A post-conflict vaccination campaign, Central African Republic

Bull World Health Organ 2018;96:540–547 | doi: http://dx.doi.org/10.2471/BLT.17.204321 Research 540 A post-conflict vaccination campaign, Central African Republic Nicolas Peyraud,a Michel Quéré,a Geraldine Duc,a Corinne Chèvre,a Theo Wanteu,a Souheil Reache,a Thierry Dumont,a Robin Nesbitt,a Ellen Dahl,a Etienne Gignoux,a Manuel Albela,a Anna Righetti,a Marie- Claude Bottineau,a Jean-Clément Cabrol,a Micaela Sarafini,a Samuel Nzalapan,b Pauline Lechevalier,c Clotilde Rambauda & Monica Rulla Introduction The conflict in the Central African Republic between the end of 2012 and 2015 had a negative effect on the health of the population. By 2015, the mortality rate in children younger than 5 years was 130 per 1000 live births, one of the highest in the world.1 The situation developed into both an acute and a chronic humanitarian emergency, which severely disrupted the health system and limited access to medical care and vac- cination programmes.1,2 By 2015, the overall security situation in the country had improved.3,4 Before the conflict, the logistic, financial and human resources available for the Expanded Programme on Immu- nization (EPI) in the country were already insufficient, with only 395 (48%) of the 815 peripheral health centres providing immunization for the population.5 During the armed conflict, parts of the vaccine cold chain and vehicles were plundered and about a quarter of health facilities were destroyed, which further aggravated a precarious situation.5 The consequent im- mediate decline in vaccination coverage resulted in the measles epidemics.1,5,6 Moreover, as the epidemiological surveillance system was also weakened, it is highly likely that diseases targeted by the EPI were under-reported. In 2015 and 2016, vaccination activities by the country’s health ministry were conducted according to the EPI and targeted only children younger than 1 year: (i) at birth, bac- ille Calmette–Guérin (BCG) vaccine and oral poliomyelitis vaccine (OPV) were given; and (ii) at 6, 10 and 14 weeks of age, infants received three doses of a combination of vaccines against diphtheria, tetanus and pertussis (DTP), Haemophilus influenzae type B (Hib) and hepatitis B (DTP–Hib–hepatitis B) as well as OPV and pneumococcal conjugate vaccine (PCV). In addition, one dose each of measles vaccine and yellow fever vaccine were given between 9 and 12 months of age. The country’s comprehensive multiyear vaccination plan for 2011 to 2015 included periodic supplementary immunization activities against poliomyelitis, measles, yellow fever, neonatal tetanus and meningitis.5 Despite a measles vaccination catch- up campaign in children aged 6 months to 15 years in 2005 and 2006 and follow-up campaigns in children aged 6 months to 5 years, which were planned for every 2 years, reactive vaccina- tion campaigns were also necessary in response to outbreaks, as occurred in May 2014 in Mambéré-Kadéï prefecture.5 At the national level, vaccination coverage in 2011 and 2012, before the conflict, was: (i) 47% for the third OPV dose; (ii) 69% for the first DTP–Hib–hepatitis B vaccine dose and 47% for the third dose; (iii) 49% for the first measles vaccine dose; (iv) 48% for yellow fever vaccine; and (v) 74% for BCG vaccine. In the first year of the conflict (2013), coverage dropped dramati- cally to: (i) 23%; (ii) 35% and 23%; (iii) 25%; (iv) 24%; and (v) 37%, respectively.7 In Mambéré-Kadéï prefecture, where Médecins Sans Frontières was working, the internally displaced population was estimated at 22 200 in May 2015.8 Although the secu- rity situation had become less tense after the majority of the Muslim population fled to Cameroon, it remained fragile.3,4 In 2014, the prevalence of severe acute malnutrition was over 2%.9 Before the conflict, in 2012, Mambéré-Kadéï ranked sixth lowest among all 16 prefectures for the third dose of DTP–Hib– hepatitis B vaccine. Vaccination coverage that year was: (i) 46% for the third OPV dose; (ii) 59% for the first DTP–Hib–hepa- titis B vaccine dose and 47% for the third; (iii) 59% for the first Objective To rapidly increase childhood immunization through a preventive, multi-antigen, vaccination campaign in Mambéré-Kadéï prefecture, Central African Republic, where a conflict from 2012 to 2015 reduced vaccination coverage. Methods The three-round campaign took place between December 2015 and June 2016 using: (i) oral poliomyelitis vaccine (OPV); (ii) combined diphtheria, tetanus and pertussis (DTP) vaccine, Haemophilus influenza type B (Hib) and hepatitis B (DTP–Hib–hepatitis B) vaccine; (iii) pneumococcal conjugate vaccine (PCV); (iv) measles vaccine; and (v) yellow fever vaccine. Administrative data were collected on vaccines administered by age group and vaccination coverage surveys were carried out before and after the campaign. Findings Overall, 294 054 vaccine doses were administered. Vaccination coverage for children aged 6 weeks to 59 months increased to over 85% for the first doses of OPV, DTP–Hib–hepatitis B vaccine and PCV and, in children aged 9 weeks to 59 months, to over 70% for the first measles vaccine dose. In children aged 6 weeks to 23 months, coverage of the second doses of OPV, DTP–Hib–hepatitis B vaccine and PCV was over 58% and coverage of the third doses of OPV and DTP–Hib–hepatitis B vaccine was over 20%. Moreover, 61% (5804/9589) of children aged 12 to 23 months had received two PCV doses and 90% (25933/28764) aged 24 to 59 months had received one