Work status and productivity costs due to ankylosing spondylitis: comparison of three European countries

Abstract

Objective: To compare work disability, sick leave, and productivity costs due to ankylosing spondylitis (AS) of three European countries.

Methods: 216 patients with AS from the Netherlands, France, and Belgium participated in a two year observational study. Employment and work disability rates at baseline were adjusted for age and sex. Productivity costs were calculated by both the friction cost method and the human capital approach. The adjusted contributions of country to employment, work disability, and having an episode of sick leave were assessed by logistic regression and the contribution of the country to days of sick leave and costs by Cox proportional hazard analysis.

Results: 209 patients completed the two years' follow up with sufficient data for cost analysis. Adjusted employment was 55% in the Netherlands as compared with 72% in both other countries and only in the Netherlands was it lower than expected in the general population. Adjusted work disability was 41%, 23%, and 9% in the Netherlands, France, and Belgium and in all countries was higher than expected in the general population. In those with a paid job, the mean number of days of sick leave per patient per year because of AS was 19 (range 0-130), six (range 0-77), and nine (range 0-60 ) in the Netherlands, France, and Belgium respectively. Applying the friction cost method to those with a paid job resulted in mean costs per patient per year of 1257 euros (range 0-7356), 428 euros (range 0-5979), and 476 euros (range 0-2354) in the Netherlands, France, and Belgium. Applying the human capital approach to the whole group resulted in mean costs per patient per year of 8862 euros (range 0-46 818), 3188 euros (range 0-43 550), and 3609 euros (range 0-34 320) in the three countries, respectively. After adjusting for sociodemographic and disease characteristics, living in the Netherlands, as compared with both other countries, was associated with a higher chance of being work disabled (odds ratio (OR)=3.82; 95% confidence interval (CI) 1.33 to 11.01), but not with the risk of having an episode of sick leave. Similarly, living in the Netherlands contributed independently to the number of days sick leave (OR=0.65; 95% CI 0.43 to 0.97), a higher amount of friction costs (OR=0.63; 95% CI 0.42 to 0.96), and a higher amount of human capital costs (OR=0.46; 95% CI 0.32 to 0.68).

Conclusion: There are remarkable differences in work status and productivity costs between the three European countries. This has implications for the generalisability of health economic studies.

Figure 1

Kaplan-Meier curve of patients with AS with a paid job, illustrating the proportion of patients incurring a specific level of friction costs in each country separately.

Figure 2

Kaplan-Meier curve of the active working population, illustrating the proportion of patients with AS incurring a specific level of human capital costs in each country separately.