02820nas a2200241   4500000000100000008004100001260005300042653003600095653003100131653001800162653003900180100001700219700001400236700001300250700001600263700001300279245015700292856007300449300001000522490000600532520202600538022001402564       2026                            d  c10/2026bSpringer Science and Business Media LLC10amodel-based geostatistics (MBG)10aSoil-transmitted helminths10aSurvey design10aNeglected tropical diseases (NTDs)1 aFronterrè C1 aJohnson O1 aGiorgi E1 aMontresor A1 aDiggle P00aGeostatistical analysis of baseline prevalence and treatment history drastically reduces the cost of soil-transmitted helminth impact assessment surveys  uhttps://link.springer.com/content/pdf/10.1186/s44263-026-00304-8.pdf  a1 - 90 v43 a<p><strong>Background</strong></p>

<p>Impact assessment surveys (IAS) are central to monitoring and evaluating the effectiveness of mass drug administration (MDA) programs targeting soil-transmitted helminths (STH). Current survey designs often follow guidelines that lack flexibility and spatial representativeness. There is a need for more efficient and context-adaptable methodologies that can guide programmatic decision-making using available data.</p>

<p><strong>Methods</strong></p>

<p>We developed a three-step geostatistical framework for optimal IAS design that integrates baseline prevalence data, environmental covariates, and treatment history. A geostatistical model is used to estimate baseline prevalence. A Markov model, coupled with a local spatial thinning algorithm, is then applied to forecast post-MDA prevalence surfaces. These projections guide the simulation and evaluation of candidate survey designs across different sampling intensities and spatial configurations. Design performance was assessed based on classification accuracy and survey cost.</p>

<p><strong>Results</strong></p>

<p>The methodology was applied to datasets from Kenya, Sierra Leone, and Zimbabwe. Compared to conventional survey designs, our model-based approach achieved similar classification accuracy with 60–90% lower survey costs. These gains were consistent across settings with different endemicity levels and deworming histories. The results also highlight the importance of leveraging spatial correlation and historical program data to inform efficient design.</p>

<p><strong>Conclusions</strong></p>

<p>Our approach offers a flexible and cost-effective solution for designing IAS under real-world constraints. By incorporating spatial modeling and intervention history, it enables evidence-based decision-making and resource allocation. This framework is disease-agnostic and can be adapted to other neglected tropical diseases to support progress toward control and elimination goals.</p>
  a2731-913X