dose. Conclusion A preventive, multi-antigen, vaccination campaign was effective in rapidly increasing immunization coverage in a post-conflict setting. To sustain high coverage, routine immunization must be reinforced. a Médecins Sans Frontières, 78 rue de Lausanne, Case Postale 1016, 1211 Geneva, Switzerland. b Health Prefecture, Mambéré-Kadéï, Central African Republic. c Médecins Sans Frontières, Paris, France. Correspondence to Nicolas Peyraud (email: peyraud7@hotmail.com). (Submitted: 12 October 2017 – Revised version received: 29 May 2018 – Accepted: 29 May 2018 – Published online: 20 June 2018 ) Bull World Health Organ 2018;96:540–547| doi: http://dx.doi.org/10.2471/BLT.17.204321 541 Research Post-conflict vaccination in Central African RepublicNicolas Peyraud et al. measles vaccine dose; (iv) 60% for yel- low fever vaccine; and (v) 59% for BCG vaccine (Dumolard LB, World Health Organization [WHO], personal com- munication, 2018). Coverage dropped during the conflict and, by 2015, was: (i) 37%; (ii) 58% and 38%; (iii) 47%; (iv) 42%; and (v) 46%, respectively (Dumolard LB, WHO, personal com- munication, 2018). Children who were younger than 1 year between 2013 and 2015 during the conflict and who were not vaccinated during this period did not have the opportunity to catch up on vaccinations afterwards because, in the Central African Republic, the EPI was limited to children younger than 1 year.5 There was, therefore, a cohort of children aged 2 to 3 years in the pre- fecture who had completely missed the opportunity to be vaccinated.5 In 2015 and 2016, Médecins Sans Frontières worked with the health ministry to quickly improve immuni- zation coverage in Mambéré-Kadéï: a preventive multi-antigen mass vaccina- tion campaign that targeted all eligible children younger than 5 years was orga- nized. Here we report on the feasibility and operational details of the campaign. Methods The preventive, multi-antigen, mass vaccination campaign targeted 323 towns and villages in five of the seven subprefectures of Mambéré-Kadéï: Amada Gaza, Berbérati, Dédé Mokouba, Gamboula and Sosso-Nakombo. The total area of the five subprefectures was 16 991 km2 and their total estimated population in 2015 was 274 332, giv- ing a density of 16 inhabitants/km2.10 Another Médecins Sans Frontières unit was operating in the two remaining subprefectures. In the five campaign subprefectures, the number of children aged 6 weeks to 5 years targeted for vac- cination was estimated at 45 647 from census data. Vaccination strategy We opted to conduct a preventive, mass vaccination campaign because very low vaccination coverage had to be increased rapidly and several birth cohorts of children younger than 5 years had completely missed the EPI. In ad- dition, this strategy was in line with recommendations of WHO’s framework for vaccination in acute humanitarian emergencies.11 An alternative approach focusing solely on improving the EPI would not have been sufficient because, in the Central African Republic, it would have covered only children younger than 1 year.5,6 After several meetings between Médecins Sans Frontières and EPI managers at national and district levels, an overall strategy was proposed by Mé- decins Sans Frontières and approved by the health ministry and EPI managers. Médecins Sans Frontières took respon- sibility for the campaign’s logistics and financial costs. Vaccines for children younger than 12 months were partially provided by the health ministry, PCV was donated by industry and the re- mainder were provided by Médecins Sans Frontières. Vaccination teams were composed of Médecins Sans Frontières and health ministry staff. We applied WHO’s decision-mak- ing framework for vaccination in acute humanitarian emergencies to select the final five vaccines for the campaign (Fig. 1):11 OPV, combination DTP– Hib–hepatitis B vaccine, 13-valent PCV, measles vaccine and yellow fever vaccine. This selection corresponded to all EPI antigens in the national im- munization schedule except for BCG. MenAfriVac and rotavirus vaccines were not included in the EPI calendar at that time and could not be provided. More- over, inactivated polio vaccine could not be included, because national and international stocks were limited and its introduction into the EPI had been delayed. The campaign followed EPI’s recommended vaccination schedule for infants, but also included catch-up vaccination with the same antigens for children aged 12 to 59 months, in accor- dance with WHO’s recommendations.11 The first round of the campaign took place between 18 December 2015 and 31 January 2016. Three vaccines were offered to children aged 6 weeks to 59 months: OPV, DTP–Hib–hepatitis B and PCV. In the second round, which took place between 23 February and 4 April 2016, measles vaccine was added for children aged 9 to 59 months. In the third round, between 24 April and 6 June 2016, yellow fever vaccine was added but restricted to children aged 9 to 11 months, because of national and worldwide shortages. In addition, vac- cination teams distributed vitamin A in the first round, conducted nutri- tional screening in the second round and provided soap in the second and third rounds. Caregivers also received paracetamol for use with side-effects, such as fever and pain at the injection site. An interval of 8 weeks was main- tained between each of the three rounds. The campaign was selective, it included all children without a record of immunization on their vaccination card, regardless of