ILO is a specialized agency of the United Nations
[an error occurred while processing this directive]

272nd Session
Geneva, June 1998


Report of the Meeting of Experts on Labour Statistics:
Occupational Injuries

(Geneva, 30 March-3 April 1998)

1. In accordance with the decision taken by the Governing Body at its 270th Session (November 1997), and in preparation for the Sixteenth International Conference of Labour Statisticians, to be held in October 1998, the Meeting of Experts on Labour Statistics: Occupational Injuries was held in Geneva from 30 March to 3 April 1998, chaired by Dr. E. Yränheikki (Government expert, Finland). Eighteen experts were invited to the meeting: six nominated after consultations with governments, six after consultations with the Employers' group, and six after consultations with the Workers' group of the Governing Body. The meeting was also attended by nine observers from intergovernmental and non-governmental international organizations.

2. The agenda of the meeting, as approved by the Governing Body, was as follows:

1. Methods of measurement of occupational injuries.

2. Classifications of occupational injuries.

3. The report of the meeting is appended.

4. The Governing Body may wish to take note of the report, in particular the conclusions in paragraphs 80 and 81.

Geneva, 22 May 1998.

Point for decision: Paragraph 4.



International Labour Organization


Meeting of Experts on Labour Statistics

Geneva, 30 March - 3 April 1998

Report of the Meeting of Experts
on Labour Statistics:

Occupational Injuries



Annex: List of participants


1.  At its 270th Session (November 1997), the Governing Body of the International Labour Office (ILO) decided to convene a Meeting of Experts on Labour Statistics. The Meeting was held in Geneva from 30 March to 3 April 1998.

Agenda of the Meeting

2.  The agenda of the Meeting was the following:

I.Methods of measurement of occupational injuries.

II.Classifications of occupational injuries.


3.  Eighteen experts were invited to the Meeting, six following consultation with governments, six nominated by the Employers' group and six nominated by the Workers' group of the Governing Body. Nine observers and four representatives of intergovernmental organizations also attended the Meeting. The list of participants is annexed to the report.

Opening address

4.  The Meeting was opened by Mr. Ashagrie, Director of the ILO Bureau of Statistics, on behalf of Mr. Taqi, Assistant Director-General, who was unavoidably absent. It was noted that occupational safety and health had been a major concern of the International Labour Organization since its inception, and that guidelines on statistics of industrial accidents had been adopted as long ago as the first International Conference of Labour Statisticians (ICLS) in 1923. There was an increasing need for information on the occurrence of occupational accidents and their causes, but adequate data were still lacking in many areas. A significant step towards rectifying this situation had been taken in 1994, when a Meeting of Experts convened by the Governing Body drew up and adopted a code of practice on the recording and notification of occupational accidents and diseases. The proposals which the present Meeting of Experts on Labour Statistics would discuss were intended to build on the provisions of the code of practice, in order to provide more practical guidance for the compilation of statistics of occupational injuries. The role of the Meeting would be of great importance in helping the ILO to prepare a report on new international standards for statistics of occupational injuries. This report would be placed on the agenda of the Sixteenth ICLS in October 1998, subject to the approval of the Governing Body of the ILO.

5.  Mr. Ashagrie referred to the absence of formal rules of procedures for the Meeting, which was meant to facilitate a free exchange of views among the experts present. Delegates were participating at the Meeting as individual experts, not necessarily representing views of their respective governments or organizations. Document MELSOI/1998/1 had been prepared by the Office for discussion by the Meeting, containing proposals for a draft resolution on the subject. He hoped that participants would provide constructive comments and suggestions concerning the proposals that should be put to the 16th ICLS later in the year. The report of the Meeting would be submitted to the Governing Body for its approval in June 1998.

Election of the Chairman

6.  Dr. Erkki Yrjänheikki, the expert nominated by the Government of Finland, was elected Chair of the Meeting. He stressed the need for more and better statistics on occupational injuries at the national level especially for prevention purposes. Harmonization of these statistics between counties would also be useful for comparisons.

Presentation of the working document

7.  In the presentation of the report to the Meeting, attention was drawn to the need for up-to-date guidance for statistics of occupational injuries. The ILO Code of Practice on the Recording and Notification of Occupational Accidents and Diseases set out principles underlying a system for their reporting and recording at the establishment and for notification to central authorities, and set out the types of statistics which should be compiled through such a system. The proposals to be discussed by the Meeting were intended to build on the provisions of the code of practice, by providing practical guidance for statistical purposes, as well as the updating and expansion of the existing international standard classifications of occupational injuries and accidents. They also provided a link with the Labour Statistics Convention, 1985 (No. 160), which covered statistics of occupational injuries.

8.  A comprehensive programme for statistics of occupational injuries was put forward, providing for a progressive approach in which a basic data set would be the first priority. This was necessary in order to allow for the fact that countries throughout the world differed greatly as to levels of development and availability of resources. Subsequently, more detailed information could be compiled in order to give a better understanding of how accidents happened and how injuries occurred. The programme would allow for statistics to be compiled from different types of sources, in addition to the traditional sources of accident compensation schemes and labour inspection.

9.  It was important that the issue of gender should be adequately reflected in the guidelines. The types of data collected, how they were collected and how they were classified should reveal areas of importance to women as well as to men.

