Assessing the provision of occupational health services in the construction industry in Hong Kong

Assessing the provision of occupational health services in the construction industry in Hong Kong T.-S. I. Yu, F. F. K. Cheng, S. L. A. Tse and T. W. Wong Department of Community and Family Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China A survey was conducted to review the provision of occupational health services (OHS) in the construction industry, the most hazardous industry in Hong Kong. A questionnaire was used to collect information on various aspects of OHS from a sample of construction companies. OHS provision was estimated by an overall score, with the various components weighted for their importance regarding prevention. Factors affecting the provision of services were explored by multiple linear regressions. Only 58 of the 183 establishments (32%) performed environmental assessment; 37 (20%) offered medical examinations to their workers and 70 (38%) provided health and safety talks. Scores for the provision of OHS were generally low, especially for the component of surveillance concerning workers health. In general, larger establishment size and having safety and/or health policies were the important factors leading to high scores. Key words: Construction industry; Hong Kong; occupational health services. Received 25 September 2001; revised 2 April 2002; accepted 9 July 2002 Introduction Correspondence to: Dr Tak-sun Ignatius Yu, Department of Community and Family Medicine, The Chinese University of Hong Kong, 4/F, School of Public Health, Prince of Wales Hospital, Shatin, New Territories, Hong Kong. Tel: +852 2252 8773; fax: +852 2606 3500; e-mail: iyu@cuhk.edu.hk Occupational health services (OHS) are unequally distributed, with great variations between countries. In typical developing regions, the coverage by employee health services ranges from 5 to 10% at best. Even in countries where coverage rates are high, there are still gaps, with small-scale enterprises, construction, agriculture and certain mobile workers and the self-employed being under-served [1]. The types of OHS described in previous studies have been quite diverse and included monitoring of work environment and management of working practices and health (health examination, health counselling, health education and exercise/fitness programmes) [2 6]. Services could be offered on site or through multiple channels such as occupational health organizations, hospitals, medical associations, trade associations and health insurance societies [3,7]. OHS were considered to be needed by and beneficial to both employers and workers [2,8 12], and have been shown to be an important feature of many workplace safety and prevention programmes [5,10]. Many factors affect the provision of OHS, including resources [4], policy or statutory requirements [3,4,7], financial implications such as increased health care cost [4,11,12], size of establishment, high turnover rate, and the organizations that provide OHS [3,6,9]. Hong Kong is one of the most economically developed areas in Asia, but anecdotal reports suggested that OHS were underdeveloped and not matching economic development. There has been no systematic study on OHS in Hong Kong. A good starting point would be to look at OHS provision in the construction industry. Construction is one of the largest employment sectors for blue-collar workers and is the most dangerous trade, with the highest accident rate (~250/1000/year) [13] and the largest numbers of reported occupational diseases [14] and, hence, would be expected to benefit most Occup. Med. Vol. 52 No. 7, pp. 375 382, 2002 Copyright Society of Occupational Medicine. Printed in Great Britain. All rights reserved. 0962-7480/02

376 Occup. Med. Vol. 52, 2002 from OHS. We therefore conducted a survey to look into the types, patterns and levels of OHS provided in the construction industry in Hong Kong, and to identify the factors that might influence the provision of such services. The results would serve as a reference for future comparisons and the information obtained would be useful to guide the formulation of strategies for future improvement. Methods A cross-sectional survey on a sample of construction establishments in Hong Kong was carried out. The target population was all construction establishments in Hong Kong. Construction establishments registered in the Central Register of Establishments of the Hong Kong Census and Statistics Department (CRE) were used as the sampling frame. A total of 1017 construction establishments were identified after excluding 166 duplicated entries (establishments that were operating in two or more sites, or those operating under a different name but all belonging to the same head office or group). All establishments employing 200 or more workers were included (n = 56), plus a 30% random sample of the rest (288/961), giving a total of 344 establishments. Sixteen replied that they were no longer engaged in any construction business and were excluded, leaving only 328 eligible establishments. All member establishments from the Hong Kong Construction Association (HKCA the construction trade association) were included after excluding 16 with incorrect addresses and two duplicated entries (n =308). The selected establishments were first sent a letter of introduction, informing them of the forthcoming interview. A telephone interview using a structured questionnaire was conducted 1 week after the dispatch of the letter. The main interviewees or respondents were safety managers, safety officers and managers. The questionnaire was based on the recommendations on OHS of the International Labour Organization [15]. It was first pilot tested in a random sample of 20 establishments and minor changes were made to improve clarity of presentation. According to the recommendations of the International Labour Organization [14], OHS have five main functions: 1. surveillance of the working environment; 2. surveillance of the health of workers; 3. information, education, training and advice; 4. first aid, treatment and health programmes; 5. other functions. The functions of OHS were grouped under four categories in the current study, according to the actual situation in Hong Kong, as follows. A. Surveillance of the working environment 1. Identification and evaluation of the environmental factors that might affect the workers health. 2. Supervision and provision of personal protective equipment. B. Surveillance of the health of workers 1. Health assessment of workers before their assignment to specific tasks that might involve a danger to their health or that of others. 2. Health assessment at periodic intervals during employment that involved exposures to particular hazards to health. 3. Health assessment on resumption of work after a prolonged absence for health reasons to determine the worker s suitability for the job and needs for reassignment and rehabilitation. 4. Health assessment after the termination of assignments involving hazards that might cause or contribute to future health impairment. C. Education and training 1. Health and safety education. 2. Job and safety training. D. Curative medical treatment, first aid and record keeping 1. Provision of curative medical services. 2. Provision of first aid services and first aid personnel. 3. Keeping records and statistics on illnesses and injuries. The provision of OHS was measured using scores. The various components were given scores weighted for their contribution towards the prevention of occupational ill-health by a panel of occupational health experts through consensus in a meeting. The maximal score for an establishment was 100. Table 1 shows the four major components of OHS and the items included under each in the current study, as well as the score allocation. Data analyses were performed using the Statistical Package for Social Sciences (SPSS) v. 7.5. The establishments were classified into three groups according to their size. Small ones employed <50 workers, medium ones employed 50 199 workers and large ones had 200 workers. Descriptive statistics were used to summarize the data obtained. χ 2 trend tests and analysis of variance were used in the analyses for establishment size. Factors affecting the provision of services were explored by multiple linear regression using the forward stepwise strategy, and included employment size (medium and large compared to small), mean age of employees, one/ more than one site, turnover rate (%), with/without health policy, safety policy, and health and safety committee.