their caretaker’s re- call. As EPI schedules are complex, with children receiving 14 different vaccines during five or more visits, relying on a caretaker’s recall of a child’s vaccination history may have resulted in missed op- portunities.12,13 The locations of the vac- cination sites were selected on the basis of population density and geographical distribution. Fig. 1. Decision-making steps on vaccine use in acute humanitarian emergencies Step 1. Determine and grade risk of the vaccine-preventable disease Is there an increased risk of the vaccine-preventable disease? Step 2. Assess vaccines and amenability to mass campaigns Are relevant vaccines available in sufficient quantities? Do the vaccine’s characteristics lend themselves to mass campaigns? Step 3. Assess contextual constraints and facilitators, alternative interventions and competing needs Is there political stability, security and adequate staff for mass campaigns? Is funding available for mass vaccination? Is there a consensus between all key stakeholders? Implement vaccination intervention Note: The steps are based on the World Health Organization’s framework for vaccination in acute humanitarian emergencies.11 Bull World Health Organ 2018;96:540–547| doi: http://dx.doi.org/10.2471/BLT.17.204321542 Research Post-conflict vaccination in Central African Republic Nicolas Peyraud et al. Data collection and analysis The principal outcome was vaccination coverage of children aged 6 weeks to 59 months. Administrative data were collected during the campaign as the number of vaccines administered by age group, data on routine EPI activities running in parallel were not included. Data were collected on tally sheets and android applications and analysed us- ing Excel 2003 (Microsoft Corporation, Redmond, United States of America). In addition, two vaccination coverage surveys were carried out in the five sub- prefectures included in the vaccination campaign: the first took place before the campaign, in December 2015, and the second, after the campaign, in June and July 2016. Gender and wealth dis- parities were not assessed. The surveys used a two-stage, cluster random sample method, as outlined in WHO’s Vaccina- tion Coverage Cluster Surveys Reference Manual.14 In the first stage, clusters (i.e. villages) were selected with a probability proportional to size; in the second stage, households in each village were ran- domly selected according to the modi- fied WHO–EPI method.14 Data on each child’s vaccination history were based on the caretaker’s recall and the vaccination card, if available. Reasons for nonvac- cination were also noted. For the base- line survey, data were transferred from paper to Excel 2003 and analysed in Epi Info (Centers for Disease Control and Prevention, Atlanta, USA). Data from the postcampaign survey were collected electronically and analysed using Stata v. 12.0 (StataCorp LP., College Station, USA). Data from the caretaker’s recall and the vaccination cards took sampling weights into account and were reported with 95% confidence intervals (CIs). Both surveys were conducted with the permission and cooperation of the lo- cal public health ministry. In addition, postcampaign vaccination coverage was also estimated separately by adding precampaign survey results to coverage derived from administrative data. Results On average, 176 vaccination sites were involved in each vaccination round. The number of sites was increased slightly in each round to improve accessibility be- cause of low attendance at locations that were too far from children’s homes. Each round took 6 weeks to complete. There were no major delays in implementation and no stock ruptures or major adverse events were recorded. The campaign employed 316 staff and administered 294 054 vaccine doses (Table 1) in 98 729 consultations. Average attendance dur- ing the three rounds was 72% and 97% of children who presented at a vaccination site received at least one vaccine. Of the 98 729 children attending the campaign, 16 361 (17%) were younger than 1 year. A mean of nine vaccine doses were admin- istered per child during the three rounds. Administrative data collected dur- ing the campaign showed that immu- nization coverage was greater than 85% for the first doses of OPV, DTP–Hib– hepatitis B vaccine and PCV in children aged 6 weeks to 59 months and greater than 70% for the first dose of measles vaccine in children aged 9 weeks to 59 months. In children aged 6 weeks to 23 months, coverage of the second doses of OPV, DTP–Hib–hepatitis B vaccine and PCV increased to over 58% and coverage of the third doses of OPV and DTP–Hib–hepatitis B vaccine increased to over 20% (Table 2). In addition, 61% (5804/9589) of children aged 12 to 23 months received two doses of PCV and 90% (25933/28764) of those aged 24 to 59 months received one dose, which represented full vaccination of children in these age groups. Table 2 and Fig. 2 show that the postcampaign vaccination coverage survey yielded similar results to the ad- ministrative data. According to survey data, the campaign increased coverage for all antigens in targeted children ex- cept for yellow fever vaccine (Table 2). Moreover, the postcampaign survey found that 56% (159/284) of children aged 12 to 23 months had received complete PCV vaccination, as had 76% (272/357) of those aged 24 to 59 months. In that survey, the main reasons given for children’s nonattendance for all vac- cinations except yellow fever vaccination were: (i) “child or household absent dur- ing