Cross-sectional studies
Cross-sectional studies look at something that is going on at a particular moment in time. If you do a cross-sectional study, you would collect information about the current situation, for example the current state of a particular disease in a particular section of your community, or the current exposure of a certain group of individuals in your community to an infection. Therefore, in cross-sectional studies the information on the possible cause and the possible effect is collected at the same time. However, the problem with cross-sectional studies is that there is no way of knowing whether the possible cause is really responsible for the effect you measured – or could it be the other way around? The following example illustrates what we mean by this problem.
A cross-sectional study has been set up in your village to see if children who are underweight are more likely to get measles than other children of normal weight. The study finds that more children who are underweight have measles than those who are of normal weight. Does this mean that underweight children are more likely to get measles, or that children become underweight because they have been ill with measles?
You would not know whether they had become infected with measles because they were underweight, or whether they were underweight because they had been ill with measles. A cross-sectional study cannot resolve this uncertainty because the information was collected at just at one point in time.
Cross-sectional studies aim at describing the distribution of certain variables in a study population at a certain time, for example in the last month. Examples might include studies of:
- Physical characteristics of people: e.g. surveys of the occurrence of tuberculosis, HIV, malnutrition, high blood pressure, diabetes, anaemia, etc.
- Physical characteristics of the environment: e.g. evaluation of latrine usage, access to safe water supply, distance from the nearest health facility, etc.
- Demographic or socio-economic characteristics of people: e.g. their age, education, marital status, number of children and income.
- The behaviour or practices of people and their knowledge, attitudes, beliefs or opinions, which may help to explain that behaviour: e.g. attitudes to contraception, or harmful traditional practices, healthcare-seeking behaviour if a child has measles or diarrhoea, etc.
- Adverse events that occurred in the population, e.g. the occurrence of road traffic accidents or burns, and natural disasters like flooding or drought.
Cross-sectional studies usually cover a selected sample of the population. For example, it would be impossible to study all children with measles in a country, so a sample of children with measles is chosen randomly without any bias in the selection. Data are collected at the same time from the sample of children on the risk factor or characteristic (in the example above, it is being underweight) and the disease condition (having measles). If a cross-sectional survey covers the total population of a country it is called a census. A cross-sectional survey of a sample of the population is cheaper than doing a census because it consumes fewer resources. A cross-sectional survey may be repeated at a later date in order to measure changes over time in the characteristics that were studied.
Case Study: A cross-sectional study on household characteristics
A questionnaire was sent to 250 individuals who had visited a health facility. They were asked information about the layout of their home, such as type of floor covering and number of windows. The investigators wished to determine the number of people affected with fever and cough in the previous month and whether there was any relationship between household characteristics and the risk of having these symptoms, which may possibly indicate tuberculosis.
What feature of the case study above indicates that the study design is cross-sectional?
A population sample was surveyed and asked questions about diseases suffered during a short period of time (the last month). So this is a cross-sectional study design.