T.-S. I. Yu et al.: Occupational health services in the Hong Kong construction industry 377 Table 1. Components of occupational health services and the score allocation (maximum score: 100) Component and items Score Component and items Score Component and items Score Component and items Score Surveillance of the 40 Surveillance of workers 30 Education and training 20 Curative services and 10 working environment health record keeping Frequency of assessment [12] a Pre-employment check-up [12] Frequency of talks [8] Curative medical service [2] At the start of work only 4 General 3 Weekly 8 First aid services [2] Regular 12 Chest X-ray 3 Monthly 4 First aid personnel (F.A.P.) [3] Change of work process 8 Audiometry 3 Others 2 Employees/F.A.P. 100 3 Type of assessment [12] Spirometry 3 Provision of job training [6] Employees/F.A.P. > 100 1 Dust 4 Others 1 CITA b 4 Compiling statistics [3] Noise 4 Periodic check-up [12] In-house 2 Chemicals 4 Chest X-ray 3 Both 6 Providing PPE for hazards [6] Audiometry 3 Others 1 Height 1 Spirometry 3 Provision of safety training [6] Hands 1 Others 1 Eyes 1 Special (blood test) 3 Dust 1 Return-to-work check-up [3] Noise 1 Pre-retirement check-up [3] Feet 1 Supervising PPE [10] Distribution 3 Instruction 4 Maintenance 3 a Values in brackets denote maximum score allocated to each sub-category. b Construction Industry Training Authority. Although contract value and medical service accessibility were factors that might affect the provision of services, they were not included in the model as there were too many missing data on these two items. Results One hundred and eighty-eight establishments responded to the questionnaire, giving a response rate of 30%. [We failed to reach 317 (50%) of the establishments after repeated telephone calls and letters and 129 companies (20%) refused to be interviewed after successful contacts.] After excluding five incomplete questionnaires, 183 questionnaires were used for subsequent analysis. About half (93) of the establishments employed <50 employees and less than a quarter (44) of the establishments employed >200 workers. Nearly three-quarters of them (136) had operations in more than one site. There was a total of 43 900 employees, comprising 37 383 (85%) males and 6517 (15%) females, with ages ranging from 16 to 87 years around a mean of 38.5. The modal age was 40. The average employee turnover rate was 18%. Sixty-seven per cent of the establishments had easy access to public medical facilities nearby, 36% had health policies, 49% had safety policies, and 30% had health and safety committees. Only 75 establishments (41%) provided their annual contract values, with a mean of HK$65 million and a range from 3 million to 8.1 billion. Only 32% of establishments performed environmental assessment and dust (57%) was most frequently involved (Table 2). Most establishments (96%) provided some personal protective equipment (PPE). Helmets (90%), gloves (86%) and safety harnesses (84%) were the most frequent items (Table 3). Surveillance of workers health was relatively poor. For pre-employment check-up, only 20% provided general check-up, 14% provided chest X-ray and <5% provided audiometry or spirometry. Only five large companies provided periodic chest X-ray. Periodic audiometry and spirometry were provided by only one company (Table 4). For education and training, 38% of establishments provided health and safety talks to their workers; most of these were irregular and infrequent, and about half of them were given by safety officers (Table 5). Job training and safety training were provided by 127 (69%) and 115 (63%) establishments, respectively. Safety training was provided mainly by the safety officer or provision was contracted out. Medical services were provided by 109 (60%) establishments, mainly in the form of curative treatment and/ or first aid (Table 6). Only 87 (48%) establishments employed any first aid personnel. The majority (84%) of the establishments kept records for injuries and about half of them kept illness records as well. Eighty-seven (47%) establishments compiled statistics on injuries and/or illnesses, and 57 (66%) of them use the statistic for report purposes.

378 Occup. Med. Vol. 52, 2002 Table 2. Performance of environmental assessment according to size of establishment Small [No. (%)] Medium [No. (%)] Large [No. (%)] Overall [No. (%)] P a Establishments performing environment assessment b 11 (12) 22 (48) 25 (57) 58 (32) <0.001 Frequency of assessment n.a. c At the start of work only 0 4 (9) 0 4 (2) Change of work process 8 (9) 11 (24) 14 (32) 33 (18) Regular 3 (3) 7 (15) 11 (25) 21 (12) Type of assessment Noise 10 (11) 18 (39) 24 (54) 52 (28) n.a. Dust 3 (3) 13 (28) 17 (39) 33 (18) 0.034 Chemicals 1 (1) 2 (4) 10 (23) 13 (7) n.a. Dust + noise 2 (2) 9 (20) 16 (36) 27 (15) 0.009 Dust + chemicals 1 (1) 2 (4) 10 (23) 13 (7) n.a. Noise + chemicals 1 (1) 1 (2) 10 (23) 12 (7) n.a. Dust + noise + chemicals 1 (1) 1 (2) 10 (23) 12 (7) n.a. b Performing environment assessment means undertaking assessment for at least one of three major hazards: dust, noise and chemicals. c χ 2 test for trend not applicable due to inadequate expected cell frequencies. Table 3. Provision and supervision of PPE according to size of establishment Hazards and PPE provision Small [No. (%)] Medium [No. (%)] Large [No. (%)] Overall [No. (%)] P a Providing PPE 87 (94) 44 (96) 44 (100) 175 (96) n.a. c Distribution b 40 (43) 31 (67) 32 (73) 103 (56) 0.001 Instructions on use b 49 (53) 39 (85) 42 (95) 130 (71) <0.001 Maintenance b 29 (31) 29 (63) 32 (73) 90 (49) <0.001 Height Helmet 79 (85) 42 (91) 44 (100) 165 (90) n.a. Safety harness 71 (76) 39 (85) 44 (100) 154 (84) 0.002 Hands Gloves 78 (84) 37 (80) 42 (95) 157 (86) n.a. Eyes Goggles 66 (71) 32 (70) 42 (95) 140 (76) 0.018 Welding shield 55 (59) 24 (52) 35 (80) 114 (62) 0.127 Dust Mask 63 (68) 30 (65) 37 (84) 130 (71) 0.217 Respirator 32 (34) 24 (52) 37 (84) 93 (51) <0.001 Noise Ear muffs 53 (57) 29 (63) 41 (93) 123 (67) <0.001 Ear plugs 49 (53) 34 (74) 39 (89) 122 (67) <0.001 Feet Safety shoes 44 (47) 33 (72) 39 (89) 116 (63) <0.001 Safety boots 37 (40) 32 (70) 33 (75) 102 (56) <0.001 b By designated persons such as safety officers or foremen. c χ 2 test for trend not applicable due to inadequate expected cell frequencies. The overall mean scores for all the four component services were <50% of the maximum scores and the performance in the surveillance of workers health was poorest (Table 7): surveillance of working environment, 15.6/40; surveillance of workers health, 1.5/30; education and training, 8.0/20; curative services and record keeping, 3.6/10; overall, 28.8/100. For the different components, larger establishment size and those having safety and/or health policies were predictive of higher scores. For the overall score, medium

T.-S. I. Yu et al.: Occupational health services in the Hong Kong construction industry 379 Table 4. Performance of medical examinations according to size of establishment Types of medical examinations Small [No. (%)] Medium [No. (%)] Large [No. (%)] Overall [No. (%)] P a Pre-employment General 10 (11) 10 (22) 17 (39) 37 (20) <0.001 Chest X-ray 2 (2) 9 (12) 14 (32) 25 (14) <0.001 Audiometry 2 (2) 2 (4) 4 (9) 8 (4) n.a. d Spirometry 1 (1) 1 (2) 4 (9) 6 (3) n.a. Others 5 (5) 1 (2) 4 (9) 10 (6) n.a. Periodic Chest X-ray 0 (0) 0 (0) 5 (11) 5 (3) n.a. Audiometry 0 (0) 0 (0) 1 (2) 1 (1) n.a. Spirometry 0 (0) 0 (0) 1 (2) 1 (1) n.a. Others 0 (0) 0 (0) 1 (2) 1 (1) n.a. Special (blood test) b 2 (2) 1 (2) 0 (0) 3 (2) n.a. Return-to-work c 0 (0) 0 (0) 0 (0) 0 (0) n.a. Pre-retirement 0 (0) 0 (0) 0 (0) 0 (0) n.a. b Blood lead level for workers doing rust proofing, painting or welding. c Return-to-work examination referred to any check upon return to work after injury or sick leave of >1 week. d χ 2 test for trend not applicable due to inadequate expected cell frequencies. Table 5. Provision of education and training according to size of establishment Small [No. (%)] Medium [No. (%)] Large [No. (%)] Overall [No. (%)] P a Establishments with health and 17 (18) 21 (46) 32 (73) 70 (38) <0.001 safety talks Frequency of talks n.a. c Weekly 0 (0) 1 (2) 3 (7) 4 (2) Monthly 2 (2) 2 (4) 2 (4) 6 (3) Others 15 (16) 18 (39) 27 (61) 60 (33) Establishments with job training 51 (55) 37 (80) 39 (89) 127 (69) <0.001 Provision of job training n.a. CITA b 5 (5) 6 (13) 4 (9) 15 (8) In-house 19 (20) 7 (15) 3 (7) 29 (16) Both 25 (27) 24 (52) 32 (73) 81 (44) Others 2 (2) 0 (0) 0 (0) 2 (1) Establishments with safety training 39 (42) 36 (78) 40 (91) 115 (63) <0.001 b Construction Industry Training Authority. c χ 2 test for trend not applicable due to inadequate expected cell frequencies. and large establishment size and having a safety policy were the significant factors, explaining >50% of the total variance (Table 8). Discussion Surveys on OHS in different industries and different countries have been reported [2,3,4,6,7,16]. Different researchers have used different definitions and tools in their surveys. A study of National Health Service (NHS) OHS in England and Wales used medical staffing levels rather than service content to measure provision [17]. The lack of a standardized approach did not facilitate comparisons between studies. More recently, Bråtveit et al. [18] attempted to use the ILO recommendation as a standard to compare the activity profiles of OHS at different locations of a multinational company and found that approach to be useful. We tried to make a further improvement by quantifying OHS provisions through assigning weights to the different components of OHS in the ILO recommendation according to their importance in fulfilling the primary preventive role of OHS. A panel of occupational health experts, consisting of three occupational physicians and one occupational health nurse,

