread of vaccine virus may have occurred but could not be demonstrated because of limitations in the study design. Mass campaigns appear to be highly effective in raising the dose-related poliovirus type-specific immunity of the population above that achieved by the routine immunization programme. Our findings support the continued use of mass campaigns as an adjunct to routine programmes in order to both enhance and catalyse current efforts to achieve the global eradication of poliomyelitis by the year 2000. . Part of the information described in this article was presented at the 33rd Interscience Conference on Antimicrobial Agents and Chemotherapy, New Orleans, 17-20 October 1993 (Abstract No. 1237). Requests for reprints should be sent to Dr D.J. Wood, Divi- sion of Virology, National Institute for Biological Standards and Control, Blanche Lane, South Mimms, Potters Bar, Herts EN6 3QG, England. 1 University of Wales College of Medicine, Cardiff, Wales. Cur- rent address: Mid-Glamorgan Health Authority, Pontypridd, Wales. 2 Centers for Disease Control and Prevention, Atlanta, GA. Dr Patriarca is currently at Center for Biologics Evaluation and Research, Food and Drug Administration, Rockville, MD, USA. 3 University College London and Middlesex School of Medicine, London, England. Current address: University of Hertfordshire, Hatfield, Herts, England. 4 National Institute for Biological Standards and Control, Potters Bar, Herts, England. 5 Thanet District General Hospital, Margate, England. Current address: Ashgabad, Turkmenistan. 6 Charing Cross Hospital, London, England. Current address: Portsmouth and SE Hants Health Commission, Portsmouth, England. 7 National Immunization Programme, Ministry of Public Health, Rabat, Morocco. 8 Avenzoar Hospital, Marrakech, Morocco. Reprint No. 5658 Introduction In 1988, the Forty-first World Health Assembly adopted a resolution that committed WHO to the goal of global eradication of poliomyelitis by the year 2000 (1). In addition to the routine administra- tion of four doses of oral poliovirus vaccine (OPV) at birth and at 6, 10, and 14 weeks of age, mass immunization campaigns for all children less than 5 years of age are recommended by WHO for coun- tries where poliomyelitis is endemic (2-4). Although mass campaigns and house-to-house vaccination of children in high-risk areas have been highly success- ful in eliminating wild poliovirus infection from Latin America (3, 4), the specific reason for the effectiveness of such campaigns has not been identi- fied. The following mechanisms have been postula- ted: further reductions in the number of unimmuni- zed children owing to more aggressive approaches in vaccine delivery; higher immunogenicity of OPV given during mass campaigns; repeated boosting of secretory immunity in the overall population (chil- dren and adults) because of more intensive spread of vaccine virus to susceptible persons in all age groups; and "displacement" of wild poliovirus and Bulletin of the World Health Organization, 1995, 73 (6): 769-777 © World Health Organization 1995 769 G. Richardson et al. non-polio enteroviruses following massive and widespread use of OPV (5, 6). A cross-sectional serosurvey of children in Morocco conducted before and after a mass immunization campaign in 1987 provided us with a unique opportunity to evaluate the immunogenicity of OPV administered in mass campaigns and to compare the seroprevalences to poliovirus types 1, 2 and 3 with those elicited by routine immunization. Trivalent oral poliovirus vaccine (OPV) with a potency of 106, 105, and 1055 (300 000) TCID50 for poliovirus types 1, 2, and 3, respectively, has been used in Morocco since 1963 and was integrated into the routine immunization programme during the 1970s. Although the annual number of reported cases of paralytic poliomyelitis declined from a peak of 577 in 1975 to 22 in 1984,a immunization coverage of 1-year-old children with three doses of OPV still remained relatively low at 45% (7). At this time the Moroccan Ministry of Health was using a two-point strategy aimed at vaccinating all newborns by 1 year of age, as well as children aged <36 months who had previously missed at least one dose of OPV. In urban populations and in rural groups located within a 2-km radius of a health clinic, three doses of OPV were administered at monthly intervals beginning at 3 months of age. In rural populations outside the 2-km radius, mobile services administered OPV on a quarterly basis at the age of 3, 6, and 9 months (7). Despite implementation of this strategy between 1981 and 1985, a follow-up