Killed or Seriously Injured

From Wikipedia, the free encyclopedia

Killed or Seriously Injured (KSI) is a standard metric for safety policy, particularly in transportation and road safety. As the name implies it is the total figure for people killed or seriously injured over a period of time.

In casualty statistics there is a hierarchy of reliability. Figures for fatalities are usually (in industrialised countries at least) highly reliable: few if any fatalities go unrecorded. Measures of seriously injured are less reliable, and are vulnerable to arbitrary classification by non-medical professionals (e.g. police officers attending incidents). Measures of slightly injured are considered highly unreliable, largely due to under-reporting (many minor injuries are self-treated and never reported at all). ^[1]

Although the seriously injured figure is significantly less robust than the fatality figure, fatality figures are typically very small and subject to considerable stochastic variation (for example, British rail passenger fatalities are typically in single figures but one train crash might kill some tens or even hundreds of people). By adding together the two figures, with seriously injured typically being at least an order of magnitude larger, a figure is arrived at which is statistically significant, less subject to this noise, and therefore more amenable to analysis. It also collects together all those casualties which are nominally serious enough to be of legitimate public concern, an important factor in setting policy.

KSI trends should always be viewed with some scepticism: figures are vulnerable to changes in reporting procedures and practices for handling serious incidents, and occasionally to the agenda of those collecting the statistics. ^[2] In some cases these changes are noted, an example being the British Department for Transport's Road Casualties Great Britain series, which has a discontinuity in 1996 due to changes in reporting procedures.

The tendency to use KSI in many widely-cited safety metrics has led to an unjustified public perception that these figures are unbiased.