Learning lessons from the investigation of "accidents"

Should the requirement on employers to investigate injuries, ill health and dangerous occurrences be more explicit? The HSC thinks so. Chris Dyer reports.

Unless the causes of workplace injuries and ill health are understood, lessons are not learned and measures are not taken to prevent their recurrence. Although many employers do investigate incidents, this is not universal and practices vary across different industries and businesses.

In 1999, the HSC consulted on whether or not to make more explicit in law requirements on employers to carry out these investigations. Having considered the responses, the HSC has now published proposals to amend the law to introduce an explicit duty to investigate accidents.1 The consultation document also describes HSE-commissioned research into employers' current investigation of incidents. This points to considerable shortcomings, with many employers mistakenly believing their actions to be adequate.

The responses to its discussion document, A new duty to investigate accidents (HSB 275), showed nine in 10 respondents thought the HSC should take further action to encourage employers to investigate accidents, and that two-thirds thought this should be achieved through a new legal duty. If there were to be a change in the law:

  • three-quarters of all respondents thought the change should involve a new duty to investigate accidents with a corresponding duty to revise the risk assessment; and

  • nearly two-thirds of respondents favoured achieving this change by amendment of the Management of Health and Safety at Work Regulations 1999 (MHSWR).

    Even respondents who believed the existing implicit duties were sufficient, and did not want a change in the law, generally supported a change to the MHSW Approved Code of Practice (ACoP) or more general guidance.

    The proposed "simple" duty, which received broad agreement from respondents, would require:

  • employers to investigate all accidents so as to learn lessons and be able to take corrective action;

  • the investigation to be proportionate to the scale or complexity of the incident; and

  • the results to lead to the amendment of relevant risk assessments.

    Nine in 10 respondents who expressed a view said that the "simple" duty would fit well with existing practice and procedures at their workplace. There was not, however, the same degree of support for an elaborated duty that additionally included:

  • production of a report on the accident;

  • the involvement of safety representatives;

  • employers following directions on investigation by the enforcing authorities;

  • sending the report to the enforcing authority on request; and

  • investigation of all incidents, rather than those that are currently statutorily reportable.

    Although none of these additional elements was supported by a majority of respondents, the HSC's proposals in the consultation document are wider in scope than the simplest duty option. They do not, however, include all of the additional duties set out in the discussion document, notably the last two in the list above. Consultation on the HSC's proposals ends on 3 September.

    PROPOSALS FOR A NEW DUTY

    The HSC proposes the investigation duty be introduced through new Regulations, the Management of Health and Safety at Work (Amendment) Regulations 2001. These Regulations would amend reg. 3 (risk assessment) of the MHSWR so as to introduce a new schedule on the investigation of accidents, dangerous occurrences and diseases into the Regulations.

    The MHSWR are not the only legislative vehicle that could be used to introduce a new duty. The HSE has drafted the proposed Regulations so that they can be slotted as a Schedule into other Regulations. If incorporated into the Reporting of Injuries, Diseases and Dangerous Occurrences Regulations 1995 (RIDDOR), the HSC believes the Schedule would make clear the linkage between reporting and investigation and also clarify the link with employers' overall management of risk. Alternatively, the proposals could form a stand-alone set of Regulations. The HSC says there is a strong case for following the MHSWR route because this promotes investigation as an integral part of health and safety management requirements, but it will listen to views on other legislative options.

    HSC rejects all-incident approach

    The definitions in the draft schedule establish a clear link with RIDDOR by requiring investigation of all accident injuries, dangerous occurrences and diseases reportable under RIDDOR. This limitation was supported by a third of respondents, whereas 48% of respondents thought that the scope of the new duty should go beyond RIDDOR to encompass all accidents. But the HSC believes that "accident" needs to be qualified if the seriousness of the duty is to be clear and not an unnecessary burden. It would be difficult to define "accident" to cover all workplace unplanned events with the potential for causing harm, without including many minor incidents that probably should not be within the scope of the legal duty. The HSC is also concerned that a duty to investigate all incidents will lead to an employer response that is bureaucratic and process-dominated.

