Classification system for work in occupational health services

WANNAG A. Classification system for work in occupational health services. Scand J Work Environ Health 1993;19:390-3. Even though the World Health Organization has repeatedly advocated re search on occupational health services to increase knowledge of its priorities and efficiency, no em pirical study on work in such services is registered in major data bases. A common classification sys tem for work in occupational health services is not available, but would facilitate comparison between studies. This report presents a feasible, reliable, and relevant classification for this purpose. Central among the policy-relevant questions on occupational health services is whether or not they work in accordance with their aims. Therefore, the project started from the aims of such services (in Norway and in the convention proposed by the International Labour Office in 1985) to obtain congruence be tween the instrument's eight categories of work and the aims of occupational health services. The classification was constructed in cooperation with occupational physicians and proved to be easy to use.

Or To xlin e data base s. The few studies which have been publi shed in Norwegian and Swedi sh registered the time spent by occupational health service personnel on various activities (2,3,(11)(12)(13)(14). Ho wever, the activities chosen for registration (eg, meetings, health co ntrol, etc) are not sufficientl y specific for a study of the priorities, effectiveness, and efficiency of occupatio nal health ser vices . For th is purpose knowledge of the subjec t matter, the co ntent of the work done by the per sonnel, is needed . Unfortunately no interna tionally accepted classificati on sys tem ex ists for the work cont ent of occupation al health services at the mom en t. Neither WHO (15) nor th e Internatio nal Lab our Office (lLO), in the conventio n and reco mmendatio n on occup ation al health services proposed in 1985 ( 16) , hav e discu ssed or developed guidelines on the subject.
My goal in thi s article is to present a system for clas sify ing the work content of occupational health services for use in future empirical st udies. Such a sys tem need s to be rele vant , reliable, and feasi ble . Each of these aspec ts have been taken int o co nsideration .

Relevance
The classification must divid e and categorize the different work tasks of occupational health ser vices in a way that en ables questions to be answered on relevant poli cy . Th ese que stions often concern the relevance of the work to the aims of occupational health serv ices. In fact , it is natural to start fro m the ai ms of serv ices whe n such a sys tem is construc ted, to ensure that a work ca tegory has relevance for o nly one of the aims of occupational health serv ices . However, the objectives of occupational health services are often stated in general term s, and it has limited practical value to make one general category for all work with relevance to a general aim. One solution is to split a general aim into natural recog nizable subaims and construct corres ponding work catego ries for these. Such subaims are often described as tasks for occ upational health services in official document s with co mments and spec ifications on the general aims of such services. As different co untries have different sets of obje cti ves for their occupational health services, it might seem as if a system for categor izing the work must be a national affair. Thi s problem can be overcome if the classification is constructe d with enough categories to match the most extensive set of national aims. When the classification is used in a country with fewer objectives, some work categ ories simply collect information on work which is not related to the objectives of the occupation al health services of that country. Such information can be valuable in the evaluation.
Therefore, it is likely that a system for categorizing the work of occupational health services can be constructed which can be applied internationally, if agreement can be reached on defin itions of cen tral concepts of the work done in occupational health services. However, the releva nce of the work in a category to an objecti ve of occupational health services must be established on a national basis.
It must also be taken into consideration that ILO, in its work with the convention and recommendation concerning occupational health services proposed in 1984 and 1985 (16)(17)(18), views occ upational health services from the position of "the needs of the hired employees at work," while WHO considers occ upational health services as services to "the seg ment of the genera l populati on that is empl oyed" ( 15). These two perspectiv es have resulted in somewhat different approac hes concerning prevention and rehabilitation in the setting of occupational health services.
Rehabilitation, as used by the Norwegian labour inspecto rate (19) and ILO (16) , describes activities restructuring work conditions for ill or disabled employees in order to prevent further deterioration of their health from work strain . Thu s occupational rehabilitation is in the realm of secondary and terti ary prevention. The primary preventive activities recommended for occ upational health services in Norway and by ILO nearly all concern specific risks to employee s at work. WHO ( 15), on the other hand , advoca tes the broad "health prom otion" approac h for occupational health services. Thi s is to a large extent primary prevention aimin g more at the health strain of employees in general than at strai n fro m work.
For a categorization of work in occupational health services it seems better to keep the co ncept of occupational rehabilitation (as has already been defined by the Norwegian labour inspectorate and ILO), and split the primary preventi ve activities into those con-Scand J Work Environ Heal th 1993, vol 19. no 6 cerned with risks from work and those co ncerned with health risks in ge neral.
Furthermore WHO has an ambiguous defin ition of "work-related illness" ( 15). It comprises both the fact that work (to a varying extent ) can cause illness and the fact that work can influen ce (most often agg ravate) many illnesses once they are established whether factors in work or outside work are the causative agent s. While fully recognizing that all illne sses and disabiliti es can be influ enced by work (as reflected in the descript ion of occupational rehabilitation already given), it seems better for a categorization of work in occ upational health services to restrict the concept of work-related illness to those co nditio ns in which a factor at work is the major cause (eve n if there is no good operational definition available at the moment).
The construction of the classification system started from the three general objectives of occup ational health services in Norway as given by the labor inspectorate ( 19). These three aims also cove r the function s of occupational health service s as stated in the ILO 1985 conve ntion (16): (i) improving the work environment, (ii) health surveillance of employees relevant to the ex posures from work, and (iii) rehabilitation.
The genera l aims for occupational health services were split into subaims, and categ ories were devised for these subaims. Three occup ational physicians individually evaluated the drafts of the categories and their definitions. The following eight categories were finall y agreed upon .
I. Work which co nce rns the establi shment of new work environments, for example, part icip ation in project s deciding on new machinery, processes, buildings, and organizations and giving information on the health consequences of planned changes in the work environment. Th e category is called sec uring future work environments.
2. Wor k which concerns the existing work environment, for exa mple, surveys of the workpl ace, evaluation of existing workplace hazards, document ation of previou s workplace hazards, improvin g condition s in the workpl ace, education of employers and employees on workpl ace hazards, protection aga inst workplace hazards which can not be remo ved, and assis tance with solving conflicts in the workplace. The category is ca lled surveyi ng and securi ng the current work environment.
certificates and similar requirements of work, and preemployment medical examinations (health requirements for specific jobs, eg, being a pilot). The category is called individual work-related aid and surveillance.
4. Work which is done to give employees with illnesses and disabilities work conditions that will not aggravate or strain their health condition. (The illness or disability can originally have been caused by the work or it can be unrelated to the work.) Examples are workplace and work practice modifications to match the employee's capacity and activities to promote the return of employees to work after sick leave (treatment for illness and addiction, as part of general rehabilitation, excluded and assigned to category 3 or 6). The category is called occupational rehabilitation of individuals.