380 Occup. Med. Vol. 52, 2002 Table 6. Provision of medical services and record keeping according to size of establishment Small [No. (%)] Medium [No. (%)] Large [No. (%)] Overall [No. (%)] P a Providing medical services 31 (33) 40 (87) 38 (86) 109 (60) <0.001 Type of medical services Curative 29 (31) 37 (80) 38 (86) 104 (57) n.a. c First aid 5 (5) 6 (13) 9 (20) 20 (11) 0.399 Employing first aid personnel 27 (29) 26 (56) 34 (81) b 87 (48) <0.001 Employee/first aid personnel ratio 100 27 (29) 25 (54) 28 (64) 80 (44) n.a. Keeping records on illnesses 30 (32) 32 (70) 25 (57) 87 (48) 0.001 Keeping records on injuries 64 (69) 45 (98) 44 (100) 153 (84) <0.001 Compiling statistics 34 (37) 24 (52) 28 (64) 86 (47) 0.002 b Data missing for two large establishments. c χ 2 test for trend not applicable due to inadequate expected cell frequencies. Table 7. Provision of OHS measured by scores according to size of establishment Component services Maximum score Small [Mean (95% CI) a ] Medium [Mean (95% CI)] Large [Mean (95% CI)] Overall [Mean (95% CI)] P b Surveillance of working environment 40 10.2 (8.6 11.9) 19.0 (16.1 21.8) 23.6 (20.5 26.6) 15.6 (14.1 17.2) <0.001 Surveillance of workers health 30 0.6 (0.2 1.0) 1.5 (0.6 2.4) 3.2 (1.8 4.8) 1.5 (1.0 2.0) <0.001 Education and training 20 5.1 (4.1 6.2) 9.8 (8.6 11.0) 12.3 (11.0 13.6) 8.0 (7.2 8.8) <0.001 Curative services and record keeping 10 2.2 (1.7 2.7) 4.6 (3.8 5.4) 5.7 (5.0 6.4) 3.6 (3.2 4.1) <0.001 Overall scores 100 18.2 (15.4 21.0) 34.8 (30.6 39.0) 44.8 (40.1 49.4) 28.8 (26.1 36.4) <0.001 a Confidence interval. b P-value for analysis of variance. confirmed the face validity of the questionnaire used and assigned the scores for the components. Cronbach s α for the overall score (32 items) was 0.84, and ranged between 0.50 (four items for curative medical treatment, first aid and record keeping ) and 0.74 (13 items for surveillance of the working environment ) for the four sub-scores, indicating satisfactory to good internal consistency. We performed sensitivity analysis by assigning equal weight to all items and found that the results were very similar to our original results. If further validated, the instrument and the scoring system may have wider applications for assessing and evaluating OHS in other trades and countries. The response rate from the eligible establishments was disappointingly low and this precluded the direct generalization of the findings to the whole industry without making some qualifications. We sent two letters through the post and made at least three phone calls before giving up. However, we still failed to reach 317 of the establishments (50%) after repeated telephone calls and letters. It was likely that these companies were no longer doing business and hence were inactive. One hundred and twenty-nine companies (20%) refused to be interviewed after successful contacts; they provided no reason for refusal or any information about their company. The respondents consisted of a disproportionately large number of large establishments (24%) when compared with the Central Register of Establishments (6%). It is possible to assume that companies responding to the survey were better organized and had a higher concern for occupational health and safety. Hence, it is justifiable to say that the results from the current study might have overestimated the level of provision of OHS in the construction industry. Most of the information solicited in the survey was factual data and did not require much recall, but it was believed that respondents were more likely, for the benefit of the company, to over-report than to under-report on the provision of OHS. Hence, we believe that the situation regarding OHS in the construction industry reported in this study represents the better end of the spectrum. The actual situation would be more likely to be further away from satisfaction. Despite the likely overestimation of the level of provision, we still believe that the results reflect the pattern of OHS provision in the construction industry in Hong Kong, as the pattern of provision was found to be quite consistent across different employment sizes. The provision of medical examinations was particularly poor; only ~20% provided any form of medical examination and just a few provided specific examinations.

