Last Updated: 15/04/2024
Estimating disease prevalence in a sentinel site catchment population to complement case-based suveillance and develop decision-making tools
Objectives
To leverage established sentinel surveillance in northern Papua New Guinea (PNG) to accurately delineate the sentinel site catchment area, conduct a cross-sectional survey to estimate prevalence, and use data generated to explore the development of surveillance tools to help local health authorities to make better decisions on service, resource allocation and response.
Papua New Guinea Institute of Medical Research (PNGIMR), Papua New Guinea
Papua New Guinea (PNG) accounts for 80% of malaria cases in the World Health Organisation (WHO) Western Pacific Region2. Transforming malaria surveillance into a core intervention is a pillar of the WHO’s Global Technical Strategy for malaria 2016-2017,1 and strengthening surveillance in PNG is an important ongoing component of moving towards malaria elimination. Through the STRIVE PNG project, PNGIMR have established febrile illness sentinel surveillance at 8 health facilities in PNG. Sentinel surveillance strategies provide longitudinal incidence data for a defined geographic area and can often provide richer data than passive surveillance due to the focusing of resources. These systems are, however, still subject to underreporting and bias due to health seeking behaviours in local communities that often generate incidence trends that underestimate the ‘true’ magnitude of disease in the community.
The study presents an opportunity for strengthening local expertise and leadership in conducting research with Provincial health authorities and potentially provides a ‘proof of concept’ to replicate this type of research not only at sentinel sites, but in settings where sentinel surveillance is not established but could still benefit from local disease prevalence and population estimates.
The study will be a cross-sectional household survey, with a mapping step to validate the sampling strategy and provide data for analysis. The sampling frame will be the catchment area population. Initially, using the latest 2011 National Census data providing village locations and population estimates allowed for an estimation of the catchment area ranging between 3910 – 5020 persons across 15 villages. The 2011 census reported the average household size for the district where Baro CHP sits was 6.2 persons. The median distance between villages and the sentinel site health facility was 11.9km (range: 1.8km – 25.4km).
Aim 1: To map the catchment area, a desk review of all publicly available literature (academic and grey) will inform on the delineation of health facility catchment areas in PNG. WSPHA, local health authorities and community leaders will be consulted to elucidate formal and informal definitions. Estimated catchment area maps will be drawn using appropriate geospatial software and populations will be estimated using methods to adjust available population data previously described. Before Household surveys, efforts will be made to map all villages identified in catchment area.
Aim 2: The sampling procedure for the household survey will follow similar practices to previous PNGIMR Malaria Indicator Surveys, whereby a random sample of 6 (3 high, 3 low incidence based on surveillance data ) villages will be selected from the catchment area within each selected village and a maximum of 20 households will be randomly sampled from each village. Within each selected household, a structured questionnaire will be used to interview an adult member acting as the household head, women aged 15-49 years and parents of recently sick children, while all household members will be eligible for providing a finger-prick blood sample. A trained research 13 nursing officer will collect blood samples by finger-prick from each member of selected households who is six months of age or older. From the finger-prick, the nurse will prepare a mRDT (type), dried blood spot (DBS) (x2) and a microcuvette sample will be used to measure haemoglobin (Hb) levels using a handheld HemoCue Hb 201+ analyser. DBS samples will be transported to the Molecular Hub in Port Moresby where the samples will be split, with one DBS used for DNA extraction and real-time PCR for QMAL, Malaria species, Dengue subtypes, Ross River Virus and Barmah Forest virus and K13 C580Y marker detection. The other DBS will be processed and sent to Melbourne for serological testing using a multi-pathogen serology assay. Village meetings will be held as a ‘tok save’ with the community to convey information of the planned activities and obtain collective community approval prior to activity commencement. A survey team of 4 PNGIMR officers, with support from the PHA, is estimated to require approximately 30 days.
Aim 3: Data generated from Aim 1 & 2, in addition to surveillance incidence data, will be used to calculate disease prevalence estimates, representativeness of surveillance and estimate under-reporting (incidence vs. prevalence). Further, we will explore how the data will lend itself to validation of surveillance thresholds as early warning signals for disease trend variance and develop spatial incidence and prevalence risk modelling.
Nov 2021 — Oct 2023
$15,000
