Skip to 0 minutes and 8 secondsHi, and welcome to this video lecture on health system coverage and access to health care. My name is Ente Rood. And I work at the Royal Tropical Institute in The Netherlands. In this lecture, I will explain some of the basic concepts of access to health care and introduce a few methods you could consider when you plan to use GIS to determine health system coverage.
Skip to 0 minutes and 29 secondsSo why do we want to determine access to health care? Well, access to health care is one of the key components of universal health coverage. The goal of universal health coverage is to ensure that all people obtain health services they need without suffering financial hardship when paying for them. All people aspire to receive quality, affordable health care. Access to health services enables people to be more productive and active contributors to their own families and communities. It also ensures that children can go to school and learn. At the same time, financial risk protection prevents people from being pushed into poverty when they have to pay for for their health services out of their own pockets.
Skip to 1 minute and 8 secondsUniversal health coverage is thus a critical component of sustainable development and poverty reduction and a key element of many efforts to reduce social inequities. So equitable health care depends on the ability of each to receive the care they need. This, again, will depend on the vulnerabilities and needs of specific groups of populations and will not be equal across regions and populations.
Skip to 1 minute and 37 secondsSo, from this, we conclude that equitable access is not the same as equal access to health care. Health inequalities can be defined as the difference in health status or in a distribution of health determinants between different population groups, for example, differences in mobility between elderly people and younger populations or differences in mortality rates between people from different social classes. It is important to distinguish between inequality in health and inequity in health. Some health inequalities are attributable to biological variations or free choice, and others are attributable to external environment, the conditions mainly outside the control of the individuals concerned.
Skip to 2 minutes and 16 secondsIn the first case, it may be impossible, or ethically, or ideologically acceptable to change to health determinants, and so the health inequalities are unavoidable. In the second, though, the uneven distribution of, for example, the availability and access to health services may be unnecessary and avoidable as well as unjust and unfair so that the resulting health care inequalities will also lead to the inequities of health.
Skip to 2 minutes and 45 secondsSo, when we speak of access to health care, there are many aspects which come to play. Access to health is not merely defined by the physical accessibility to care but depends on the interplay of the availability, affordability, and appropriateness of the services delivered by the system. On the other hand, the ability to seek, reach, and pay for these services will determine whether populations or individuals have access to services. Here the matter of equity determines whether each individual is equally likely to find the health care which they need. While access to health care is being quantified by means of geographic access, we usually, not exclusively, focus on the availability and physical accessibility of health care services.
Skip to 3 minutes and 28 secondsTo conduct a full analysis, detailed data on the demand side are also needed in order to match the supply and demand.
Skip to 3 minutes and 39 secondsIn the following slide, I'll explain a bit further how DIS can be used to measure access to health care.
Skip to 3 minutes and 49 secondsMeasuring geographic access to health care. So what do we mean by this? Determining the physical access to health cares are commonly measured as the travel time needed to reach a certain facility or specific health care service. In order to do this, we need a contact specific conceptual model of travel, including assumptions on the cost of travel time. Such abstractions raise questions regarding the representational model you would like to use.
Skip to 4 minutes and 16 secondsFor example, people might be traveling in a car at 100 kilometers an hour over a motorway. Or it might be more likely that people will travel by foot across an unpaved track. These are different representational models of the reality. It's important here to realize that any abstraction of reality will influence the outcome and is, therefore, important to identify what differences might exist and how these will influence your outcome.
Skip to 4 minutes and 46 secondsIn our GIS environment, we generally encounter two types of models which can be used to represent the reality. The first one is shown here and is defined by a network of lines which represent routes along which people can move between destinations. Each line segment in this network has a certain travel speed associated with it and, hence, a defined travel time. In this representation, people will move between nodes of the network, and travel times are calculated as the sum of associated travel times. This type of representation is similar to what's used in present day's navigation software, such as Google Maps. The second representation consists of a raster data, which is a continuous service of equally-sized cells.
Skip to 5 minutes and 33 secondsEach cell of the surface has a certain travel speed or cost associated with it. In this model, the travel time is calculated by following the least crossed path between two destinations. In the raster data model, travel occurs through cell to cell movement, wherein a specific course is designated to each cell, representing the time required to tranverse that cell. In most GIS packages, movement occurs in only cardinal directions, which is called rooks, or in both cardinal and diagonal directions, which is a queen's case. However, other software packages might offer more flexible options, such as the knight's case movement. Travel time is calculated using the cell dimensions and the travel speed assigned to the cell.