    Near misses

    Some respondents suggested the new duty could apply to "near misses", as well as to RIDDOR-defined dangerous occurrences. The HSC is cautious about extending the range of the duty, as drawing up a legal definition would be difficult and employers might be uncertain about coverage. Employers could again be led into introducing overly bureaucratic systems, and the extended criteria would not sit easily with the restriction of the scope of the duty under RIDDOR. Instead, the HSE proposes publishing guidance setting out good practice that employers could follow in investigating "near misses", in addition to the proposed new duty.

    Diseases

    The new duty to investigate would apply to diseases reportable under RIDDOR. Currently, if an employer receives, in respect of an employee, a doctor's written diagnosis of an occupational disease listed in Schedule 3 of RIDDOR, it must promptly report the case to the enforcing authorities. Before making the report, the employer must consider whether the ill employee's current job involves the corresponding work activity specified in the Schedule. This in itself may involve making enquiries.

    The HSC recognises there are difficulties associated with the investigation of workplace health, that there is significant underreporting and that in investigating the cause of some diseases or conditions (and when developing solutions), employers may need help. Despite these acknowledged difficulties, responses to the discussion document showed strong support for an explicit duty on employers to investigate cases of disease reportable under RIDDOR.

    Proportionate response

    The duty to investigate will fall on the responsible person, as defined in RIDDOR, who has the legal duty to report injuries, dangerous occurrences and diseases under those Regulations. The HSC believes this person will be best placed to carry out an investigation and will, in most cases, be the employer.

    In carrying out an investigation, the duty-holder will be required to take steps that are practicable and proportionate to the scale of the incident, to investigate the underlying causes, and to take the findings into account in the review of any risk assessment. The investigation must start as soon as possible, and in any event within three days of notification to the enforcing authority. This means any investigation of a death, major injury to people at work (and, in certain circumstances, not at work) and a dangerous occurrence should normally start within four days.

    In the case of over-three-day injuries, the duty-holder would again be required to start an investigation as soon as possible, and in all circumstances within three days of the event having to be reported to the enforcing authority. This means that the investigation should be started at the latest within 13 days of the injury occurring. Investigation of a disease should commence at the same time that it is reported to the enforcing authority.

    In practice, investigations should start as soon as possible after the reportable event has occurred, and in most cases immediately after the event. The HSC has included time limits in the draft proposals - based on the RIDDOR notification and reporting timescales - as a means of ensuring that there is an obligation to commence an investigation within a reasonable period.

    The self-employed, who will also be the duty-holder, may not always be able to commence an investigation within the prescribed time limits and in this situation the duty allows for the investigation to start as soon as practicable.

    Enforcing authority investigations

    When HSE inspectors and local authority officers decide to investigate incidents themselves, they will have the authority to direct the duty-holder about their investigation process to ensure confusion does not arise from parallel investigations. But the HSC does not intend that duty-holders will seek advice routinely from inspectors on how to go about investigations in particular cases - the supporting guidance will provide details on carrying out an investigation.

    Duty-holders will have to involve the appointed safety representatives or the workforce's elected representatives in any investigation. The HSC believes a team approach, involving line managers and safety representatives and supported where appropriate by safety professionals, can only lead to more informed investigations, which should help to establish underlying causes.

    Record-keeping

    To demonstrate compliance with the requirements, employers will need to keep records. The HSC wants to ensure that a record is kept when an investigation has been carried out, proportionate to the scale of the incident, and that any changes to the risk assessment(s) are recorded. Records will need to be retained for three years - consistent with the period specified in the Social Security (Claims and Payments) Regulations 1979 for the retention of accident-book entries. The aim is to avoid overly bureaucratic record-keeping requirements, but to ensure there is an identifiable audit trail for inspectors to follow when necessary.

    One alternative would be for the duty-holder to keep a record of the significant findings of an investigation only in companies where there are five or more employees, which would be consistent with the other MHSW record-keeping requirements. There is also a case for all duty-holders, irrespective of size, keeping a record of the significant findings of an investigation, as this would be consistent with the record-keeping requirements of RIDDOR.

    A more extensive option would be to require all investigation findings to be recorded and, where appropriate, acted upon to prevent a recurrence. As well as amendment of the risk assessment, this could also include changes in the arrangements for planning, organising and monitoring the health and safety arrangements in the workplace (as required by reg. 5 of the MHSWR).