survey conducted by the Ministry of Health in 1987 revealed that only half of Moroccan children had been adequately vaccinated against poliomyelitis, diphtheria, tetanus, pertussis, measles, and tuberculosis, the six target diseases of the Expanded Programme on Immunization (EPI). A national campaign was therefore organized to immu- nize all children under 5 years of age who had not yet received the recommended number of doses of each vaccine. Women aged 15-44 years from rural areas were also targeted for immunization with teta- nus toxoid. Three rounds of immunization took place during the campaign, separated by periods of approx- imately 1 month: the first occurred on 2-4 October; the second on 13-15 November; and the third on 18-20 December 1987. An estimated 41% of Moroc- can children aged 0-5 years received one or more OPV doses during this mass campaign. aExpanded Programme on Immunization. Information sys- tem: summary for the WHO Eastern Mediterranean Region. Unpublished WHO document, 1991. Methods Study population and survey design In the month preceding the first round of the mass campaign, stratified multistage random cluster sam- pling methods (8) were used to obtain samples of children aged 1-4 years from three Moroccan pro- vinces. Because immunization coverage and the prevalence of antibody against polioviruses vary with the degree of urbanization (9-11), each province was stratified into urban (cities with >25 000 inhabi- tants), semi-urban (2500-25 000 inhabitants), and rural areas (<2500 inhabitants). The areas sampled included the cities of Casablanca and Marrakesh, and the semi-urban and rural areas of Marrakesh and Ouarzazate provinces, including the High Atlas mountain and pre-Saharan desert regions (Fig. 1). The samples for the survey were proportionate to population sizes estimated in the 1982 census. A total of 30 clusters of eight or more children were sampled from each urban area (8). From semi-urban and rural areas, 15 clusters of eight or more children were sampled due to constraints in collecting speci- mens within the time available for the survey. During the month following the third round of the mass campaign, the survey areas were revisited and attempts were made to locate the original chil- dren sampled. When children could not be located, replacements were chosen randomly from the same cluster site. Thus, only part of the original group of subjects was included in both the pre- and post-mass campaign surveys. Fig. 1. Map showing the study areas in Morocco. Legend ( Countries CED Morocco * Cities Tangier Ceuta * Study areas tIMountains RABAT Casablanc Fes Marrakzhn Agadi r 0|1 neyoKn 100 200 770 WHO Bulletin OMS. Vol 73 1995 Immunogenicity of OPV Enrolment procedures In both surveys, a supervised nurse who had been previously trained by the study investigators ex- plained the study procedures to the parent or guard- ian and obtained their informed consent. Pre-coded questionnaires were used to solicit information on each child's OPV immunization history (either child health cards held by the parent, if available, or by verbal report); the number of siblings in the family (both alive and dead); and type of housing, water supply, and sewage disposal. Samples of blood (3-5 ml) were collected from each child by vene- puncture, regardless of their vaccination status. The blood samples were stored in test-tubes containing serum separation gel and transported in insulated cold-boxes to the local laboratory for initial proces- sing within 72 hours of being collected. Laboratory procedures After being centrifuged the sera were separated, inactivated at 56 °C for 30 minutes, and then frozen and stored at -24 °C before being sent on ice-packs to the National Institute for Biological Standards and Control, England, for testing. The specimens were numerically coded so that laboratory staff were un- aware of the number of doses of OPV each child had received during the routine programme or mass campaign. The methods used to determine the neutralizing antibody titres to poliovirus types 1, 2, and 3 have been described previously.b Briefly, replicate two- fold dilutions of sera (from 1:4 to 1:512) were incu- bated for 3 hours at 35 OC with 100 TCID50 of the corresponding Sabin-type poliovirus. Hep2C cells were added and the assay was scored after 7 days. The serum titre recorded was the highest dilution that inhibited 100% cytopathic effect; a reference antiserum was included in each assay. Titres of .8 for poliovirus types 1 and 2 and .6 for type 3 were considered to be seropositive; these titres were