Near-accident reporting

3 Near-accident reporting 13.3.3.1 Why report near accidents  

The results of more recent studies contradict this hypothesis. Studies of the consequences of incidents have shown that the severity is dependent on the energy involved (Shannon and Manning, 1980 Salminen et al., 1992). Fatalities, for example, are more likely when the victim is struck by a falling object or hit by a moving vehicle. It follows that there is no generally valid ratio between the number of accidents and the number of near accidents at a workplace. This is because the distributions of incidents differ between different types of workplace, mainly depending on the types of energies involved in production. 

Near accidents with the potential of resulting in severe injuries are important sources of information in accident control. Near-accident reporting also has other advantages (Carter and Menckel, 1985 Van der Schaaf et al., 1991)  

These results have been interpreted as proof of the benefits of near-accident reporting. There are other possible explanations. A change, for example, in top management attention may explain both effects. There is also another concern. The plant did not experience a similar reduction in the frequency of severe accidents, indicating that near-accident reporting will not necessarily pick up the causes of these types of accidents. 

In Section 13.3.2 we identified different incentives and disincentives in the reporting of accidents. In principle, the same types of incentives and disincentives influence the reporting of near accidents. This may be illustrated by an example from civil aviation. 

---

The two reasons mentioned above are of a fundamental nature they aim at an improvement of the quality of (near-) accident reports. The third reason is more practical and should increase the quantity of reported cases. 

In action research, scientists participate in the introduction of changes in the SHE information system of a company. They evaluate how these changes affect the behaviour of its organisation and the company s accident rates. Many studies report reductions in accident rates following the introduction of such measures as improved accident investigation and workplace inspection routines and introduction of near-accident reporting (Adams et al., 1981 Menckel, 1990 Komaki et al., 1978). 

In evaluation research, scientists study effects of programme changes made inside companies, without being involved in the changes themselves. An aluminium plant had introduced a documented SHE information system, involving improved routines for accident and near-accident reporting, workplace inspections and follow-up of SHE results. An evaluation study showed that the introduction of internal control, in combination with other measures such as improved maintenance, contributed to an improved control of production. This resulted not only in substantially reduced losses from accidents but also in reduced overall operational expenditures (Kjellen et al., 1997). 

In our model of a SHE information system in Chapter 1, we identified four different subsystems, i.e. data collection, data analysis or processing, a memory and distribution of information. We will here first focus on the data-collection subsystem. Chapter 12 presents an overview and Chapters 13 and 14 give details about different methods of data collection in accident and near-accident reporting and investigation, workplace inspections and SHE audits. We will then proceed in Chapter 15 to go through some basic principles for the establishment of a memory (database) on accident risks. After that, we will review different methods for data analysis based on data from accidents and near-accidents in particular. In practice, it is not always easy to separate data collection from data analysis or processing, because these activities often interact. 

At the beginning of the 1980s, the author conducted an evalnation of six large Swedish companies accident and near-accident reporting and workplace-inspection rontines (Kjellen, 1982). The study was performed before any of the companies had been affected by the recent trends within corporate SHE management. The data that actually were collected in these activities depart from the ideal scope according to Figure 12.1. There were barriers and filters in the data collection at the six companies that suppressed certain types of data on accident risks. 

Only two of the six companies conducted near-accident reporting on a routine basis. At those two companies, the near-accident reports mainly contained data on technical incidents, deviations and causal factors. The lack of data on human errors was explained by the fact that the reporter s anonymity could not be guaranteed and by fear of disciplinary action. 

Near-accident reporting had improved dramatically, from in the order of one near accident per lost-time accident in the two Swedish companies reporting near accidents to in the order of 100 near accidents per lost-time accident in the oil company. The near-accident reports of the oil company were still dominated by technical events. The reporting of unsafe conditions was a new phenomenon. A closer analysis of the accident, near-accident and unsafe condition reports revealed the intentions of the reporters. The oil company had introduced an efficient system for follow-up of reports with remedial action. This fact was known among the employees, who used the near-accident and unsafe-condition reporting systems as a means of getting working-environment problems solved. 

Creation of positive safety attitudes and alertness. Severe accident occurrences will shape the accident perceptions of those concerned. A well-functioning programme for accident and near-accident reporting, investigation and follow-up will support the development of a positive climate for safety-related behaviour and awareness. 

Near-accident reporting creates an increased knowledge and awareness among the employees about accident risks. 

Experiences from industry are used as evidence on the positive effects of near-accident reporting. Figure 13.4 shows a typical example. In the course of a seven-year period, the yearly number of lost-time accidents went down by almost a factor of ten at an industrial plant. During the same period, the near-accident reporting frequency increased by about the same factor of ten. 

