oa Determinanten van geregistreerd ziekteverzuim;

Determinants of registered long-term sickness absence; differences between a prospective and a retrospective design.

This study examines the differences in the determinants of registered long-term sickness absence when using prospectively or retrospectively gathered absence data. Two databases, the long-term sickness absence registration of the Dutch Workers Insurance Authority (UWV) and data from the Netherlands Working Conditions Survey (NWCS) 2005 and 2006 were both prospectively and retrospectively linked together. Using two regression analyses we examined the relations between personal characteristics, job characteristics, working conditions, health and retrospectively and prospectively gathered data on long-term sickness absence. The underlying question is whether these relations differ between prospective and retrospective analyses. Personal characteristics, job characteristics, working conditions and health appear associated with long-term sickness absence in the past as well as with long-term sickness absence in the future. Some working conditions are differently related with long-term sickness absence when using a retrospective or a prospective research design. Health appears to be more strongly related with long-term sickness absence history than with long-term sickness absence in the future. It is concluded that a prospective research design gives a more valid image of the predictors of long-term absenteeism than a retrospective design, but the differences are small.