We have reviewed our 'Tools' pages and this new page brings together the most useful
statistical tools for public health analysis.
DSR Batch Calculation spreadsheet This spreadsheet allows the user to calculate directly
standardised rates with 95% confidence intervals for many areas or groups in one go. The
spreadsheet uses the European Standard Population for standardisation or you can enter
other standard population data for comparison.
Commonly used public health statistics and their confidence intervals
This APHO spreadsheet provides Excel formulæ for calculating the more common types
of statistic used within public health intelligence. These include rates, proportions,
means and age-standardised rates and ratios. Formulæ are also given for calculating
confidence intervals for all of these statistics.
The traditional method of calculating confidence intervals produces negative CIs for
small numbers and symmetrical intervals at small numbers which are incorrect.
For this reason Altman et al recommend the score test also known as Wilson's method
instead of the traditional method. It gives wider confidence intervals than exact methods
and is thought to be statistically more robust. The APHO spreadsheet uses Wilson's
method for confidence intervals around rates.
Funnel plot resources on the APHO website allow users to enter their own data
to produce funnel plots.
Health inequality tools include tools for calculating Gini coefficient and the Lorenz
curve, and the slope index of inequality.
Life expectancy calculator enables users with death and population data for small areas
(such as LA or ward level) to calculate life expectancy figures.
Decrement life tables What happens to life expectancy at birth when death rates
from accidents, or cancer or heart disease fall? Use this spreadsheet to find out
the impact on life expectancy of changes in mortality - useful for modelling, planning
or health inequalities work. With thanks to Steve Salzano.
Tools for manipulating datasets include a tool from Tableau to turn Excel datafiles into
raw data and the Data Wrangler from Stanford University which makes it simpler
to tidy and clean datasets.
The West Midlands Public Health Observatory has a useful tools collection, including an
Excel spreadsheet to generate spine charts, and add-ins to calculate confidence intervals
for rates, proportions and percentages. There is also an add-in to calculate
directly standardised rates and confidence intervals.