equivalent to 340, 640, and 190 milli-intemational units (mIU) for types 1, 2, 3, respectively, when cali- brated against the second International Standard for anti-poliovirus antibodies (12). Data management To identify children who participated in both the pre- and post-mass campaign surveys, we initially bDetermination of neutralizing antibodies to polioviruses. In: Manual for the virological investigation of poliomyelitis. Unpub- lished document WHO/EPI/CDS/POLIO/90.1, 1990. selected 544 children of the same sex and from the same sampling area and cluster as potential matches. Matching criteria were then established to compare information from pre- and post-mass cam- paign survey questionnaires for each of these poten- tial matches. An agreement score was calculated by weighting the similarity and completeness on the fol- lowing matching criteria: name (40 points); date of birth (25 points); dates of previous doses of OPV (20 points); and street address (15 points). Children with identical information for all matching variables thus received scores of 100; those children with non-iden- tical information were assigned scores according to pre-established degrees of agreement, e.g., children with similar, but not identical names, received 35 rather than 40 points. Agreement scores were made independently by two persons not directly involved in other aspects of the study, and disagreements were resolved before final score assignment. A new questionnaire combining each of the original data items was then produced, and the infor- mation from both the pre- and post-campaign sur- veys was consolidated. If the vaccine histories for individual children in the two surveys differed, reports confirmed by vaccination card were used. All information that did not match was coded as "unknown". This consolidated questionnaire was then double-entered into a computer database, and the two data sets were compared. All discordant values were corrected. Because of the methods used for the pre-cam- paign survey, participants were considered to be representative of all children in the areas sampled. To determine whether participants in both surveys (pre- and post-campaign) were also representative of the general population, the demographic and other characteristics of these children were compared with those of children whose participation was limit- ed to the initial (pre-campaign) survey, based on data collected from the initial survey. All subsequent analyses were based on data consolidated from both surveys, and the OPV history derived from doses verified by card or verbal report. Statistical analysis To compare the immunogenicity of OPV adminis- tered during the mass campaign with doses adminis- tered during the routine immunization program- me, type-specific seroprevalence rates were calcula- ted according to the number of doses reported in the routine (pre-campaign) programme and compared with post-campaign seroprevalences in children who received the corresponding number of OPV doses exclusively in the mass campaign. This analysis esti- mated the direct effect of 1-3 doses of OPV on sero- WHO Bulletin OMS. Vol 73 1995 771 G. Richardson et al. prevalence rates as well as the seroprevalence rates associated with zero OPV doses. Type-specific geo- metric mean titres (GMTs) were also compared for children who were seropositive before and after the mass campaign, according to the number of OPV doses reported in the routine programme and the number of doses received exclusively in the mass campaign. To account for other (confounding) fac- tors that may have influenced seroprevalence rates and GMTs in the two groups, additional analyses were carried out by stratifying children according to the following variables: age (<30 months and 30-50 months) - to account for differences in the length of potential exposure to wild polioviruses; the number of doses administered during the rainy seasons, April-May and October-November (13), versus dry seasons (all other months) - to account for differ- ences in immunogenicity according to the season when OPV was administered (14); and location of the child's residence (urban, semi-urban, or rural) - to account for differences in seroprevalence rates according to the degree of urbanization (9-11). Potential differences in the rate of decay of neutrali- zing antibody between the two groups were assessed by comparing dose-specific seroprevalence rates for children who received OPV exclusively in the mass campaign with those for children vaccinated in the routine programme who had received their last dose of OPV within 126 