Figure 13.4 The development in the frequency of lost-time and first-aid accidents and near accidents at an industrial plant after the introduction of new near-accident reporting routines.

A successful reporting of human errors has been accomplished by giving pilots and air-traffic controllers benefits when reporting such events. Most important is the fact that they receive immunity from punitive actions as soon as they have filed a near-accident report. This limited immunity covers negligence but not criminal acts such as drug trafficking. They also receive feedback on the results of their reporting. 

The lessons from these examples are that the different incentives and disincentives must be carefully analysed when designing near-accident reporting routines. The willingness of the workers to report near accidents is dependent on the extent to which there is openness and trust between workers... 

It is difficult to measure near-accident reporting reliability, because the definition of what is a reportable near accident is fuzzy. It is mainly left to the discretion of the reporter. Instead, we apply measures of the reporting propensity, i.e. the employees willingness to report. 

The reliability of near-accident reporting may be improved by applying the same general principles as those for accident reporting. Measures directed especially at the improvement of near-accident reporting in industry include ... 

Efficient follow-up of reports with remedial action and feedback to the reporters with visible results. The employees will use the near-accident reporting system for the purpose of achieving improvements in their working environment if they know that their reports are taken seriously. 

Analyse factors that affect near-accident reporting in a positive and negative way seen from the point of view of (a) the supervisors and (b) the operators. 

Propose measures to improve near-accident reporting. 

The employees self-reporting of unsafe conditions by filling in a form is complementary to the traditional informal notification of the immediate supervisor. This self-reporting follows the same basic principles as those of near-accident reporting, and usually the same form is employed. Information on deviations and shortcomings in design and routines at the workplace are... 

Experiences of the British Airway Safety Services (BASIS) near-accident reporting system support the use of a form for self-reporting with open-ended questions according to the first principle (Reason, 1997). BASIS first tried a form with questions concerning types of human errors and contributing factors, where the answers were given in a multiple-choice format. The resulting data suffered from poor validity and reliability. 

Enhanced quality of accident and near-accident reports and remedial actions by ensuring input and quality checks from people at different levels of the organisation and geographical locations. 

Accident reports may easily be distributed to a large number of people as e-mails. This distribution will add to the general information load within the organisation and may create problems of overload and inexpedient reactions. The problem may occur in large companies and in companies with an efficient reporting of near accidents. It is important to establish routines for the distribution of accident and near-accident reports inside the companies, based on the criticality of the events and the receiver s actual needs. 

A procedure for accident and near-accident reporting and investigation... 

The routines for accident and near-accident reporting and investigation need to be documented in a procedure. Below is a proposed outline of such a procedure. 

A common solution is to allow for widespread permission to read the accident and near-accident reports with the exception of those fields containing personal data on the victim. The possibility of commenting on or changing the document should be restricted to persons with special authority. A dedicated document owner should be given the responsibility to approve all changes in the report itself. 

A bar chart showing the distribution of events by type of event and year for the last three years, i.e. a bi-variate distribution. It showed a decrease in the number of LTIs during the last year (9), as compared to the previous two years (14 and 17 respectively). The near-accident reporting frequency developed in a positive direction. 

SHE performance measurement utilises all four sub-systems of a SHE information system. The performance indicators apply data from the different sources presented in Chapter 12 such as accident and near-accident reporting, workplace inspections and SHE audits. The data is analysed and presented in a specific way, enabling comparison with goals, previous results, results of other companies, etc. 

SHE performance indicators based on near-accident reporting... 

Accident and near-accident reports from the actual area (or from a similar area at another plant if the analysis is carried out during the design of new workplaces). 

Traffic-conflict techniques are used to improve the statistical basis for decisions on safety measures, i.e. the aim is similar to that of near-accident reporting (cf. Chapter 13). There is one important difference, however. Whereas near accidents are usually reported by the involved personnel, traffic-conflict techniques utilise specially trained observers located directly in the traffic. There are differences between the traffic-conflict techniques developed in different countries. A comparison between eight different European techniques showed that they all evaluated the severity of the conflict in a similar way although there were differences in the scales applied (Englund et al., 1998). The Swedish traffic-conflict technique, which is a representative example, applies two central concepts ... 

This data set appUes to reports on accidents and near accidents. Decisions must be made at an early stage on the totality of types of reports that are to be stored, even when the development work proceeds in steps and the immediate focus is on accident and near-accident reports. A generalised data model should be apphcable to the storage of reports on the following SHE activities ... 

---