Planning surveys

There is a balance to be struck between the amount of information you would like to know and the resources you have to collect the data, clean the data, analyse the data and report findings. Based upon what is already known and identified information gaps it should be possible to frame an informational objective for your potential survey such as β€œwe wish to understand the disease experience and treatment need of care home residents in this area”.

With that statement in mind you can consider the potential sampling frame, clinical disease indicators/tools, oral health impact indicators and other information which will help you address the objective. With a computer there is less work on data cleaning, analysis and reporting for a narrow data set for many individuals than for a wide data set for a few.

Sample size needs to be large enough to draw conclusions with reasonably narrow confidence intervals. Consider the analyses you would like to undertake (e.g. by quintile of deprivation) as part of planning sample size.

Typically for a clinical survey data to be collected includes demographic data, data related to risk of developing the disease of interest, disease data, disease impact data and possibly treatment need opinion.

The demographic data such as age, and postcode allows you to ensure that the individuals are those you intended to sample and to be able to profile the sample geographically and potentially socio-economically.

Disease data is useful for highlighting those with high disease experience. Alongside monitoring of disease trends this assists in prioritising future prevention activity and planning treatment services.

Impact data from socio-dental indicators, cross referenced with other disease information can assist in identifying priorities for action. Treatment need opinion can provide a feel for an appropriate service to commission to address the priorities.

Use of validated indicators previously used in other surveys enables wider comparison and will assist extrapolation of findings to other areas should any novel data be collected.