Overview

Methodology for Adaptive Malaria Control, Uganda

Adaptive malaria control is a methodology for managing malaria with high quality, robust analytics that includes two essential components:

Adaptive vector control is a kind of applied ecology with important parallels in the management of natural resources. In developing adaptive malaria control, we were thus motivated by adaptive management in fisheries sciences, forestry, pest management, conservation biology, and similar activities. We have thus borrowed many concepts and practices from adaptive management, including a reliance on mathematical theory and models. Malaria is also a public health problem. Human health and health systems pose different challenges, including medical ethics, and some big data problems associated with health management information systems (HMIS), including health facility data.

The concept of adaptive malaria control is not entirely new to malaria, but it had not been formally described as a methodology or implemented by a national malaria program. This vignette (Overview) gives a very high-level overview. We introduce the core elements and guiding principles for adaptive malaria control. To learn more, including the technical details, we encourage the reader to read the vignettes that follow OVERVIEW in the sidebar. A more abstract discussion for academics is found in a related website RAMP & Adaptive Malaria Control

Fig 1: Over the past few years, we have defined an adaptive malaria control program in Uganda for the National Malaria Elimination Division (NMED) and the Department of Health Information (DHI). This schematic diagram presents an overview. The processes and supporting concepts are described in detail in vignettes, easily navigated using the sidebar.

Core Elements

Human malaria is a complex adaptive system. The complexity can be overwhelming, especially for managers who must make trade offs involving aspects of malaria that involve different domains of knowledge.

The challenge of using evidence to guide malaria policy presents a different kind of challenge that it is equally daunting. The data sources are large and problematic. Data curation creates enormous technical challenges. Malaria analytics uses the same methods as statistical analysis for malaria research, but the analyst must learn how to give advice that is based on the best evidence available, even if the evidence weak. In a policy setting, the advice can help guide development of policies that can help to resolve some of the uncertainty. To make all this work, we designed Adaptive Malaria Control as an integrated decision support system. It includes data science and conventional and simulation-based analytics.

Our presentation of the methodology is organized around six core elements:

  1. Data – data should be organized in systems that make clean, high quality data and digital objects available to malaria analysts, to NMED, and to the Ministry of Health

  2. Information – information (i.e., data that has been given meaning through some sort of statistical analysis) should be routinely generated, examined, and discussed by NMED and the Ministry of Health

  3. Malaria Intelligence – malaria intelligence (i.e., information that has been developed and organized for decision support) is developed in a predictive statistical framework to bridge scientific research and surveillance and to set the stage for studies that test predictions

  4. Analytics – conventional and simulation-based analytics are organized around malaria intelligence to develop timely decision support and policy advice that is robust to uncertainty

  5. Policy – malaria analytics generate timely and robust advice to support strategic planning, early warning systems, outbreak planning and response, national stratification, and advocacy

  6. Adaptive Surveillance – the advice is critically examined to identify and prioritize critical gaps that could be filled through various kinds of studies

These six elements are also used to structure our discussion, making it possible to navigate the discussion.

Guiding Principles

To support adaptive malaria control, we are guided by some core principles:

  1. Replicability – To ensure that the analyses are stable over time, routine analyses should have well-defined protocols and procedures. The algorithms, protocols, and procedures should thus be well defined, version-controlled, archived, and documented, so that they can be run iteratively, on a schedule dictated by malaria program activities.

  2. Accountability – Analysts should be accountable for their analysis and the policy advice they give. The analysis should thus be transparent, and it should be possible to replicate any analysis done at any point in the past. If policy advice changes, it should be possible to explain why: did the data change? did the software change? or did the analytical procedure change? For accountability, all data, software, and the protocols and procedures defining each analysis should be archived and version controlled.

  3. Robustness – The analysis should be rigorous and aligned with best practices for malaria. The analysis should be designed to characterize and quantify uncertainty. The uncertainty should be propagated through the entire analytics pipeline, including each inferential step in a structured workflow that starts with data and ends with malaria policy advice. An important part of robust analytics is simulation – mathematical models are used for synthesis, evaluation, scenario planning, forecasting, gap analysis, and various forms of decision support. A goal is to understand malaria in context as a changing baseline that has been modified by control

  4. Updating – the methods should be sbable over time, but there is also a need to critically evaluated and occasionally update protocols and procedures. Methods development should not disrupt operations. Each step in an analytics pipeline should thus have a well-defined production version (prod). Meanwhile, the production version is critically tested and revised and new features are explored in a development pipline (dev). Any new feature from dev is thoroughly tested before integrating it into prod (see Production vs. Development).

  5. Learning – A goal for adaptive malaria control is thus to have a way of storing and retrieving information about malaria in its local context. The information should be routinely examined to identify critical gaps in our knowledge. New studies should be designed to fill critical reduce uncertainty about a policy. Over time, as we learn about malaria in Uganda, uncertainty about malaria should tend to decline.