Work content of Norwegian occupational physicians

WANNAGA, NORDE. Work content of Norwegian occupationalphysicians. ~cand J W?rk Environ Health 1993;19:394-8. There is little empirical knowledge of the work done I.n occupationalhealth services. In a study of the work of 50 Norwegian occupatio~a1 physic!ans,their work on 249 work days was recorded and classified into nine different categories accordmg to the work co.nt.e~L Most (40%)of the workconcernedworkenvironmentmatters,26% wentto nonwork~related aCtIVI~I~S: 18% was for administration, 4% was for personalproficiency,and 12% we~t to vanous o~her aC~I":ltIes. A total of 36 % of the work concerned preventive activities, and 30% mvolved Curative a~tIVltIeS, of which two-thirds (22%) was for nonwork-related illnesses.This latter figure seems large m compan son with the preventive aims of occupational health services given by the Nor:weglan labor mspec torate. The physicians spent little time on educatingemployeesabout work envI.r~nm.ent m~tters and on new projects to ensure good work environmentstandards.The amount of activity mvolvmgoccu pational rehabilitationwas fair.

The World Health Organization has repeatedly advocated evaluation research on occupational health services (1-3), but no empirical studies of the work in such services is registered in Medline, Excerpta Medica, Sociological Abstracts, or Toxline data bases. In Norway, only two studies (with limited information on the work content and some information on activities in occupational health services) have been published previously (4,5). This lack of empirical knowledge has hampered the reforms of Norwegian occupational health services and is one of the reasons discussions of the efficiency of Norwegian occupational health services have not been very constructive (6,7).
In order to develop occupational health services according to changing concepts, resources and tasks, we need knowledge of the ongoing work done by such services. To obtain this information, we must be able to determine the content -the subject matter -of the different work tasks of occupational health services. Such data will make it possible to answer policy-relevant questions about occupational health services, particularly questions regarding priority setting and effectiveness.
This study was undertaken to obtain data on the content of occupational physicians' work by studying the distribution of their work time in different work tasks.

SUbjects and methods
Fifty occupational physicians working half-to fulltime in two cities in Norway were randomly chosen to be interviewed. The interviews took place between September 1990 and February 1991. No major influenza epidemic or other events which could have substantially influenced the requests for services occurred during the registration period.
The physicians were registered for gender, age, workhours in occupational health services, if working alone or with other occupational physicians, their total workhours in occupational health services, total workhours of all personnel in the services, and the number of employees attended by the services. These last figures were used to calculate the worktime resources of the occupational health services, defined as the number of employees attended by one work year of personnel. The figures were calculated both for the physicians only and for all personnel. Last, the organizational type (of which there were nine) and the number of professions (an indication of knowledge resources) in the services were recorded.
Each physician was interviewed on the telephone at the end of five workdays. The same interviewer made all of the interviews. Altogether 249 days with 1858 workhours were covered. For each physician the interviews were spread over all the different weekdays worked. The days were picked arbitrarily and without any knowledge in advance of what the physician had been doing that day. Each interview lasted from 15 to 45 min. The physicians recalled their different work tasks during the day, and the time spent on them, in chronological order. They were urged to use available records to support their mem-ory. Specifi cally, they were asked to state whether they saw a work task as work-related or not. All time within office hours, including lunch, and all additional work for the occupational health services outside office hours was recorded for each day. All of the information was recorded in detail by the interviewer, who later assigned the recorded work tasks and the time spent on them to nine different categories using the classification system constructed by Wannag (8). The system is easy to use and has good interrater and test-retest reliability (8). The categories defined in the instrument are as follows : (i) securing future work environments, (ii) survey ing and securing the current work environment, (iii) individual work-rel ated aid and surveillance, (iv) occupational rehabilitation of individuals, (v) influencing individuals ' life-styles, (vi) individual nonwork-related aid, (vii) admini stration , and (viii) education of occupational health service personnel. We added a ninth category, other, which included time for travel between office and workpl aces, lunch without discussions of medical or admini strative subjects, contacts not related to work in occupational health services, and time for private affair s.
The concept of an illness or an event being workrelated is central to an evaluation of the activity of occupational health services. However, the definition used by the World Health Organization for work-related illne ss is ambiguous (9). It compris es both the fact that work (to various extents) can cause illness and the fact that work can influence (most often aggravate) many illnesses once they are establi shed, whether factors in work or outside work are the causative agents. While it should be fully recogni zed that all illness es and disabilities may be influenced by work, for a categorization of the work in occupational health services, it seems better to restrict the concept of work-related illne ss to tho se conditions in which a factor at work is the major cause. We used this, more restricted , view of work-related illne ss (which is common among Norwegian occupational physicians) in our study. However, due to the lack of an accepted objective standard, each physician had to use his or her own subjective standard when deciding wheth er an illness was work-related or not.
Events other than illness were considered workrelated if they (i) arose during the perform ance of work or (ii) came about because of specific demands from the job.
In the assessment of reliability, 210 work tasks classified in September 1990 were reclassified by the same inter viewer three and six months later. The 95% confidence interval for the combined kappa was 0.91-0.95 (10).
The physicians were able to account for more than 99% of their work time. The time not account ed for was added to the recorded work of the same days on a percentage basis.
The length of the workdays ranged from 2 to 12 h, the average being 7.5 h. To be able to treat work-Scand J Work Environ Heal th 1993, vo l 19, no 6 days of different length in the same analysis, we calculated the time in the different categories for each workd ay as a percentage of the total work time of that day. From these percentages the distribution of the work time amon g the categories and among the work tasks could be calculated for a constructed "average workday. " We sampled the recorded workday s by first randomly selecting 50 occupational physicians and then five workdays for each physician. This two-step procedure increased the possibility that the sample was not repre sentative of the universe of workda ys of occupat ional physicians. The stand ard error of the result s was therefore adju sted by a factor of the square root of 1.5 -the correction factor commonly used by the Central Bureau of Stati stics of Norway for two-step sampling ( I I) .
Simple cross-tabulations of the variables recorded for the physicians and for the occupational health services were used when we searched for indication s that these variables could explain the differences in the distribution of the occupational physicians' work. Table I shows the distribution of the work time in the work categories for an average workda y, while table 2 shows the compo sition of selected catego ries. As ca n be seen from the standard deviations (table  I), the day-to-day variation in the work done by occupational physicians is large. The largest categories were individual non work-related aid (22%) and administration (18%).

