1. Epidemiology and Global Health
  2. Microbiology and Infectious Disease

Estimating probabilities of malaria importation in southern Mozambique through modelling P. falciparum genomics and mobility patterns

  1. Arnau Pujol  Is a corresponding author
  2. Arlindo Chidimatembue
  3. Clemente da Silva
  4. Simone Boene
  5. Henriques Mbeve  Is a corresponding author
  6. Pau Cisteró
  7. Carla García-Fernández
  8. Arnau Vañó-Boira
  9. Dário Tembisse
  10. José Inácio
  11. Glória Matambisso
  12. Fabião Luis
  13. Nelo Ndimande
  14. Humberto Munguambe
  15. Lidia Nhamussua
  16. Wilson Simone
  17. Andrés Aranda-Díaz
  18. Manuel García-Ulloa
  19. Neide Canana
  20. Maria Tusell
  21. Júlia Montaña
  22. Laura Fuente-Soro
  23. Khalid Ussene Bapu
  24. Maxwell Murphy
  25. Bernardete Rafael
  26. Eduard Rovira-Vallbona
  27. Caterina Guinovart
  28. Bryan Greenhouse
  29. Sonia Maria Enosse
  30. Francisco Saúte
  31. Pedro Aide
  32. Baltazar Candrinho
  33. Alfredo Mayor
  1. ISGlobal, Spain
  2. Centro de Investigação em Saúde de Manhiça, Mozambique
  3. Facultat de Medicina i Ciències de la Salut, Universitat de Barcelona (UB), Spain
  4. EPPIcenter Research Program, Division of HIV, Infectious Diseases and Global Medicine, Department of Medicine, University of California, San Francisco, United States
  5. Malaria Consortium, Mozambique
  6. National Malaria Control Program, Mozambique
  7. Instituto Nacional de Saúde (INS), Mozambique
  8. Faculty of Medicine, Universidade Eduardo Mondlane, Mozambique
  9. Spanish Consortium for Research in Epidemiology and Public Health (CIBERESP), Spain
https://doi.org/10.7554/eLife.107136.4
Share this article
Cite this article
  1. Arnau Pujol
  2. Arlindo Chidimatembue
  3. Clemente da Silva
  4. Simone Boene
  5. Henriques Mbeve
  6. Pau Cisteró
  7. Carla García-Fernández
  8. Arnau Vañó-Boira
  9. Dário Tembisse
  10. José Inácio
  11. Glória Matambisso
  12. Fabião Luis
  13. Nelo Ndimande
  14. Humberto Munguambe
  15. Lidia Nhamussua
  16. Wilson Simone
  17. Andrés Aranda-Díaz
  18. Manuel García-Ulloa
  19. Neide Canana
  20. Maria Tusell
  21. Júlia Montaña
  22. Laura Fuente-Soro
  23. Khalid Ussene Bapu
  24. Maxwell Murphy
  25. Bernardete Rafael
  26. Eduard Rovira-Vallbona
  27. Caterina Guinovart
  28. Bryan Greenhouse
  29. Sonia Maria Enosse
  30. Francisco Saúte
  31. Pedro Aide
  32. Baltazar Candrinho
  33. Alfredo Mayor
(2026)
Estimating probabilities of malaria importation in southern Mozambique through modelling P. falciparum genomics and mobility patterns
eLife 14:RP107136.
https://doi.org/10.7554/eLife.107136.4
  2. Download BibTeX
  3. Download .RIS
9 figures, 9 tables and 2 additional files

Figures

Figure 1
Download asset Open asset
Map of the study health facilities.

Provinces were colours according to their region: north (red), centre (green), or south (blue). White dots show the locations of the health facilities included in the study, with a zoom-in to show the ones included in Magude and Matutuine districts. Maps used OpenStreetMap data, available under the Open Database License.

Figure 2
Download asset Open asset
Flowchart of the P. falciparum samples and data from Magude and Matutuine districts collected in 2022.
Figure 3
Download asset Open asset
Statistics of travel reports.

