"Doubts over road casualty figures" - FinancialTimes
Discussion
Although the inconsistency of police v hospital figures has been remarked on previously, according to this report (from today's FT), the Statistics Commission has now called attention to the anomaly.
Doubts over road casualty figures
By Simon Briscoe, Statistics Editor
Claims that the casualty toll on Britain’s roads is falling may be based on questionable statistics, raising the prospect that government safety measures such as speed cameras and traffic calming are not working.
At the centre of the controversy is the gap between information collected by the police at the scene of an accident and figures for road accident victims released by hospitals.
A “briefing note” published on Friday by the Statistics Commission, the government watchdog, voices “particular concern” that it is the police statistics that are used to judge whether the government is meeting its targets on reducing road accidents. The Office of National Statistics said that Karen Dunnell, the National Statistician had “expressed her concern” over the issue.
According to the accident-scene statistics that ministers favour, the number of people killed or seriously injured in Britain has fallen from the mid-90s average of 47,700 to 34,400 in 2004.
The final figures for 2005 are expected to fall to about 32,200 when they are published next week. Under the police figures, about 54 people per 100,000 are killed or seriously injured, down from 80 per 100,000 a decade ago.
However, figures for hospital admissions suggest that the figures have remained broadly constant in the past decade at about 90 per 100,000.
If the hospital statistics were used as the benchmark, the government would be missing its 2010 target of a 40 per cent reduction in road deaths and serious injuries by a long way.
The Department for Transport said it was “continuing to examine the issue”.
Let's hope that this doesn't end up being just another Government spin-cycle whitewash of the truth.
Without wishing to sing the praises of the police sourced stats which I suspect do have their problems, given the state of systems in general in the NHS I really cannot see that they are likely to be more accurate. Indeed quite the opposite.
For a start the NHS can claim costs from the insurance companies. Combine that with the people who walk in to A&E claiming whiplash (some no doubt real, some not) a day or so after an accident and in need of a record on which to base a compo claim and I think there is scope for doubt about the veracity of the numbers.
As with all stats the cut off points may be significant. Live for 30 days or less after the event and the casualty becomes a fatality. Live for 31 days and the victim will be a 'Serious Injury'. And then of course there is always the question about whether a death related to the original incident.
Such things can always be debated but may not mean much in terms of measurement for relative performance purposes. The sort of things we seem to be fixated by these days.
So many variables change year by year that I am not entirely sure the comparisons of any figures are very valid.
For a start the NHS can claim costs from the insurance companies. Combine that with the people who walk in to A&E claiming whiplash (some no doubt real, some not) a day or so after an accident and in need of a record on which to base a compo claim and I think there is scope for doubt about the veracity of the numbers.
As with all stats the cut off points may be significant. Live for 30 days or less after the event and the casualty becomes a fatality. Live for 31 days and the victim will be a 'Serious Injury'. And then of course there is always the question about whether a death related to the original incident.
Such things can always be debated but may not mean much in terms of measurement for relative performance purposes. The sort of things we seem to be fixated by these days.
So many variables change year by year that I am not entirely sure the comparisons of any figures are very valid.
The NHS are very likely to be closer to the truth. There are examples known to the ABD of political pressure operating on completion of police stats forms. VH has of course asked for names 
and cast doubt on the claim when the obvious reply was forthcoming. In addition, what qualifications and reliability do police have, compared to medical staff, for judging severity of injury?
Hospitals have no particular axe to grind whereas the politicsed police definitely do. Our road safety policies are failing and wrapped up in spin to save incompetent politicians' skins. The politicised senior BiB dancing to the piper's tune share blame.
and cast doubt on the claim when the obvious reply was forthcoming. In addition, what qualifications and reliability do police have, compared to medical staff, for judging severity of injury? Hospitals have no particular axe to grind whereas the politicsed police definitely do. Our road safety policies are failing and wrapped up in spin to save incompetent politicians' skins. The politicised senior BiB dancing to the piper's tune share blame.
Edited by turbobloke on Sunday 24th September 00:45
So do we think there is no politicising and no fanatical axes to grind in the NHS?
Do we think that the consistency of reporting through NHS systems and the following of the STATS19 reporting guidelines will be better and more consistently followed by the NHS bureaucracy?
Will the NHS figures, going back 15 years or so, further if possible, be anywhere near consistent given changes in policy and practice which have occurred in those years 9in both sources).
