Methods for estimating population coverage of mass distribution programmes: a review of practices in relation to trachoma control

Methodological quality of the review: Low confidence

Author: Cromwell EA, Ngondi J, McFarland D, King JD, Emerson PM

Geographical coverage: Not reported

Sector: Trachoma

Sub-sector: Population coverage surveys, best practice

Equity focus: None specified

Review type: Literature review

Quantitative synthesis method: Narrative synthesis

Qualitative synthesis method: Not applicable

Background:

In the context of trachoma control, population coverage with mass drug administration (MDA) using antibiotics is measured using routine data. There are currently two approaches commonly used to estimate the population coverage of MDA for trachoma control.

In the first, post-MDA monitoring of antibiotic distribution performance currently relies on calculating reported coverage using antibiotic distribution registers and population data (doses distributed divided by the total population). The second approach, via district-level trachoma prevalence surveys (usually implemented every 3-5 years), uses distribution records and household survey responses to measure participation in each round of MDA. However, there is currently no other standard method to assess population coverage with MDA for trachoma control programmes.

Objectives:

The objective of this literature review was to compare coverage estimation practices employed by other public health programmes that use a similar community-level implementation strategy to deliver commodities and determine whether there is a ‘gold standard’ approach that can be applied to trachoma control.

Main findings:

The literature review identified six published sources for assessing population coverage developed by either the WHO or a collaborating entity: one for EPI, one for lymphatic filariasis and one for human helminthic infections (covering onchocerciasis, lymphatic filariasis and schistosomiasis). Two published guidelines for onchocerciasis control have been issued by the African Programme for Onchocerciasis Control and one manual describes the monitoring and evaluation of schistosomiasis control programmes. Overall, authors reported limitations of methods used by each included study. Authors reported recall bias as the main limitation of included surveys.

Authors identified several methods, including the 30×7 survey method for the Expanded Programme on Immunisation (EPI 30 ×7), other cluster random sampling (CRS) methods, lot quality assurance sampling (LQAS), purposive sampling and routine data. When compared with each other, the EPI and other CRS methods produced similar population coverage estimates, whilst LQAS, purposive sampling and use of administrative data did not generate estimates consistent with CRS.

Authors concluded that CRS methods present a consistent approach for MDA coverage surveys despite different methods of household selection. They merit use until standard guidelines are available. Authors noted that CRS methods should be used to verify population coverage derived from LQAS, purposive sampling methods and administrative reports.

Methodology:

Authors conducted a search on PUBMED for published studies and searched the websites of the WHO and implementing organisations responsible for mass administration of these programmes to identify whether any standard methods to estimate population coverage exist for the following public health programmes: EPI, lymphatic filariasis elimination, onchocerciasis control and elimination, and schistosomiasis control. Authors included studies written in English only which were published after 1980.

Applicability/external validity:

Authors noted that “we recommend the use of CRS methods because they provide an estimate of the proportion of participation in MDA generalisable to the implementation unit (such as a health district).”

Geographic focus:

Authors did not report the geographic focus of included studies. Nevertheless, authors noted that “however, whilst there is evidence that routine data introduces bias into population coverage estimates, there is a need for programmes to ensure that their choice of coverage survey methodology will provide results that are programmatically relevant and implement a method that is most appropriate for the local context.”

Summary of quality assessment:

Overall, there is low confidence in the conclusions about the effects of this study as important limitations were identified. Authors did not conduct a thorough search of the literature to ensure that all relevant studies were included in the review, indicating the presence of publication bias within the review. Methods used to screen and extract data were not reported. In addition, authors do not mention critically appraising included studies, so we are not clear if the surveys included in the review clearly follow guidelines; as such it’s not clear which studies were subject to high or low risk of bias. Authors did not acknowledge the limitations of the review.