1. Epidemiology and Global Health

Using step selection functions to analyse human mobility using telemetry data in infectious disease epidemiology: a case study of leptospirosis

  1. Pablo Ruiz Cuenca  Is a corresponding author
  2. Fábio N Souza
  3. Roberta Coutinho do Nascimento
  4. Ariane Goncalves da Silva
  5. Max T Eyre
  6. Juliet O Santana
  7. Daiana de Oliveira
  8. Emile V Ribeiro de Souza
  9. Fabiana G Palma
  10. Diogo C de Carvalho Santiago
  11. Priscyla dos Santos Ribeiro
  12. Priscilla Elizabeth Ferreira dos Santos
  13. Hussein Khalil
  14. Jonathan M Read
  15. Cleber Cremonese
  16. Federico Costa
  17. Emanuele Giorgi
  1. Centre for Health Informatics, Computing, and Statistics (CHICAS), Lancaster Medical School, Lancaster University, United Kingdom
  2. Institute of Collective Health, Federal University of Bahia, Brazil
  3. Instituto Gonçalo Moniz, Fundação Oswaldo Cruz, Ministério da Saúde, Brazil
  4. Environmental Health Group, London School of Hygiene and Tropical Medicine, United Kingdom
  5. Federal University of Bahia, National Institute of Science and Technology in Interdisciplinary and Transdiciplinary Studies in Ecology and Evolution, Brazil
  6. Amsterdam University Medical Centre, Leptospirosis Reference Center, Medical Microbiology and Infection Control, Netherlands
  7. Department of Wildlife, Fish and Environmental Studies, Swedish University of Agricultural Sciences (SLU), Sweden
  8. Department of Epidemiology of Microbial Diseases, School of Public Health, Yale University, United States
https://doi.org/10.7554/eLife.107153.3
Share this article
Cite this article
  1. Pablo Ruiz Cuenca
  2. Fábio N Souza
  3. Roberta Coutinho do Nascimento
  4. Ariane Goncalves da Silva
  5. Max T Eyre
  6. Juliet O Santana
  7. Daiana de Oliveira
  8. Emile V Ribeiro de Souza
  9. Fabiana G Palma
  10. Diogo C de Carvalho Santiago
  11. Priscyla dos Santos Ribeiro
  12. Priscilla Elizabeth Ferreira dos Santos
  13. Hussein Khalil
  14. Jonathan M Read
  15. Cleber Cremonese
  16. Federico Costa
  17. Emanuele Giorgi
(2025)
Using step selection functions to analyse human mobility using telemetry data in infectious disease epidemiology: a case study of leptospirosis
eLife 14:RP107153.
https://doi.org/10.7554/eLife.107153.3
  2. Download BibTeX
  3. Download .RIS
7 figures, 5 tables and 4 additional files

Figures

Figure 1
Download asset Open asset
Map showing location of each study area in Salvador.

Each area includes symbology for stream (blue line), open sewer points (purple diamond), and domestic rubbish piles (orange triangle).

Figure 2
Download asset Open asset
Schematic diagram showing what data sources are used in which model, and how models are linked with each other.

The blue sections represent phase 1, the individual-level models, whilst the orange section represents phase 2, the population-level model.

Figure 3
Download asset Open asset
Descriptive diagram of step selection functions.

(A) Step lengths (sl) and turning angles (θ) are used to characterise an individual’s movements. (B) These parameters are used to create a set of available steps (grey dots) for every used step (black dots).

Figure 4
Download asset Open asset
Graph showing results of final analyses.

(A) Results for distance-based rasters; values above 1 interpreted as increasing distance to points of reference. (B) Results for 20 m buffer-based rasters; values above 1 show movement within buffer zones. Each horizontal band represents a specific time period (right hand side y-axis label): all day (5 am to 9 pm, Tables 2 and 3), morning (5 am to 9 am), midday (9 am to 1 pm), afternoon (1 pm to 5 pm), and evening (5 pm to 9 pm). All data points include their corresponding 95% confidence intervals, some of which are too narrow to show up clearly.

Appendix 1—figure 1
Download asset Open asset
Distribution of telemetry data provided by each individual across 24 hr periods (x-axis), separated into each of the four study areas (1: NVS, 2: ARE, 3: JSI, 4: CAL).

Overlapping areas represent multiple days. Vertical bars represent 5 am (left-hand bar) and 9 pm (right-hand bar), the period of analysis.

Appendix 1—figure 2
Download asset Open asset
Distribution of Leptospirosis antibody status (serological status) by gender and age.
Appendix 1—figure 3
Download asset Open asset
Distribution of nearest distance to each of the environmental factors being analysed (central stream, open sewer points, and domestic rubbish piles) by serological status (x-axis) and study area (1: NVS, 2: ARE, 3: JSI, 4: CAL).

NA represents the rest of households in the study area that did not take part in movement analysis.

Tables

Table 1
Summary table comparing parent study participants and movement study participants.
Parent study participantsMovement study participants
n=1086%/meann=128%/mean
Study area
1—Nova Sussuarana29727.3%3225.0%
2—Arenoso24622.7%2821.9%
3—Jardim Santo Inacio27825.6%3527.3%
4—Calabetao26524.4%3325.7%
Gender
Female45441.8%5946.1%
Male63258.2%6953.9%
Age (mean±SD)32.2±19.739.4±15.4
Leptospira antibody status
Positive948.7%1310.2%
Negative99291.3%11589.8%
Table 2
Proportion of tracked time (full-day period, 9 am to 5 pm) spent within each buffer.

Mean (standard deviation).

