Statistics - How to find them and use them.
Guidance for health and safety representatives
Statistics are an important tool for any health and safety representative. They can tell us a lot about what is happening in an industry or area and give us the information that we need about injuries or illnesses that we need for prevention.
Statistics do not need to be complicated. They are just numbers, and a health and safety representative does not need to know about any of the complex ways that numbers are collected and analysed to be able to use them.
Statistics will be able to tell you how many people are injured in your workplace every year, or how many people are off sick and for how long. This will allow you spot problem areas and make proposals for how they can be reduced. You can also compare figures with previous years to see if there are any trends on new problems arising.
These are a few sources for simple straightforward figures about injuries and illnesses that you may find useful.
The first stop will be your employer. They have to give health and safety representatives information on injuries and sickness if you request it and also let you inspect any documents that relate to health and safety. Although there are exceptions for some personal information and health records you should be able to get:
- RIDDOR Reports - These are details of any injury and illness reports that have been made to the HSE under the RIDDOR Regulations.
- Accident Book entries - You can examine them at any time although the employer can remove information that would identify the individual if they ask.
- Sickness absence records - Again you cannot get the records of any individual but you should get reports on the level of sickness absence and the causes where they are known. In larger workplaces you should also get it broken down by work area.
- Reports - If the employer produces any reports from their records they normally have a duty to give you a copy if you ask for it. This should include any regular reports they receive from any occupational health provider you have.
Where the employer has more than one site it is important to get information across the organisation so that you can compare performance. Safety Committees are a useful way of getting information on the whole organisation as they should receive regular reports.
Information will also be of use if you can compare it to the rates of injury or illness in similar type employers or just against the national average. This information may be quite easy to get. The HSE site has a huge number of statistics on both injuries and occupational illnesses. These are collected from a range of studies but mainly the RIDDOR reports that employers make and a large government survey.
The main page is at http://www.hse.gov.uk/statistics/
There you will find the current annual statistics but also breakdowns by industry, occupation, region and disease or type of injury. The site also has European figures for comparison.
There is even a tool that lets you break down some of the information and make your own report at https://handson.hse.gov.uk/hse/public/home.aspx
For figures on violence you should try the British Crime Survey. The most recent one is at http://www.homeoffice.gov.uk/publications/science-research-statistics/research-statistics/crime-research
The official site for government statistics is the Office of National Statistics at: http://www.statistics.gov.uk/default.asp
For European statistics there are two main places.
The first is the European Agency for Occupational safety and Health at: http://osha.europa.eu/en/statistics
The other is the massive Eurostat which contains figures on just about everything. Go to: http://epp.eurostat.ec.europa.eu/portal/page/portal/statistics/themes
Click on Health (health and safety at work) to get into the health and safety information.
There are very few reliable sources of comparable international data. The International Labour Organisation keeps a number of databases but, because international statistics are often collected in very different ways, very little of it will be of use to health and safety representatives. http://www.ilo.org/global/statistics-and-databases/lang--en/index.htm
However do not forget that your union health and safety section may have information on injury or illness rates in your sector. Have a look at their website. Plus of course you can always get your own figures by running your own survey on issues like stress or bullying.
Statistics can show you very clearly what the level of injury and disease is, how it has changed over a period of time and how likely a person is to be killed or injured in a workplace. They can also be used to confuse.
The first question that any health and safety representative has to ask is 'are you comparing like with like'. Comparing the rate of reported injuries in agriculture with the number of injuries reported in manufacturing would be very misleading if farmers report a far lower proportion of their injuries.
Also there is a difference between numbers and rates (see below). There may be a decline in the number of injuries in an industry, but if the number of workers in that industry has fallen by a half, while the number of injuries have only been cut by a quarter it actually means that the rate of injury has gone up by 50%.
Another example of how statistics had make things look safer than they are is using a short time period or single occurrence. An employer may claim that the chance of a person falling from a tower crane when climbing up and down is only 1 in 100,000. That sounds pretty low but if a person is going up and down twice a day every day for a year they will have made the journey at least 1,000 times. If they work as a crane operator for 25 years then the odds begin to look a bit less good. It would mean that every crane operator would have a 1 in 4 change of being killed by a fall during their working life. Even a one in a million chance can be pretty worrying if the activity that is being done often or there are a lot of people doing it.
Official statistics can be misleading as well. Not deliberately, but because of the difficulties in calculating numbers in some areas, especially diseases which can be caused by a range of different things, not just work. The TUC has always claimed that the official statistics for the number of cases of cancers caused by work is too low. This is because it is usually impossible to say what the cause of any individual cancer is and often they are not blamed on work. Many lung cancers for instance are automatically blamed on smoking despite the fact the worker may have worked with a wide range of dusts at work that can also cause cancer.
A few terms you may come across.
This is not intended to be comprehensive but these are some terms that are often used in health and safety statistics.
Mean. The mean is what most people think of as the average, where you add up all the numbers and then divide the total by the number of numbers.
Median. The median is the middle point in a list of numbers. To find the median, the numbers have to be listed in numerical order,
Rate. This is a number measured against another measure, usually a period of time, so it could be a figure like 200 a year, but it can also be measured against another number, such as an exchange rate
Ratio. This is the same as a fraction. If the ratio of A to B is 2 to 3, that means for every 2 of A there is 3 of B. It can be written as a ratio of 2:3 or as a fraction 2/3.
Prevalence and Incidence. You will often find both these given for occupational disease statistics. Prevalence refers to actual number of cases (the number actually living with the disease) while incidence refers to the new cases in a period - usually a year.
Confidence levels and confidence limits. These are level of confidence you can have in the results, so a report could say that 33% of the workforce witnessed bullying at work in the last year but it attaches a 95% confidence level to the figure of 33% plus or minus 3%. That is, it very likely that between 30% and 36% will witness bullying with 33% being an average. The confidence limits would therefore be 30% and 36%. The bigger the survey the more confidence there is so the limits get narrower.
Issued: 14 July, 2011