Estimating probabilities of malaria importation in southern Mozambique through P. falciparum genomics and mobility patterns
Curation statements for this article:-
Curated by eLife
eLife Assessment
This study introduces a useful method to estimate the probability that a malaria case is imported and to identify the geographic origin of parasites by using a Bayesian approach that integrates epidemiological, travel, and genetic data. The authors provide convincing evidence that the approach can reliably identify the main sources of malaria imports. This work will be of great interest to the area of genomic epidemiology and public health strategies aiming to eliminate malaria.
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
Abstract
Abstract
Imported malaria is a critical obstacle to achieving elimination in low transmission settings. Characterising malaria importation and transmission sources using human mobility and parasite genomics has the potential to inform elimination strategies, but tools combining both types of data are lacking. We developed a novel Bayesian approach that provides individual importation probabilities and geographic origin of P. falciparum cases by combining epidemiological, human mobility and parasite genetic data. Spatial genetic structure and connectivity were assessed using microhaplotype-based genetic relatedness (identity-by-descent) from 1605 P. falciparum samples collected from 9 provinces in Mozambique during 2022, including two very-low transmission elimination-targeted districts (Magude and Matutuine) in the south. Travel reports were combined with genetic relatedness metrics to classify clinical cases as local or imported. Genetic relatedness between parasites from southern and northern/central Mozambique was lower (0.021) than average (0.034, p<0.001). 42% (88/207) of infections in elimination-targeted districts were classified as imported, had a higher genetic complexity (OR=1.3) and originated mainly from Inhambane (63% [55/88]). Significant differences in importation rates were found between the two studied districts (OR=6.6), with Magude district (10.71%, 3/28) showing lower importation rates than Matutuine (48.60%, 87/179) district. Differences in importation rates observed between both elimination districts suggest the need for fine-scale analysis to tailor cost-effective elimination strategies. Importation is playing a crucial role in sustaining transmission in Matutuine district, and increasing efforts to reduce malaria burden in their sources of transmission (especially in Inhambane province), as well as targeting travelers to central and northern Mozambique, could significantly contribute to malaria elimination in the south.
Article activity feed
-
-
-
eLife Assessment
This study introduces a useful method to estimate the probability that a malaria case is imported and to identify the geographic origin of parasites by using a Bayesian approach that integrates epidemiological, travel, and genetic data. The authors provide convincing evidence that the approach can reliably identify the main sources of malaria imports. This work will be of great interest to the area of genomic epidemiology and public health strategies aiming to eliminate malaria.
-
Reviewer #1 (Public review):
Summary:
This study presents a new Bayesian approach to estimate importation probabilities of malaria, combining epidemiological data, travel history, and genetic data through pairwise IBD estimates. Importation is an important factor challenging malaria elimination, especially in low-transmission settings. This paper focuses on Magude and Matutuine, two districts in southern Mozambique with very low malaria transmission. The results show isolation-by-distance in Mozambique, with genetic relatedness decreasing with distances larger than 100 km, and no spatial correlation for distances between 10 and 100 km. But again, strong spatial correlation in distances smaller than 10 km. They report high genetic relatedness between Matutuine and Inhambane, higher than between Matutuine and Magude. Inhambane is the main …
Reviewer #1 (Public review):
Summary:
This study presents a new Bayesian approach to estimate importation probabilities of malaria, combining epidemiological data, travel history, and genetic data through pairwise IBD estimates. Importation is an important factor challenging malaria elimination, especially in low-transmission settings. This paper focuses on Magude and Matutuine, two districts in southern Mozambique with very low malaria transmission. The results show isolation-by-distance in Mozambique, with genetic relatedness decreasing with distances larger than 100 km, and no spatial correlation for distances between 10 and 100 km. But again, strong spatial correlation in distances smaller than 10 km. They report high genetic relatedness between Matutuine and Inhambane, higher than between Matutuine and Magude. Inhambane is the main source of importation in Matutuine, accounting for 63.5% of imported cases. Magude, on the other hand, shows smaller importation and travel rates than Matutuine, as it is a rural area with less mobility. Additionally, they report higher levels of importation and travel in the dry season, when transmission is lower. Also, no association with importation was found for occupation, sex, and other factors. These data have practical implications for public health strategies aiming for malaria elimination, for example, testing and treating travelers from Matutuine in the dry season.
