Objectives: In economic evaluation of health care programmes, productivity costs may be overestimated, because compensation of lost work is neglected. This study tested the conclusions of previous studies which indicated that absenteeism is often compensated for during normal work hours, thereby limiting the actual production losses incurred. In this study the following compensation mechanisms are considered: 1) Lost productivity is made up for by the absent employee (after returning to work) or by colleagues during their normal working hours, 2) Work is taken over by extra workers who were hired especially for this, 3) Lost productivity is made up for by the absent employee (after returning to work) or by colleagues during extra hours. Methods: Absenteeism and compensation mechanisms were measured in a randomised trial of 1266 smokers. The study looked at the cost-effectiveness of reimbursement for smoking cessation treatment. By means of a mail-based questionnaire compensating mechanisms were asked for. A distinction was made between the above-mentioned possible compensating mechanisms. In the questionnaire also a distinction was made between the duration of the longest and the shortest period of absence, and the relevant compensation mechanism. Next to that, the questionnaire assessed baseline characteristics (age, gender, educational level, main (labour) situation), cause of absenteeism (.e.g. mental disorders, musculoskeletal diseases, respiratory diseases, circulatory diseases, flu and cold, pregnancy, etc). Furthermore as an indicator for the work situation, patients' occupational and job characteristics were evaluated. Results: Six months after the intervention had ended, questions were asked about absenteeism and compensation mechanisms. The response on this questionnaire was 66% of the initial respondents. In the group who where absent during a short or long period around 80% of lost work hours were compensated. About 60% was compensated by the absentee or colleagues during normal working hours (mechanism 1), only a small part was compensated by extra workers or in extra hours (mechanism 2 and 3). We did not find any difference in the type of compensation mechanism between short-term and long-term absenteeism. Due to presence of compensation mechanism there is a difference in the amount of the productivity costs as calculated by the classical friction costs method as opposed to method in which productivity costs were only calculated when extra efforts (mechanism 2 and 3) were needed. In our study populations, compensating mechanisms differed with occupational characteristics, like physical or mental work, managerial work. Conclusion: We advise researchers to take into account a correction factor for compensation mechanisms as including compensation reduces productivity costs due to absence from work to a considerable extent. Opposed to earlier studies we did not find any difference in the type of compensation mechanism revealed between short-term and long-term absenteeism. Nevertheless, we advise researchers to differentiate between short-term and long-term absenteeism, and take notice of our finding that the consequences for productivity costs may vary between due to occupational characteristics. Moreover, more research is required regarding the costs of compensation mechanisms