days of the initial (pre-campaign) survey, which is the median interval between the pre- and post-campaign survey dates. To estimate the relative degree of secondary spread of vaccine virus during the mass campaign versus the routine immunization programme, pre- campaign seroprevalence rates among previously unvaccinated children (n = 113) were compared with the post-campaign rates among the same group of unvaccinated children who received zero doses of OPV during the mass campaign (n = 47). This subset of 47 children who did not receive vaccine in either the routine or mass-vaccination campaigns were rep- resentative of the entire group of 113 children who were not vaccinated in the routine programme, as indicated by a comparison of their demographic pro- files and pre-campaign seroprevalence levels (data not shown). "Seroconversion" was defined as an increase in antibody titre from <8 in the pre-campaign survey to >8 in the post-campaign survey for poliovirus types 1 and 2, and an increase from a titre of <6 in the pre- campaign survey to 26 in the post-campaign survey for type 3. Mantel-Haenszel and McNemar's X2 tests, Fisher's exact test, Mantel-Haenszel x2 test for linear trend, and Student's t-test were used to detect statistically significant differences between the study groups for all the comparisons described above (15). Results Number and characteristics of the study participants A total of 913 children were enrolled in the pre-cam- paign study, 536 (59%) of whom were considered to have participated in both the pre- and post-cam- paign serosurveys. Approximately 50% of these chil- dren had identical and complete information for all matching criteria in the pre- and post-campaign questionnaires. An additional 43% had agreement scores of at least 65 on matching criteria. The remain- ing 7% of children were considered matches, but were given agreement scores of 50-64 because infor- mation was missing for some matching criteria. Res- triction of the analyses to the 325 children with agreement scores of 90 or higher did not substantial- ly alter any of the study findings. A total of 221 (41%) of the 536 children were vaccinated during the mass campaign, 144 of whom received one dose of OPV, 13 two doses, and 64 three doses; 66 of the vaccinated children received OPV exclusively during the mass campaign. Of the 315 children who were not vaccinated, 259 (82%) had received three or more doses of OPV during the routine programme. Pre-campaign seroprevalence rates for the 536 children who participated in both surveys were 74%, 78%, and 54% for poliovirus types 1, 2, and 3, respectively. Post-campaign sero- prevalence rates in this group increased by 14% for each poliovirus type, i.e., to 88% (P <0.05), 92% (P <0.05), and 68% (P <0.05) for types 1, 2, and 3, respectively. Selected characteristics of the 536 children who participated in both surveys and of the 377 children whose participation was limited to the pre-campaign survey are compared in Table 1. The children who participated in both surveys were more likely to have a greater number of siblings; to live in semi-urban and rural areas of the country; to use open sewer sys- tems; and to have been seronegative to poliovirus type 2 prior to the campaign. There were no signifi- cant differences between the two groups for the following variables: the child's age at interview; primary place of vaccination; proportion of vac- cine histories confirmed by card; number of OPV doses received before the pre-mass campaign survey; source of water; type of toilet; or seropositivity to poliovirus types 1 and 3 prior to the mass campaign (data not shown). The distribution of OPV doses received by the 536 children who participated in both surveys is shown in Table 2. WHO Bulletin OMS. Vol 73 1995772 Immunogenicity of OPV Table 1: Comparison of the characteristics of children who participated only in the pre- campaign survey with those in both the pre- and post-mass campaign surveys No. in pre-campaign only: No. in pre and post-campaigns: Characteristic (n = 377) (n = 536) P-value No. of children alive in family 1 46 (13)a 66 (13) 2 82 (22) 93 (18) 3 81 (22) 85 (16) 4 64 (18) 80 (16) 2.5 92 (25) 188 (37) 0.01 No. of children who had died in family 0 250 (68) 317 (62) al 115 (32) 195 (38) 0.05 Area Marrakesh Urban 127 (34) 120 (22) Semi-urban 45 (12) 80 (15) Rural 52 (14) 75 (14) Ouarzazate Semi-urban 31 (8) 91 (17) Rural 36 (9) 84 (16) Casablanca Urban 86 (23) 86 (16) 0.00 Type of sewer Closed 252 (74) 326 (67) Open 72 (21) 118 (24) Other/none 16 (5) 45 (9) 0.02 P2 seropositiveb Yes 315 (84) 