vaccination” (mean: 37 to 46%; 95% CI: 31 to 53) and (ii) “child’s caretaker did not have time” (mean: 22 to 24%; 95% CI: 15 to 34). Discussion In a chronic humanitarian emergency in the Central African Republic, our preventive, multi-antigen, vaccination campaign resulted in a substantial and rapid increase in vaccine coverage after three vaccination rounds, thereby potentially decreasing the burden of targeted diseases and the risk of epidemics.11,15,16 In addition, most of the beneficiaries were aged 12 to 59 months and thus benefited from catch-up vac- cination that would have otherwise been unavailable to them. It is possible that actual vaccination coverage might have been higher than we found from administrative data. Some doses recorded as first doses were probably, in reality, second or third dos- es because children presenting without a vaccination card may have been pre- viously immunized. Hence, coverage of the first dose may have been lower than that indicated by administrative data and coverage of second or third doses may have been higher. The absence of an improvement in yellow fever vaccine coverage occurred, because vaccination was indicated only for children aged 9 to 11 months. Table 1. Vaccinations by age group, preventive mass vaccination campaign, Mambéré- Kadéï prefecture, Central African Republic, 2015–2016 Vaccine No. vaccinations administered Age group 6 weeks to 11 months 12 to 23 months 24 to 59 months 6 weeks to 59 months OPV 13 823 16 920 62 594 93 337 Combined DTP–Hib– hepatitis B vaccine 14 034 26 257 44 155 84 446 PCV 14 107 15 934 30 744 60 785 Measles vaccine 2 625 9 520 42 848 54 993 Yellow fever vaccine 408 57 28 493 Total 44 997 68 688 180 369 294 054 DTP: diphtheria, tetanus and pertussis; Hib: Haemophilus influenzae type B; OPV: oral poliomyelitis vaccine; PCV: pneumococcal conjugate vaccine. Bull World Health Organ 2018;96:540–547| doi: http://dx.doi.org/10.2471/BLT.17.204321 543 Research Post-conflict vaccination in Central African RepublicNicolas Peyraud et al. Ta bl e 2. Va cc in at io n co ve ra ge , p re ve nt iv e m as s v ac cin at io n ca m pa ig n, M am bé ré -K ad éï p re fe ct ur e, Ce nt ra l A fr ica n Re pu bl ic, 2 01 5– 20 16 Va cc in e an d do se Va cc in at io n co ve ra ge Ad m in ist ra tiv e da ta co lle ct ed d ur in g th e ca m pa ig n, n o, (% )a Su rv ey s b ef or e an d af te r t he ca m pa ig nb Es tim at ed co ve ra ge fo r c hi ld re n ag ed 1 2 to 2 3 m on th s a ft er th e ca m pa ig n, % c Ch ild re n ag ed 6 w ee ks to 11 m on th s n = 7 29 5 Ch ild re n ag ed 12 to 2 3 m on th s n = 9 58 9 Ch ild re n ag ed 24 to 5 9 m on th s n = 28 76 4 Be fo re th e ca m pa ig n in ch ild re n ag ed 1 2 to 2 3 m on th s n = 29 1 Af te r t he ca m pa ig n in ch ild re n ag ed 1 2 to 2 3 m on th s n = 28 3 Ch an ge in co ve ra ge fr om b ef or e to a ft er th e ca m pa ig n No . % (9 5% CI ) No . % (9 5% CI ) Pe rc en ta ge p oi nt s ( 95 % CI ) O PV D os e 1 8 06 5 (1 10 .6 ) 8 27 1 (8 6. 3) 30 9 71 (1 07 .7 ) 50 17 .2 (1 0. 6 to 2 4. 4) 21 9 77 .4 (6 9. 2 to 8 3. 9) 60 .2 (4 8. 7 to 7 1. 77 ) 10 3. 8 D os e 2 4 23 6 (5 8. 1) 5 70 2 (5 9. 5) 21 9 15 (7 6. 2) 44 15 .1 (9 .3 to 2 1. 6) 17 0 59 .9 (5 2. 5 to 6 6. 9) d 44 .7 (3 2. 7 to 5 6. 8) 74 .9 D os e 3 1 52 2 (2 0. 9) 2 94 7 (3 0. 7) 9 70 8 (3 3. 8) 41 14 .1 (8 .4 to 2 0. 4) 10 4 36 .7 (3 0. 2 to 4 3. 8) 22 .7 (1 0. 6 to 3 4. 7) 45 .1 Co m bi ne d D TP –H ib –h ep at it is B v ac ci ne D os e 1 8 18 5 (1 12 .2 ) 8 39 2 (8 7. 5) 34 6 50 (1 20 .5 ) 49 16 .8 (1 0. 3 to 2 3. 6) 21 7 76 .7 (6 8. 4 to 8 3. 3) 59 .8 (1 0. 6 to 3 4. 7) 10 4. 5 D os e 2 4 30 6 (5 9. 0) 5 74 5 (5 9. 9) 8 65 8 (3 0. 1) 42 14 .4 (8 .5 to 2 0. 6) 16 9 59 .5 (5 2. 0 to 6 6. 6) d 45 .1 (3 3. 3 to 5 7. 1) 74 .4 D os e 3 1 54 3 (2 1. 2) 2 18 3 (2 2. 8) 82 3 (2 .9 ) 37 12 .7 (7 .0 to 1 8. 6) 71 25 .1 (1 9. 2 to 3 2. 1) 12 .4 (0 .5 to 2 4. 2) 35 .6 PC V D os e 1 8 11 6 (1 11 .3 ) 8 30 4 (8 6. 6) 25 9 33 (9 0. 2) 48 16 .5 (1 0. 4 to 2 2. 8) 21 6 76 .3 (6 8. 3 to 8 2. 8) 59 .8 (4 8. 2 to 7 1. 4) 10 3. 2 D os e 2 4 27 9 (5 8. 7) 5 80 4 (6 0. 5) 4 02 7 (1 4. 0) e 43 14 .8 (8 .8 to 2 0. 9) 15 9 56 .0 (4 8. 5 to 6 3. 2) d 41 .2 (2 9. 1 to 5 3. 3) 75 .4 D os e 3 1 71 2 (2 3. 5) 1 82 6 (1 9. 0) e 78 4 (2 .7 )e 41 14 .1 (8 .4 to 2 0. 0) 66 23 .3 (1 7. 6 to 3 0. 2) e 9. 2 (− 2. 5 to 2 1. 0) e 33 .2 e M ea sl es v ac ci ne D os e 1 2 62 5 (1 25 .9 )f 6 82 9 (7 1. 2) 26 8 36 (9 3. 3) 75 25 .8 (1 6. 9 to 3 5. 2) 20 1 70 .8 (6 4. 3 to 7 6. 6) d 45 .0 (3 3. 8 to 5 6. 2) 97 .2 D os e 2 0 (0 .0 )e, f 2 69 1 (2 8. 0) 16 0 12 (5 5. 7) 50 17 .2 (8 .9 to 2 5. 9) 99 35 .0 (2 9. 1 to 4 1. 4) 17 .8 (6 .0 to 2 9. 6) 59 .5 Ye llo w fe ve r va cc in e 40 8 (1 9. 6) f 57 (0 .6 )e 28 (0 .1 )e 33 11 .3 (6 .7 to 1 6. 8) 23 8. 4 (4 .8 to 1 4. 2) e, g − 3. 