Distinguishing annual measles vaccine dose gaps from cohort dropout

To the Editor

Dzeha and Amponsah report that the annual dropout indicator between the first and second doses of measles-containing vaccine (MCV1 and MCV2) exceeded 10% in four years in Chereponi District and declined to 8.6% in 2024 [1]. Interpreting this improvement warrants a distinction between an administrative monitoring signal and the proportion of first-dose recipients who subsequently miss their second dose.

The indicator was calculated as [(MCV1 doses − MCV2 doses)/MCV1 doses] × 100 using annual totals [1]. Ghana's schedule places MCV1 at 9 months and MCV2 at 18 months [2]. Doses recorded within one calendar year therefore concern overlapping, but nonidentical, birth cohorts. Even with accurate recording, their ratio is not a direct estimate of completion among the same children.

The authors recognize that negative values can arise when dose counts do not represent a closed cohort, and they appropriately caution against causal interpretations of the surveillance and coverage trends [1]. The narrower issue is that a positive value does not resolve this mismatch. A value above 10% may reflect incomplete vaccination, but also changes in cohort size or first-dose uptake and the timing of catch-up vaccination. Falling below 10% likewise does not, by itself, establish improved retention within a birth cohort.

Aggregate dropout measures are used in regional programme analyses [3]; the routine calculation itself is therefore not the objection. Its interpretation as cohort dropout, however, requires assumptions about comparability across cohorts and vaccination periods. Peck et al. reported percentage-point differences between MCV1 and MCV2 coverage rather than dropout rates because the estimates represented different birth cohorts [4]. That coverage difference is not numerically interchangeable with the relative dose-count gap reported here, but the distinction illustrates the need to specify what each indicator measures.

For example, 1000 first doses and 800 s doses yield a 20% annual gap; they do not demonstrate that 200 of those 1000 children missed MCV2. These illustrative counts are not study data. Conversely, catch-up vaccination of older children may narrow the annual gap while some recent first-dose recipients remain overdue. The direction and magnitude of any resulting discrepancy in Chereponi cannot be determined from the published aggregates.

This does not weaken the rationale for improving second-dose access. It matters when comparing subdistrict retention or deciding where defaulter tracing should be intensified. The annual gap can flag a need for investigation, but cannot directly quantify the number of children overdue for MCV2. Interpreting differences between subdistricts as differences in cohort dropout could therefore misdirect follow-up activity.

A useful clarification would describe the measure explicitly as an annual relative dose-count gap, explain the operational meaning of the <10% benchmark, and apply the cohort caveat to positive as well as negative values. Where underlying registers permit linkage, completion could be assessed among documented MCV1 recipients by a prespecified age after MCV2 becomes due, with adequate follow-up and consideration of migration. Where linkage is unavailable, a targeted register audit could assess whether a large annual gap corresponds to more children overdue for their second dose. This would preserve the study's programme-monitoring contribution while providing a more direct basis for defaulter tracing and subdistrict microplanning.

CRediT authorship contribution statement

Ercan Yılmaz: Conceptualization, Writing – review & editing.

Ethical approval

Ethical approval was not required because this Letter to the Editor does not report new research involving human participants or animals.

Funding

No specific funding was received for this work.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.