General discussion

10.  Recognizing the need for revised standards for statistics on occupational injuries, many participants also emphasized the importance of statistics on all areas of occupational health, including occupational diseases, dangerous occurrences and the effects of stress and repetitive movement on the health of workers and their families. These were not included in the present proposals, partly because they presented different measurement problems from those covered in the proposed statistical standards, but also because few countries had yet developed expertise or systems for measuring them. The Meeting considered that, in the future, the Office should undertake research in these fields, with a view to developing relevant statistical standards. In this connection, the Worker experts noted that upper limb disorders, repetitive traumas and similar conditions would not be covered in the statistics of occupational injuries. They proposed that future ILO development of standards on statistics of occupational diseases should also cover these types of injury. However, it was considered that the developmental work on statistics of occupational injuries should be completed before similar work could begin on other areas.

11.  The Meeting welcomed the proposals to extend the coverage of statistics of occupational injuries to all workers. Traditional sources of data often excluded certain groups of workers, occupations or economic activities. Often agriculture was excluded, as well as the self-employed and, in some countries, even workers in the public sector (railways, post office, police, etc.). These accounted for a very large proportion of total employment, and were exposed to risks as high, if not higher, than those for private sector employees. It was regarded as important to have additional information on these and other similar groups. It was not clear how evolving types of work, such as teleworking, would be covered by accident compensation schemes, for example.

12.  Several participants supported the proposal to supplement the traditional data sources by linking them with other types of sources, such as household (including labour force) or establishment surveys. In developing countries where compensation schemes often covered only a very limited part of total employment, this could provide greatly needed information, particularly for the informal sector and the self-employed. Guidance should be provided in particular on ways of linking data from administrative sources to those from other sources. Other participants defended the use of administrative sources as providing a minimum level of data, which should be accessible to most countries.

13.  Concern was expressed by some Employer and Worker experts that the proposed definitions could lead to the use of two sets of data, and an increased burden on employers for reporting. This was because the proposals did not necessarily correspond to what was required by the legislation covering, in particular, compensation of occupational injuries.

14.  The Meeting considered that it was important for the draft proposals to recognize the need for statistics on occupational safety and health, of which occupational injuries were a significant component. Information on all aspects of safety and health would be a long-term objective, for which the Office should work in order to provide statistical guidance. The proposals under discussion during the Meeting would, however, concern only statistics of occupational injuries.

15.  It also stressed that the principles set out in the proposals should be intended as minimum standards for good practice, and could be further developed by countries in accordance with their own national needs and circumstances. They should not be used to limit or undermine the current national standards, nor inhibit their improvement.

Objectives and uses

16.  The Meeting felt that a clear distinction should be made between the objectives of the statistics, on the one hand, and the uses to which they could be put, on the other. The principal objective of the programme of statistics was to provide comprehensive information required for the purposes of accident prevention. The data should identify occupations and economic activities where occupational injuries occur, along with their extent and severity. Support was voiced for the need for information about the causes of accidents. However, it was generally agreed that it was not possible to distinguish the real causes of accidents, i.e. the full chain of events leading to an accident, and there could in fact be multiple causes. Nonetheless, it was feasible to collect information as to the ways in which accidents and injuries occur, which was necessary for accident prevention. At this stage, it was agreed that the role of statistics was to indicate important areas to which attention should be paid. These areas would be the subject of more detailed investigation, in which more information could be gathered regarding the chain of events leading to an accident or injury.

17.  Support was expressed for most of the proposed uses enumerated in the report, although those relating to the estimation of the cost of occupational injuries and the determining and adjustment of insurance premiums should be modified. Some experts felt the latter use should be deleted, since it could lead to the under-reporting of occupational injuries. Some others considered it inappropriate to include financial uses in the guidelines. Other views put forward supported the inclusion of financial aspects, since one of the reasons for undertaking preventive action was to reduce costs. It was suggested that the proposed uses should include the fact that the data could be used as a basis for policy-making aimed at enhancing financial incentives for employers, employers' associations, workers and trade unions to introduce accident prevention measures. The cost of occupational injuries should be expressed either in terms of days lost or days of absence from work, or other variables should be included, such as production losses, in order to quantify the full cost. Additional uses put forward included assisting in developing more education and training programmes for occupational safety, and in identifying possible areas for future research.

Terms and definitions

18.  A number of experts expressed concern that there might be contradictions between the definitions included in the code of practice and those proposed in the report. It was generally recognized that the definitions should take into account existing reporting systems, and be consistent with them. A suggestion was made that any clarifications of these definitions could be included in footnotes which would avoid the need to modify them. The secretariat pointed out that the proposed definitions were intended for statistical purposes, not merely for purposes of reporting, and that they did not contradict those included in the code of practice. If data were to be compiled from sources other than reporting systems, the definitions used also needed to be suitable for use with those sources. One observer noted that in his country, data were compiled from a number of different sources, including reporting systems and surveys, and that definitions were used for purely statistical purposes. The representative from EUROSTAT stated that the proposed definitions were in line with those used in the European Union project for statistics of accidents at work.

19.  Several Employer experts suggested that commuting accidents should be removed from the proposals, as they did not constitute occupational accidents. If they were included, they should be counted separately from the statistics of occupational injuries. It was noted that the inclusion of commuting accidents derived from legal provisions regarding accident compensation. In addition, the transportation of workers to the workplace was traditionally part of social policy in some countries, and was the subject of negotiation between employers and workers. Several experts proposed that the definition of a commuting accident should be modified in such a way as to refer to the "ordinary" or "habitual" route rather than the "direct way" between the place of work and the locations listed. The list of locations should also be expanded to include places of training. A question was raised as to whether it was practicable to consider accident prevention in the context of commuting accidents.

20.  A number of experts drew attention to the need to distinguish commuting accidents from travel accidents. The latter should be included as occupational accidents, but there should be a clearer definition in the proposals.

