Last updated: December 26, 2019


Authors: Gabriel Carrasco-Escobar*, Edgar Manrique, Kelly Tello-Lizarraga, J. Jaime Miranda

* corresponding author


Aim

This study sought to estimate the travel time to the most proximate health facility in rural and urban areas across heterogeneous land coverage types in Peru as a means to help resources prioritization, disease surveillance, as well as prevention and control strategies.

Methods summary

  • Multiple sources of geospatial data were fitted with a novel cloud-based geospatial modeling approach (Weiss et al., 2018) to produce high-resolution (30 m) estimates of travel time to the most proximate health facility across the country. These estimates were then compared between urban and rural settings and across 16 major land coverage types in Peru.

  • The estimation of travel time were conducted in Google Earth Engine (GEE) (Gorelick et al., 2017). A surface grid was constructed using the information about land coverage, road infrastructure, and river network.

  • To calculate the travel time from the villages to the most proximate health facility, the cumulative cost function was used in GEE to generate the accessibility map.

  • The minimum travel time to the most proximate health facility was computed for each pixel in the grid, then the median travel time was summarized in a 500m-radius from the geolocation of each city or village.

  • The computed travel time was then summarized per district, province or department; by urban/rural areas; and across 16 major land coverage types defined by the Ministry of Environment. Urban/rural status was defined based on the MODIS land coverage satellite images.

Figure 1

Figure 1. Study area. A) Major ecological areas (coast, andes, and jungle) in Peru. Solid lines represent the 25 Departments (administrative level 1). B) Spatial location of primary, secondary, and tertiary health facilities. C) Spatial location of villages. Maps were produced using QGIS, and the base map was obtained OpenTopoMap (http://www.opentopomap.org), under CC BY427 SA 3.0.

Figure 2

Figure 2. Country-wide map of travel time to health facilities for 2018. District-level average travel time to each category of healthcare facilities. A) Primary healthcare. B) Secondary healthcare. C) Tertiary healthcare. Color scale in logarithmic scale.

Figure 3

Figure 3. Distribution of travel time to most proximate health facility. Estimates across the 16 eco-regions defined by the Peruvian Ministry of environment and rural/urban settings for primary, secondary and tertiary healthcare. Y-axis in logarithmic scale.

Figure 4

Figure 4. Median travel time to each health facility category relative to the proportion of population with unsatisfied basic needs per department. Y-axis in logarithmic scale.

Travel time to health facilities

For this study, we gathered geo-referenced data on 145,134 villages (Figure 1B) and 8,067 health facilities (Figure 1C) across the 1,722 districts in the Peruvian territory. The health facility density (number of health facilities divided by the total population) in Peru was 2.58 per 10,000 inhabitants with variations between major ecological areas, from 1.35 in the coast, 4.56 in the highlands, to 5.21 in the jungle.

Friction and travel time maps were reconstructed in Google Earth Engine using the described local datasets at a spatial resolution of 30 meters per pixel. Country-wide median travel time from each village to the most proximate health facility varies according to category: primary healthcare = 39 minutes (IQR=20 – 93), secondary healthcare = 152 minutes (IQR=75 – 251), and tertiary healthcare = 448 minutes (IQR=302 – 631). Importantly, maximum travel time reached 7,819, 12,429, and 35,753 minutes for primary, secondary, and tertiary levels, respectively (Figure 2).

Urban/rural and ecological settings

High heterogeneity was observed in contrasting land coverage areas. The median travel time was 5.3 fold higher in rural (85 minutes; IQR=11–7,819) than in urban settings (16 minutes; IQR = 11–835) to a primary healthcare facility; 3.2 fold higher in rural (226 minutes; IQR = 11–12,429) than in urban settings (70 minutes; IQR = 11–3,386) to a secondary healthcare facility; and 2.4 fold higher in rural (568 minutes; IQR = 11–35,753) than in urban settings (235 minutes; IQR = 11–10,048) to a tertiary healthcare facility. A larger variation in travel time to primary healthcare was observed in rural compared to urban areas, and conversely, a larger variation in travel time to tertiary healthcare was observed in urban compared to rural areas (Figure 3). The district197 level stratified averages in Figure 2 show that there were also strong heterogeneities within major ecological regions. The north-east part of the Amazon Region, which borders with Colombia and Brazil, presented the largest country-wide travel times to the most proximate health facilities. The largest travel times to the most proximate health facilities within the Highlands Region was observed in the southern areas of the Andes, and in the coast was observed in the southern coast. Contrasting distributions of travel time to the most proximate health facility was observed between the 16 eco-regions defined by the MEnv (Figure 3).

Travel time to health facilities relative to UBN

When the travel time to most proximate health facilities was distributed relative to the proportion of the population with unsatisfied basic needs at department level (administrative level 1), a positive trend was observed (Figure 4). The slope of this relation was increased in geographical accessibility to tertiary health facilities in comparison to primary or secondary health facilities.

Limitations

It is important to highlight that the analysis conducted in this study did not take into account variability due to climatic factors that may prevent displacement to health facilities (i.e. floods or landslides). However, Highlands and Jungle areas are more prone to this kind of natural disaster, leading to a conservative estimation of travel time in these areas. Traffic, which may greatly influence the estimates in the large cities, was not considered in the analysis and potentially cause an underestimation of the travel time to health facilities. In addition, seasonal variability may greatly affect some displacement routes such as rivers; however, only navigable rivers were considered in this approach and the availability to displace through this rivers are less affected by seasonality. Another important consideration about the least-cost-path algorithm used in this analysis is that we infer the lowest travel time boundary to reach a health facility. This consideration relies on the assumption that the villagers opt for this route despite the cost and danger of the route in addition to its availability, as explained above.

In addition, the data reported here was generated at a meso-scale, with a spatial resolution of 30 meters. At this scale and resolution, some important details could be lost and affect the travel time estimations. For instance, in some settings the travel time might be increased due to meandering rivers or roads that follow the morphology of the terrain. The model assumes that transit flows in a direct manner, meaning that zigzagging routes may cause our approach underestimate the real travel time to reach a health facility. Despite these possible shortcomings, the proposed approach provided conservative yet useful estimates of travel times to health facilities that are important for planning of prevention and control strategies for multiple health268 related events.