(A) Pie chart showing the distributions of the travel destination provinces in sampled individuals from Maputo province (Magude and Matutuine). Colours show the provinces from blue (south) to red (north). (B) Spatial connectivity based on travel history. Line widths are proportional to the number of travels reported from Maputo province to their destination province, with the same colours as in A. Maps used OpenStreetMap data, available under the Open Database License.

Figure 4
Download asset Open asset
Genetic relatedness (identity-by-descent [IBD]) of P. falciparum infections between regions in Mozambique.

(A) Fraction (R) of IBD-related sample pairs (IBD>0.1 with p<0.05) within and between different provinces, represented in colours. (B) Spatial genetic connectivity between provinces. Line widths and point sizes are proportional to R, and colours show the ranking in R values (from blue to red, using turbo colourmap). (C) R between and within different regions (south, centre, and north), combining samples from centre and north (D) and across Magude and Matutuine districts (E). N=1467. Maps used OpenStreetMap data, available under the Open Database License.

Figure 5
Download asset Open asset
p-Values of identity-by-descent (IBD)-relatedness results.

(A) p-Values of the deviation of the fraction of IBD-related pairs (IBD>0.1 with p<0.05) of the samples within and between different provinces with respect to the average across all pairs. (B) The same, but comparing sample pairs within and across regions (south, centre, and north). (C) The same, but comparing sample pairs across south and centre/north. (D) The same, but across Magude and Matutuine districts.

Figure 6
Download asset Open asset
Genetic relatedness as a function of spatial distance.

(A) Fraction of identity-by-descent (IBD)-related pairs (using different thresholds of IBD [always with p<0.05]), shown in different colours as a function of the pairwise geographical distance, using a range of distances from 100 km to 1400 km (typically inter-province samples). (B) The same as A but for distances between 20 km and 100 km (typically within the same province but distinct health facilities). (C) The same but for distances between 0 km (same household) and 20 km, which is impacted by the significantly higher fraction for cases from the same household. N=1467.

Figure 7
Download asset Open asset
Distribution of importation probabilities by district.

Distribution of the individual probabilities of being imported for the studied clinical cases from Magude (orange) and Matutuine (blue) districts.

Figure 8
Download asset Open asset
Odds ratio statistics of factors associated with importation and travel.

Odds ratio of importation (A, B), reporting travel (C, D), and importation for cases with travel reports (E, F) for different factors in univariate (A, C, E) and multivariate (B, D, F) models, for all P. falciparum clinical cases recruited in Magude and Matutuine (n=200 for A–D, n=105 for E and F).

Appendix 1—figure 1
Download asset Open asset
Distribution of importation probabilities using different combinations of factors.

Probability distributions of the clinical cases with travel reports of being imported, depending on the inclusion of factors included in the estimation. The black histogram (A) shows the estimations from the formula described in Methods sections, while the orange (B), purple (C), and green (D) distribution show the estimations when removing the genetic relatedness (R’), the travel duration (T), or the transmission intensity (PR) metrics, respectively.

Tables

Table 1
Sample size and characteristics of the P. falciparum clinical cases in Magude and Matutuine with travel history data.
CategoryMagudeMatutuinep-Value (Chi2)p-Value (t-test)
N27173NANA
Dry season (%)4 (14.8)60 (34.7)0.066
Mean age (SD)21.56 (14.1)24.12 (17.6)0.473
Female (%)10 (37.0)91 (52.6)0.194
Occupation, n (%)Domestic3 (11.1)41 (23.7)0.021
Student9 (33.3)16 (9.2)
Agriculture3 (11.1)19 (11.0)
Other11 (40.7)47 (27.2)
Unemployed1 (3.7)50 (28.9)
Reported travel within previous 28 days (%)3 (11.1)102 (59.0)<0.001
Province travel destinationGaza2110.374
Inhambane156
Maputo010
Zambézia015
Manica01
Maputo City02
Sofala01
Nampula05
Cabo Delgado01
Table 2
Travel destinations among the 105 P. falciparum clinical cases in Magude and Matutuine that reported having travelled during the previous 28 days.