21.  Regarding the proposed definition of fatal occupational injury, several experts felt that the introduction of a limit of one year would not permit the capturing of deaths from illnesses with long latency periods. Others pointed out that deaths occurring long after the date of the accident could be difficult to identify as occupational injuries; in many cases, they were probably treated as deaths due to occupational diseases, and thus excluded from the statistics of occupational injuries.

22.  The proposed definitions of incapacity to work, and permanent and temporary incapacity to work, were the subject of a number of comments. It was generally felt that incapacity for work should be defined as in the code of practice. The distinction between temporary and permanent incapacity for work needed to be developed. Several experts gave examples of the practices in their countries. In some, the distinction was defined in terms of days of absence from work, in others in terms of percentage disability. Partial and total disability were also defined in a number of ways.


Occupational injuries

23.  The Meeting supported the proposal that the statistics should cover all occupational injuries, as defined earlier, without specifically mentioning the inclusion of fatal and non-fatal injuries. It was however generally agreed that only those injuries resulting in a loss of at least one day excluding the day of the accident should be included. The issue of what constituted lost work time was also raised. The observer from EUROSTAT explained that, in the harmonized European Union statistics, a cut-off point of three days was applied. This was partly because of the national practices in Member States, where compensation schemes applied this limit, and partly because it was very difficult to obtain reliable data on injuries involving shorter periods of absence from work.

24.  One observer drew attention to the fact that, in his country, the statistics showed that occupational injuries involving days of absence from work were decreasing in number, while those with restricted working days, i.e. days during which the workers carried out lighter duties than normal, were increasing. If the statistics only covered injuries with lost work time, there would not be a complete picture. Several experts also expressed the need to take account of injuries involving days of restricted work activity.

25.  It was noted that statistics on commuting accidents were generally collected along with those on occupational injuries, even if the data relating to them were published separately. The proposals should refer to the separate compilation of the statistics, rather than separate collection.

Persons, economic activities and geographic areas

26.  The Meeting considered that the statistics should cover all workers regardless of their status in employment, in principle, although this was not always practical. It was agreed that the proposal in the draft that, as a minimum, all employees should be covered should be deleted as unnecessary. The need to cover the self-employed and the various forms of non-regular employment was emphasized, since these were generally excluded from the existing statistics.

27.  The Meeting considered that statistics of occupational injuries should be collected in connection with child labour. However, this inclusion should not be interpreted as condoning child labour. The ILO Convention concerning Minimum Age for Admission to Employment (No. 138) and the UN Convention on the Rights of the Child were recalled, and it was hoped that all countries would ratify them.

28.  Regarding special population groups, such as child labour, informal sector workers and homeworkers, it was suggested that these could be the subject of special studies. Useful sources of information in this respect would be trade unions and non-governmental organizations.

29.  There was general agreement that the statistics should cover all economic activities, sectors and the whole country. Certain groups of workers posed particular problems, for example seafarers, workers on oil rigs, persons travelling for their work, frontier workers, and workers on detachment. For these workers, a de jure approach was suggested, i.e. that they should be covered by the country in which they were normally resident. It was also proposed that seafarers could be covered in the statistics of the country in which the ship was registered. Persons injured while travelling abroad should be included in the statistics for the country in which they were covered by the legal rules governing compensation or accident prevention.

Types of data

30.  The proposal for a progressive approach regarding the types of data to be collected was supported. The minimum data set would constitute the first stage, which many countries had already achieved. The extended data set would provide the more detailed types of information required about the sequence of the accident and its consequences. The Meeting agreed that, for purposes of clarity, each of the variables to be included in the data sets should be explained in the resolution.

31.  It was suggested that information about the work process, the activity of the person at the time of the accident and the item associated with the activity of the injured person should be included in the basic data set, rather than in the extended data set.

32.  Attention was drawn to the importance of information about the employment size of the establishment, since this often determined whether there was an accident prevention committee or safety and health department. It was proposed that the basic data set should include an indication as to whether the establishment had such a body. One expert noted that there was a growing tendency for larger businesses to contract out certain activities to smaller establishments. These were often activities requiring low levels of skill and involving more dangerous processes. Consequently, employment size of establishment was also very important for accident prevention.

33.  The following additional variables were proposed for inclusion in the minimum data set:

(a) whether the injured person had been examined by a physician, nurse or health worker;

(b) number of hours worked by the injured person up to the time of the accident, so as to give an indication since the beginning of the shift of the level of fatigue of the person injured;

(c) geographical location of the accident.

34.  In addition to the proposed variable for the length of service with the present employer, it was considered useful to include the length of time in the present job, since the latter could give an idea of the experience of the person in the type of work performed.

35.  Regarding the consequences of the injury in terms of incapacity for work, attention was drawn to the fact that this would require information about regular hours of work and whether the person normally worked full time or part time. The Meeting agreed that a variable indicating whether the person normally worked full time or part time should be included. It was also suggested that information should be collected on whether the person was a permanent or casual worker, and the duty status of the worker at the time of the accident, such as on duty, on a rest break, a road traffic accident and so on.

36.  It was suggested that the following additional variables be included in the extended data set:

(a) race or ethnic group, since this could help in identifying groups at risk. For example, language problems could cause difficulties in understanding safety rules or instructions for the use of equipment; however, some objections were raised in view of possible discriminatory aspects;

(b) whether the injured person had received training in the job;

(c) the highest level of education attained.


37.  The Meeting supported the proposals regarding the measurement of an occupational injury, although it was suggested that the term to be used for the unit of observation should be the "occupational injury" rather than the "case of occupational injury". The addition of a point concerning the separate enumeration of each case of injury resulting from a single occupational accident was approved. Also, a recurrence of an injury due to a single accident should be treated in the statistics as a continuation of the same injury, rather than a new injury.