0 (− 14 .3 to 8 .3 )e 12 .4 e CI : c on fid en ce in te rv al ; D TP : d ip ht he ria , t et an us a nd p er tu ss is; H ib : H ae m op hi lu s i nfl ue nz ae ty pe B ; O PV : o ra l p ol io m ye lit is va cc in e; P CV : p ne um oc oc ca l c on ju ga te v ac ci ne . a Fo r a dm in ist ra tiv e da ta , e st im at ed c ov er ag e co ul d be m or e th an 1 00 % in se ve ra l s itu at io ns . F or e xa m pl e, v ac ci na tio n do se s m ay n ot h av e be en c or re ct ly a ttr ib ut ed a s fi rs t, se co nd o r t hi rd d os es b ec au se p re vi ou sly v ac ci na te d ch ild re n m ay h av e re tu rn ed w ith ou t v ac ci na tio n ca rd s, w ith th e re su lt th at th ei r s ec on d or th ird d os es w er e re ga rd ed a s fi rs t d os es . b Va cc in at io n hi st or y in th e su rv ey s w as b as ed o n th e ca re ta ke r’s re ca ll an d th e va cc in at io n ca rd , if a va ila bl e. c Fr om su rv ey a nd a dm in ist ra tiv e da ta . V ac ci na tio n co ve ra ge a fte r t he c am pa ig n w as e st im at ed b y ad di ng p re ca m pa ig n su rv ey re su lts to c ov er ag e de riv ed fr om a dm in ist ra tiv e da ta . d Th e de no m in at or w as 2 84 . e Th es e da ta a re n ot re pr es en ta tiv e be ca us e va cc in at io n w as n ot in di ca te d fo r a ll ch ild re n in th e ag e gr ou p. f Th e de no m in at or w as 2 08 5. g Th e de no m in at or w as 2 75 . N ot e: In co ns ist en ci es a ris e in so m e va lu es d ue to ro un di ng . Bull World Health Organ 2018;96:540–547| doi: http://dx.doi.org/10.2471/BLT.17.204321544 Research Post-conflict vaccination in Central African Republic Nicolas Peyraud et al. The study had several limitations. The vaccination coverage surveys were based on vaccination cards and caretaker recall, which may not have been reliable. A 2013 systematic review found that coverage estimates based on vaccination cards tended to be too low (with a high positive and a low negative predictive value), whereas estimates based on recall tended to be too high, coverage based on the combination could be either overes- timated or underestimated.17 Possession of a vaccination card has been associated with higher coverage, possibly because it reflects greater commitment, or access, to vaccination. However, a substantial proportion of the population does not keep cards.12,13,17 Moreover, replacement cards provided for lost cards may omit recall data, thereby contributing to underestimates of vaccination cover- age. In addition, the interpretation of survey results depends on the timing of vaccination relative to the child’s age: children older than 12 months require fewer doses than those younger than 12 months (for instance, the second and third PCV doses are not indicated for children older than 23 months). We were unable to take this into account in our surveys. Caution is also required when comparing our coverage estimates with those of other surveys, because of differences in methods.18 Another limitation is that the denominator we chose for calculating coverage and at- tendance (i.e. the population in last census in 2003) may have been inaccu- rate, because of population movements during and after the armed conflict.2,5 However, we do not believe that possible population movements would have had a substantial effect on our estimates, given the location and timing of our vaccination campaign. Achieving high vaccination cover- age for vaccines requiring three doses (i.e. OPV, DTP–Hib–hepatitis B and PCV) in a population with very low baseline vaccination coverage was chal- lenging, given that there were only three vaccination rounds and attendance was around 70%. Although our campaign covered a large area, some inequity in accessing vaccinations might have re- mained and there could have been some low-coverage pockets where people were still vulnerable to vaccine-preventable diseases. Our strategy was to optimize performance by adapting the campaign design to take into account experience gained during each vaccination round. We added vaccination sites, improved communication and collaboration with the community, and provided paracetamol for use only to treat fever or pain in the days following vaccina- tion. To account for situations in which families had access to both the EPI and our preventive vaccination campaign, we adapted our health promotion mes- sages to communities. In addition, our campaign was selective: we took into account vaccinations administered by the EPI that were recorded on cards. We also encouraged families to continue visiting their closest EPI centre regularly for further vaccinations as the campaign was only temporary. Collaboration between Médecins Sans Frontières and the health ministry during the planning, implementation and evaluation of the campaign was excellent. One challenge was a lack of personnel with sufficient experience of the EPI. Another was a misunder- standing about the vaccines Médecins Sans Frontières needed from the health ministry, which led to the third round of PCV immunization being slightly adapted. Finally, there was the challenge of not depleting the EPI’s personnel by employing health workers seeking addi- tional income, in light of the poor remu- neration offered by the health ministry. As resources are never unlimited, devising a vaccination strategy for re- mote populations with poor access to the EPI and to medical care presented a dilemma: should overall vaccination coverage be optimized by concentrating efforts in the most populated areas and vaccinating the maximum number of people or should vaccination activities be distributed equitably across the geo- graphical area? Targeting hard-to-reach populations requires more resources per person, but is crucial because these populations have limited access to medi- cal care for disease and because pockets of low vaccine coverage are potential Fig. 2. Vaccination coverage of children aged 12 to 23 months before and after preventive mass vaccination campaign, Mambéré-Kadéï prefecture, Central African Republic, 2015–2016 Va cc in at io n co ve ra ge (% ) 100 90 80 70 60 50 40 30 20 10 0 First OPV dose Second OPV dose Third OPV dose First DTP–Hib– hepatitis B vaccine dose Second DTP–Hib– hepatitis B vaccine dose Third DTP–Hib– hepatitis B vaccine dose