The Role Of Disease Mapping In Health Boards

Disease mapping can play an important role in monitoring the health of a community. Plotting new cases of disease on a map is a frequently used technique for monitoring the spread of infectious diseases. The dot map drawn by John Snow in the 1850s is possibly the most renowned example,1 but countless maps exist in offices of the Directorates of Public Health which are charged with monitoring the spread of infectious diseases such as dysentery, meningitis and flu. Dot maps facilitate the search for links between cases. In the case of Snow, dot maps indicated the Broad Street pump as a potential source for the outbreak of cholera. Dot maps of cases of dysentery can elucidate whether the cases are related by residential proximity, by attendance at a particular school, or by some other type of community activity. But dot maps are not restricted to the monitoring of disease. They may be used effectively for monitoring, for example, the uptake of vaccinations or health service usage, or for locating black-spot areas for road traffic accidents. Figure 7.1 shows the sites of road traffic accidents to child pedestrians for one year within a small area of Dundee City. The association between accident location and proximity to schools is clearly demonstrated around several schools shown on the map.

Dot maps of non-infectious diseases, while valuable, are generally not as helpful as dot maps of infectious diseases for establishing the precise cause of disease. Many non-infectious diseases, such as cancer, cardiovascular and cerebrovascular diseases, have multifactorial causes, which makes it

Hull Street Map
Figure 7.1. Dot map showing locations of fatal road traffic accidents to child pedestrians in Dundee City. Map copyright Geographia Ltd, adapted by permission

extremely difficult to establish cause and effect. When used with such diseases, dot maps are valuable for generating hypotheses about disease causation or for identifying clusters of disease but they do not provide information about specific disease aetiology.

Where the aim is to elucidate causation, the non-infectious diseases which are suited to dot mapping are those which are primarily caused (or triggered) by one factor. For instance a study in Barcelona looked at the onset of an epidemic of asthma by time and by geographical clustering and was able to identify the unloading of soybean in the city's harbour as the cause of the asthma epidemic days.2 In another study, plotting the residential addresses of cases of lung cancer was one technique by which an association was demonstrated between residential proximity to a steel foundry and lung cancer.3 (Lung cancer is one of the very few cancers which is caused primarily by a single factor, namely tobacco smoking. It is estimated that tobacco smoking causes 80-85% of lung cancer cases.)

Dot maps require careful interpretation. Because they represent cases spatially, it is essential that the person interpreting the significance of the spatial pattern is familiar with the underlying population structure as this allows an estimate of the population at risk. The interpreter requires local knowledge about the population density, the age and sex structure that the map represents. The higher the population density the higher the number of expected cases. There are several ways in which the demographic characteristics of the population may be assessed. The approach depends on the level of interpretation required. The cases could be converted to incidence rates. But this approach negates the use of a dot map as it forces the use of some (arbitrary) denominator, which is commonly based on a postcode or enumeration district. Another approach is to re-draw the map subdivided into areas that are proportional in size to the population density, the so-called population-based map. Each case, represented by one dot, is placed evenly throughout the area (Figures 7.2a and 7.2b).

It is generally more helpful to monitor the course of diseases with multifactorial causation adjusted for the main confounding factors. Dot maps are not ideal for this purpose as they cannot be adjusted easily. The most typical characteristics, which need to be adjusted for, irrespective of the disease, are the age and sex distributions of a community. Additional factors (such as deprivation indices, smoking and nutritional status) which are disease specific may then need to be considered. A more appropriate way to monitor diseases with multifactorial origins would be to display them using some summary statistic such as the standardized mortality ratio. The standardized mortality ratio is typically adjusted for age and sex but can be adjusted also for disease-specific factors. Where indicated the standardized mortality ratio may be truncated to represent particular groups within a community (for instance, the school ages, working population, young adults, retired population or the very elderly). Of course, caveats concerning the use of SMRs (of4'6) must also be considered here.

After deciding whether to use dot maps or adjusted summary statistics for

Figure 7.2a. Dot map showing Salmonella cases on a routine geographical map. Redrawn from Dean (1976)4 by permission

displaying the geographical characteristics of a disease, the next step is to decide on the geographical unit of measurement. To enable a comprehensive review of a health status of a community the data need to be presented in sufficiently small geographical units to allow any potential variations in health to be observed. But the units need to be large enough to enable some sort of statistical interpretation. Health Boards in the UK have access to routinely collected data from the decennial censuses. This information allows the adjustment by many characteristics (for example deprivation indices, ethnicity, age and sex) and is presented at the level of the postcode sector. Table 7.1 describes the hierarchy of units in UK postcodes.

The subdivisions of postcodes are designed to maintain roughly equal populations within each category. For instance postcode units typically represent about 16 households irrespective of which part of the country is represented by the postcode. Because of this principle, postcodes cover very different geographical areas. Densely populated communities will have more postcodes than sparsely populated communities. Postcode sectors

Redrawn from Dean (1976)4 by permission
Table 7.1 Postcode hierarchy

Example

Approximate number

Postcode area

FK

1 (i.e there is 1 area starting FK)

Postcode district

FK1

21

Postcode sector

FK1 2

42

Postcode unit

FK12ES

> 6000

(Table 7.1) are thus only a reasonable unit for monitoring the health status of communities. By their nature they are more suited to monitoring the health of cities than rural communities. Particular care must be taken if the aim of health monitoring is to assess the impact of pollution on the health of a community. The use of postcode data in such instances may seriously weaken the observed effect because of dilution of the health effect.

Was this article helpful?

0 0
Coping with Asthma

Coping with Asthma

If you suffer with asthma, you will no doubt be familiar with the uncomfortable sensations as your bronchial tubes begin to narrow and your muscles around them start to tighten. A sticky mucus known as phlegm begins to produce and increase within your bronchial tubes and you begin to wheeze, cough and struggle to breathe.

Get My Free Ebook


Post a comment