River bufferOpen sewer bufferDomestic rubbish buffer
Total0.54 (0.33)0.33 (0.31)0.07 (0.15)
Area1—Nova Sussuarana0.54 (0.29)0.43 (0.29)0.15 (0.18)
2—Arenoso0.65 (0.25)0.28 (0.29)0.03 (0.13)
3—Jardim Santo Inacio0.36 (0.35)0.41 (0.31)0.08 (0.18)
4—Calabetao0.67 (0.31)0.21 (0.28)0.01 (0.02)
GenderFemale0.56 (0.32)0.37 (0.32)0.07 (0.15)
Male0.53 (0.34)0.30 (0.29)0.06 (0.16)
Leptospirosis serological statusNegative0.55 (0.33)0.35 (0.31)0.07 (0.16)
Positive0.48 (0.29)0.25 (0.30)0.02 (0.05)
Table 3
Estimated differences (γ) in selection coefficients (β) for each environmental factor using distance-based rasters.

Values >1 represent increasing distance from points of reference.

Community streamOpen sewersDomestic rubbish piles
Estimate95% CIEstimate95% CIEstimate95% CI
Gender*
Male(Ref)–(Ref)–(Ref)–
Female0.980.97, 0.991.041.02, 1.060.990.98, 1.01
Age†
Per year increase1.001.00, 1.001.001.00, 1.001.001.00, 1.00
Leptospira serological status ‡
Negative(Ref)–(Ref)–(Ref)–
Positive0.990.96, 1.011.031.00, 1.071.000.98, 1.02

  1. (Ref) is the referrence group for the Odds Ratio.

  2. *

    Adjusted for age and study area.

  3. †

    Adjusted for gender and study area. Values represent increases by 1 year of age.

  4. ‡

    Adjusted for gender, age, and study area.

Table 4
Estimated differences (γ) in selection coefficients (β) for each environmental factor using 20 m buffers around each point of reference.

Values>1 represent movement within the buffer zone for each point of reference.

Community streamOpen sewersDomestic rubbish piles
Estimate95% CIEstimate95% CIEstimate95% CI
Gender*
Male(Ref)–(Ref)–(Ref)–
Female1.221.02, 1.460.950.80, 1.140.920.66, 1.27
Age†
Per year increase1.001.00, 1.000.990.98, 1.001.000.99, 1.01
Leptospira serological status ‡
Negative(Ref)–(Ref)–(Ref)–
Positive0.890.67, 1.190.640.47, 0.870.850.48, 1.49

  1. (Ref) is the referrence group for the Odds Ratio.

  2. *

    Adjusted for age and study area.

  3. †

    Adjusted for gender and study area. Values represent increases by 1 year of age.

  4. ‡

    Adjusted for gender, age, and study area.

Appendix 1—table 1
Demographic details of excluded individuals due to having less than 50 relocations.
ID (anonymised)Relocation below 50PeriodGenderAge groupLeptospirosis serological status
60TRUE05–21Male50–54Neg
91TRUE05–21Male>55Neg
15TRUE05–09Female45–49Neg
60TRUE05–09Male50–54Neg
81TRUE05–09Female50–54Neg
91TRUE05–09Male>55Neg
108TRUE05–09Female50–54Neg
109TRUE05–09Male20–24Neg
128TRUE05–09Male25–29Neg
129TRUE05–09Male40–44Pos
15TRUE09–13Female45–49Neg
24TRUE09–13Female50–54Neg
60TRUE09–13Male50–54Neg
70TRUE09–13Male35–39Neg
71TRUE09–13Female35–39Neg
76TRUE09–13Male35–39Neg
91TRUE09–13Male>55Neg
108TRUE09–13Female50–54Neg
109TRUE09–13Male20–24Neg
128TRUE09–13Male25–29Neg
129TRUE09–13Male40–44Pos
24TRUE13–17Female50–54Neg
60TRUE13–17Male50–54Neg
71TRUE13–17Female35–39Neg
76TRUE13–17Male35–39Neg
91TRUE13–17Male>55Neg
5TRUE17–21Male35–39Neg
7TRUE17–21Male50–54Neg
18TRUE17–21Male30–34Neg
22TRUE17–21Male45–49Neg
24TRUE17–21Female50–54Neg
27TRUE17–21Male30–34Neg
30TRUE17–21Female>55Neg
60TRUE17–21Male50–54Neg
71TRUE17–21Female35–39Neg
91TRUE17–21Male>55Neg
114TRUE17–21Female40–44Neg

Additional files

MDAR checklist
https://cdn.elifesciences.org/articles/107153/elife-107153-mdarchecklist1-v1.docx
Source data 1

Data used in population-level models, showing step selection coefficients for each individual and each environmental factor.

https://cdn.elifesciences.org/articles/107153/elife-107153-data1-v1.zip
Source code 1

R code used for individual-level step selection functions.

https://cdn.elifesciences.org/articles/107153/elife-107153-code1-v1.zip
Source code 2

R code used for population-level models.

https://cdn.elifesciences.org/articles/107153/elife-107153-code2-v1.zip

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. Pablo Ruiz Cuenca
  2. Fábio N Souza
  3. Roberta Coutinho do Nascimento
  4. Ariane Goncalves da Silva
  5. Max T Eyre
  6. Juliet O Santana
  7. Daiana de Oliveira
  8. Emile V Ribeiro de Souza
  9. Fabiana G Palma
  10. Diogo C de Carvalho Santiago
  11. Priscyla dos Santos Ribeiro
  12. Priscilla Elizabeth Ferreira dos Santos
  13. Hussein Khalil
  14. Jonathan M Read
  15. Cleber Cremonese
  16. Federico Costa
  17. Emanuele Giorgi
(2025)
Using step selection functions to analyse human mobility using telemetry data in infectious disease epidemiology: a case study of leptospirosis
eLife 14:RP107153.
https://doi.org/10.7554/eLife.107153.3