Provinces of the travel destinations reported (left column) and the total number of travels (and percentage) reported to each province (right column).

Travel destination provinceNumber of travels n (%)
Maputo City2 (1.9%)
Maputo Province10 (9.5%)
Gaza13 (12.4%)
Inhambane57 (54.3%)
Sofala1 (1.0%)
Manica1 (1.0%)
Zambézia15 (14.3%)
Nampula5 (4.8%)
Cabo Delgado1 (1.0%)
Total105
Table 3
Convenience sampling from selected health facilities in nine provinces: number of samples and patient characteristics.

Province: province of the health facilities from where the samples were collected. N: number of samples collected in each province. Mean age (SD): mean and standard deviation of the participant ages from each province. Female fraction (%): fraction of females of the participants from each province.

ProvinceNMean age (SD)Female fraction (%)
Maputo9520.55 (18.78)47.87
Inhambane3456.17 (2.67)54.44
Sofala805.5 (2.4)50
Manica655.55 (2.34)47.69
Tete832.3 (1.49)43.9
Zambézia724.55 (2.17)56.94
Nampula441.45 (0.91)43.18
Niassa795.68 (2.68)43.04
Cabo Delgado862.22 (1.28)51.16
p-Value (*=t-test, **=Chi-square test)<0.001*0.442**
Table 4
Travel destination (by province) of the imported cases reported in Magude and Matutuine.

Weighted number of cases: the total number of cases (weighted by their probability) imported from each province (n=200). Imported cases: the total number of cases classified as imported if p(imported)>50% (n=200). % cases with travel reports: The fraction of imported cases from each province with respect to the total number of cases reporting travel in the previous 28 days (n=105). % all reported: the total contribution of imported cases from each province with respect to all reported cases in the Maputo province (n=200). Imported to Magude: total of imported cases residing in Magude (n=27). Imported to Matutuine: total of imported cases residing in Matutuine (n=173).

Importation provinceWeighted number of casesCases classified as imported% imported among cases with reported travel (imported/105)% imported among all reported (imported/200)Imported to MagudeImported to Matutuine
Cabo Delgado1.011.00.501
Gaza11.31211.46.0210
Inhambane51.95451.427.0153
Manica0.911.00.501
Maputo2.321.91.002
Maputo City0.300.00.000
Nampula4.043.82.004
Sofala1.011.00.501
Zambezia11.91211.46.0012
Total84.58782.943.5383
Appendix 1—table 1
Importation probability statistics using different data sources in the estimations.

The weighted number of cases (second column, n=2007) and their percentages with respect to all mRDT+ cases with travel reports (third column, n=1057) and with respect to all studied cases (last column, n=2007) obtained for different combinations of data sources (specified in the first column). R’: the estimation of probability of importation (see Methods) includes the R’ in the travel destination (fraction of samples from the travel destination province genetically related to the case) and origin (fraction of samples from the local province – Maputo – genetically related to the case). Otherwise, equal R’ is assumed between origin and travel destination. T: the estimation of probability of importation includes the metrics of the travel duration and time spent at home. Otherwise, equal time is assumed. PR: the mRDT positivity rates (PRs) per province in children reported in the last Health Demographic Survey 2022–2023 is used as a proxy of transmission intensity for each province. Otherwise, equal PR is assumed.

Variable used for p(importation)Weighted number of cases% with travel reports% all reported
R', T, and PR84.580.541.3
T and PR88.284.044.1
R' and PR91.887.445.9
R' and T46.944.723.5
R'54.551.927.3
T45.743.522.8
PR96.391.748.2
Appendix 2—table 1
Comparison of travel reports and genetic relatedness with Maputo province (local) and travel destination provinces.

