3.1. Elements of a PMTCT surveillance assessment
The decision to transition to a PMTCT-based system of HSS among pregnant women should be supported by a robust and comprehensive evidence base. In order for PMTCT data to serve as the basis for HSS, PMTCT HIV testing services need to be consistently available at all ANC HSS sites, PMTCT HIV testing needs to be accurate, PMTCT data need to be of high quality, and PMTCT-based HIV prevalence estimates need to be unbiased. The objective of a surveillance assessment is to evaluate the ability of PMTCT HIV testing and programme data to meet the needs of HSS.
To achieve this objective, a comprehensive assessment would address five areas of PMTCT HIV testing and data quality:
Agreement between ANC HSS and PMTCT HIV test results;
The magnitude of selection bias inherent in PMTCT HIV testing data compared to ANC HSS data;
The proportion of ANC HSS sites that provide PMTCT HIV testing services;
The quality of routinely collected PMTCT programme data at ANC HSS sites, including the minimum dataset of variables for surveillance;
The state of QA practices for PMTCT HIV testing at ANC HSS sites.
outlines the process flow for conducting a PMTCT surveillance assessment. The three principal assessment activities are as follows:
Process flow for conducting a PMTCT surveillance assessment.
Routine ANC HSS with PMTCT HIV testing variables added to the ANC HSS data collection form (addresses assessment areas 1 and 2);
A data quality assessment (DQA) of PMTCT data and data recording practices at ANC HSS sites (addresses assessment areas 3 and 4). The DQA has two components:
“site assessment”: a questionnaire to rapidly assess if PMTCT HIV testing and data collection procedures at the ANC HSS site are of high quality, standardized, and appropriate to ensure complete and valid PMTCT data for surveillance;
“data abstraction” or “rapid data review”: systematic examinations of the completeness and validity of routinely collected PMTCT data at ANC HSS sites;
A QA assessment of PMTCT HIV testing at ANC HSS sites (addresses assessment area 5).
Assessing the utility of PMTCT data for HSS involves a cycle of assessment, actions to improve programme performance and further assessment. The results of each assessment serve to identify limitations in PMTCT HIV testing and data quality, and inform recommendations for improvement. Strategies to address these limitations can be developed and implemented before conducting another assessment. This cycle may continue until assessment findings show that PMTCT programme data are suitable for surveillance.
Assessing and improving PMTCT programme data for HIV surveillance requires a collaborative effort among all relevant agencies and stakeholders, including surveillance and monitoring and evaluation (M&E) staff, maternal and child health (MCH) and PMTCT programmes, the national HIV/AIDS control programme, the national HIV reference laboratory and key partners. To ensure that the results of the assessment are translated into actions to improve PMTCT programme performance and data, it is advisable that all of these stakeholders form a technical body charged with directing the assessment. In this way, all important stakeholders will be actively engaged in the design and implementation of the assessment, interpretation of assessment results, and generation and implementation of recommendations for programme improvement.
3.2. Deciding to conduct a PMTCT surveillance assessment
It is suggested that all countries currently conducting ANC HSS implement a PMTCT surveillance assessment. In countries where PMTCT HIV testing or programme data are known to be of substandard quality, an assessment will help to identify and quantify programme gaps and inform strengthening measures. In countries where PMTCT services in ANC HSS sites are limited but expanding, an assessment can inform the immediate improvement of existing sites in anticipation of future transition to PMTCT-based HSS.
3.3. Preparing to conduct a PMTCT surveillance assessment
3.3.1. Site selection
A PMTCT surveillance assessment seeks to understand the true condition of PMTCT HIV testing services and programme data at ANC HSS sites. It is suggested that all ANC HSS sites offering PMTCT HIV testing services be included in the assessment. Restricting selection to those sites with well-established or high-performing PMTCT programmes would bias the assessment.
Some countries may face significant challenges to including all ANC HSS sites offering PMTCT HIV testing services in the assessment (due to resource constraints or the number or accessibility of ANC HSS sites). These countries may consider an alternative site selection strategy in which activities B (DQA of PMTCT data) and C (QA assessment of PMTCT HIV testing) of the assessment are conducted at a subset of ANC HSS sites (). Such a subset could be selected to achieve a representative sample of the various settings in which ANC HSS is conducted (urban/rural, geographical regions, etc.). Conducting activities B and C at a sample of ANC HSS sites can provide an estimate of the overall quality of PMTCT data and QA of HIV testing at ANC HSS sites. However, it is suggested that the final assessment before making the decision to transition to PMTCT-based HSS include all three activities at all ANC HSS sites to ensure the readiness of all sites for transition.