T.-S. I. Yu et al.: Occupational health services in the Hong Kong construction industry 381 Table 8. Factors affecting the provision of OHS (scores) Component services Significant factors Regression coefficient (B) 95% confidence intervals for B P Surveillance of working environment (R 2 = 0.404) Medium establishments 5.72 2.55 8.90 <0.001 Large establishments 5.85 2.10 9.60 0.002 With safety policy 6.74 3.57 9.91 <0.001 Turnover rate (%) 0.06 0.00 0.12 0.037 Surveillance of workers health (R 2 = 0.056) Large establishments 1.83 0.66 3.00 0.002 Education and training (R 2 = 0.447) Medium establishments 2.32 0.66 3.97 0.006 Large establishments 2.93 1.08 4.79 0.002 With safety policy 4.86 3.42 6.30 <0.001 Curative services and record keeping (R 2 = 0.416) Medium establishments 1.59 0.70 2.47 0.001 Large establishments 1.45 0.44 2.46 0.005 With health policy 1.61 0.56 2.66 0.003 With safety policy 1.65 0.67 2.62 0.001 Mean age of employees 0.10 0.16 0.02 0.011 Overall scores for occupational health services (R 2 = 0.506) Medium establishments 11.09 6.30 15.88 <0.001 Large establishments 14.15 8.84 19.46 <0.001 With safety policy 15.63 11.32 19.93 <0.001 This was in contrast to many other previous studies on OHS, which reported medical examinations to be the predominant activity [2,3,4,6,7,15]. The under-provision of medical examinations and medical surveillance could be attributable to the lack of legislative requirements for such programmes in Hong Kong. On the other hand, the law requires employers to provide a safe and healthy workplace, and to provide information to the employees under the general duties. There are also specific requirements for assessment of noise, PPE and first aid provision in the workplace. None of the OHS in the construction industry is headed by an occupational physician and this might be another reason for the low level of surveillance for workers health. In fact, there are just over 10 accredited specialists in occupational medicine in Hong Kong. A limited number of factors leading to better provision of OHS were identified. It is likely that other important factors might have been missed, as the R 2 of the multiple regression models were not very high. Having safety and/or health policies is important, and this is potentially modifiable with the will of the management. Employment size is also important. Presumably, larger companies were more likely to provide OHS because they could afford them better and because economies of scale made them more likely to be fully utilized. However, employer size itself is not modifiable and might only be a surrogate for other amenable factors having a more direct impact on OHS. These need to be explored in future studies. Our findings confirmed that the construction industry is usually under-served in OHS [1,19]. Policy or statutory requirements, the size of establishment, and a high turnover rate were found to be important in determining the provision of OHS, in agreement with previous studies [3,4,9]. Financial implications, such as increased health care cost, appeared not to be important factors in the current survey. A likely explanation is the widely available public medical services in Hong Kong, which can take care of injured workers or workers suffering from various occupational diseases at almost zero cost to the employers. In conclusion, construction establishments in Hong Kong in general provided a low level of OHS to their employees, except in certain areas governed by current legislation (e.g. the provision of certain PPE). This may be one of the main reasons for the high rates of injuries and the large number of occupational diseases reported. Improvements will depend on better legislative coverage, financial incentives, professional assistance, management commitment and other factors to be identified in future studies. Acknowledgements The authors would like to thank the Hong Kong Construction Association for providing the membership list and all the construction companies that participated in the survey. This study was supported in part by a grant from the Occupational Health Advisory Committee of the Hong Kong Occupational Safety and Health Council.

