Deciding on the Surveillance Design and Data Collection Strategy
Step 3
As defined in the introduction, SMSS is sample-based but requires continuous data collection to generate near real-time mortality and cause-of-death data.
Recommended Surveillance Strategy: Continuous monitoring of vital events in randomly selected geographic clusters.
The selection of the clusters must be rigorously done from a complete sampling frame to ensure representativeness at the national level (and if desired for sub-national areas). Within each geographic cluster, a resident community worker will be recruited, trained and equipped to monitor the community. Monitoring will include active frequent household visits and the reporting of vital events such pregnancies, pregnancy outcomes and deaths. The community workers will collaborate with the community leaders to ensure completeness of reporting of vital events, enrollment of new households, and linkage with the community health workers.
When deaths occur, Next-of-kin is identified and followed up with verbal and social autopsy interviews for cause of death determination. A digital and information technology solution can be implemented to facilitate data capture and transfer and allow near real-time access to the data for analysis and result generation.
To ensure the completeness of events recorded, the community workers may be engaged to conduct a retrospective event data recapture on an annual basis. The data recapture may be conducted by an external team if resources are available. During an annual recapture period (preferrably a short period of 1-2 months) data collectors will visit every household in a clusters to collect data on births and deaths that have occurred in the past twelve months. These events will then be matched with the event data reported by the community workers and analyzed to assess the level of completeness of the community reporting. These data are analyzed together to obtain more accurate estimates of annual mortality.
Every 2-3 years, it may also be necessary to conduct a complete census of all clusters to update the population by age and sex. During this update, retrospective data on births and deaths can also be collected to further assess the surveillance data.
The clusters under surveillance can be easily linked with the existing CRVS to help strengthen the completeness of the CRVS. One strategy to facilitate linkage with CRVS is to incorporate CRVS formatted questions into the SMSS questionnaires. Addtionally, linking each SMSS cluster to a designated administrative civil registration post and health facility will also help facilitate data linkage and flow.
Hidden benefit:
By establishing a continuous sample of geographic clusters, it is easy to mobilize these clusters for additional data collection, monitor outbreaks and add-on surveillance of other programs. This approach is implemented by most countries that are currently implementing a SMSS.
Challenges for SMSS: SMSS requires a well-motivated community worker, a clear identification of the boundaries of the geographic clusters, and continuous active support for all data collectors. Another major challenge resides in the management of household individual IDs and the matching of events over time. Use of names of heads of households, GIS data, and additional household characteristics can help with this matching.
The Viva docs site assumes that the recommended Continuous approach is selected for the SMSS. All steps described refer to this option. In summary, the surveillance and data collection strategy approach involves:
- Establishing a nationally and sub-nationally representative sample of clusters, with their maps and population by age and sex.
- Continuous community surveillance of pregnancy outcomes and deaths in each cluster by a resident worker.
- Annual retrospective data recapture by the resident worker or an external team, implying a dual recording approach.
- Follow-up of all deaths identified to implement verbal and social autopsy interviews for cause of death determination.
- Update of the cluster population every 2-3 years.
- Implementation of a digital and information technology solution to facilitate real-time data collection, analysis and dissemination.
- Development of strategies for linking/integrating with existing surveillance systems such the CRVS and RHIS
Other Surveillance strategies to be considered:
-
Civil Registration and Vital Statistics Systems in combination with the RHIS data This must be carefully reviewed and discussed. But, most likely it will be ruled out as CRVS and RHIS suffer from incompleteness and a lack of representativeness. Efforts to correct these issues will take more resources and time to reach the level of functionality for accurate mortality data at national and sub-national level. Still, it is important to discus how linkages between the SMSS system and existing CRVS or RHIS systems could strengthen both efforts.
-
Repeated annual household mortality surveys: This approach involves conducting a national mortality survey at regular intervals of no more than twelve months. The survey will identify births and deaths in the past twelve months, conduct verbal autopsies on deaths identified and collect information on the population by age and sex. The same sample may be visited annually in a panel survey strategy, or a different sample drawn every time. While this approach may provide reasonable data, it has several limitations. It is costly as a rigorous survey will require careful mapping of selected clusters followed by the survey. It is often more top down as it engages less with communities in a continuous collaboration. The timeliness of data will suffer, and real-time data will most-likely not be possible. Furthermore, it does not establish a stable sample of communities on which to rely for rapid data collections and does not allow continuous monitoring for disease outbreaks and pandemics.
To Learn More:
| Last updated |
|---|
| June 26, 2025 |