38.  Support was also voiced for the proposals for the measurement of time lost, although it was recognized that they could be difficult to apply. It should be measured in calendar days, for each period of absence from work due to a single accident. Time lost should be collected for all injuries, including those leading to permanent incapacity for work and fatalities. These data were particularly important for calculating severity rates. Guidance should be provided for estimating the time lost by the two latter groups.

Reference period

39.  The experts agreed to the proposals in principle, but more precision was needed in the case of fatalities, which should be included in the statistics for the period in which the occupational accident occurred. It was noted that seasonal trends could be identified from the date of accidents.

Comparative measures

40.  The Meeting considered that the proposals relating to comparative measures represented the minimum set required. The following additional measures were suggested for inclusion:

(a) fatality rates;

(b) risk rate: the number of days lost per 100 workers; this represented a combination of severity and incidence rates, and was useful for analysing levels of risk, particularly in individual enterprises;

(c) median number of days lost per occupational injury.

41.  Some experts did not agree to the use of the term "persons exposed to risk" in connection with the denominator used to calculate the proposed measures. The expression "workers in the reference group" or "workers in the group under consideration" reflected more closely the intended meaning.

42.  There was considerable discussion about the measures to be used as denominators for the different rates proposed, i.e. the total number of hours or days worked or the total number of persons "exposed to risk". Hours worked could be estimated on the basis of normal hours of work, excluding periods of absence. Hours worked would be a better measure than the total number of days worked for the calculation of severity rates, since the latter did not take into account the lower number of working days of some part-time workers.

43.  Regarding the number of workers "exposed to risk", it was noted that, with a long reference period, an average would need to be calculated for the whole period. Also, it would be appropriate to use full-time equivalents for part-time workers, otherwise the rates would be biased downwards.

44.  Several experts emphasized the fact that the proposals constituted minimum standards, and that other more detailed analyses would be necessary to identify certain risks. While recognizing the desirability of using data on hours worked and persons "exposed to risk" that were as accurate as possible, a number of experts noted that in reality the information required was sometimes missing or, if it did exist, was not always of the preferred level of quality. It was generally necessary to compromise, and make use of the best data available. Simplicity and practicability should be borne in mind.

45.  It was decided that the example provided concerning the coverage of the numerator and denominator should be expressed in a more positive manner, without reference to exclusions of certain groups of workers. This would provide a clear guide as to the approach to be adopted and avoid any implied exclusion from the statistics of groups such as the self-employed.

46.  The secretariat pointed out that the proposed rates should be calculated using the number of new cases of occupational injury occurring during the reference period.

Publication (dissemination)

47.  The Meeting agreed that "publication" included all means of dissemination, whether on paper or in electronic or another form and that this section should be termed "Dissemination" as a result. It recommended that, where possible, the relevant national competent authority in each country should make data available on the Internet so as to facilitate analysis by users throughout the world, provided the confidentiality of these data were ensured. Several experts noted that their countries' data were already available on the Internet.

48.  The proposals relating to publication were supported, with suggestions for certain clarifications or additions. The published statistics should be accompanied by methodological information, including an explanation of any changes over time in the methods used and an indication that data for past years may have been revised. The data should not be limited to the latest reference period, but should include time series.

49.  The need for timely statistics was stressed, and guidance for the time-limit by which countries should release their data was sought. Suggestions were made that aggregate provisional figures should be released rapidly, with final data appearing later. The guidelines should include the need to make published data available to trade unions, employers' organizations and professional associations.

50.  The need to protect the confidentiality of data on occupational injuries was recalled. In certain circumstances, however, and for purely statistical purposes, the experts recognized that individual data could be made available.

51.  The Meeting also approved the following proposals:

(a) that countries should send their data on occupational injuries to the ILO, and that the ILO should disseminate the data received each year; and

(b) that the ILO should provide assistance to countries in establishing their programmes for compiling and publishing these statistics.

In addition, paragraph 4 of the proposals should be included in this section.

Sources of data

52.  The proposals for using different sources of information for the programme of statistics on occupational injuries were welcomed by the Meeting. It was noted that, even if several different sources were used, it would not necessarily be possible to achieve full coverage of all occupational injuries. Nonetheless, sources such as labour force surveys, hospital records, establishment surveys, household budget surveys and population censuses could provide useful data to complement those collected through compensation schemes or labour inspectorates. This was particularly important for certain countries, where compensation schemes were limited in coverage. It was agreed that the idea of developing new sources of information should be introduced, while the programme of statistics should use existing sources of data.

53.  Combining data on occupational injuries from different sources could lead to lower costs of producing statistics, reduced reporting burden on enterprises and information on more variables. Linking the data from different sources could be difficult, however. It would be useful at the national level to establish a coordinating committee comprising representatives of both state and private bodies, so as to avoid the use of different terminology and classifications.

Demonstration of a system for
accident recording and analysis

54.  On the second day of the meeting, Ms. Jørgensen gave a demonstration of the system used in Denmark for the recording and analysis of occupational injuries. The computerized system incorporated the complete classification schemes for each of the variables to be recorded, with useful search features. This greatly facilitated recording the necessary information, and each variable could be used for analysis. In-depth studies could show which types of accident occurred most frequently and caused the most serious injuries; all accidents of a particular type or involving a particular agency could be analysed; and so on. Such analyses were only possible with a computerized system. The system was intended mainly for central authorities, but could also be used by large enterprises.

