Table 1.
List of variables, their spatial and temporal scales, and transformations.
Table 2.
Age and gender of study patients.
(Q1 and Q3 indicate the first and third quartiles, respectively).
Table 3.
Distribution of distances (in km) to the nearest major road, by diagnosis type.
Counts of villages are from within 3 standard deviational ellipses (SDEs) of all LP villages (referred to as the “study area” in text). In some cases, multiple patients came from the same village meaning that counts of villages will be smaller than counts of total patients. (Q1 and Q3 indicate the first and third quartiles, respectively).
Fig 1.
Spatial distributions of the home villages of study patients, for A: all study patients, B: study patients with JEV infections, and C: with cryptococcal infections, D: scrub typhus infections, E: with dengue virus infections, F: with leptospiral infections, and G: with murine typhus infections. SDDs and SDEs are weighted by case numbers, with some patients coming from the same village. Maps were created using QGIS version 3.4.9. All layers were created by the authors of this manuscript.
Fig 2.
Environmental indices for villages with study patient homes for the duration of the study period (January 2003 through August 2011) for all study patient villages, non study patient villages in the study area, and for major diagnoses (JEV = Japanese Encephalitis virus; Crypto = cryptococcal infection; ST = scrub typhus; MT = murine typhus; dengue = Dengue virus; lepto = Leptospira spp. infection).
Bar values are mean values and the error bars are 95% confidence intervals, using the t-distribution. NFI values here have a constant (0.25) added to them for visualization only.
Fig 3.
Environmental indices for study patients by major diagnosis and at different times leading up to the date of admission.
JEV = Japanese Encephalitis virus; Crypto = cryptococcal infection; ST = scrub typhus; MT = murine typhus; dengue = Dengue virus; lepto = Leptospira spp. infection. Bar values are mean values and the error bars are 95% confidence intervals, using the t-distribution. NFI values here have a constant (0.25) added to them for visualization only.
Table 4.
Logistic regression and model adjusted odds ratios (AOR) for village level analysis of LP villages.
Villages in which a CNS infection was diagnosed were compared to villages with no diagnosed infections.
Table 5.
Logistic regression and model adjusted odds ratios (AOR) for village level analysis of JEV villages.
Villages in which a JEV infection occurred are compared to villages with no diagnosed CNS infections.
Table 6.
Mixed effects logistic regression and model adjusted odds ratios (AOR) for individual level analysis.
Individuals with JEV infections were compared to individuals with other CNS infections.