More to the point, are any of the retrievable figures worth anything at all frm an analysis point of view, and if they are why is policy dominated by such a small factor in the overall picture?
Do we think that the consistency of reporting through NHS systems and the following of the STATS19 reporting guidelines will be better and more consistently followed by the NHS bureaucracy?
Will the NHS figures, going back 15 years or so, further if possible, be anywhere near consistent given changes in policy and practice which have occurred in those years 9in both sources).
More to the point, are any of the retrievable figures worth anything at all frm an analysis point of view, and if they are why is policy dominated by such a small factor in the overall picture?
I think the very fact that doubts are being aired about the accuracy of the statistics and which source is better shows how complicated the real issue is and how inappropriately simple the Government's campaign is.
I had a conversation with a road safety engineer recently who was talking about the pending A77 SPECS review. It compares the last 2 years to the previous 5 years and does not take into account factors such as the 6-7 mile stretch that has been converted from 4 lane single carriage way to Motorway, the removal of several 90 degree junctions onto NSL sections and improved lighting.
Nor does it take into account improved passive safety devices on cars or response times of the emeregency services, which I would expext have improved due to the upgraded road design. Throw in this debate over whose statistics to use and I would say it is a useless comparison.
I had a conversation with a road safety engineer recently who was talking about the pending A77 SPECS review. It compares the last 2 years to the previous 5 years and does not take into account factors such as the 6-7 mile stretch that has been converted from 4 lane single carriage way to Motorway, the removal of several 90 degree junctions onto NSL sections and improved lighting.
Nor does it take into account improved passive safety devices on cars or response times of the emeregency services, which I would expext have improved due to the upgraded road design. Throw in this debate over whose statistics to use and I would say it is a useless comparison.
Plotloss said:
flemke said:
A sad state of affairs when we don't trust the Police to produce accurate statistics about such an important matter, we don't trust the NHS to do so, and we especially don't trust the Government.
Sadly, they've made their bed with that one.
KB_S1 said:
I think the very fact that doubts are being aired about the accuracy of the statistics and which source is better shows how complicated the real issue is and how inappropriately simple the Government's campaign is.
I had a conversation with a road safety engineer recently who was talking about the pending A77 SPECS review. It compares the last 2 years to the previous 5 years and does not take into account factors such as the 6-7 mile stretch that has been converted from 4 lane single carriage way to Motorway, the removal of several 90 degree junctions onto NSL sections and improved lighting.
Nor does it take into account improved passive safety devices on cars or response times of the emeregency services, which I would expext have improved due to the upgraded road design. Throw in this debate over whose statistics to use and I would say it is a useless comparison.
Don't know the present state of affairs, but not that long ago a TrafPol friend told me that, at the time, the RTA forms that his constabulary used for recording the causes of an accident did not have a category of "Speeding". I had a conversation with a road safety engineer recently who was talking about the pending A77 SPECS review. It compares the last 2 years to the previous 5 years and does not take into account factors such as the 6-7 mile stretch that has been converted from 4 lane single carriage way to Motorway, the removal of several 90 degree junctions onto NSL sections and improved lighting.
Nor does it take into account improved passive safety devices on cars or response times of the emeregency services, which I would expext have improved due to the upgraded road design. Throw in this debate over whose statistics to use and I would say it is a useless comparison.
They had "Excessive Speed", which had to cover doing 10 mph on black ice, doing 65 on a DC in dense fog, tailgating (at too high a speed to be able to brake in time), et al. In other words, "Excessive Speed", as in movement, could have been said to be a cause in the overwhelming majority of RTAs.
In reporting the national statistics for road safety, "Excessive Speed" would be transmuted into "Speeding".
Edited by flemke on Sunday 24th September 11:37
Plotloss said:
Short term, yes.
Lack of faith in a government is a far more damning long term trait though.
The downside is of course though that successive government very rarely reverse the policies of their predecessors.
Lack of faith in a government is a far more damning long term trait though.
The downside is of course though that successive government very rarely reverse the policies of their predecessors.
Of course it is not the Police per se that we do not trust, nor the NHS.
Rather, it is that successive Governments have sought to bend to their will the Police, the NHS and other public services that ought to be protected from the perversions of self-interested, mendacious politicians.
LongQ said:
So do we think there is no politicising and no fanatical axes to grind in the NHS?