Classification of occupational injuries

55.  The Meeting noted that the various classification schemes proposed were intended to provide countries with a basis for establishing or for revising their own classification schemes if necessary, and also to promote the comparability of the data at the international level. As a minimum, data would be classified at the highest level of aggregation of the individual schemes, with further detail decided at the national level on the basis of requirements and circumstances. Several experts recommended that the schemes should be simple, clear and easy to understand and apply. They should also constitute an improvement over any previous international classifications. The need to field test the proposed classification schemes was underlined by several experts. In addition, consistent coding should be used in the different schemes.

56.  A number of experts expressed the view that the guidelines should incorporate clear explanations of the various classification schemes, including the criteria for classification and broad indications as to how they should be applied. More detailed information on the application of the schemes, as well as more detailed groupings, should be developed for inclusion in a manual. It was recommended that, where relevant international standard classifications existed, these should be used. During the discussion of the proposed classification schemes, questions were raised as to how they should be applied, by whom and at which level, as well as how the relevant information could be collected and recorded. It was suggested that the ILO should develop guidelines for these aspects.

57.  The Meeting was informed that a number of the proposed classification schemes followed closely those developed or currently under development for the EUROSTAT/European Commission project on European Statistics of Accidents at Work (ESAW). Some of these draft classifications would be tested later in the year, and could be subject to modifications. It was recalled by several experts that, although considering proposals based on the EUROSTAT drafts, this should not be interpreted as the Meeting endorsing them. Nonetheless, they formed a sound basis for discussion, and the experts were grateful for them.

58.  The Meeting agreed in principle to the proposal that the data should be classified at least according to major branch of economic activity and as far as practical according to significant characteristics of persons injured, of establishments, of occupational injuries and of occupational accidents.

59.  Regarding the classification of economic activities, it was agreed that, where practical, the latest version of the International Standard Industrial Classification of All Economic Activities (ISIC) should be used. Several experts felt that the inclusion of only the tabulation categories of ISIC Revision 3 in the guidelines was not sufficient; the next level of detail should also be shown, since this comprised more meaningful categories. If this level were not included, the guidelines should merely include a reference to ISIC Revision 3.

60.  The Meeting recommended that the primary activity of the establishment should be used for classification. As a result, information collected on the economic activity of the establishment should refer to the main activity, and this should be reflected in the minimum data set. Attention was drawn to the availability of more detailed levels of aggregation in the full document concerning ISIC Revision 3, which also included guidance on its application.

61.  The proposed classification according to the employment size of establishments was based on the experience of four countries, after examination of all the employment size classes used by countries in the European Union. A number of experts indicated that the proposed classes were not relevant to their national circumstances, and it was generally agreed that it would be difficult to determine a set of employment size classes which would be applicable in all countries. The proposals could however serve as a basis for international comparisons. Another, simpler classification might also be appropriate, with small, medium and large establishments as the groups, with another group for "micro" establishments. The secretariat proposed to find out whether there were any international standards in this connection.

62.  Support was voiced for the suggestion that the size of an establishment should be measured in terms of full-time equivalent workers. The classes should be determined in terms of workers employed, so as to allow for the inclusion of the self-employed and other workers who were not employees of the establishment. A question was raised as to whether it should be the establishment or enterprise which should be classified by size. The establishment was preferred in general as the unit as its size was usually easier to obtain than that of the enterprise to which it belonged.

63.  The Meeting supported the proposals regarding the classification of occupations, recommending that the guidelines include the International Standard Classification of Occupations (ISCO-88) at the second, i.e. sub-major, group level. It was noted that there was no major group for the residual category of occupations.

64.  The proposed classification of status in employment was not considered suitable for the guidelines. The Meeting preferred the International Classification of Status in Employment (ICSE-93), which would enable the identification of workers in precarious employment and other groups which might be at risk.

65.  A number of modifications were put forward to the proposed classification of the consequences of injuries: deletion of the categories "temporary incapacity to work" and "less than one day lost", and separation of the proposed group "permanent incapacity to work or 183 days or more lost" into "more than six months lost and up to one year" and "permanent incapacity to work". The observer from EUROSTAT explained that the proposed groups corresponded to the most frequent categories observed in the European Union countries. Less that 5 per cent of occupational injuries resulted in a loss of more than three months. It was noted that a definition of permanent incapacity for work would be required. Several experts drew attention to different national approaches, including reference to the payment of a pension, physical incapacity and inability to return to work or to normal work.

66.  The Meeting considered that, if several injuries were incurred by a worker in an occupational accident, the injury to be classified according to type of injury should be the most serious one, and that this should be reflected in the title of the classification, which should also refer to "occupational injury". Several additional categories were put forward for inclusion in the proposed classification scheme: internal injuries, psychological shock, loss of a sense (sight, hearing, etc.), acute effects of radiation, overexertion injuries, contagious diseases, hernias. It was noted that the proposal did not include categories for diseases resulting from occupational accidents, and some experts suggested that the International Classification of Diseases and Related Health Problems (ICD-10) should be used for this purpose. Others proposed that a list of diseases either be included in the guidelines as an additional classification scheme, or added to the classification according to nature of injuries. Several experts stated that the latter approach had been adopted in their countries.

67.  In response to a proposal that the ICD-10 should be used for classifying types of injury, the ILO consultant drew attention to the difficulties this would involve. The ICD-10 would require a medical diagnosis of the injury, which employers and workers would not generally be able to provide. Concern was also expressed by one expert that the guidelines should avoid increasing the burden of information to be provided, otherwise they might not be widely accepted. The Meeting supported the secretariat's proposal that the next stage in its developmental work should take into consideration the different national and international classifications relating to types of injury, including ICD-10, in collaboration with WHO.