Skip to 6 minutes and 20 secondsUnlike the network model, the length of the individual steps in the route is based on the cell resolution of the data and, thus, constant throughout the entire raster grid. It is possible to switch from line networks to raster-based networks. Line networks can be converted into raster representations, if required. The resolution of the raster image will, however, have a strong effect on the accuracy of the final outcome. The abstraction process necessitates decision rules for assigning a travel speed to cells in which multiple roads with varying speed limits might fall and, or cells in which no roads are present.
Skip to 6 minutes and 58 secondsWhen the vector roads are converted in cells, the roads cease to exist as unique and individual entities and become a service of travel speeds, instead. In the raster data model, the strict topology that governs real world travel along roads is replaced by predefined, directional movement among cells. Thus, in routing applications, the raster data model has the potential to produce unexpected results. Furthermore, travel time estimates may be either overestimated or underestimated, depending on the geometric complexity of the road network and the cell resolution, as shown in this figure.
Skip to 7 minutes and 36 secondsThere has been considerable debate in scientific literature about which model, factor or raster, should be used on the various circumstances. Although no consensus has been reached on this, it's generally agreed that network models perform better when high quality network data is available in order to estimate travel times with high accuracy. On the other hand, when such data is not available and whether topographic or other landscape factors are believed to influence travel times, raster-based analysis provide a more robust methodology to estimate travel times and costs.
Skip to 8 minutes and 13 secondsThe cost algorithm used to calculate cost distances creates an output raster in which each cell is assigned accumulative cost to the closest source cell. The algorithm utilizes the nodes and links our representation using graph theory. In the node and the link representation, each center of a cell is considered a node, and each node is connected to adjacent nodes by multiple links. When moving from a cell to one of its four directly-connected neighbors, the cost to move across these links to a neighboring node is 1 times the value of cell 1 plus cell 2 divided by 2 in the figure shown.
Skip to 8 minutes and 51 secondsCreating accumulative cost raster using graph theory can be viewed as an attempt to identify the lowest cost cell and added to an output list. It is an interactive process that begins with a source cell. The goal of each cell is to be assigned quickly to the output cost-distance raster.
Skip to 9 minutes and 9 secondsThe final step after calculating an accumulative cost raster is to define those areas which fall within certain travel distance to the health care facility or service of interest. Often, the question is raised what can be considered an acceptable distance. The answer to this question can only be found if a solid understanding exists about the uses and perceptions of health care, which depend on the context within which the analysis is conducted. This brings back to the conceptual model of access to equitable health care, which showed that in order to have good access to health care, the service of supply and demand system side should match and agree with the demand for services coming from individuals living within the population being served.
Skip to 9 minutes and 51 secondsThis is the end of this video lecture. I would like to thank you for your attention. I hope you enjoyed the lecture. Further on in this course, you will practice the use of QGIS to calculate the raster based cost distances. I hope to see you there.
Universal health coverage
Access to health care is one of the key components of universal health coverage. The goal of universal health coverage is to ensure that all people obtain the health services they need without suffering financial hardship when paying for them. Access to health is not merely defined by the physical accessibility to care but depends on the interplay of the availability, acceptability, affordability, and appropriateness of the services delivered by the system. Within GIS the focus lies on the availability of health care facilities and the measurement of physical accessibility.
Measuring geographic access to health care
Physical accessibility to health care can be measured in two different ways, using network analysis and raster modelling. This further elaborates on what has already been discussed during week 1.
Network analysis measures distance or travel time over a network that represents the infrastructure that people use to reach health care. People can walk over a road, use a bus or train, or use a combination of these. In a network model, the railway tracks, the bus routes and the roads are all represented as line segments.
The raster model divides the complete study area into small cells (normally square) that represent a small part of the actual area. The cells are like the tiles in your bathroom and the total area is covered by these cells. We measure accessibility by measuring the distance between each cell and the health care facility. In this model we assume people can walk everywhere (not only via the roads). In some countries a raster representation is more realistic as people walk to health centers using small footpaths, or even travel across fields. In this example we will use the raster mode to represent accessibility.
Outputs that can be generated
When conducting accessibility studies, we can generate different types of output: 1. Distance layer showing the distance from a cell to the nearest health facility, 2. Catchment of a health facility, and 3. Route between a person (patient) and a health facility.
© KIT (Royal Tropical Institute), Amsterdam, the Netherlands