382 Occup. Med. Vol. 52, 2002 References 1. Rantanen J, Fedotov IA. Standards, principles and approaches in occupational health services. In: Stellman JM, ed. Encyclopaedia of Occupational Health and Safety, 4th edn. Geneva: International Labour Office, 1998; Vol. 1, 16.2 16.18. 2. Isah EC, Asuzu MC, Okojie OH. Occupational health services in manufacturing industries in Nigeria. Occup Med 1996; 46: 333 336. 3. Muto T, Higashi T, Mizoue T, et al. Multiple channels for occupational health services to small-scale enterprises in Japan. Occup Med 1995; 45: 268 272. 4. Mizoue T, Higashi T, Muto T, et al. Activities of an occupational health organization in Japan, in special reference to services for small and medium-scale enterprises. Occup Med 1996; 46: 12 16. 5. Weel ANH, Slotboom MAJ. Quality improvement by means of differential occupational health care: an experiment in a regional occupational health centre. Occup Med 1996; 46: 20 24. 6. Muto T, Hsieh SD, Sakurai Y. Status of health promotion programme implementation in small-scale enterprises in Japan. Occup Med 1999; 49: 65 70. 7. Roh J. Overall view of occupational health services in Korea. Asia-Pacific J Public Health 2000; 12(Suppl.): S41 44. 8. Dryson E. Preferred components of an occupational health service for small industry in New Zealand: Health protection or health promotion? Occup Med 1995; 45: 31 34. 9. Dryson E. Occupational health needs in small industry in New Zealand: preferred sources of information. Occup Med 1993; 43: 176 179. 10. Pickvance S. Towards multidisciplinary prevention services. Occup Health Rev 1996; 63: 27 32. 11. Ritchie KA, McEwen J. Employee perceptions of workplace health and safety issues. Can a questionnaire contribute to health needs assessment? Occup Med 1994; 44: 77 82. 12. Kocks DJ, Ross MH. The financial components and contribution of occupational health services to the provision of primary health care service in a developing country such as South Africa. Occup Med 1995; 45: 156 158. 13. Occupational Safety & Health Branch. Accident Statistics Bulletin, Issue No. 1. Hong Kong: Labour Department, October 1997. 14. Occupational Safety & Health Branch. Occupational Safety and Health Statistics Bulletin, Issue No. 1. Hong Kong: Labour Department, June 2001. 15. International Labour Organization (ILO). Occupational Health Services Recommendation, No. 171. Geneva: ILO, 1985. 16. Jeebhay MF, Lopata AL, Robins TG. Seafood processing in South Africa: a study of working practices, occupational health services and allergic health problems in the industry. Occup Med 2000; 50: 406 413. 17. Hughes A, Philipp R, Harling K. Provision and staffing of NHS occupational health services in England and Wales. Occup Environ Med 1999; 56: 714 717. 18. Bråtveit M, McCormack D, Moen BE. Activity profiles of the occupational health services in a multinational company. Occup Med 2001; 51: 168 173. 19. Gyi DE, Haslam RA, Gibb AG. Case studies of occupational health management in the engineering construction industry. Occup Med 1998; 48: 263 271.