418 (78) No 62 (16) 118 (22) 0.04 a Figures in parentheses are percentages. b Titre 28 to OPV 2. Seroprevalence levels in the routine programme and mass campaign Fig. 2 shows the type-specific pre-mass campaign seroprevalence levels in the 482 children who received 0-3 doses of OPV in the routine immuni- zation programme (based on card or verbal history) as well as the post-mass campaign seroprevalence rates in the 111 children who received the corre- sponding number of doses exclusively in the mass campaign (two children were excluded owing to miss- ing data). While few of the differences were signifi- cant at the P<0.05 level, the seroprevalence levels associated with the mass campaign were consistently higher than those of the routine programme. Similar differences were observed when the data were restricted to the 68% of children with OPV doses that were confirmed on their vaccination cards. The dose-specific geometric mean titres (GMTs) for seropositive children before and after the mass immunization campaign were consistent with these findings (data not shown), although none of the differences was statistically significant. Potential confounding variables Examination of other variables that could have influenced seroprevalence rates in the two groups Table 2: Distribution of OPV doses among children who participated in both the pre- and post-mass campaign surveys No. of mass campaignNo. of routine OPV doses programme OPV doses 0 1 2 3 Total 0 47 13 3 50 113 1 5 2 9 4 20 2 4 17 0 2 23 3 220 97 1 8 326 4 39 15 0 0 54 Total 315 144 13 64 536 WHO Bulletin OMS. Vol 73 1995 773 G. Richardson et al. Fig. 2. Type-specific seroprevalence rates, b' for the routine programme and the mass Morocco, 1987 (figures above each column arE seropositive/total number tested; signifies P s( Type 1 2p2100 - -- _ Al 80- _ 11/t3 15/23A ailf |!!1212/2o 40a0.'l o 'drepdo i---A y OPV dose, remained when the children were stratified according s campaign, to the number of doses received during the rainy the number and dry seasons and according to whether the child 0.05). resided in an urban, semi-urban, or rural location (data not shown). Aggregate (multivariate) analysis controlling for each of these variables simultaneous- 262n26 39/49 ly was not possible because of the low number of AeAe children in many of the strata. .eAe.e Potential differences in secondary spread of ArAe vaccine virus .0 -e-e In an effort to determine whether increased faecal AeAe shedding and secondary transmission of vaccine Ae? virus may have occurred in the mass campaign rela- AeA tive to the routine programme, the pre-campaign 3 seroprevalence rates among the 113 unvaccinated children were compared with the post-campaign Al 80 - E ror jo Ie 144/20 o l o oP o. e e t,60 W / S J - // / / / S No.edofe doses p doodd 60471113 'Peee 13 J2ipd ede- p a 40Img o 'OOPeP'PJP' e Jr de d Jo e'p'p'p' 'p'p'p'p i i i Cl) 20-NoofIFII s es -dp 'p *p c0.05p _p 'Routin ee pee Mass 'pip r pam pep camp aign JPIf.? reeaednoipotndifrcebtwn them. A shown riFi3a No. edosese 20vlec raeswee conistnl higer formas Iof poeni lrexposure to wild Jriviu in eto(ae3 moth versus 30-5 monhs and dferent Cosstn dpifferne beteen the twrusas Type 3 10 60 24/4op--op 1/ 4l0 ,.e?PPd ePd No.ree of do es l rd i005 routin e dreMass poe CO20 oprog dramm e campaig revealedno importantdiferencesObetweene the.,A sOwninFi. an dFig 4 t edos-pecifcAsroprevalencerates wer~~eeonistntl higer fo mas capig ossafe cnrllingfr varidou legh (ae<0moth eru 305 mots3nifrn Conisealdn motntdifferences betweenthemto.rop Also Fig. 3. Type-specific seroprevalence rates, by OPV dose, for the routine programme and the mass campaign stratified by age (figures above each column are the num- ber seropositive/total number tested; * signifies P s0.05). 100 - Type 1 Al 80 /4 1 rs12_ 1 lE Dw i5l -j; 60 22113 41 0X 31 aa1 Ca 40 < 11_1 l CO 20 |U_l|____ o 1 2 3 0 1 2 3 Age, <30 months No fdss Age, 30-50 months Type 3 n00. 9/9/~~~~~~~~~~~~~~~~s 'R. 80 _ 3f4 113mo ._~~ ~ ~ ~ ~ ~ A_e. _0- s9t40 5m141 41I_ I elm 20 0 1 2 3 0 1 Age, c30 mronths No fdssAge, 30-5 *P50.05 _ Pro-campaign _ Post-campaign WHO Bulletin OMS. Vol 73 1995 2W 774 Immunogenicity of OPV Fig. 4. Type-specific seroprevalence rates, by OPV dose, for routine programme doses given within 126 days of enrolment and mass campaign doses (figures above each column are the number seropositive/total num- ber tested; * signifies P <0.05 by Fisher's exact test). Sero- positivity was defined as titres z8 for OPV1 and OPV2 and ;6 for OPV3. 100 11/13 80 _ 4) 0. 