Antenatal Care as a Platform for Malaria Surveillance: Utilizing Community Prevalence Measures From the New Nets Project to Validate ANC Surveillance of Malaria in Mozambique
Antenatal Care as a Platform for Malaria Surveillance: Utilizing Community Prevalence Measures From the New Nets Project to Validate ANC Surveillance of Malaria in Mozambique
This study outlines a plan for conducting a routine assessment of malaria infection prevalence and intervention coverage using antenatal care (ANC) attendees. This will be a non-randomized assessment of the potential to use pregnant women attending their first ANC visit as a pragmatic sentinel population to monitor prevalence of malaria and the coverage of malaria control interventions. The use of a questionnaire, to include standard malaria rapid diagnostic testing, will be piloted with consenting women attending their first ANC visit at 21 individual health facilities across three of the New Net Project pilot study districts in western Mozambique: seven facilities each from Changara, Chemba, and Guro Districts. The results of the ANC questionnaires will be analyzed to see how well they correlate to similar malaria prevalence and intervention coverage estimates obtained during the contemporary community-based cross-sectional surveys administered during New Net Project pilot evaluation activities.
As part of the New Nets Project, Mozambique is deploying next-generation ITNs through mass campaigns in pre-determined provinces. The present study aims to leverage planned New Nets Project cross-sectional surveys and strengthened routine case surveillance data in three of the study districts (Changara, Guro, and Chemba) to assess (1) whether the malaria infection prevalence data collected during ANC surveillance correlates with the cross-sectional survey estimates of community infection prevalence in children 6 to 59 months and (2) if intervention coverage data (particularly ITN ownership and use) collected from ANC surveillance are valid and representative of the population as a whole. These additional data could catalyze a new model of surveillance for malaria, and greatly simplify evaluation of the impact of new interventions, as ANC surveillance could potentially replace or supplement cross-sectional household surveys and provide more granular and timely data.
All pregnant women attending first ANC visit at seven health facilities in each study district will be eligible for enrollment. Potential participants will be approached during their visit by a health facility worker. During group counselling sessions at initial intake, women will be informed of this pilot surveillance activity, and written informed consent will be obtained from each woman individually prior to routine ANC testing. All consenting women attending ANC first visit at a participating health facility will be tested for malaria using an RDT and asked to complete a study questionnaire which will include questions about the participant's net use, and care seeking behavior. It is expected to take 15 minutes to complete. Women who test positive for malaria will be given treatment according to national guidelines. There is no additional benefit to individual participants.
Data from public health surveillance systems are used for multiple purposes, including measuring the burden and monitoring the trends of a disease; guiding the planning, implementation, and evaluation of prevention and control programs; and prioritizing the allocation of health resources. In addition to routine health surveillance activities, public health programs often collect additional data on health indicators, as well as intervention coverage and program utilization, via nationally representative household surveys, such as the Demographic and Health Survey and the Malaria Indicator Survey (MIS).
These household surveys provide data to help measure the burden of and monitor trends in malaria, including data on the prevalence of symptomatic and asymptomatic infections in the population. Additionally, because the surveys provide data on monitoring and coverage impact indicators, they help guide the planning, implementation, and evaluation of prevention and control programs, and prioritization of the allocation of national malaria control program (NMCP) resources. Indeed, these representative surveys have become key tools for benchmarking progress towards malaria control and elimination.
Nevertheless, the surveys are expensive to conduct and are limited by long periodic intervals (e.g., every two to five years) and broad sampling strategies (i.e. generating national or regional estimates of disease burden and intervention coverage) that fail to capture seasonal trends or sub-regional heterogeneity, limiting their use for real-time or locally targeted planning and response.
Routine health information systems, on the other hand, may provide a continuous, granular data source representative of all age groups, primarily to help measure the burden and monitor the trends of symptomatic malaria and other diseases. Many countries have implemented standardized, electronic-based health management information system (HMIS), on platforms such as District Health Information System 2 (DHIS2), which provide monthly health facility data in real time. Despite being a potential source of valuable public health information, several factors limit the utility of HMIS data and should be considered carefully. One key limitation is the quality and representativity of the data being reported into HMIS, which is affected not only by general data issues, but also by factors such as case management practices and stockouts. These limitations require monitoring and health systems strengthening to overcome. HMIS data on case numbers may also be biased because some symptomatic patients may not present to a facility for care or may not seek care at all, and it can be difficult to interpret the incidence because of uncertainties surrounding regional and local population estimates. Further, not all health facilities report into HMIS; in many countries, the data derive predominantly from public health facilities. If the population using alternative sources of care for diagnosis and treatment of malaria (e.g., private facilities or traditional healers) varies over time, or differs from those using public facilities, the data trends observed in the HMIS may not be representative. Finally, as HMIS data are limited to patients presenting with symptomatic illness, it is not possible to assess the burden and epidemiology of asymptomatic infections, which are believed to represent the majority of malaria infections in many populations. Representativeness and validity of routine data may also be affected by diagnostic testing practices, which in turn may vary by health care provider performance, seasonality, and availability of diagnostic commodities. Finally, data on monitoring and coverage of malaria control interventions, such as insecticide treated bed nets (ITNs), are not typically collected during routine health visits or reported in HMIS.
To build on the strengths and address the limitations of these two complementary surveillance approaches, collecting data on parasite prevalence and coverage of malaria control interventions during antenatal care (ANC) clinic visits may provide an easily accessible, reliable, and representative source for monitoring population trends in malaria control with a higher degree of granularity than is possible in nationally representative household surveys. ANC coverage is high across sub-Saharan Africa; about 90% of women make at least one ANC visit in each pregnancy. Only six countries reported that fewer than 80% of women attended ANC at least once during their most recent pregnancy. Thus, pregnant women attending their first ANC visit are likely to be representative of the overall population of pregnant women. Pregnant women attending their first ANC visit are an ideal population for monitoring trends in malaria prevalence over time, as monthly ANC attendance remains relatively constant over time (minimal effects of seasonality) and is independent of malaria symptoms, thus giving a general infection prevalence estimate similar to that of a community survey. It has been demonstrated that malaria parasite prevalence among pregnant women correlates with the prevalence among children under five years of age. This highlights the possibility that pregnant women could be a good sentinel population for representative measures of trends in malaria prevalence and raises the question of whether pregnant women would be a good source of data on coverage of malaria control interventions that could be readily tracked over time. Once scaled up, these systems may also provide these data more frequently, at a finer spatial scale, and at a lower cost than household surveys.
In Mozambique, the NMCP is distributing new ITNs in mass campaigns. The New Nets Project is funding community surveys in six provinces, including rapid diagnostic test (RDT) testing in these provinces, to look at the impact of ITNs. This presents an important opportunity to validate whether the coverage of malaria control interventions and measures of parasite prevalence among pregnant women at first ANC visits is representative of the population at large in Mozambique.
In three districts, pregnant women presenting for their first ANC visit will be consented and tested for malaria using an RDT and asked to complete a study questionnaire. A comparison of results obtained at first ANC visits will be made to results from concurrent household surveys conducted by the New Nets Project. If validated and found to be a sustainable change to ANC visit procedures, data from the ANC sentinel population could be used to augment or even replace much of the data collected through the use of Malaria Indicator Surveys, which are expensive and infrequently conducted and generally only powered to the regional level. This data could then be triangulated with malaria burden data from the outpatient department to understand variation in community prevalence (as measured through ANC) and malaria clinical incidence, thereby guiding programmatic decision-making and response.
Inclusion Criteria:
Exclusion Criteria:
• Women with signs of severe malaria as determined by the ANC clinician