Strengths:
The strength of this study lies in the combination of different sources of data - epidemiological, travel, and genetic data - to estimate importation probabilities, and the statistical analyses.
Weaknesses:
The authors recognize the limitations related to sample size and the biases of travel reports.
-
Reviewer #2 (Public review):
Summary:
Based on a detailed dataset, the authors present a novel Bayesian approach to classify malaria cases as either imported or locally acquired.
Strengths:
The proposed Bayesian approach for case classification is simple, well justified, and allows the integration of parasite genomics, travel history, and epidemiological data. The work is well-written, very organized, and brings important contributions both to malaria control efforts in Mozambique and to the scientific community. Understanding the origin of cases is essential for designing more effective control measures and elimination strategies.
Weakness:
While the authors aim to classify cases as imported or locally acquired, the work lacks a quantification of the contribution of each case type to overall transmission.
The Bayesian rationale is …
Reviewer #2 (Public review):
Summary:
Based on a detailed dataset, the authors present a novel Bayesian approach to classify malaria cases as either imported or locally acquired.
Strengths:
The proposed Bayesian approach for case classification is simple, well justified, and allows the integration of parasite genomics, travel history, and epidemiological data. The work is well-written, very organized, and brings important contributions both to malaria control efforts in Mozambique and to the scientific community. Understanding the origin of cases is essential for designing more effective control measures and elimination strategies.
Weakness:
While the authors aim to classify cases as imported or locally acquired, the work lacks a quantification of the contribution of each case type to overall transmission.
The Bayesian rationale is sound and well justified; however, the formulation appears to present an inconsistency that is replicated in both the main text and the Supplementary Material.
-
Reviewer #3 (Public review):
The authors present an important approach to identify imported P. falciparum malaria cases, combining genetic and epidemiological/travel data. This tool has the potential to be expanded to other contexts. The data was analyzed using convincing methods, including a novel statistical model; although some recognized limitations can be improved. This study will be of interest to researchers in public health and infectious diseases.
Strengths:
The study has several strengths, mainly the development of a novel Bayesian model that integrates genomic, epidemiological, and travel data to estimate importation probabilities. The results showed insights into malaria transmission dynamics, particularly identifying importation sources and differences in importation rates in Mozambique. Finally, the relevance of the …
Reviewer #3 (Public review):
The authors present an important approach to identify imported P. falciparum malaria cases, combining genetic and epidemiological/travel data. This tool has the potential to be expanded to other contexts. The data was analyzed using convincing methods, including a novel statistical model; although some recognized limitations can be improved. This study will be of interest to researchers in public health and infectious diseases.
Strengths:
The study has several strengths, mainly the development of a novel Bayesian model that integrates genomic, epidemiological, and travel data to estimate importation probabilities. The results showed insights into malaria transmission dynamics, particularly identifying importation sources and differences in importation rates in Mozambique. Finally, the relevance of the findings is to suggest interventions focusing on the traveler population to help efforts for malaria elimination.
Weaknesses:
The study also has some limitations. The sample collection was not representative of some provinces, and not all samples had sufficient metadata for risk factor analysis, which can also be affected by travel recall bias. Additionally, the authors used a proxy for transmission intensity and assumed some conditions for the genetic variable when calculating the importation probability for specific scenarios. The weaknesses were assessed by the authors.