First PCV dose Second PCV dose First measles vaccine dose Second measles vaccine dose DTP: diphtheria, tetanus and pertussis; Hib: Haemophilus influenzae type B; OPV: oral poliomyelitis vaccine; PCV: pneumococcal conjugate vaccine. Notes: Vaccination history in the surveys was based on the caretaker’s recall and the vaccination card, if available. Administrative data collected during the vaccination campaign did not include data on parallel activities under the Expanded Programme on Immunization. Bull World Health Organ 2018;96:540–547| doi: http://dx.doi.org/10.2471/BLT.17.204321 545 Research Post-conflict vaccination in Central African RepublicNicolas Peyraud et al. sources of epidemics. Médecins Sans Frontières chose an equitable approach for the campaign and covered remote areas as best as possible, which led to higher costs. Consequently, costs were probably too high for the health ministry to replicate the campaign. It is important, then, that different vac- cination strategies are complementary. One option is to increase the number of immunization rounds in some areas ac- cording to vaccination coverage data. In addition, linking some supplementary immunization activities to other vacci- nation programmes could be explored. Deciding on which population groups to target should take into account local epidemiology, vaccination coverage data and the resources available. In the Central African Republic, a child must attend an EPI site at least four times during the first year of life to complete the EPI schedule (apart from birth vaccines). However, limiting the EPI to children younger than 1 year is unrealistic and fails to achieve satisfac- tory vaccination coverage.5 Extending the EPI to children aged 2 years or more is essential as a catch-up strategy for those with incomplete immunization. The approach adopted for this cam- paign and for some others organized by Médecins Sans Frontières in 2014 and 2015 differed from previous campaigns despite similar needs for catch-up vac- cinations. Previous campaigns were not preventive; they were primarily reactive campaigns against measles or meningitis outbreaks. The recent development of preventive, multi-antigen campaigns has been made possible by better un- derstanding of the feasibility of such campaigns in humanitarian emergencies and other undervaccinated contexts, by Médecins Sans Frontières’ experience with mass reactive campaigns and by WHO’s recommendations. In situations where the EPI is weak- ened, such as in the recent humanitarian crisis in the Central African Republic, a targeted, subnational, preventive, multi- antigen, mass vaccination campaign, as recommended by WHO, can be effective in rapidly increasing immunization coverage and decreasing the accumu- lation of susceptible children. As a consequence, the risk of an epidemic in the wake of a humanitarian crisis and the burden of disease could be reduced. However, routine EPI activities must be strengthened in parallel and over the long term if increased vaccination cover- age is to be sustained. The sustainability of increased coverage could be assessed by regular surveys, which would help guide decisions on more affordable and innovative campaigns. ■ Acknowledgements The authors thank Iza Ciglenecki, Tanja Ducomble, Kate Elder, Tania Kapoor and Emilie Macher, all Médecins Sans Frontières and health ministry team members and WHO. Funding: The study was funded as part of Médecins Sans Frontières routine operations and the Ministry of Health of Central African Republic provided some vaccines for children younger than 1 year. Competing interests: None declared. صخلم ىطسولا ايقيرفأ ةيروهجم ،عاصرلا دعب ام تانيصتح ةلحم ةيئاقو ميعطت ةلحم للاخ نم ةعسرب لافطلأا عينتم ةدايز فدلها ايقيرفأ ةيروهمجب ،ييداك ييربمام ةعطاقم في تادضتسلما ةددعتم ماع لىإ 2012 ماع نم راد يذلا عاصرلا ىدأ ثيح ،ىطسولا .تمايعطتلا ةيطغت نم دلحا لىإ 2015 نوناك ينب ام تلاوج ثلاث نم نوكتت ةلحم قلاطإ مت ةقيرطلا )1( :مادختساب 2016 وينوي/ناريزحو 2015 برمسيد/لولأا طلتخلما حاقللا )2(و ؛)OPV( يومفلا لافطلأا للش حاقل ةيمدتسلما حاقلو ،)DTP( يكيدلا لاعسلاو سوناتيتلاو اييرتفدلاب -DTP-Hib( )ب( دبكلا باهتلاو ،)Hib( ب عونلا نم ةيلزنلا ؛)PCV( نترقلما ةيوئرلا تاروكلما حاقل )3(و ؛)ب دبكلا باهتلا تانايبلا عجم مت .ءارفصلا ىملحا حاقل )5(و ؛ةبصلحا حاقل )4(و ءارجإ مت ماك ،ةيرمعلا ةعومجملل ًاقفو يتلا تاحاقللا نع ةيرادلإا .ةلملحا دعبو لبق ميعطتلا ةيطغت حوسم .حاقللا نم ةعرج 294054 ميدقت مت ،لياجمإ لكشب جئاتنلا 6 ينب مهرماعأ حواترت نيذلا لافطلأل ميعطتلا ةيطغت تداز حاقل نم لىولأا تاعرجلل 85٪ نم رثكأ لىإ اًرهش 59و عيباسأ فيو ،PVCو ،ب دبكلا باهتلا-DTP-Hib حاقلو ، OPV تداز ،اًرهش 59 لىإ عيباسأ 9 ينب مهرماعأ حواترت نيذلا لافطلأا نيذلا لافطلأا في .ةبصلحا نم لىولأا ةعرجلل 70٪ نم رثكأ لىإ تاعرلجا ةيطغت تناك ،اًرهش 23و عيباسأ 6 ينب مهرماعأ حواترت ،ب دبكلا باهتلا-DTP-Hib حاقلو ،OPV حاقل نم ةيناثلا حاقل نم ةثلاثلا تاعرلجا ةيطغت تناكو ،58٪ نم رثكأ PCVو .20٪ نم رثكأ ،ب دبكلا باهتلا-DTP-Hib حاقلو ،OPV لافطلأا نم )9589/5804( 61٪ ىقلت ،كلذ لىع ةولاعو ،PCV نم ينتعرج ،اًرهش 23و 12 ينب مهرماعأ حواترت نيذلا حواترت نيذلا لافطلأا نم )28764/25933( 90٪ ىقلت ماك .ةدحاو ةعرج ،اًرهش 59 لىإ 24 ينب مهرماعأ في ةلاعف تادضتسلما ةددعتم ةيئاقولا ميعطتلا ةلحم تناك جاتنتسلاا لىع ظافحلل .عاصرلا دعب ام ةترف في ةعسرب عينمتلا ةيطغت ةدايز .ينيتورلا عينمتلا زيزعت بيج ،ةيلاع ةيطغت 摘要 中非共和国：冲突后的疫苗接种运动 目的 通过在中非共和国曼贝雷 - 卡代省实行一次预防 性的多抗原疫苗接种运动，以快速提高儿童疫苗接种 量。曼贝雷 - 卡代省曾于 2012 年至 2015 年间由于冲 突而减少了疫苗接种覆盖率。 方法 2015 年 12 月至 2016 年 6 月举行了三轮疫苗接种 运动，分别为 ：(i) 口服脊灰减毒活疫苗 (OPV) ；(ii) 百 白破疫苗 (DTP) 、乙型流感嗜血杆菌疫苗 （Hib） 和 乙型肝炎疫苗 (DTP–Hib–hepatitis B) 的混合疫苗 ； (iii) 肺炎球菌结合疫苗 (PCV) ；(iv) 麻疹疫苗 ；(v) 黄热 病疫苗。