The table shows the number of cases that reported (Yes) or not (No) a travel to each province (columns ‘Destination province’) and how many of these cases show a fraction of genetically related pairs (identity-by-descent [IBD]>0.1 with p<0.05) higher when compared with the samples collected in Maputo province (third column) or when compared with the travel destination province (fourth column) (n=200). The p-value shows the result of a chi-square consistency test under the hypothesis that the genetic relatedness with Maputo or the other province does not depend on whether a travel has been reported or not.

Destination provinceTravel reportedMore genetically related to MaputoMore genetically related to destination provincep-Value
InhambaneNo57860.018
Yes1245
SofalaNo107921.000
Yes10
ManicaNo119800.846
Yes01
ZambéziaNo114710.171
Yes69
NampulaNo122730.143
Yes14
Cabo DelgadoNo103960.976
Yes01
Appendix 2—table 2
Factors associated with imported malaria among the 200 cases reported in Matutuine and Magude: distribution by malaria case classification (local versus imported) and odds ratios from univariate and multivariate logistic regression.

SD = standard deviation; GM = geometric mean; CI = confidence interval.

CategoryLocalImported% importedUnivariateMultivariate
Odds ratio (95% CI)p-ValueOdds ratio (95% CI)p-Value
District of residenceMagude24311.1%11
Matutuine898448.6%6.6 (2.3, 25.4)<0.0014.5 (1.5, 17.6)0.005
SeasonDry293554.7%1.9 (1.1, 3.5)0.0291.6 (0.9, 3.0)0.147
Rainy845238.2%11
SexFemale515049.5%1.6 (0.9, 2.9)0.0851.5 (0.8, 2.9)0.194
Male623737.4%11
PregnantYes200.0%0.2 (0.0, 2.7)0.246NA
No353449.3%1NA
OccupationDomestic251943.2%0.5 (0.2, 1.1)0.0890.5 (0.2, 1.4)0.177
Student17832.0%0.3 (0.1, 0.8)0.0190.4 (0.2, 1.2)0.094
Farmer14836.4%0.4 (0.1, 1.0)0.0580.3 (0.1, 1.0)0.051
Other372136.2%0.4 (0.2, 0.8)0.0110.4 (0.1, 1.4)0.164
Unemployed203160.8%11
AgeAdult735543.0%0.9 (0.5, 1.7)0.8371.6 (0.7, 3.9)0.311
Minor403244.4%11
Province travel destinationGaza11292.3%28.3 (3.9, 382.5)<0.001NA
Inhambane35494.7%52.9 (10.3, 390.8)<0.001NA
Maputo8220.0%1NA
Zambézia31280.0%12.1 (2.2, 94.1)0.004NA
Manica01100.0%10.2 (0.4, 1703.3)0.146NA
Maputo City200.0%0.7 (0.0, 13.2)0.668NA
Sofala01100.0%10.2 (0.4, 1703.3)0.157NA
Nampula1480.0%10.2 (1.2, 148.8)0.034NA
Delgado01100.0%10.2 (0.4, 1703.3)0.150NA
Parasite densityGM (SD)3.18 (1.1)3.08 (1.0)NA0.9 (0.7, 1.2)0.529NA
eCOIMean (95% CI)1.53 (1.0, 3.4)1.80 (1.0, 4.2)NA1.4 (1.0, 1.9)0.038NA
Polyclon.Polyclonal545048.1%1.5 (0.8, 2.6)0.176NA
Monoclonal593738.5%1NA
COIMean (95% CI)2.21 (1.0, 5.2)2.80 (1.0, 7.8)NA1.3 (1.1, 1.5)0.009NA
Prob. polyclonalMean (SD)0.54 (0.5)0.65 (0.4)NA1.8 (0.9, 3.3)0.075NA
Naive polyclon.Polyclonal676248.1%1.7 (0.9, 3.1)0.081NA
Monoclonal462535.2%1NA
Appendix 2—table 3
Factors associated with travel history among the 200 cases reported in Matutuine and Magude: odds ratios from univariate and multivariate logistic regression.

