Table 1 Challenges in creating, using, and sustaining UCRs with recommendations based on our lessons learned

Design of UCR

The degree of data complexity available to end-users

• A high-level summary offers an entry point for novice/casual user.

• Linking to and/or archiving source material satisfy users who need access to greater detail without adding complexity to data structure.

• Storing and providing access to source documents within UCR (or linking to externally located source documents when direct access is not legally allowed) also creates an audit trail and increases the transparency of the extracted data.

Data extraction

• Having a streamlined data extraction form provided users with an easier interface, possibly increasing the likelihood that people would use the UCR.

Sustainability

Short-lived interest in cost data

• IRB reviews could mandate that researchers coordinate with a UCR so that a condition for publishing articles with cost data would be uploading results to a global repository.

Staffing requirements for UCR development and maintenance

• Donors may need to prioritize longer-term activities that promote sustained local capacity.

• Absent dedicated staff, some countries are trying to task-shift current government workers (e.g., national AIDS council staff) to cover this human resource gap.

Data types and quality

Matching the needs of decision-makers with appropriate data

• The development and use of a UCR should not be viewed as the only necessary step to strengthening a costing “ecosystem” (i.e., inclusive of an organization or country’s data/tools, skills/knowledge, and processes/governance).

• The selection of a unit cost requires careful consideration so that the best cost estimates can be applied to match the needs of those decision-makers.

Tradeoffs between data quality, purpose, and amounts

• In some cases, it may be better to use cost data from a comparable setting (e.g., peer country), that it is higher quality, more recent, or better matched to the intervention being costed in the current costing exercise.

• Expertise is required to appropriately identify and handle quality issues when using data from UCR in costing exercises.

Documentation of data (i.e., sources, methodologies for determining unit costs)

• Poor documentation of methods in source material can limit the ability to identify quality issues, and ultimately diminish the usability of findings.

• UCRs could include a feature for users to flag potential quality concerns.

Data can become obsolete

• Expertise is required to know (1) whether a unit cost estimate is obsolete due to innovation in service delivery, and (2) to appropriately fill gaps in data when country-specific estimates are outdated.

Data inclusion in UCRs

Time lag in the availability of unit cost data

• Access to the most recent cost estimates should be easier for governments, donors, and researchers when global UCRs are updated regularly.

• Research review boards and funders could create policies that set expectations that costing results are shared to UCRs expeditiously (e.g., within 30 days of acceptance of publication).

• A global technical working group for UCRs could establish best practices for dissemination, coordinate agreements with journals so that a condition for publishing articles with cost data would be uploading complete results to one or more global repositories.

Poor incentives for researchers to include their data in UCRs

• UCR administrators may need to implement a method of embargoing data it has obtained from authors/journals until publication date.

• Potential economies of scope include an interactive data input form which could be developed for authors of costing studies to submit information to a global repository—with some additional quality control (i.e., a “Wiki-UCR”).