040 Co 1 2C C 43/49 11/13 I 1" 2 3 1* 2 3 Type 1 Type 2 No. of doses lI 1* 2 3* Type 3 _ Routine programme Mass campaign seroconversion rates among the subset of 113 chil- dren who did not receive any OPV during the mass campaign. Seroconversion in the latter group was presumably due to secondary transmission of vaccine virus. As shown in Table 3, the seroconversion rates among children who received zero doses of OPV during the mass campaign were 50%, 64% and 33% for types 1, 2, and 3, respectively; the seroprevalence rates in children who received zero doses of OPV during the routine programme were similar (54%, 62%, and 42% for types 1, 2, and 3, respectively) even though these children, by virtue of their age (mean, 2.4 years), were far more likely to have been exposed to wild poliovirus infection than the mass campaign group (3-month period). Estimates of the secondary transmission of vaccine virus due to the mass campaign appeared to be greatest for poliovirus type 2 and least for type 3, which is consistent with the known excretion patterns of Sabin viruses (9). Discussion Doses of OPV given during the 1987 mass campaign in Morocco were consistently associated with higher type-specific seroprevalence rates than the doses administered during the routine programme. Al- though the findings were generally not statistically significant, no explanation for these differences could be found when the two groups were evaluated for other factors such as the interval between vacci- nation and specimen collection, degree of urbaniza- tion, or variation in the seasonality of vaccine ad- ministration. Although no information on the effect of exposure to wild polioviruses on pre- and post- campaign seroprevalence rates was available, the cumulative risk of natural infection was probably con- siderably higher during the 5-year period represen- ted by the pre-campaign seroprevalence rates than for the 3-month period reflected in post-campaign seroprevalence rates. In Morocco the number of cases of poliomyelitis declined from 22 in 1984 to nine in 1987; two cases were reported in 1989;c con- sequently, differences between the two groups were probably underestimated. Although the cohort of children included in the analysis differed in some respects from the children who participated in the original (pre-campaign) survey, the potential non- representativeness of the cohort would affect only the generalizability of these findings rather than the validity of the pre- and post-mass campaign survey comparisons. Although we were unable to identify directly factors that could explain the improved immuno- genicity of OPV when administered in mass cam- paigns, indirect evidence suggests that enhanced secondary spread of vaccine virus could account for c See footnote a, p. 770. Table 3: Estimates of type-specific secondary transmission of OPV in the routine programme and mass immunization campaign Poliovirus Routine programme: Mass campaign: type No. seropositive Totala No. who seroconverted Totaib 1 61 1 13 (54)C 10 20 (50) 2 70 113 (62) 11 16 (64) 3 47 113 (42) 10 30 (33) a Total number of children tested. b Seronegative to the respective poliovirus type at the time of enrolment (pre-campaign). cFigures in parentheses are percentages (uncorrected for differences in the interval of potential exposure to wild poliovirus infections in the mass campaign (3 months) versus the routine programme (average age, 2.4 years)). WHO Bulletin OMS. Vol 73 1995 775 G. Richardson et al. the differences we observed. Although there were no important differences between the routine pro- gramme and the mass campaign groups in terms of the estimated degree of secondary spread, any differ- ences could have been obscured because the routine group was much more likely to have been exposed to wild poliovirus infection, with attendant increases in seroprevalence rates among unvaccinated children. Overall, the rate of seroconversion by unvaccinated children during the mass campaign was substantial and supports this hypothesis. In addition, differences in dose-specific seroprevalence rates remained after stratifying the data by age, a proxy for wild-virus exposure. Previous studies have also shown that while vaccine virus can spread among household and community contacts following routine administration of OPV (16), massive and prolonged excretion occurs when the vaccine has been administered to large numbers of children simultaneously (17). Another potentially important factor that could not be assessed directly was the effect of potential differences in the maintenance of the cold chain. Prior to the mass campaign, cold chain