68.  In the discussion of the proposed classification according to the part of body injured, it was noted that classification should relate to the part of body relevant to the major injury incurred. One expert proposed the inclusion of a separate category for urogenital organs. Another suggested that categories for multiple locations, such as "hand and fingers" would be useful. For certain injuries, these body parts could not always be distinguished from each other, as in the case of crushing. It was also suggested that separate categories should be included for each finger. Closer specification of vertebrae in the neck and back as cervical vertebrae and thoracic and lumbar vertebrae respectively was also suggested. It was noted that two groups appeared to include the same body parts: shoulder joints.

69.  The Meeting did not support the proposals for the classification according to the type of disability. In general, disability was defined by compensation schemes, and there were considerable variations between countries and between insurance schemes in the criteria used. While recognizing the usefulness of information on the degree of permanent disability, particularly for showing the impact of an injury on the earning capacity of a worker, it was considered too subjective and difficult to measure. In addition, it caused long delays in the finalization of the statistics.

70.  The classification according to the work environment was considered as useful for developing practical accident prevention measures by several experts, although the title should be more meaningful, for example by using "type of workplace". A number of participants supported the proposed scheme, but suggested modifications, including the development of basic categories which would fall between those in the first and second levels of the draft. Terminology would also need to be refined or clarified, and apparent duplications of categories removed.

71.  The Meeting was informed that the new proposed classification of work processes was still very much at the draft stage. No country had a classification of this kind, which was intended to cover the work process being performed by the worker when the accident occurred. A number of experts considered the proposals confusing, as they seemed to overlap the classification of activity. It was generally agreed that considerable development and testing of both the general framework and the designations would be required. A proposal was made for an alternative structure, which would relate categories such as "things, plants, animals, humans, information and others" with processes such as "used, processed, produced, handled (manually and mechanically), stored and transported".

72.  Before the discussion of the next four proposed classifications, the ILO consultant explained the underlying framework that was intended, as illustrated by figures 1, 2 and 3 at the end of this report. The activity of the injured person was the action of the person at the time of the accident. The deviation from the normal described what went wrong as a result of which the accident occurred. The mode of injury described the way in which the person was injured. Each of these was associated with an item or agent. The proposed classifications were based on those used by the Danish Work Environment Service since 1992.

73.  The Meeting found the approach interesting, and considered that it could provide information that would be very useful for accident prevention purposes. Several experts noted that international classifications already existed for the mode of injury, type of accident and item or agent, and preferred the structure of these to the ones proposed. It was suggested that the current classifications should be updated, if necessary, but that this should be achieved without altering the structure.

74.  The proposed classification according to the activity of the injured person was considered to provide a good starting-point for the guidelines, though too complex in its present form. It was suggested that the terminology should be re-examined, and a few experts drew attention to similar schemes used in their countries.

75.  In the discussion of the proposed classification of deviations from the normal, attention was drawn to the fact that the explanation and some of the categories appeared to assign responsibility for accidents, which should be avoided. Also, the use of "deviation" was not necessarily appropriate, since normal work itself could cause an accident. However, a suggestion by one expert to use "cause of accident" was opposed by others. It was also noted that the failure to provide or the failure to use protective equipment or safety equipment were also factors which should be taken into account. The experts recommended that the terms used and explanations given should be reviewed carefully with a view to eliminating any tendency to allocate blame for accidents. Several experts observed that there was often not a single deviation leading to an accident but several, which could cause classification problems. The ILO consultant explained that, generally, it was the last in the chain of events which was classified. As a result, some precision might be lost, but a compromise had to be found. It was important, however, that a consistent approach be applied. The Meeting considered that the proposed framework provided a sound basis for future development.

76.  It was noted that these classification schemes were intended primarily for statistical purposes and could not be used on their own for identifying the causes of accidents. They could only be used to identify through statistical analysis accidents that needed further investigation to determine their causes. It was suggested that the ILO should develop guidelines for the investigation of individual accidents.

77.  Several experts felt the proposed classification of the mode of injury would be useful for understanding the way in which the person was injured, but that some editing and modification of categories would be necessary. Some others preferred the simplicity of the 1962 classification according to type of accident.

78.  The Meeting noted that information about the item or agency was very important. The proposed classification was supported by a number of experts. Others considered the structure of the 1962 classification according to agency was more practical, and could usefully be updated.

79.  The experts were informed that the proposed classifications of commuting accidents, which were drawn from the Nordic Medico-Statistical Committee (NOMESCO) classification, were close to the classifications of traffic accidents which it was expected WHO would adopt in the near future. Several experts considered that the proposals could provide a good understanding of commuting accidents, but felt they should be reduced and simplified. Others reiterated their view that commuting accidents should not be covered in the statistics.


80.  It was considered that the Meeting's discussions would provide a sound basis for the Office's work leading to the preparation of a draft resolution on statistics of occupational injuries. There had been general agreement on many points relating to the measurement of occupational injuries. Regarding classifications, however, more developmental work and field testing of the various new or revised schemes would be required. It was recommended that, where relevant international standard classifications existed, these should be used.

81.  Many experts recommended that, once the ILO's developmental work on statistics of occupational injuries was completed, similar work should be taken up in respect of occupational diseases.

Adoption of the report

82.  The participants examined the draft report containing a summary of the Meeting's discussions and conclusions and, following the introduction of a few amendments, adopted it.

83.  The final report will be presented to the Governing Body at its June 1998 Session.

Figure 1

Figure 2

Figure 3

Action (verb)

Item (noun)

Stage 1 - Action

Activity of the victim at the time of the accident

What was the injured person doing at the time of the accident?

