LTIFR – A Measure of Safety Performance?
By Tom Gardener
Look at almost any company’s annual report and you’ll find some comments about what the business is doing to improve their safety performance – usually measured by the number of injuries they’ve had that year and more specifically how many of those injuries have required time off work for recovery. So, we get told what a good job they’re doing to prevent injuries and they measure they’re performance by how many times they’ve failed. Bizarre… definitely but not quite as bizarre as the wide spread use of the Lost Time Injury Frequency Rate or LTIFR which is an almost universal statistic used to report on safety performance and compare the performance of one company or part of a company with another.
What is the LTIFR? It’s a measure of the number of lost time injuries (LTI) per so many hours worked. For example, an LTIFR of 3 may mean 3 LTIs for every million hours worked or it could be 10 000 or 50 000 hours. So what does that tell the reader or the business? Think about it. If this years LTIFR is less than last year’s does this mean the company has a more effective safety program in place? Or, conversely, does a higher LTIFR indicate a less effective program? If one company has an LTIFR of 2.5 and another company an LTIFR of 5 does this mean that the latter company is less committed to safety or has a worse safety program?
Let’s break this down a little further. The almost universal definition of an LTI is a work related injury that results in the injured employee having at least one full shift off work – by definition this excludes the shift on which the injury happened. Converting this to a million hour LTIFR involves multiplying the number of LTIs by 1 000 000 and diving the result by the number of hours worked. So the LTIFR is influenced not only by the number of LTIs that happened but also by the number of hours worked. So 2 companies with the same number of LTIs but substantially different hours will end up with totally different LTIFRs – the larger the number of hours the lower the LTIFR.Have a look at the table below to see the calculation for different combination’s of hours and LTIs (hours across the top, LTIs on the left side):
…100K…200K…300K…400K…500K…600K…700K…800K…900K…1000K
1…10……..5…….3.3……2.5…….2……1.67…..1.43….1.25….1.1 1…….1
2…20…….10……6.67…….5……..4……3.33…..2.86…..2.5…..2.22…….2
3…30…….15…….10…….7.5…….6……..5……..4.29….3.75….3.33…….3
4…40…….20…..13.33……10…….8…….6.67…..5.71……5……4.44…….4
5…50…….25…..16.67….12.5……10……8.33…..7.14….6.25….5.55…….5
So if we return to our 2 companies, company A had an LTIFR of 2.5 and company B had 5. As you can see from the table there are a number of different combination’s that can produce these results and without knowing either the hours worked or the number of injuries, the LTIFR by itself doesn’t really say anything and reveals nothing about either company’s safety performance.
So, does the LTIFR reveal anything else? Not really. Since only one day off work is required for an injury to be classified as an LTI, the LTIFR doesn’t tell us how serious the injuries were or whether they resulted in a day, a week, a month or even a year off work. Since a fatality would not be considered an LTI they are similarly excluded from the calculation.
If the LTIFR reveals nothing of value why is it used so much? Quite simply because it’s easy. It’s easy to calculate and it’s easy to comprehend. As well, because it’s a frequency rate it provides an appearance of a standardised statistic that can be used to compare against other businesses or different parts of the same business. Generally, businesses that measure their safety performance using the LTIFR are businesses that have only a superficial commitment to improving safety performance. If they were serious they would quickly come to realise that this statistic provides them with nothing to evaluate their safety performance and it would be quickly abandoned in favour of other more meaningful sources of information.
So the next time you hear someone bragging about their great safety performance because of the declining LTIFR think about what you’re being told and ask some questions to see what they’re doing, how they’re doing it and what (if any) other measurements they are using to assess performance.
Tom Gardener has worked as a full time health and safety professional for more than 30 years in both government and private sectors. This has enabled him to gain a great deal of knowledge and experience in the practical management of health and safety in modern workplaces.
If you have a comment on this article or wish to read and/or comment on his other articles visit his blog at http://www.thesafetybloke.com
Article Source: http://EzineArticles.com/?expert=Tom_Gardener
http://EzineArticles.com/?LTIFR—A-Measure-of-Safety-Performance?&id=5752258
I’m sorry, I am not a safety professional. I’m just a guy trying to understand these concepts.
I don’t quite get this. The LTIFR is similar to a negative binomial distribution, if you try and reverse the equation, it is basically saying that it took x amount of manhours for an injury to happen. For work environments like in construction, the greater the manhours went on without any injuries will make it less likely a chance and more likely due to the safety management system. How is this not a measure of performance? I would really appreciate a clarification on this.
Zack, the issue is not in HOW it is measured, but in WHAT is measured. The argument is therefore that an injury, or lack of injury, is not a measure of how safe a workplace is. It is only a measure of how many injuries were recorded. If you would use the same measurement for measuring how many tonnes of cement was poured (constructively, not wasted) or how many bricks were laid, this would have been a measurement of performance. On the other hand, an argument can be made that LTIFR is a measure of failure rate. However, even then, it does not measure SAFETY performance, it measures INJURY performance. While it is true that the prevention of injuries is one of the reasons for safety management, the existence of injuries tell you almost nothing about safety. I trust this answers the question to some extent. (Please note, the capital letters were just used to highlight, not to shout.)
Visit a construction site and see how many walking wounded and makeshift bandages you see yet absolutely no injuries recorded. If you swim across a crocodile infested river and make it to the other side without an incident was it therefore a safe thing to do? Imagine if the only measure of a good marriage was zero affairs? Then we’d see the redefinition of affair – it was only a one night stand or a contractor so doesn’t count…..worse still, what would that measure do to the marriage in terms of trust, fear, control, checklists, rules, audits etc? Scary huh? Safety is no different – we need much higher order goal and measures for something so important. Of course, the higher you go the harder is to measure and control and safety could never abide that
The application of the hierarchy of controls to marriage or long term relationships opens up an interesting can of worms:
Elimination (It’s quite extreme if you or your partner are always arguing but the consequences are somewhat daunting)
Substitution (Quite expensive, especially if you get caught)
Isolation (Sleeping in separate bedrooms)
Engineering controls (Changing door locks)
Administrative controls (AVOs)
PPE (I won’t delve any further)
Hah! Sensational. Everyone one of those controls comes with unintended byproducts. Heinrich could have had fun with marriage data as well
The elephant in the room is luck. Noone told me 45 years ago when I made the leap of faith (like everyone who gets married), that it all rests on the shoulders of Fortuna.
45 years – you must have been a perfect husband 🙂
Ha, Dave. Fundamental Attribution Error. There is no link between 45 years and whatever is attributed.
This issue is not what or how its measured but the philosophy of measurement itself. It’s such a selective process and most of what is made of data is attributed, there is actually no connection between the data and what is asserted. Another one of the grand myths of the industry.