storage and transport facilities had been increased fivefold. In addition, vaccine administered in the mass campaign was probably more rapidly used after receipt from the manufacturer than was the OPV in the routine programme. Because these factors were not studied, we cannot exclude the possibility that changes in the cold chain during the mass campaign also contribu- ted to improvements in vaccine immunogenicity. In summary, the findings of the present study support the use of mass campaigns as a means of enhancing the effectiveness of routine immunization programmes. The effectiveness of mass campaigns may prove to be even greater if they are conducted according to recent WHO guidelines (which call for the administration of two doses of OPV to all chil- dren aged <5 years, regardless of their vaccination historyd), since the mass campaign was targeted at incompletely vaccinated children and only 41% of children in our survey were given OPV during the campaign. While care must be taken to ensure that primary health care services are not affected detri- mentally during the planning and implementation phases of the campaign (18), the gains in terms of rapid improvements in the overall immune status of the population were clearly evident. The use of repeated campaigns also appears justified, in view of suboptimal antibody responses to poliovirus types 1 and 3 (9), and the relatively large proportion of chil- d Eradication of poliomyelitis: report of the third consultation, Geneva, 3-6 September 1990. Unpublished document WHO/EPI/POLIO/90.3, 1990. dren who remained seronegative to one or more types following the mass campaign in Morocco. The augmentation of ongoing health programmes by national vaccination days has been a critical factor in the successful eradication of poliomyelitis from the Region of the Americas (2-4). The main conclusion from the present study is that mass campaigns must continue to play an important role in the thrust towards global eradication of poliomyelitis, especial- ly if it is to be accomplished by the year 2000. Acknowledgements Support for the mass immunization campaign was pro- vided by UNICEF, the U.S. Agency for International Development (USAID), and Rotary International. The Moroccan Ministry of Health, Save the Children Fund, and the Cardiff Guild of Graduates supported the survey. Ethical approval for the study was obtained from South Glamorgan Ethical Committee in accordance with University of Wales guidelines and from the Moroccan Ministry of Health. All participants in this study gave their free and informed written consent. The following individ- uals are also thanked: P. Elwood, M. Burr, J. Marple, E. Duncan, R. Keegan, S. Gaze, N. Stott, A. John, J. Howlett, J. Hicks, and B. Davies. Resume Immunogenicite comparee du vaccin antipoliomyelitique oral administre lors des campagnes de masse et de celui des programmes de vaccination classiques Les campagnes de vaccination de masse par le vaccin antipoliomy6litique oral (VPO) font mainte- nant partie int6grante des strat6gies visant a obte- nir l'6radication mondiale de la poliomy6lite d'ici l'an 2000. Cependant, le m6canisme sp6cifique de l'efficacite de ces campagnes n'a pas 6t6 6lu- cid6. En octobre 1987, une campagne nationale a ete lanc6e au Maroc pour vacciner les enfants de moins de 5 ans n'ayant pas encore requ au moins trois doses de VPO. En recueillant des echan- tillons de s6rum appari6s dans une cohorte d'enfants avant et apres la campagne, nous avons pu 6tudier l'immunog6nicit6 du VPO admi- nistr6 lors de la campagne de masse et la compa- rer a celle du vaccin administr6 dans le cadre d'un programme de vaccination classique. Au cours du mois pr6c6dant le premier cycle de la campagne de masse, on a proc6d6 a un sondage par grappes stratifi6, en plusieurs 6ta- pes, chez les enfants de 1 a 4 ans des r6gions 776 WHO Bulletin OMS. Vol 731995 Immunogenicity of OPV urbaines, semi-urbaines et rurales de trois pro- vinces marocaines. On a recueilli des echantillons de s6rum et des renseignements sur l'6tat vacci- nal et les autres facteurs pouvant influer sur la reponse en anticorps au VPO. Quatre mois plus tard, a la suite du troisieme (et dernier) cycle de vaccination de masse, on a pr6lev6 des 6chan- tillons de sang chez 536 enfants de la cohorte d'origine, qui ont ensuite fait l'objet d'un question- naire. On a mesur6 les titres d'anticorps neutrali- sants dirig6s contre les poliovirus de type 1, 2 et 3 au moyen d'une technique de dosage normali- s6e. Dans toutes les r6gions urbaines, semi- urbaines et rurales 6chantillonn6es, les doses