SD = standard deviation; GM = geometric mean; CI = confidence interval.

CategoryNo travelTravel% travelledOdds ratio (95% CI)p-ValueMultivariate
Odds ratio (95% CI)p-Value
District of residenceMagude24311.1%11
Matutuine7110259.0%10.0 (3.5, 38.5)<0.0016.7 (2.3, 26.2)0.000
SeasonDry234164.1%2.0 (1.1, 3.7)0.0251.6 (0.8, 3.0)0.178
Rainy726447.1%11
SexFemale435857.4%1.5 (0.9, 2.6)0.1611.3 (0.7, 2.4)0.487
Male524747.5%11
PregnantYes02100.0%0.1 (0.0, 1.8)0.136NA
No412840.6%1NA
OccupationDomestic182659.1%0.7 (0.3, 1.7)0.4490.5 (0.2, 1.4)0.190
Student16936.0%0.3 (0.1, 0.8)0.0120.4 (0.2, 1.2)0.099
Farmer121045.5%0.4 (0.2, 1.2)0.0930.5 (0.2, 1.4)0.184
Other322644.8%0.4 (0.2, 0.9)0.0230.4 (0.1, 1.4)0.151
Unemployed173466.7%11
AgeAdult587054.7%1.3 (0.7, 2.3)0.4111.9 (0.8, 4.7)0.166
Minor373548.6%11
Parasite densityGM (SD)3.18 (1.1)3.11 (1.0)NA0.9 (0.7, 1.2)0.658NA
eCOIMean (95% CI)1.48 (1.0, 3.6)1.80 (1.0, 4.1)NA1.5 (1.1, 2.2)0.012NA
Polyclon.Polyclonal634139.4%2.0 (1.1, 3.5)0.018NA
Monoclonal425456.3%1NA
COIMean (95% CI)2.13 (1.0, 5.6)2.78 (1.0, 7.4)NA1.3 (1.1, 1.6)0.003NA
Prob. polyclonalMean (SD)0.50 (0.5)0.67 (0.4)NA2.4 (1.3, 4.4)0.007NA
Naive polyclon.Polyclonal765341.1%2.1 (1.2, 3.7)0.015NA
Monoclonal294259.2%1NA
Appendix 2—table 4
Factors associated with importation among the 105 cases that reported having travelled during the previous 28 days: distribution by malaria case classification (local vs imported) and odds ratios from univariate and multivariate logistic regression.

SD = standard deviation; GM = geometric mean; CI = confidence interval.