(e.g. operating a machine, working with a hand tool, driving, walking)

Item associated with the activity of the victim

What item was associated with the activity?

(e.g. power press, screwdriver, vehicle, floor)

Stage 2 - Mechanism

Deviation from the normal

What went wrong, what deviant event occurred?

(e.g. explosion, fall, lost control of machine or hand tool)

Item associated with the deviation (A & B)

What items were associated with the deviation?

(e.g. vessel under pressure, wall, power press, drilling machine)

Stage 3 - Contact

Mode of injury (action leading to injury)

How did the action lead to the injury?

(e.g. struck by (moving object), fall (from height), crushed, contact with sharp object, etc., contact with hot object)

Item associated with the mode of injury (agent of injury)

What item caused the injury?

(e.g. car, floor, machine, knife, stove)



List of participants
Liste des participants
Lista de participantes

Government experts
Experts gouvernementaux
Expertos gubernamentales

Mr. Henry Chikova, Statistician, National Social Security Authority, NSSA House, 2nd St. & Selous Avenue, P.O. Box CY 1387, Causeway, Harare, Zimbabwe
Tel: +263 4 722 047; Fax: +263 4 796 320

Dr. Dulce P. Estrella-Gust, Executive Director, Occupational Safety and Health Centre (OSHC), North Avenue Corner Science Road, Diliman, Quezon City, Metro Manila, Philippines
Tel: +632 928 66 90 or 928 67 28; Fax: +632 929 60 30; e-mail:

M. Charles Lefebvre, Chef, Statistiques et analyse, Division de la sécurité et de la santé au travail et de la prévention des incendies, Direction des opérations, Direction générale du travail, Développement des ressources humaines Canada, Place du Portage, Phase II, 7e étage, Ottawa, Ontario K1A 0J2, Canada
Tél.: +1 819 953 0234; Fax: +1 819 953 1743; E-mail:

Sra. Teresa Santa Cruz Romero, jefe de Sección de Estadística, Subdirección General de Estadísticas Sociales y Laborales, Ministerio de Trabajo y Asuntos Sociales, Secretaría General Técnica, María de Guzmán, 52, 28071 Madrid, España
Tel: +34 1 456 1643; Fax: +34 1 53 44 93

Mr. Tim Williams, Manager, Statistics Unit, National Occupational Health and Safety Commission, 92 Parramatta Road, Camperdown NSW 2050, Australia
Tel: +61 2 9577 9287; Fax: +61 2 9577 9300; e-mail:

Mr. Erkki Yrjänheikki, Chief of Development Unit, Ministry of Social Affairs and Health, Department for Occupational Safety and Health, P.O. Box 536, FIN-33101 Tampere, Finland
Tel: +358 3 260 8111; Fax: +358 3 260 8511

Employer experts
Experts des employeurs
Expertos de los empleadores

Sr. Victoriano Anguis Terrazas, Presidente, Comisión Nacional de Seguridad e Higiene, COPARMEX, Unión No. 359-A, Colonia Tepeyac Insurgentes, 07020 México, D.F., México
Tel: +525 577 77 37 or 577 71 06; Fax: +525 577 77 37

Mr. Eric Jannerfeldt, Medical Adviser, Svenska Arbetsgivareföreningen, Södra Blasieholmshamnen 4A, 103 30 Stockholm, Sweden
Tel: +46 8 762 60 00; Fax: +46 8 762 64 39; e-mail:

Mr. B.V. Prakash Kaunhye, President, Association of Health & Safety Managers, c/o SoniaWear Limited, Valentina Industrial Zone, Phoenix, Mauritius
Tel: +230 698 41 54 or 696 72 75; Fax: +230 686 98 64; e-mail:

Mr. Imanuel Mulder, General Manager, Occupational Health and Safety, Iscor Ltd., P.O. Box 450, Pretoria 0001, South Africa
Tel: +27 12 307 42 97; Fax: +27 12 324 38 08

Sr. Eduardo Undurraga Undurraga, Gerente General, Asociación Chilena de Seguridad, ACHS, Vicuña Mackenna, 152, Santiago, Chile
Tel: +56 2 685 20 00 and +56 2 685 25 06; Fax: +56 2 222 35 33

Mr. Bruce Waechter, Labour Affairs, Planning Manager, Ford Motor Company of Canada, Limited, P.O. Box 2000, The Canadian Road, Oakville, Ontario L6J 5E4, Canada
Tel: +1 905 845 25 11, ext. 1109; Fax: +1 905 845 01 35; e-mail:

Worker experts
Experts des travailleurs
Expertos de los trabajadores

Ms. Lea Achdut, Director, The Institute for Economic & Social Research, Histadrut-General Federation of Labour in Israel, Histadrut Building, 7 Beit Hadfus Street, 95 483 Jerusalem, Israel
Tel: +972 2 658 9868; Fax: +972 2 652 6781; e-mail:

Ms. Evette Gibson, The Barbados Workers' Union, P.O. Box 172, "Solidarity House", Harmony Hall, St Michael, Barbados
Tel: +1 246 436 6079; Fax: +1 246 436 6496; e-mail:

Mr. S. Gopee, Executive Board Member, Mauritius Labour Congress, 8, Louis Victor de la Faye Street, Port Louis, Mauritius
Tel: +230 212 4343; Fax: +230 208 8945

Mr. Godfrey Kanyenze, Zimbabwe Congress of Trade Unions, Chester House, 88, Speke Avenue, P.O. Box 3549, Harare, Zimbabwe
Tel: +263 4 794 742; Fax: +263 4 728 484

