Because specimens were previously collected, you will need to evaluate the conditions under which they have been maintained. Specimens should be clearly labeled with a unique identifier, organized in a meaningful way to facilitate identification, and listed in a database that contains freezer location information as well as links to other demographic data.
The conditions required for specimen storage will vary depending on the specimen type. For specimens that have been frozen, monitoring freezer temperatures, the number of times the specimen aliquots have been thawed and refrozen, and how long they have been stored are important in assessing specimen quality and whether serological testing is likely to yield accurate results.
The completeness of the biorepository should be assessed. Some specimens may be missing, misplaced or have insufficient sample volume remaining. If a high percentage of specimens are missing or missing specimens share characteristics (i.e., missingness is not random), this could lead to biased seroprevalence estimates. Conducting a missingness analysis can help assess whether there may be biases in the sampled specimens.

Once an appropriate biorepository has been identified and specimens can be accessed, the following steps can be taken:

Once specimen testing is complete, analysis of data would proceed as for any serological survey. Analysis considerations will vary based on study design, such as whether the specimens were collected using probability-based sampling. If probability-based sampling was done, consider analytic techniques such as weighting, propensity scores, or multiple imputation methodologies. Some aspects to consider to enhance generalizability to the target population:

• Probability of selection—Consider both the chance that a specimen was included from the original study population and the chance that a specimen from the biorepository was chosen for the serosurvey.
• Non-response—If missing specimens are a non-random sample of the biorepository, consider weighting to align the sample with the overall biorepository.
• Post-stratification weighting—Can be done to ensure the tested specimens match the characteristics of the target population.

For non-probability sampling, consider what adjustments could be made to the sample population to be more representative of the target population (e.g., adjust age distribution of the sample to reflect that of the target population).
Other analytical techniques could be appropriate for your study design. Consultation with a biostatistician or further reading can help. Refer to the WHO Vaccination Coverage Cluster Survey manual for more detailed information about sample sizes (Section 2.7) and analyses (Section 6).

There are two primary survey staff that can be involved in collection and transportation of residual specimens:

Figure 4. Roles and Responsibilities for Sampling from Health Facilities

Ideally survey staff are hired to coordinate the project and work closely with the health facility staff. Depending on the facility’s procedures and capacity, the specimen collection and data abstraction may be done by the facility staff (following training by the survey team) or by survey staff themselves.

Specimen collection procedures should be tailored to the facility so collection can be integrated within standard procedures. Below is a general overview of the steps in residual specimen collection.

Figure 5. Steps in conducting a serosurvey using residual blood specimens. Steps in orange are conducted by the health facility laboratory and steps in blue are completed as a part of the serosurvey.

Protocols will differ depending on specimen type, sampling method used and the facility-specific procedures. Refer to Toolkit: Facility Serosurvey Specimen SOP for a detailed protocol and supply list for collecting sera from a health facility using residual blood specimens.

Only residual specimens meeting selection criteria established by the serosurvey should be collected and tested. Example selection criteria include specified age range, specimen type (whole blood, serum, or plasma), minimum volume (as determined by the volume required for the assay), or absence of hemolysis. See below for additional considerations for collecting residual specimens.

Figure 6. Identifying eligible residual specimens from health facility

After specimens are collected from the health facility, the specimens should be transported to the laboratory for further processing, storage and testing. Refer to Facility Serosurvey Specimen SOP, and Specimen Processing and Assay protocols modules for detailed information on specimen processing and testing.

Data requirements may vary by survey objectives but the key variables to capture related to specimen characteristics and processing include:

Figure 7. Key variables for residual specimen collection.

The data may be captured in a paper-based register while at the health facility or laboratory and then entered into an electronic database.

Depending on the survey objectives, additional data may be valuable, including reason for specimen collection and information on village, town or district of residence to evaluate the geographic distribution of specimens and to determine whether the specimens are representative of the target population.  However, these data points may not be available at the laboratory, requiring additional steps by to gather this information.

Considerations when developing facility-specific protocols include the movement of specimens from the point of collection to testing, availability of data to identify eligible specimens, and type of data records at hospital (e.g., paper vs electronic, single or multiple databases). Refer to Toolkit: Residual Specimen Case Studies and Toolkit: Facility Serosurvey Specimen SOP for detailed facility specific protocols.