Work done to inspire and help employees to change unhealthy life-styles. The category is called influencing individuals' life-styles.
6. Work with treatment and aid for nonwork-related accidents, illnesses, and problems. The category is called individual nonwork-related aid.
7. Work with the administration of the occupational health service organization itself. (All other administration, such as project planning and report writing, is allocated to the other appropriate categories depending on the content of the work which is done.) The category is called administration.
8. Work done to increase the competence of occupational health service personnel. (Education supplied by occupational health service personnel to employers and employees is allocated to the other appropriate categories depending on the aim and content of the education.) The category is called education of occupational health service personnel.

Reliability
Persons using the classification system must be able to assign the same work task to the same category.

392
To test the interrater reliability, seven physicians were asked to categorize 39 written descriptions of work tasks in occupational health services. The overall interrater kappa was 0.59 with an approximate standard error of 0.02 (20). This is good agreement according to Fleiss, who says that kappas above 0.75 represent excellent agreement beyond chance and kappas between 0.75 and 0.40 represent good agreement beyond chance (20). None of the individual physicians disagreed markedly with the rest. All of the kappas, calculated according to Cohen (21), showed good to excellent agreement when the classifications of one rater were compared with the modal classifications of the group as a whole (table 1).
The classification system also gave good to excellent results in three test-retest situations. Two occupational physicians rescored the 39 written descriptions of work tasks after three months. This length of time was deemed sufficient for them to forget the initial scorings. One physician obtained a kappa of 0.79 with a standard error of 0.07; the other had a kappa of 0.61 with standard error of 0.09 (21).
I rescored 210 work tasks obtained by interviews with occupational physicians three and six months after the initial scoring. The 95 % confidence interval for my combined kappa was 0.91-0.95 (20).

Feasibility
To ensure that the classification system is feasible in practical work, I developed it in cooperation with three occupational physicians who made sure that the chosen work categories would be natural to occupational health service personnel. Seven physicians, two of them occupational physicians, have used the instrument. They found it simple to use and had no problem when classifying 39 written descriptions of work tasks in occupational health services.

Concluding remarks
The way this classification system was constructed seems to have ensured adequate relevance for the chosen work categories. The system has proved feasible for use in test situations and has excellent to good interrater and test-retest reliabilities.
Thus far, the classification system has been used by physicians to categorize work tasks in test situations. The procedure could easily be expanded to let occupational physicians describe how their own work time is divided between different categories. Surveys based on this approach could give valuable information on how the resources of occupational health services are used, for example, how resources are divided between work-related and nonwork-related problems, how much time is spent on preventive versus curative activities, and how official guidelines for services are met. Results from such a survey, based on tel ephone interv ie w s w ith 50 N orwegia n o ccu pational ph y s icians , can b e found e lsewh e re (22).