Mr. T. Mellish, Trades Union Congress, Congress House, 23-28 Great Russell Street, London WC1B 3LS, United Kingdom
Tel: +44 171 636 4030; Fax: +44 171 636 0632

M. Paul Palsterman, Juriste au service d'études, Confédération des syndicats chrétiens de Belgique, 121 rue de la Loi, B-1040 Bruxelles, Belgique
Tél.: +32 2 237 3111 or 237 3445; Fax: +32 2 237 3300


Ms. Kirsten Jørgensen, Arbejdstilsynet, The Danish Working Environment Service, Directorate, Landskronagade 33, 2100 Copenhagen ø, Denmark
Tel: +45 39 15 20 00; Fax: +45 39 15 25 60

Intergovernmental organizations
Organisations intergouvernementales
Organizaciones intergubernamentales

CEC, Commission of the European Communities: Mr. Fuente, Commission of the European Communities, Unit for Statistics on Occupational injuries, DG V/F/4 "Industrial accidents and injuries", Bâtiment Jean Monnet, Office EUFO 3/291, L-2920 Luxembourg
Tel: +352 4301 32739; Fax: +352 4301 34259; e-mail:

EUROSTAT, Statistical Office of the European Communities: Mr. Didier Dupré, Unit E3, Education, Health and other Social Fields, EUROSTAT, Bâtiment Jean Monnet, L-2920 Luxembourg
Tel: +352 4301 35 034; Fax: +352 4301 35 399; e-mail:

OAT, Organisation arabe du travail: Dr A. El Telawi, Organisation arabe du travail, 44 rue de Lausanne, 1201 Genève, Suisse
Tél.: +41 22 732 58 06; Fax: +41 22 732 54 17

WHO, World Health Organization: Dr. M. Mikheev, Senior Scientist, Occupational Health Unit, Division of Operational Support in Environmental Health, WHO, 20, avenue Appia, 1211 Geneva 27, Switzerland
Tel: +41 22 791 2111; Fax: +41 22 791 0746


BLS, Bureau of Labor Statistics: Dr. John Ruser, Senior Research Economist and Chief, Compensation Research and Program Development Group, BLS, US Department of Labor, 2, Massachusetts Avenue, N.E., Room 4130, Washington, D.C. 20212, USA
Tel: +1 202 606 7392; Fax: +1 202 606 63 10; e-mail:

GCTU, General Confederation of Trade Unions: Mr. Vladimir Kouvchinov, Director, Department of Protection of Health and Labour, GCTU, Leninski prospect 42, 117119 Moscow, Russian Federation
Tel: +7095 938 7215; Fax: +7095 938 2155

Mr. Gueorgui Kanaev, International Department, GCTU, Leninski prospect 42, 117119 Moscow, Russian Federation
Tel: +7095 938 7215; Fax: +7095 938 2155

ICFTU, International Confederation of Free Trade Unions: Mr. Dan Cunniah, Deputy Director, ICFTU, 46 av. Blanc, CH 1202 Geneva, Switzerland
Tel: +41 22 738 4202/3; Fax: +41 22 738 1082; e-mail: Dan.Cunniah@geneva.

ICN, International Council of Nurses: Mrs. Mireille Kingma, Consultant, Nursing and Health Policy, 3, place Jean-Marteau, 1201 Geneva, Switzerland
Tel: +41 22 908 01 00; Fax: +41 22 908 01 01; e-mail:

IOE, International Organization of Employers: Mr. Ousmane Touré, Regional Adviser, International Organization of Employers, 26 Chemin de Joinville, CH 1216 Cointrin-Geneva, Switzerland
Tel: +41 22 798 16 16; Fax: +41 22 798 88 62; e-mail:

IOHA, International Occupational Hygiene Association: Ms. Riitta Viinanen, President, Neste oy, Environment and Industrial Hygiene, P.O. Box 320, FIN-06101 Porvoo, Finland
Tel: +358 204 50 2301; fax: +358 204 50 7658; e-mail:

IOS, International Organization for Standardization: Mr. Timothy J. Hancox, Technical Programme Manager, Standards Department, ISO Central Secretariat, 1, rue de Varembé, CP 56, 1211 Geneva 20, Switzerland
Tel: +41 22 749 01 11; Fax: +41 22 749 73 49; e-mail:

WFTU, World Federation of Trade Union: Mr. Albert Potapov, Permanent Representative, 10 rue Fendt, 1201 Geneva, Switzerland
Tel: +41 22 733 12 50; Fax: +41 22 734 75 50

ILO secretariat
Secrétariat du BIT
Secretaría de la OIT

Bureau of Statistics
Bureau de statistique
Oficina de Estadística

Mr. K. Ashagrie, Director
Ms. K. Taswell
Mr. A.S. Young
Ms. M.-T. Dupré
Mr. L.J. Johnson
Ms. A. Mata
Ms. C. Antiochus
Ms. V. Arthaud
Ms. D. Boile
Ms. C. Gondrand
Ms. L.A. Hua
Ms. A. Laurie

Bureau for Workers' Activities
Bureau des activités pour les travailleurs
Oficina de Actividades para los Trabajadores

Mr. R. Kyloh

Bureau for Employers' Activities
Bureau des activités pour les employeurs
Oficina de Actividades para los Empleadores

Mr. R. Gijon von Kleist

Occupational Safety and Health Branch
Service de la sécurité et de la santé au travail
Servicio de Seguridad y Salud en el Trabajo

Dr. J. Takala
Dr. J. Serbitzer
Mr. P. Baichoo
Ms. R. Sanchez Sanz

Updated by VC. Approved by NdW. Last update: 26 January 2000.