Not in this particular arena. Others maybe, not this. LongQ said:
Do we think that the consistency of reporting through NHS systems and the following of the STATS19 reporting guidelines will be better and more consistently followed by the NHS bureaucracy?
I think I believe medics and don't believe the police, with good reason. fluffnik said:
Plotloss said:
The downside is of course though that successive government very rarely reverse the policies of their predecessors.
I, for one, would support a party that stood on a platform of only repealing laws and removing regulation.
Anyone else?
Oh yes; of a certainty.
fluffnik said:
Plotloss said:
The downside is of course though that successive government very rarely reverse the policies of their predecessors.
I, for one, would support a party that stood on a platform of only repealing laws and removing regulation.
Anyone else?
How about a platform of committing to expunge or repeal ten or twenty laws for every one that it enacted?
turbobloke said:
LongQ said:
So do we think there is no politicising and no fanatical axes to grind in the NHS?
Not in this particular arena. Others maybe, not this. LongQ said:
Do we think that the consistency of reporting through NHS systems and the following of the STATS19 reporting guidelines will be better and more consistently followed by the NHS bureaucracy?
I think I believe medics and don't believe the police, with good reason. You sound very certain about that.
What are your experiences of the police and the NHS that lead you to be so certain in each case?
My basis is as follows:
- reading and reviewing past reports from or involving or commissioned by both
- information from inside sources (police)
- a common sense view that medics can diagnose SI in KSI better than police
- an appraisal of previous statements from police and NHS top bods, not least regarding the politicisation of public services we both mention
What is your basis for an opposing view?
- reading and reviewing past reports from or involving or commissioned by both
- information from inside sources (police)
- a common sense view that medics can diagnose SI in KSI better than police
- an appraisal of previous statements from police and NHS top bods, not least regarding the politicisation of public services we both mention
What is your basis for an opposing view?
I would like to add an additional piece of information:-
There are a lot less police patrols now than 10 years ago...they've been replaced with GATSO's and talivans. As a result, the police MAY not come to all accidents now, even if someone is mildly injured (bear in mind the 'SI' stat includes simple cuts and mild concussion...neither of which I consider serious and certainly wouldn't waste a copper's time with). So are the police still collecting the stats with the same degree of rigour...after all, if they've a target to reduce KSI, how better than by not attending all accidents reported to them?!?
My gut feel says the NHS stats will be more reliable.
There are a lot less police patrols now than 10 years ago...they've been replaced with GATSO's and talivans. As a result, the police MAY not come to all accidents now, even if someone is mildly injured (bear in mind the 'SI' stat includes simple cuts and mild concussion...neither of which I consider serious and certainly wouldn't waste a copper's time with). So are the police still collecting the stats with the same degree of rigour...after all, if they've a target to reduce KSI, how better than by not attending all accidents reported to them?!?
My gut feel says the NHS stats will be more reliable.
havoc said:
I would like to add an additional piece of information:-
There are a lot less police patrols now than 10 years ago...they've been replaced with GATSO's and talivans. As a result, the police MAY not come to all accidents now, even if someone is mildly injured (bear in mind the 'SI' stat includes simple cuts and mild concussion...neither of which I consider serious and certainly wouldn't waste a copper's time with). So are the police still collecting the stats with the same degree of rigour...after all, if they've a target to reduce KSI, how better than by not attending all accidents reported to them?!?
My gut feel says the NHS stats will be more reliable.
There are a lot less police patrols now than 10 years ago...they've been replaced with GATSO's and talivans. As a result, the police MAY not come to all accidents now, even if someone is mildly injured (bear in mind the 'SI' stat includes simple cuts and mild concussion...neither of which I consider serious and certainly wouldn't waste a copper's time with). So are the police still collecting the stats with the same degree of rigour...after all, if they've a target to reduce KSI, how better than by not attending all accidents reported to them?!?
My gut feel says the NHS stats will be more reliable.
Most collisions aren't reported by trafpol, they are reported by other uniform officers with traffic dealing with the more serious ones. All personal injury collisions coming to the notice of Police must be reported (it's one of the criteria), whether the Police attend or they don't attend the scene & someone subsequently reports them.
Of course if people have a minor collision, exchange details, then think there might be a bit of money in making an injury claim so go to the hospital or their doctor after the event (without informing the Police), they may not be aware of those.
Gassing Station | Speed, Plod & the Law | Top of Page | What's New | My Stuff