Comprehensive and alternative site selection approaches for the three principal activities of the PMTCT surveillance assessment.
PMTCT-based HSS has the potential to facilitate expansion of the number of HSS sites to additional sites providing PMTCT HIV testing services. However, it is suggested that initial efforts focus on assessing and transitioning existing ANC HSS sites. Expansion of PMTCT-based HSS to include additional sites may be considered at a future date. Additional sites would need to be rigorously assessed to ensure that PMTCT HIV testing and data are of sufficient quality to support surveillance.
3.3.2. Standards for interpreting assessment results
It is necessary that PMTCT HIV testing and programme data meet high standards to be considered ready to serve as the basis for HSS. These standards should reflect the consensus of the technical body charged with directing the PMTCT surveillance assessment, including government public health actors, partners and subject matter experts.
As general guidance, this document suggests the following general standards for the five assessment areas (described in section 3.1). Though not strict targets, these standards represent a high level of programme performance that may serve as a point of reference to aid countries in the interpretation of assessment results and guide local discussions on the readiness to transition to PMTCT-based HSS.
The agreement between ANC HSS and PMTCT HIV test results
It is important to achieve a high level of agreement between individual-level ANC HSS and PMTCT HIV test results. As a general standard, positive per cent agreement and negative per cent agreement between ANC HSS and PMTCT HIV test results which approximate the benchmarks described in
section 4.1.2
at all above-site levels at which HIV surveillance estimates are generated (typically district/regional/provincial and national levels) may be considered high. At the ANC HSS site level, sites with one or less discrepant results in cell b (negative as per ANC HSS and positive as per PMTCT) and one or less discrepant results in cell c (positive as per ANC HSS and negative as per PMTCT) may be considered to have achieved a high level of agreement (section 4.1.2). At sites with a larger sample size and a higher HIV prevalence (sites with a sample size of 400 and an HIV prevalence of ≥18%, and sites with a sample size of 500 and an HIV prevalence of ≥14%), two or less discrepant results in cell b (negative as per ANC HSS and positive as per PMTCT) and two or less discrepant results in cell c (positive as per ANC HSS and negative as per PMTCT) may be considered to have achieved a high degree agreement (section 4.1.2).
The magnitude of selection bias inherent in PMTCT HIV testing data compared to ANC HSS data
It is important that the bias inherent in PMTCT data be low at all ANC HSS sites. As a general standard, if the selection bias is less than 10% and more than ‒10% at all levels at which HIV surveillance estimates are generated (typically site or district, region/province and national), the bias inherent in PMTCT data may be considered low. For the purpose of these guidelines, selection bias is defined as the per cent relative change (positive or negative) from the total HIV prevalence (among pregnant women who do and do not receive PMTCT HIV testing) to the observed HIV prevalence (among pregnant women who receive PMTCT HIV testing). To ensure the continuing ability of PMTCT-based HSS to limit bias to acceptable levels, it is important that the uptake of PMTCT HIV testing be high at all ANC HSS sites. As a general standard, an uptake of PMTCT HIV testing of 90% or greater may be considered high (Appendix C). The proportion of ANC HSS sites that provide PMTCT HIV testing services
It is important that 100% of ANC HSS sites provide PMTCT HIV testing services.
The quality of routinely collected PMTCT programme data, including the minimum dataset of variables for surveillance (age, date of visit and HIV test result)
It is important that all surveillance variables of interest, including the minimum dataset of variables for surveillance (age, date of visit and PMTCT HIV test result), be of high quality in site records at all ANC HSS sites. As a general standard, a variable for which 90% of site records are complete and valid may be considered to be of high quality.
The state of QA practices for PMTCT HIV testing
It is important to maintain a robust system of QA supporting routine PMTCT HIV testing at all HSS sites. As a general standard, an assessment score of at least 80% in each of the three phase categories (pre-testing phase, testing phase and post-testing phase) on the QA checklist for PMTCT HIV testing” to be consistent. (Appendix E) May be considered to be within acceptable QA parameters.
Should the assessment findings show that standards are not met to the satisfaction of the technical body charged with directing the assessment, the country would not be considered ready to transition to PMTCT-based HSS. In this scenario, the results of the assessment can be used to generate recommendations for improvement of PMTCT programmes and data quality. Strategies to realize these improvements may be implemented before the next PMTCT surveillance assessment. This process of assessment, programme improvement and re-assessment would continue until all standards are met. shows PMTCT surveillance assessment activities, assessment areas, general standards for interpreting assessment results and a decision tree for transitioning from ANC HSS to PMTCT-based HSS among pregnant women.
PMTCT surveillance assessment activities, assessment areas, general standards for interpreting results and transition decision tree.