搜集根据年龄段管理的疫苗相关管理性数据， 并在接种运动前后对疫苗接种覆盖率进行了调查。 结果 共计接种 294,054 剂疫苗。首次接种口服脊灰 Bull World Health Organ 2018;96:540–547| doi: http://dx.doi.org/10.2471/BLT.17.204321546 Research Post-conflict vaccination in Central African Republic Nicolas Peyraud et al. 减毒活疫苗 (OPV)、百白破、乙型流感嗜血杆菌和 乙型肝炎 (DTP–Hib–hepatitis B) 混合疫苗和肺炎球 菌结合疫苗 (PCV) 的 6 周至 59 个月儿童的疫苗接种 覆盖率增长至 85% 以上。首次接种麻疹疫苗的 9 周 至 59 个月儿童的疫苗接种覆盖率增长至 70% 以上。 二次接种口服脊灰减毒活疫苗 (OPV)、百白破、乙型 流感嗜血杆菌和乙型肝炎 (DTP–Hib–hepatitis B) 混 合疫苗和肺炎球菌结合疫苗 (PCV) 的 6 周至 23 个 月儿童的疫苗接种覆盖率为 58% 以上。三次接种 的 覆 盖 率 为 20% 以 上。 另 外，12 至 23 个 月 儿 童 中 有 61% (5804/9589) 接 种 过 两 次 肺 炎 球 菌 结 合 疫 苗 (PCV) ，24 至 59 个月儿童中有 90% 接种过一次此 疫苗。 结论 预防性的多抗原疫苗接种运动能够快速有效地提 高冲突后的疫苗接种覆盖率。为了保持高覆盖率，必 须加强常规免疫接种。 Résumé Campagne de vaccination post-conflit en République centrafricaine Objectif Augmenter rapidement le taux de vaccination des enfants grâce à une campagne de vaccination combinée préventive dans la préfecture de Mambéré-Kadéï, en République centrafricaine, où le conflit qui a sévi de 2012 à 2015 a fait diminuer la couverture vaccinale. Méthodes Cette campagne en trois phases s’est déroulée entre décembre 2015 et juin 2016 avec les vaccins suivants: (i) vaccin antipoliomyélitique oral (VPO); (ii) vaccin combiné contre la diphtérie, le tétanos et la coqueluche (DTC), vaccin contre l’Haemophilus influenza type B (Hib) et l’hépatite B (DTC–Hib–hépatite B); (iii) vaccin pneumococcique conjugué (VPC); (iv) vaccin contre la rougeole; et (v) vaccin contre la fièvre jaune. Des données administratives ont été recueillies au sujet des vaccins administrés par tranche d’âge, et des enquêtes sur la couverture vaccinale ont été réalisées avant et après la campagne. Résultats Au total, 294 054 doses de vaccins ont été administrées. La couverture vaccinale des enfants âgés de 6 semaines à 59 mois a dépassé 85% pour les premières doses de VPO, de vaccin DTC–Hib– hépatite B et de VPC et, chez les enfants âgés de 9 semaines à 59 mois, elle a dépassé 70% pour la première dose du vaccin contre la rougeole. Chez les enfants âgés de 6 semaines à 23 mois, la couverture assurée par la deuxième dose de VPO, de vaccin DTC–Hib–hépatite B et de VPC était supérieure à 58% et celle de la troisième dose de VPO et de vaccin DTC–Hib–hépatite B était supérieure à 20%. Par ailleurs, 61% (5804/9589) des enfants âgés de 12 à 23 mois avaient reçu deux doses de VPC et 90% (25 933/28 764) de ceux âgés de 24 à 59 mois en avaient reçu une dose. Conclusion Une campagne de vaccination combinée préventive s’est avérée efficace pour augmenter rapidement la couverture vaccinale dans une région sortant d’un conflit. La vaccination systématique doit être renforcée pour maintenir cette couverture élevée. Резюме Кампания по вакцинации в постконфликтных ситуациях, Центрально-Африканская Республика Цель Быстрый рост уровня иммунизации детей посредством превентивной кампании с использованием многоантигенной вакцины в префектуре Мамбере-Кадеи, Центрально-Африканская Республика, где конфликт, длившийся в период с 2012 по 2015 год, сократил охват вакцинацией. Методы В период с декабря 2015 года по июнь 2016 года проводилась трехэтапная кампания, в рамках которой использовались следующие вакцины: (i) оральная полиомиелитная вакцина (ОПВ); (ii) комбинация вакцины против дифтерии, столбняка и коклюша (АКДС) и вакцины против гепатита В и гемофильной палочки типа В (Haemophilus influenza type B, Hib) (АКДС-Hib-гепатит B); (iii) пневмококковая конъюгированная вакцина (ПКВ); (iv) противокоревая вакцина; (v) вакцина против желтой лихорадки. Были собраны административные данные о введенных вакцинах в разных возрастных группах и обследованиях охвата вакцинацией, которые проводились до и после кампании. Результаты В целом было введено 294 054 дозы вакцин. Охват вакцинацией для детей в возрасте от 6 недель до 59 месяцев увеличился до более 85% для первых доз ОПВ, АКДС-Hib- гепатит B и ПКВ, а у детей в возрасте от 9 недель до 59 месяцев — до более 70% для первой дозы противокоревой вакцины. У детей в возрасте от 6 недель до 23 месяцев охват второй дозой ОПВ, АКДС-Hib-гепатит B и ПКВ составил более 58%, а охват третьей дозой ОПВ и АКДС-Hib-гепатит B составил более 20%. Кроме того, 61% детей (5804/9589) в возрасте от 12 до 23 месяцев получили две дозы ПКВ, и 90% детей (25933/28764) в возрасте от 24 до 59 месяцев получили одну дозу. Вывод Профилактическая кампания с использованием многоантигенной вакцины была эффективной для быстро растущего охвата иммунизацией в постконфликтных ситуациях. Для поддержания высокого уровня охвата необходимо укрепить плановую иммунизацию. Resumen Una campaña de vacunación posterior a los conflictos en República Centroafricana Objetivo Incrementar rápidamente la inmunización infantil a través de una campaña de vacunación preventiva multiantigénica en la prefectura de Mambéré-Kadéï, República Centroafricana, donde un conflicto entre 2012 y 2015 redujo la cobertura de vacunación. Métodos La campaña de tres rondas tuvo lugar entre diciembre de 2015 y junio de 2016 utilizando: (i) vacuna oral contra la poliomielitis (OPV); (ii) vacuna combinada contra la difteria, el tétanos y la tos ferina (DTP), vacuna contra Haemophilus influenza tipo B (Hib) y hepatitis B (DTP-Hib-hepatitis B); (iii) vacuna antineumocócica conjugada (PCV); (iv) vacuna contra el sarampión; y (v) vacuna contra la fiebre amarilla. Se recopilaron datos administrativos sobre las vacunas administradas por grupo de edad y se realizaron encuestas de cobertura de vacunación antes y después