CategoryLocalImported% importedUnivariateMultivariate (without travel destination)Multivariate (including travel destination)
Odds ratio (95% CI)p-ValueOdds ratio (95% CI)p-ValueOdds ratio (95% CI)p-Value
District of residenceMagude03100.0%111
Matutuine188482.4%0.7 (0.0, 7.2)0.7700.64 (0.0, 11.5)0.7512.7 (0.01, 78.38)0.577
SeasonDry63585.4%1.3 (0.5, 3.9)0.6171.35 (0.47, 4.23)0.5650.99 (0.22, 4.94)0.713
Rainy125281.3%111
SexFemale85086.2%1.7 (0.6, 4.6)0.3162.19 (0.72, 7.15)0.1642.26 (0.53, 10.6)0.260
Male103778.7%111
OccupationDomestic71973.1%0.3 (0.1, 1.1)0.0680.54 (0.07, 6.22)0.5631.04 (0.07, 24.96)0.743
Student1888.9%0.6 (0.1, 7.1)0.6670.82 (0.06, 8.28)0.8563.27 (0.18, 64.28)0.399
Farmer2880.0%0.4 (0.1, 2.6)0.3000.92 (0.08, 6.36)0.7636.61 (0.41, 125.99)0.144
Other52180.8%0.4 (0.1, 1.8)0.2500.39 (0.03, 2.46)0.3291.18 (0.1, 9.71)0.839
Unemployed33191.2%111
AgeAdult155578.6%0.4 (0.1, 1.2)0.1070.66 (0.09, 7.4)0.6770.16 (0.01, 2.45)0.158
Minor33291.4%111
Province travel destinationGaza11292.3%28.3 (3.9, 382.5)<0.001NA5.58 (0.63, 84.96)0.111
Inhambane35494.7%52.9 (10.3, 390.8)<0.001NA54.13 (8.31, 661.76)<0.001
Maputo8220.0%1NA1
Zambézia31280.0%12.1 (2.2, 94.1)0.0035NA7.22 (0.13, 1788.14)0.312
Manica01100.0%10.2 (0.41, 1703.29)0.1461NA29.1 (3.01, 561.76)0.002
Maputo City200.0%0.7 (0.0, 13.2)0.6677NA8.08 (0.28, 1432.17)0.185
Sofala01100.0%10.2 (0.4, 1703.3)0.1565NA3.57 (0.1, 680.45)0.415
Nampula1480.0%10.2 (1.2, 148.8)0.034NA10.72 (1.2, 148.94)0.030
Delgado01100.0%10.2 (0.4, 1703.3)0.1496NA0.41 (0.0, 17.53)0.516
Parasite densityGM (SD)3.22 (0.9)3.08 (1.0)NA0.9 (0.5, 1.5)0.6546NANA
eCOIMean (95% CI)1.80 (1.0, 2.7)1.80 (1.0, 4.2)NA1.0 (0.6, 1.7)0.9263NANA
Polyclon.Polyclonal135079.4%0.6 (0.2, 1.6)0.2626NANA
Monoclonal53788.1%1NANA
COIMean (95% CI)2.67 (1.0, 4.6)2.80 (1.0, 7.8)NA1.0 (0.8, 1.4)0.8224NANA
prob. polyclonalMean (SD)0.75 (0.4)0.65 (0.4)NA0.6 (0.2, 2.0)0.4195NANA
Naive polyclon.Polyclonal146281.6%0.8 (0.2, 2.3)0.6368NANA
Monoclonal42586.2%1NANA

Additional files

Supplementary file 1

Table of accession numbers of sequencing data of the study samples, available at NCBI Sequence Read Archive (SRA).

https://cdn.elifesciences.org/articles/107136/elife-107136-supp1-v1.xlsx
MDAR checklist
https://cdn.elifesciences.org/articles/107136/elife-107136-mdarchecklist1-v1.docx

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Arnau Pujol
  2. Arlindo Chidimatembue
  3. Clemente da Silva
  4. Simone Boene
  5. Henriques Mbeve
  6. Pau Cisteró
  7. Carla García-Fernández
  8. Arnau Vañó-Boira
  9. Dário Tembisse
  10. José Inácio
  11. Glória Matambisso
  12. Fabião Luis
  13. Nelo Ndimande
  14. Humberto Munguambe
  15. Lidia Nhamussua
  16. Wilson Simone
  17. Andrés Aranda-Díaz
  18. Manuel García-Ulloa
  19. Neide Canana
  20. Maria Tusell
  21. Júlia Montaña
  22. Laura Fuente-Soro
  23. Khalid Ussene Bapu
  24. Maxwell Murphy
  25. Bernardete Rafael
  26. Eduard Rovira-Vallbona
  27. Caterina Guinovart
  28. Bryan Greenhouse
  29. Sonia Maria Enosse
  30. Francisco Saúte
  31. Pedro Aide
  32. Baltazar Candrinho
  33. Alfredo Mayor
(2026)
Estimating probabilities of malaria importation in southern Mozambique through modelling P. falciparum genomics and mobility patterns
eLife 14:RP107136.
https://doi.org/10.7554/eLife.107136.4