    Risk-assessment revision

    The proposed Schedule requires that the results of an investigation be brought to the attention of any person who has carried out a risk assessment relevant to the work or activity associated with the injury, dangerous occurrence, or disease. This is necessary as the duty-holder, under these proposals, may not always be the person also required to carry out an MHSWR risk assessment. It is also possible that they are not the duty-holders under risk-assessment provisions in other health and safety law, such as the Control of Substances Hazardous to Health Regulations 1999. There could also be occasions where the duty-holder is responsible for the MHSWR risk assessment, but where it might be appropriate for another duty-holder's risk assessment to be revised as well. Where this exchange of information is needed, it may be that the easiest way to provide it is through an investigation report.

    Employee representatives

    The HSC proposes to make it a requirement to inform safety representatives, or the workforce's elected representatives, of the findings of the investigation. This duty to inform would fall on anyone who has carried out a risk assessment relevant to the work activity associated with the accident, dangerous occurrence or disease, and who has been informed of the results of the investigation by the responsible person. The HSE also wants to see employers involve safety and workforce representatives in the investigation so as to ensure the active involvement of the workforce throughout the investigation process.

    A further option would be to place a legal requirement on the responsible person to inform those directly involved in the incident of the investigation findings. In some cases, this information could be provided through the safety/workforce representatives, but the HSC considers it good practice for employers generally to do it.

    Insurance companies

    Another option would be to introduce a legal requirement on duty-holders to pass the investigation findings to their employers' liability insurer. This might help to speed up the civil claims process and to protect the employer's interests under the Civil Procedure Rules. On the other hand, it could be argued that the insurance industry already has good channels of communication with companies and would find it difficult to process copies of all investigation findings. In drafting the Regulations, the HSC/E would have to take account of the Government's policy to minimise, where appropriate, the regulatory impact of new proposals on business.

    Regulatory impact assessment

    The HSE estimates that the cost to business of introducing the new duty would be £18 million each year in present-value terms. The average annual costs per employer (for all investigations they would face under the Regulations) are:

  • £2,400 for large enterprises;

  • £120 for medium-sized enterprises; and

  • £10 for small employers.

    The costs would vary from industry to industry, and would double if the duty was to cover all incidents, rather than only those reportable under RIDDOR. The new duty would need to lead to a reduction in injuries and ill health of 3% for benefits to employers to balance the costs. A reduction of 1% would be necessary for benefits to society to balance the costs.

    There are, however, gaps in the information needed to quantify the costs and benefits of introducing the new duty, particularly around the amount of time and resources currently devoted to investigations. The HSE's assumptions are based mainly on research it had carried out during the summer and autumn of 2000. This found that:

  • 42% of incidents took less than five hours of investigation time;

  • 35% took between five and 20 hours;

  • 18% took over 20 hours; and

  • in general, investigation time increases with the severity or potential severity of the incident.

    BARRIERS TO QUALITY

    The research also drew conclusions about the main barriers to improving the overall quality of incident investigation. It found that, in most companies, proactive risk assessment, performance monitoring, audit and near-miss reporting are not coupled to incident investigation processes, which typically is viewed as a standalone process. Although each process potentially provides feedback into risk assessment and the development of risk-control measures, and can conceptually be seen as a continuum (see figure on p.8), only a quarter of companies use the output from their incident investigation to update risk assessments. The problem is compounded by the processes being often based on different models, with little consistency in the taxonomy or approach used, further undermining the effectiveness of any feedback loop.

    The researchers found that approaches to incident investigation ranged from the totally unstructured to quite formal and well-supported systems. The research highlighted:

  • the importance of the individual driving the investigation;

  • the focus on incident reporting rather than investigation; and

  • the general failure to distinguish or understand the difference between immediate and underlying causes.

    The research concludes that these results suggest two deficiencies that hamper improvements in the quality of incident investigation:

  • a significant lack of awareness and understanding of system-based, rather than the more traditional accident, models. System-based models examine all potential contributory factors, including organisational factors - whereas traditional models focus on the individual concerned and the most immediate cause; and

  • a lack of practical support, in terms of usable systems and documentation, to encourage the application of such models.