de la campaña. Bull World Health Organ 2018;96:540–547| doi: http://dx.doi.org/10.2471/BLT.17.204321 547 Research Post-conflict vaccination in Central African RepublicNicolas Peyraud et al. Resultados En general, se administraron 294.054 dosis de vacuna. La cobertura de vacunación para niños de 6 semanas a 59 meses aumentó a más del 85% para las primeras dosis de OPV, vacuna DTP-Hib-hepatitis B y PCV y, en niños de 9 semanas a 59 meses, a más del 70% para la primera dosis de vacuna contra el sarampión. En niños de 6 semanas a 23 meses, la cobertura de las segundas dosis de OPV, vacuna DTP-Hib-hepatitis B y PCV fue más del 58% y la cobertura de la tercera dosis de vacuna OPV y DTP-Hib-hepatitis B fue superior al 20%. Además, el 61% (5804/9589) de los niños de 12 a 23 meses habían recibido dos dosis de PCV y el 90% (25933/28764) de entre 24 y 59 meses habían recibido una dosis. Conclusión Una campaña de vacunación preventiva, con múltiples antígenos, fue efectiva en el rápido aumento de la cobertura de inmunización en un contexto posterior al conflicto. Para mantener una cobertura alta, se debe reforzar la inmunización rutinaria. References 1. Levels and trends in child mortality. Report 2015. New York: United Nations Children’s Fund; 2015. Available from: https://www.unicef.org/publications/ files/Child_Mortality_Report_2015_Web_8_Sept_15.pdf [cited 2018 Feb 25]. 2. L’OMS et ses partenaires de l’action sanitaire font face à cinq crises prioritaires sans précédent. Geneva: World Health Organization; 2014. French. Available from: http://www.who.int/features/2014/humanitarian- emergency-response/fr/ [cited 2018 Feb 25] 3. Central African Republic: Muslims forced to flee. Christian militias unleash waves of targeted violence. New York: Human Rights Watch; 2014. Available from: https://www.hrw.org/news/2014/02/12/central-african-republic- muslims-forced-flee [cited 2018 Feb 25]. 4. Security in Central African Republic improves, but situation remains fragile – UN envoy. Geneva: UN News; 2015. Available from: https://news.un.org/en/ story/2015/08/505752-security-central-african-republic-improves-situation- remains-fragile-un-envoy [cited 2018 Feb 25]. 5. Plan pluri annuel complet du programme élargi de vaccination (PEV) en République Centrafricaine 2015–2017. Bangui: Ministry of Health; 2015. French. Available from: http://www.nationalplanningcycles.org/ sites/default/files/planning_cycle_repository/central_african_republic/ ppac_2015-2017_08-06-16_15.34.pdf [cited 2018 Feb 25]. 6. Rapport biennal d’activités 2014–2015. Bangui: World Health Organization; 2015. French. Available from: https://reliefweb.int/report/central-african- republic/rapport-biennal-dactivit-s-2014-2015 [cited 2018 Feb 25]. 7. Central African Republic: WHO and UNICEF estimates of immunization coverage: 2016 revision [internet]. Geneva: World Health Organization; 2017. Available from: http://www.who.int/immunization/monitoring_ surveillance/data/caf.pdf [cited 2018 Feb 25]. 8. Central African Republic. IDPs face significant challenges as instability and political transition continue. Geneva: Internal Displacement Monitoring Center; 2015. Available from: http://www.refworld.org/pdfid/556c2a8a4.pdf [cited 2018 Feb 25]. 9. Suivi des Objectifs du Millénaire pour le Développement. Rapport final 2015. New York: United Nations Development Programme; 2015. French. Available from: http://www.cf.undp.org/content/dam/car/docs/omd/ RAPPORT%20OMD%202015%20%20VALIDATION%20TECHNIQUE%20 DU%2010%2009%2015.pdf?download [cited 2018 Feb 25]. 10. République centrafricaine [internet]. PopulationData.net; 2018. French. Available from: https://www.populationdata.net/pays/republique- centrafricaine/ [cited 2018 Feb 25]. 11. Vaccination in acute humanitarian emergencies: a framework for decision making. Geneva: World Health Organization; 2013. Available from: http:// www.who.int/hac/techguidance/tools/vaccines_in_humanitarian_ emergency_2013.pdf [cited 2018 Feb 25]. 12. Bolton P, Holt E, Ross A, Hughart N, Guyer B. Estimating vaccination coverage using parental recall, vaccination cards, and medical records. Public Health Rep. 1998 Nov-Dec;113(6):521–6. PMID: 9847923 13. Valadez JJ, Weld LH. Maternal recall error of child vaccination status in a developing nation. Am J Public Health. 1992 Jan;82(1):120–2. doi: http:// dx.doi.org/10.2105/AJPH.82.1.120 PMID: 1536315 14. World Health Organization vaccination coverage cluster surveys: reference manual. Geneva: World Health Organization; 2015. Available from: http:// www.who.int/immunization/monitoring_surveillance/Vaccination_ coverage_cluster_survey_with_annexes.pdf [cited 2018 Feb 25]. 15. Griffiths UK, Mounier-Jack S, Oliveira-Cruz V, Balabanova D, Hanvoravongchai P, Ongolo P. How can measles eradication strengthen health care systems? J Infect Dis. 2011 Jul;204 Suppl 1:S78–81. doi: http:// dx.doi.org/10.1093/infdis/jir090 PMID: 21666217 16. Measles vaccines: WHO position paper – April 2017. Wkly Epidemiol Rec. 2017 04 28;92(17):205–27. PMID: 28459148 17. Miles M, Ryman TK, Dietz V, Zell E, Luman ET. Validity of vaccination cards and parental recall to estimate vaccination coverage: a systematic review of the literature. Vaccine. 2013 Mar 15;31(12):1560–8. doi: http://dx.doi. org/10.1016/j.vaccine.2012.10.089 PMID: 23196207 18. Suarez L, Simpson DM, Smith DR. Errors and correlates in parental recall of child immunizations: effects on vaccination coverage estimates. Pediatrics. 1997 May;99(5):e3. doi: http://dx.doi.org/10.1542/peds.99.5.e3 PMID: 9113960