Results
Forty percent of the physicians ' time was spent on work-r elated activities in the first four categories,   while 26% of their time covered nonwork-related problems in the fifth and sixth categories. The remaining 34% of their time was spent on administration, maintaining their proficiency, driving between workplaces, and activities not related to work in occupational health services.
Preventive work is found in categories 1,2,4, and 5, altogether 26% of the work time (table I), and in part also in category 3. Close inspection of the work tasks in category 3 (eg, documenting the employee's exposures from work and health surveys of employees exposed to workplace hazards) suggests that the preventive part comprised around 10%, and the curative part around 8%, of the physicians' work time (table 2). Thus altogether 36% of the physicians' work time went to preventive activities. Curative work took 8% of their work time for work-related problems in category 3, and 22% for nonwork-related problems in category 6. Altogether 30% of the physicians' work time was spent on curative activities. Table 3 shows that face-to-face consultations with employees occurred on at least four out of five work- days. A total of 12% of the average workday was spent on face-to-face consultations for nonwork-related conditions, with an additional 10% for paperwork and telephone conversations (table 2). These consultations lasted an average of about 18 min each. The average consultation for work-related problems lasted about 25 min, and the occupational physicians spent 8% of their average workday on these matters. Five percent was in face-to-face consultations, and 3% was with additional paperwork and telephone conversations (table 2).
Medical check-ups took place on less than half of the workdays (table 3), but a total of 14% of the average workday was spent on this activity (II % from categories 2 ,3, and 6 in table 2 and 3% from category 5 "influencing individuals' life-styles" in table I). Medical check-ups lasted an average of about 33 min and mainly dealt with (7%) work in category 3 "work-related aid and surveillance" -documenting the employees' exposures from the work environment and checking for possible work-related health problems (table 2). The occupational physicians also paid due attention to influencing life-style risks (3%).
Even if only a few of the physicians announced the possibility for employees to consult during medical check-ups, employees invariably used the opportunity to ask for curative help. The problems raised were nearly all non work-related matters, and the employees would probably not have raised them with the occupational physician at all if they had not been called for a medical check-up. Altogether these consultations took a quarter of the work time (3.5%) of the medical check-ups (table 2).
Few physicians seemed to use the medical checkups to collect systematic information on the work environment in category 2 (0.5% ) (table 2).
Cross-tabulations did not reveal any significant correlation between the characteristics of the occupational physicians or the occupational health services on one hand and the distribution of the work time on the other. However, the number of physicians in the different subgroups was limited, especially when one looks at the nine different organizational types of occupational health services.