    Other factors that the researchers identify as impeding high-quality incident investigations include competence and complacency. They found that many of the individuals responsible for leading or supporting incident investigation are not dedicated health and safety professionals and have minimal training in incident investigation. This is a particular problem in small companies where incidents - and even near-misses - are rare events, which makes it difficult to maintain appropriate awareness and skills in incident investigation. The study concludes that this reinforces the need for comprehensive advice, guidance and documentation to encourage incident investigation.

    The researchers also found an overall level of satisfaction with current approaches. They examined a series of company investigations and found that often the findings are limited to the immediate causes of the incident, even though most of the companies felt their approach had led them to discover the underlying causes of the incident. Most companies also feel they have a structured process that identifies the underlying causes of accidents, although they are less confident about ill health. Consequently, most companies believe that any forthcoming legislation will make little or no difference to them. The researchers conclude: "This finding implies the need for a significant shift in awareness, so that more companies recognise the need for, and value in, moving from an approach that focuses on merely describing the chain of events that happened to one that tries to identify why these events happened."

    Liability issues

    Although the researchers did not directly address the sensitive issue of potentially unfair and punitive employee claims, they believe the results show that legal pressures could be a significant issue in incident investigation. They suggest that in an increasingly litigious business climate, there is a powerful disincentive to identifying and documenting the management factors that have contributed to an incident.

    Nearly a quarter of the companies questioned said that they would change their investigation process if it were likely that an incident could lead to a legal claim. Often these changes, which seemed primarily to be driven by the requirements of insurers, resulted in what was described as a "more thorough" investigation, but not necessarily a more system-based investigation.

    The researchers also found a notable discrepancy between the profile of causes identified by companies and the profile of subsequent interventions. Typically, companies tend not to identify or document organisational issues as immediate or underlying causes of incidents but, paradoxically, are prepared to make secondary recommendations for interventions that directly address these issues. This suggests, the researchers conclude, a tacit acknowledgement that these are significant contributory factors, but a reluctance to formally highlight management or organisational deficiencies.

    The researchers believe it is critical that the HSE use its influence to decouple the issues of blame and liability from the process of objective investigation. If this problem is to be overcome, they say, the legal system must be seen to encourage, not penalise, companies that adopt a system-based approach to investigations. It is suggested that one practical measure to achieve this would be the introduction of relevant insurance incentives.

    The researchers conclude that the HSE could take a strategic role in this process by generating an active, high-level dialogue between industry, legal and insurance representatives, with the aim of identifying and committing the parties to a coherent approach to the challenge. In last year's Revitalising health and safety strategy document (Employers face major health and safety at work shake-up), the HSC committed itself to considering how best to involve the insurance industry more closely in raising health and safety standards, even though the industry has seemed to be reluctant to take a more significant role. Once again, the role of insurers in health and safety management has been raised as an issue and the HSE hopes that this consultation document and the associated research will prompt a new debate amongst stakeholders as to what that role might be in the future.

    1"Proposals for a new duty to investigate accidents, dangerous occurrences and diseases", CD169, HSE Books or www.hse.gov.uk/condocs . Comments on the consultation document must reach Neville Higham at HSE, Rose Court, 2 Southwark Bridge, London SE1 9HS, by 3 September 2001.

    The draft Regulations

    The draft Regulations amend the Management of Health and Safety at Work Regulations 1999 by introducing into them a new Schedule concerning the investigation of accidents, dangerous occurrences and diseases reportable under the Reporting of Injuries, Diseases and Dangerous Occurrences Regulations 1995 (RIDDOR).

    The provisions contained in the Schedule:

  • require the "responsible person" - the person required to make the RIDDOR report - to carry out an investigation into any RIDDOR-reportable accident, dangerous occurrence or disease;

  • empower an inspector to direct how the investigation is to be carried out;

  • require the responsible person to permit safety representatives and representatives of employee safety to participate in the investigation;

  • require the responsible person to keep a record of the investigation and to inform any person who has carried out a risk assessment relevant to the work or activity concerned of the result of the investigation; and

  • require a person who has carried out a relevant risk assessment to notify safety representatives and representatives of employee safety of the result of the investigation.