Intervene before leaving: clustered lot quality assurance sampling to monitor vaccination coverage at health district level before the end of a yellow fever and measles vaccination campaign in Sierra Leone in 2009

Consultant for the World Health Organization, Geneva, Switzerland. .
BMC Public Health (Impact Factor: 2.26). 06/2012; 12(1):415. DOI: 10.1186/1471-2458-12-415
Source: PubMed


In November 2009, Sierra Leone conducted a preventive yellow fever (YF) vaccination campaign targeting individuals aged nine months and older in six health districts. The campaign was integrated with a measles follow-up campaign throughout the country targeting children aged 9-59 months. For both campaigns, the operational objective was to reach 95% of the target population. During the campaign, we used clustered lot quality assurance sampling (C-LQAS) to identify areas of low coverage to recommend timely mop-up actions.
We divided the country in 20 non-overlapping lots. Twelve lots were targeted by both vaccinations, while eight only by measles. In each lot, five clusters of ten eligible individuals were selected for each vaccine. The upper threshold (UT) was set at 90% and the lower threshold (LT) at 75%. A lot was rejected for low vaccination coverage if more than 7 unvaccinated individuals (not presenting vaccination card) were found. After the campaign, we plotted the C-LQAS results against the post-campaign coverage estimations to assess if early interventions were successful enough to increase coverage in the lots that were at the level of rejection before the end of the campaign.
During the last two days of campaign, based on card-confirmed vaccination status, five lots out of 20 (25.0%) failed for having low measles vaccination coverage and three lots out of 12 (25.0%) for low YF coverage. In one district, estimated post-campaign vaccination coverage for both vaccines was still not significantly above the minimum acceptable level (LT = 75%) even after vaccination mop-up activities.
C-LQAS during the vaccination campaign was informative to identify areas requiring mop-up activities to reach the coverage target prior to leaving the region. The only district where mop-up activities seemed to be unsuccessful might have had logistical difficulties that should be further investigated and resolved.

143 Reads
  • Source
    • "All of the 234 individuals reported to have ever had influenza, as defined by the household, indicated a self-diagnosis. More than 96% of the 160 persons reported to have had what the household considered to be yellow fever (which had been the focus of a recent vaccination campaign [11]) reported self-diagnosis, as did more than 60% of the 445 persons reported to have had typhoid (which is a relatively common laboratory diagnosis in Bo, as per Widal tests). More than half of the 317 people reported to have had what the household designated as pneumonia were self-diagnosed. "
    [Show abstract] [Hide abstract]
    ABSTRACT: The objective of this study was to evaluate the prevalence of self-diagnosis of malaria and other febrile illnesses in Bo, Sierra Leone. All households in two neighboring sections of Bo were invited to participate in a cross-sectional survey. A total of 882 households (an 85% participation rate) that were home to 5410 individuals participated in the study. Of the 910 individuals reported to have had what the household considered to be malaria in the past month, only 41% were diagnosed by a healthcare professional or a laboratory test. Of the 1402 individuals reported to have had any type of febrile illness within the past six months, only 34% had sought a clinical or laboratory diagnosis. Self-diagnosis of influenza, yellow fever, typhoid, and pneumonia was also common. Self-diagnosis and presumptive treatment with antimalarial drugs and other antibiotic medications that are readily available without a prescription may compromise health outcomes for febrile adults and children.
    Pan African Medical Journal 05/2013; 15:34. DOI:10.11604/pamj.2013.15.34.2291
  • Source
    • "This may not be feasible in Uganda which at times relies on logistical support from WHO to carry out such evaluations. Cluster lot quality assurance sampling techniques can be used to monitor coverage as soon as the campaign ends or even before its end [35]. Since we did not have the proportion of population vaccinated per strata during the campaigns, calculating overall vaccination coverage without taking this into account may have resulted in less accurate results if there were great imbalance between strata. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Background Following an outbreak of yellow fever in northern Uganda in December 2010, Ministry of Health conducted a massive emergency vaccination campaign in January 2011. The reported vaccination coverage in Pader District was 75.9%. Administrative coverage though timely, is affected by incorrect population estimates and over or under reporting of vaccination doses administered. This paper presents the validated yellow fever vaccination coverage following massive emergency immunization campaigns in Pader district. Methods A cross sectional cluster survey was carried out in May 2011 among communities in Pader district and 680 respondents were indentified using the modified World Health Organization (WHO) 40 × 17 cluster survey sampling methodology. Respondents were aged nine months and above. Interviewer administered questionnaires were used to collect data on demographic characteristics, vaccination status and reasons for none vaccination. Vaccination status was assessed using self reports and vaccination card evidence. Our main outcomes were measures of yellow fever vaccination coverage in each age-specific stratum, overall, and disaggregated by age and sex, adjusting for the clustered design and the size of the population in each stratum. Results Of the 680 survey respondents, 654 (96.1%, 95% CI 94.9 – 97.8) reported being vaccinated during the last campaign but only 353 (51.6%, 95% CI 47.2 – 56.1) had valid yellow fever vaccination cards. Of the 280 children below 5 years, 269 (96.1%, 95% CI 93.7 – 98.7) were vaccinated and nearly all males 299 (96.9%, 95% CI 94.3 – 99.5) were vaccinated. The main reasons for none vaccination were; having travelled out of Pader district during the campaign period (40.0%), lack of transport to immunization posts (28.0%) and, sickness at the time of vaccination (16.0%). Conclusions Our results show that actual yellow fever vaccination coverage was high and satisfactory in Pader district since it was above the desired minimum threshold coverage of 80% according to World Health Organization. Massive emergency vaccination done following an outbreak of Yellow fever achieved high population coverage in Pader district. Active surveillance is necessary for early detection of yellow fever cases.
    BMC Public Health 03/2013; 13(1):202. DOI:10.1186/1471-2458-13-202 · 2.26 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Lot quality assurance sampling (LQAS) surveys are commonly used for monitoring and evaluation in resource-limited settings. Recently several methods have been proposed to combine LQAS with cluster sampling for more timely and cost-effective data collection. For some of these methods, the standard binomial model can be used for constructing decision rules as the clustering can be ignored. For other designs, considered here, clustering is accommodated in the design phase. In this paper, we compare these latter cluster LQAS methodologies and provide recommendations for choosing a cluster LQAS design. We compare technical differences in the three methods and determine situations in which the choice of method results in a substantively different design. We consider two different aspects of the methods: the distributional assumptions and the clustering parameterization. Further, we provide software tools for implementing each method and clarify misconceptions about these designs in the literature. We illustrate the differences in these methods using vaccination and nutrition cluster LQAS surveys as example designs. The cluster methods are not sensitive to the distributional assumptions but can result in substantially different designs (sample sizes) depending on the clustering parameterization. However, none of the clustering parameterizations used in the existing methods appears to be consistent with the observed data, and, consequently, choice between the cluster LQAS methods is not straightforward. Further research should attempt to characterize clustering patterns in specific applications and provide suggestions for best-practice cluster LQAS designs on a setting-specific basis.
    PLoS ONE 06/2015; 10(6):e0129564. DOI:10.1371/journal.pone.0129564 · 3.23 Impact Factor
Show more

Full-text (3 Sources)

143 Reads
Available from
May 22, 2014