de VPO administr6es exclusivement au cours de la campagne de masse ont ete r6gulierement asso- ciees a des seropr6valences sp6cifiques de type plus elevees que le meme nombre de doses administr6es lors des vaccinations de routine. Ces resultats n'ont pas pu etre attribu6s a des diff6- rences liees a d'autres facteurs, tels que l'interval- le entre la vaccination et le pr6levement de l'6chantillon, la dur6e de 1'exposition 6ventuelle a l'infection par le poliovirus sauvage, ou la variation saisonniere de I'administration du VPO. La meilleure immunogenicit6 du VPO lors de la cam- pagne de masse est peut-etre due a une plus forte propagation secondaire du virus vaccin, mais cela n'a pas pu etre confirme du fait des limites qu'imposait I'etude elle-meme. Des modifications apparues au niveau de l'entretien de la chaine du froid au cours de la campagne de masse et l'utili- sation plus rapide du vaccin ont peut-etre egale- ment contribu6 a ameliorer l'immunogenicite du VPO. Les campagnes de masse semblent donc permettre d'augmenter avec beaucoup d'efficacit6 l'immunite specifique de type (liee a la dose) de la population, au-dela de ce qu'obtiennent les pro- grammes de vaccination classiques. Ces resultats viennent donc confirmer l'utilit6 de poursuivre les campagnes de masse en compl6ment des pro- grammes classiques de fa9on a potentialiser et a acc6lerer les efforts actuels visant a obtenir l'6ra- dication mondiale de la poliomy6lite. References 1. Resolution WHA41.28. Forty-first World Health Assembly, 1988. In: Handbook of resolutions and decisions of the World Health Assembly and the Executive Board, vol. III, third edit., 1985-1992. Geneva, World Health Organization, 1993: 100. 2. Wright PF et al. Strategies for the global eradica- tion of poliomyelitis by the year 2000. New England journal of medicine, 1991, 325: 1174-1179. 3. Pan American Health Organization. Update: eradi- cation of poliomyelitis in the Americas. Morbidity and mortality weekly report, 1992, 41: 681-683. 4. de Quadros CA et al. Polio eradication from the western hemisphere. Annual review of public health, 1992, 13: 239-252. 5. Patriarca PA et al. Optimal schedule for the admin- istration of oral poliovirus vaccine. In: Kurstak K, ed. Measles and poliomyelitis-vaccines and immuniza- tion (in press). 6. Sabin AB 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. 7. Akalay 0, Louis JM. The 1987 National Vaccina- tion Campaign in Morocco. Casablanca, Moroccan Ministry of Public Health, 1988. 8. Henderson RH, Sundaresan T. Cluster sampling to assess immunization coverage: a review of experi- ence with a simplified sampling method. Bulletin of the World Health Organization, 1982, 60: 253-260. 9. Patriarca PA, Wright PF, John TJ. Factors affect- ing the immunogenicity of oral poliovirus vaccine in developing countries. Reviews of infectious dis- eases, 1991, 13: 926-939. 10. Sufter RW et al. Outbreak of paralytic poliomyelitis in Oman: evidence for widespread transmission among fully vaccinated children. Lancet, 1991, 338: 715-720. 11. Schoub BD et al. A comprehensive investigation of immunity to poliomyelitis in a developing country. American journal of epidemiology, 1986, 123: 316-324. 12. Wood DJ, Heath AB. The second International Standard for anti-poliovirus sera types 1, 2 and 3. Biologicals, 1992, 20: 203-21 1. 13. Kurian, GT, ed. Encyclopedia of the Third World, vol. 2, 4th edit. New York, Facts on File, Inc. 1992: 1306. 14. Collaborative Study Group on Oral Polio Vaccine (Linkins R). Effect of season of vaccination on ser- oconversion to oral poliovirus vaccine (Abstract). In: Thirty-first Interscience Conference on Antimicrobial Agents and Chemotherapy. Washington, DC, Ameri- can Society for Microbiology, 1991. 15. SAS/STAT user's guide, version 6, 4th ed., vol. 1. Cary, NC, SAS Institute, Inc., 1989. 16. Fox JP et al. The spread of vaccine strains of polio- virus in the household and community in Southern Louisiana. In: Poliomyelitis: papers and discussion presented at the Fifth International Poliomyelitis Conference, Copenhagen, Denmark. Philadelphia, Lippincott, 1961: 368-383. 17. Sabin AB et al. Effects of rapid mass immunization of a population with live oral poliovirus vaccine under conditions of massive enteric infection with other viruses. In: Poliomyelitis: papers and discus- sion presented at the Fifth International Poliomyelitis Conference, Copenhagen, Denmark. Philadelphia, Lippincott, 1961: 342-355. 18. Bryce JW, Cutts FT, Saba S. Mass immunization campaigns and quality of services. Lancet, 1990, 335: 739-740. WHO Bulletin OMS. Vol 731995 777