Task 1. Flag all the suspicious values. (Outliers, repetitions, etc.) (spend max 15 minutes)

Repetition of reports by district and months

District1:d: Feb-March, nov-dec

District2: feb-march

District3: feb-march

District4: feb-march-Apr

District5: feb-march

District6: feb-march

District7: feb-march

District8: feb-march

District9: feb-march

District10: feb-march

District11: feb-march

District12: feb-march

District14: feb-march,Apr-May

District15: feb-march

Outliers: +1.5*IQR

There are about 12 outliers in all the 15 districts in the 12 months period in 2017. Most of the outliers are in district 3. The biggest outlier is in district 4 for the month of May. The average vaccinated children are 740 month for the 11 months excluding May. However, in May 2017 the number of children reported are 9008, which is 12 times the monthly average for the other months.

Task 2. Review the national and subnational coverage for MR1. Your data manager produces the following tables. What can you conclude from the administrative data?

 

The national coverage for all the years except 2015 is 90% or more. The national coverage is 92% for 2017.The coverage for Remo and Grandtown is above 100%( 102% and 117% respectively). This shows there is data quality issue. In addition the coverage as well as the absolute number of children vaccinated in 2017 has decreased from previous year at national level. On the other hand the number of children for Grandtown has increased from 92,000 to 97,000 a significant increase in vaccinated children. From the 1st table, the district disaggregated coverage for Grandtown region, shows disparity in the MR1 coverage some districts have <50% coverage other have > 200% coverage.

 

Task 3. Review coverage evaluation survey data. You remember that in 2013,there was a coverage evaluation survey. You pull up the data for that. Does this change your view about coverage at national level? For any of the regions?

 

The DHS coverage survey report for 2013 shows 89.2%(86.5.1-91.9). The admin report for 2012 was 91% at national level. The admin report for 2012 is within 95% CI of the DHS coverage 2013 ( for 2012 target).However, for Remo and Grandtown the DHS survey report show relatively lower coverage despite the above 100% HMIS report for 2012 ( which was 116 and 122 respectively).

Remo 84.6%(77.8%-91.4%)
Grandtown 89.0%(84.1-94.9%)

 

Though, the overall HMIS coverage for 2012 was close to the HMIS coverage, the DHS coverage for some Regions is not consistent with the HMIS coverage (over reporting) and the coverage is not also homogenous ranging from 73% to 100% at regional level. There could be more disparity in coverage at district level in Grandtown where the HMIS shows huge admin coverage disparities among districts in the Region.

 

Task 4. Review the chart with the age distribution of measles cases. Does that tell you anything additional about coverage?

 

The measles cases chart shows that children aged 1-4 years to be the most affected age group. More than 25% of the cases are children aged 1-4 years. In addition, slightly above 10% of the cases were aged 5-9 years old. This shows there was suboptimal routine immunization performance in the past few years. The last measles SIAs was conducted following the 2011 measles out outbreak and was reported to be successful. Hence over the last 6-7 years there was accumulation of susceptible children which led to the 2018 outbreak which mainly affected children that were mostly likely not targeted during the previous SIAs.

 

 

Task 5. Brief the Minister (spend max 1/2 hour on this section). Summarize the situation in three bullet points.

Though the overall HMIS MR1 coverage is consistent with DHS report, huge disparity in HMIS and DHS coverage in some regions such as Grandtown and Remo. Hence data quality is an issue in vaccinland immunization program.
The different data sources show immunity gap at national and at subnational level. The immunization coverage which is around 92% from HMIS for 2017 or the DHS coverage 89.2% for 2013 is not high enough to stop measles outbreak which require more than 93% population coverage. In addition the immunity gap in some districts can be very high especially in Grandtown where the HMIS report shows districts with coverage less than 50%. More over the fact that 80% of the cases were unvaccinated shows that there big immunity gap indicating the problem to be failure to vaccinate and not vaccine failure.

The reason for the immunity gap is not known, Based on a 2017 KAP survey, we are aware that vaccine hesitancy appears to be growing in Vacciland. However, as the last coverage survey was done in 2013, currently the immunity gap is not well known.

 

Task 6. Brief the Minister. Propose three actions to respond to the outbreak.

House to house 9-59 or 9-179 months children registration which includes community DQS verification of MR1 vaccinated in 2017 from community registration 0-23 months age and what was reported by service providing Health facility in 2017 for the same age group.
Catch up measles SIAs in 9-59 months in all regions and up to 15 years as dictated by regional measles epidemiology.
Post MR1 coverage survey

Task 7. Formulate recommendations. List your top 3-5 recommendations specific to data strengthening you would prioritize as the EPI and surveillance teams in Vacciland

National coverage survey + DQS in catchment HFs for the same target age group
Conduct DQR at different levels of the health system and identify data quality issues related to HMIS tools, guidelines, data management capacities at different levels and data quality monitoring systems.
Develop data quality improvement plan based on the above assessments and which rewards data quality rather than over reporting of (high coverage).