Discussion
Our sample had obvious limitations. The numb er of occupational physicians was small, the physicians were all in urban area s, and we did not include physicians working less than half-time in occupational medicine. On the other hand, in terms of the number of workdays examined, the sample was substantial. This conclusion is reflected in the small standard errors. Moreover, most employees covered by occupational health services receive services from professional occup ational physician s working more than half-time in occupational medicine.
There is a danger of recall bias and strategic bias in a study of this kind. However, interviews at the end of the workday bring the registration clo se to prospecti ve (tomorrow) registration , the ideal for registering acti vity. Our impression was that the physicians were able to reconstruct their workdays and did so spontaneously, without applying censorship.
Altogether, we feel that the data reflect work content and priority setting in Norwegian occupational health service s well. Unfortunately, as previous studies (4,5) have used different categorizations and registrations, it is difficult to compare the results with older data.
The aims of Norwegian occupational health services are given by the labor inspectorate (12) and are in accordance with the convention proposed by the International Labour Office in 1985 (13). Occupational health services are supposed to be preventive and to assist management and employees in creating and securing safe work conditions by (i) improving the work environment, (ii) sur veying the health of employees in accordance with their workplace exposure s, and (iii) rehabilitation. The tasks within these general aims have , to a fair extent, also been specified in publications from the Norwegian labor inspectorate ( 12).
Other countries have broader aims for their occupational health services, also giving them tasks within the realm of general public health and curative medical treatment (9,14,15). Thu s the results of this study are mostly of interest to countries who apply their occupational health services more restrictedly to prevent ill health from expo sure s at work.
The Norwegian labor inspe ctorate has especially stressed the importance of securing the work environment when new activities and organizations are established. Howe ver, we found that only 3% of occupational physicians' time was devoted to such work . One reason may be that , in the larger occupational health services, persons other than occupational physicians handle these projects. Howev er, because of the economic recession taking place during the study, most firms might not have projects creating new work environments at all. But it is also possible that the firms were not interested in having the ir occupati onal physician s en gaged in these projects. Twenty-two percent of the occupational physician s' work time went to treatment and aid for nonwork-related health problems. This value seems high when it is considered that occupational health services should be doing preventive work . The labor inspectorate states that consultation and cure should, under no circumstances, take priority over preventive tasks in occupational health services. Substantial pressure on occupational health services for consultations and treatment has been documented from employees and management (4,16), and this pressure, in part, may explain the amount of curati ve activity . In this field comparative studies between The Netherl ands and Norw ay could be very interesting. Plomp ( 17) found that Dutch employees who rated the activities they wanted their occupational health services to provide gave medical treatment a relatively low score, presumably creating little pressure on the occupational health services to deliver treatment. In addition, occupational health services in The Netherlands have strictly preventive aims. Even so, Dick Draaisma has so far found that about 10% of the work of Dutch occupational physicians goes to general consultations (Dick Draaisma, personal communication).
In addition, Svarva (5) has shown that personnel in occup ational health services, in cooperation with employees and management, can establish occupational health services with very little curat ive activity for nonwork-related conditions.
Around 1980 Natvig (4), and later Svarva (5), found that about 5% of the work time in occupational health services is spent on rehabilitation. These activitie s seem to be increasing, as we observed that the occupational physicians alone spent 7% of their work time on occupational rehabilitation.
The labor inspectorate has repeatedly highlighted the importance of educ ating employee s to enable them to identify risks in their work environment and to take action against such risks . According to table 2, only I% of the occupational physicians' time went to this education. However, some information with educational aspects is contained in the reports and discussions following surveys of the work environment, but altogether education of employees seemed to receive little attention .
The "other" category occupied 12% of the work time . It contained various work tasks considered not to be directly invol ved in the work of occupational health services. However, some of these tasks decisively support the work of occupational health ser vices , and others are unavoidable. It is difficult to judge if 12% of an occupational phys ician's work time should be considered too much for these activities.
Eighteen percent of the average workday went to admini strative tasks. Swedish occupational physicians, generally working in larger occupational health servic es with the support of admin istrative personnel, spent around 10% of their time on admini stra-tion (18 ). It see ms as if occupational health serv ices in Norway have a potential to increase the efficiency of its administration. But it al so seems fair to suggest that th e amount of administrative work for occupational physicians has been underestimated by the government and employers.
None of the variables concerning occupational ph ysicians or the occupational health serv ices whe re they worked seemed to explain the observed differences in the work of the occupational ph ysicians. We had thought that the structure of the situation in which the ph ysicians worked, expressed as the calculated worktime re sources and the typ e of occupational health services , would have explained some of the differences. Th at thi s was not the case is interesting, but we shall need more ob servations in a larger study be fore drawing conclusions about th is possibility.

Concluding remarks
This study shows that the work of Norwegian occupational ph ysicians is oriented towards work environment matters (40 %), with emphasis on pre vention. However, a substantial amount (30 %) of their work still goes to curative activities, and two-thirds of this time is devoted to nonwork-related conditions.
Only with caution could our data be compared wi th data from other studies. Th e need to establish co mmon categories and definitions for the registration of work in oc cupational health services in future studies should therefore be stres sed . To gain further insight into the process of occupational health services and the factors which influence th eir priorities , we should be able to study the individual work of all personnel in occupational he alth serv ices. Occupational ph ysicians are only a part of the occupational health serv ice team . Such studies of the process in occupational health services ca n probably be done with instru ments si milar to the International Classification of Primary Care (19).