-
Author response:
Reviewer #1 (Public review):
Summary:
This study presents a new Bayesian approach to estimate importation probabilities of malaria, combining epidemiological data, travel history, and genetic data through pairwise IBD estimates. Importation is an important factor challenging malaria elimination, especially in low-transmission settings. This paper focuses on Magude and Matutuine, two districts in southern Mozambique with very low malaria transmission. The results show isolation-by-distance in Mozambique, with genetic relatedness decreasing with distances larger than 100 km, and no spatial correlation for distances between 10 and 100 km. But again, strong spatial correlation in distances smaller than 10 km. They report high genetic relatedness between Matutuine and Inhambane, higher than between Matutuine and Magude. …
Author response:
Reviewer #1 (Public review):
Summary:
This study presents a new Bayesian approach to estimate importation probabilities of malaria, combining epidemiological data, travel history, and genetic data through pairwise IBD estimates. Importation is an important factor challenging malaria elimination, especially in low-transmission settings. This paper focuses on Magude and Matutuine, two districts in southern Mozambique with very low malaria transmission. The results show isolation-by-distance in Mozambique, with genetic relatedness decreasing with distances larger than 100 km, and no spatial correlation for distances between 10 and 100 km. But again, strong spatial correlation in distances smaller than 10 km. They report high genetic relatedness between Matutuine and Inhambane, higher than between Matutuine and Magude. Inhambane is the main source of importation in Matutuine, accounting for 63.5% of imported cases. Magude, on the other hand, shows smaller importation and travel rates than Matutuine, as it is a rural area with less mobility. Additionally, they report higher levels of importation and travel in the dry season, when transmission is lower. Also, no association with importation was found for occupation, sex, and other factors. These data have practical implications for public health strategies aiming for malaria elimination, for example, testing and treating travelers from Matutuine in the dry season.
Strengths:
The strength of this study lies in the combination of different sources of data - epidemiological, travel, and genetic data - to estimate importation probabilities, and the statistical analyses.
Weaknesses:
The authors recognize the limitations related to sample size and the biases of travel reports.
Thank you for your review and consideration. As mentioned, we state in the manuscript the limitations related to sample sizes and travel reports. We aim to continue this study with new prospective data, aiming to address these limitations.
Reviewer #2 (Public review):
Summary:
Based on a detailed dataset, the authors present a novel Bayesian approach to classify malaria cases as either imported or locally acquired.
Strengths:
The proposed Bayesian approach for case classification is simple, well justified, and allows the integration of parasite genomics, travel history, and epidemiological data. The work is well-written, very organized, and brings important contributions both to malaria control efforts in Mozambique and to the scientific community. Understanding the origin of cases is essential for designing more effective control measures and elimination strategies.
Weakness:
While the authors aim to classify cases as imported or locally acquired, the work lacks a quantification of the contribution of each case type to overall transmission.
The Bayesian rationale is sound and well justified; however, the formulation appears to present an inconsistency that is replicated in both the main text and the Supplementary Material.
In fact, one of the questions that remains unanswered is the overall contribution of importation events to transmission in the areas. While the Bayesian classifier does not quantify this, our future analysis will focus on combining outbreak detection, genetic clustering and importation classification to quantify the contribution of imported cases to outbreak resurgence and to the overall transmission.
Thank you for pointing out the inconsistency in the final formula. In fact, the final formula corresponds to P(IA | G), instead to i>P(IA), so:
instead of
We will correct this error in a new version of the manuscript.
Reviewer #3 (Public review):
The authors present an important approach to identify imported P. falciparum malaria cases, combining genetic and epidemiological/travel data. This tool has the potential to be expanded to other contexts. The data was analyzed using convincing methods, including a novel statistical model; although some recognized limitations can be improved. This study will be of interest to researchers in public health and infectious diseases.
Strengths:
The study has several strengths, mainly the development of a novel Bayesian model that integrates genomic, epidemiological, and travel data to estimate importation probabilities. The results showed insights into malaria transmission dynamics, particularly identifying importation sources and differences in importation rates in Mozambique. Finally, the relevance of the findings is to suggest interventions focusing on the traveler population to help efforts for malaria elimination.
Weaknesses:
The study also has some limitations. The sample collection was not representative of some provinces, and not all samples had sufficient metadata for risk factor analysis, which can also be affected by travel recall bias. Additionally, the authors used a proxy for transmission intensity and assumed some conditions for the genetic variable when calculating the importation probability for specific scenarios. The weaknesses were assessed by the authors.
We acknowledge the limitations commented by the reviewer. We have the following plans to address the limitations. We will repeat the study for our data collected in 2023, which this time contains a good representation of all the provinces of Mozambique, and completeness of the metadata collection was ensured by implementing a new protocol in January 2023. Regarding the proxy for transmission intensity, we will refine the model by integrating monthly estimates of malaria incidence (previously calibrated to address testing and reporting rates) from the DHIS2 data, taking also into account the date of the reported cases in the analysis.
-
