Information Technology (IT) Checklist

These are key IT topics that should be considered for an SMSS system.

IT Planning Checklist

Personnel

In house or contractor

Having IT personnel as part of the in-house core study team can have many advantages. They can be involved in all aspects of the project. They can focus on the project and the goals of the implementing agency without distraction of other clients or pressures from their private company. If there are gaps in specific IT skills, contractors can be employed.

Roles

Programming and database developers: create, maintain and expand all custom software and features. The number of programmers needed may vary during the project. More effort will be needed during design and start up periods. Regular operation and maintenance will generally require less effort. (Estimate 2-3 FTE initially and 0.5 to 1 FTE as the project settles into regular operation).
Systems administrators: are the backbone of the system. They provide a solid foundation for managing project platforms, software, and devices. They are responsible for server and device configuration and maintenance. They work closely with programmers. There can be periods of intense demand when configuring hundreds of devices before startup. (Estimate 0.5 to 1 FTE during planning and startup with a smaller role during regular operation)
Data managers: design forms and reports, train data collectors, clean and reshape data, provide some statistical analysis. The demand for this role is high during all phases of the project. During planning and design, they design forms and reports. As launch approaches they test forms and train data collectors. During regular operation, they clean and shape data, write reports, and provide analysis. ( Estimate 3 to 6 FTE ; or 0.25 to 0.5 FTE per province)

Internet connectivity and Electrical Power

Offline data collection: Most of the IT information in this VIVA-DOCS guide assumes that a large portion of the data will need to be captured offline. If there is strong cell phone data connectivity in at least 80-90% of the clusters, an online system should be considered. Online systems have huge advantages in terms of simplicity, data validation, and longitudinal data collection.

Electrical power: It’s important to consider the consistency of electrical power. Generally, rural areas will have more electrical outages and will be the most vulnerable to interruptions. They may be seasonal or cyclical througout the year. In addtion to difficulties charging devices and running laptops, prolonged outages can cause interruptions in mobile phone connectivity as system batteries for supporting mobile phone towers shut down.

What can be done to overcome power outages?

Personal solar chargers

Personal battery packs

Charging from vehicle batteries

Servers

We have identified three key areas where servers will be needed. Data Collection, Analysis, and Sharing. These three areas can be managed by a single server or they can be spread across 2-3 servers. This can be decided based on software types, performance needs, costs, and privacy considerations. For publicly shared data it can make sense to use an isolated server that does not contain any protected data. This allows public access and reduces the risk of accidental exposure of protected data.

Data Collection Server (Key functionality)

Choose a mature and proven system with an accessible database. Key consideration: Is offline data collection needed?

Data repository for raw data

Operational Management: Provides tools to manage data collection operations (user accounts, assignments, activity monitoring, report generation, maps)

Serves blank forms to Community and Provincial data collectors, Receives and stores form data

Data Analysis Server (Key functionality)

Data collection shares some requirements with data analysis, for example: errors found in the data by supervisors or by analysts may both need to be reported back to data collection. There are several areas of overlap between operational data management and data analysis. This can be an argument in favor of combining or at least connecting data collection and data analysis platforms.

Platform for cleaning, merging, and analyzing data

Data repository for identifiable and deidentified analytic data sets

Tools for statistical calculations and visualizations

Remove barriers to access

Consider Frequency of updates to the data

Provide Codebooks and contextual information

Will a Data Use agreement be required

Track user downloads

Repositories

Server Software

Operating Systems: Either Linux or Windows systems can be used. Make the choice based on software requirements, cost, and the existing IT team’s skillset.

Cloud vs. On-Premises

In general, it is easier and cheaper for IT teams to setup and manage cloud servers, but the cost savings can dissipate over time. Here are some pros and cons to consider:

Cost: pay as you go vs. up-front investment

Security

Scalability and customizable hardware

Reliability

Resides in country. Some countries have laws requiring data be stored within their geographic boundaries.

Automated backups and recovery

Hardware

Central: Servers (see above), Laptops, printers for staff. Tablets and phones to test and monitor data collection

Provincial: Laptops, printers for staff. Tablets and phones for data collection. Estimate replacement of 1/3 of field devices every year.

Community: Tablets and phones for data collection. Estimate replacement of 1/3 of field devices every year.

Hardware for the field (phones and tablets)

It’s important to choose the best quality hardware available for the budget. Brands with a good reputation and repairability are generally preferred to devices promising the latest features at a discount.

It is highly recommended to purchase protective cases for each device. This will quickly pay off in terms of extended life.

Remember to budget for internet connectivity. This will likely be needed for each data collector for the life of the project.

Focus on these aspects:

Durability

Performance

Battery life

Availability

Timeframe for procurement

Quantity

A rule of thumb for estimating is to buy one for each data collector and expect around 30% replacement per year. With care and luck, the devices may last longer, but if conditions are rough or electricity quality is poor, the devices may need more frequent replacment.

Data Content

The core content of the data will be in three major area: Population denominator, Vital events, and Verbal Autopsy (Cause of Death).

Population: Households and members

After initial surveying and mapping of clusters. Data collectors will record every household and every person within the household. This will provide an important denominator for all future mortality calculations. Consider how this population denominator will be maintained. Will households and members be continuously added and dropped, or will the clusters be re-surveyed on a cyclical basis.

Vital Events

Community based data collectors will continuously record every pregnancy, pregnancy outcome, and death. It’s important to set standards for how often to visit each home, a starting point might be 1 visit per month. But the timeframe can vary based on project plan. The vital event form can record a visit to a household even if no vital event is found. This can help encourage data collectors to do the work, and can provide GPS evidence to confirm their activity. Try to find additional tools or resources to motivate data collectors to find and record deaths.

Verbal Autopsies

For each death recorded by the community data collector, a Verbal Autopsy will need to be collected. There are different versions of verbal autopsies available. The most commonly used is the 2016 WHO ODK Verbal Autopsy. At this time, the 2022 WHO ODK Verbal Autopsy is NOT recommended because the algorithms for processing the 2022 data structure are not yet complete. Mozambique (SISCOVE) has extended the 2016 WHO ODK Verbal Autopsy to include Social Autopsy questions. This form is called the Verbal and Social Autopsy or VASA. Because the MZ VASA is based on the 2016 WHO VA, it can be processed with the standard InterVA or Insilico algorithms.

Sampling frame and support data

The data team will need to create additional tables to support data collection and processing. This data will be merged with core content data (population, vital events, cause of death) for several purposes: field team management and error correction, contextualizing data within provinces or districts.

This data includes cluster names, and polygon data to draw boundaries. It should also include information about data collection staff and the devices they use. Device ID’s will be recorded in the data and can be used to confirm data collector identity.

Programming

A programming team can take a variety of forms, part-time or dedicated, 1 person or several. Whatever form it takes, its important that a programming team have contextual knowledge of project goals and history. During different project phases, programming demands can increase or decrease drammitcally. It can seem expedient to outsource programming tasks, but this can be a big risk for the project. Without intimate knowledge of systems, data structures, and details, programmers will be hamstrung and may make poor decisions.

Last updated
Apr 13, 2025