Background/Purpose: It is well known that patients with RA experience health-related job loss, more days absent from work (i.e., absenteeism), and to a greater extent, a decrease of at-work productivity (i.e., presenteeism). However, little is known about the effect of biologic treatments on absenteeism and presenteeism. The aim of this study was to investigate whether achieving clinical response to treatment, and changes in physical ability and quality of life (QoL) during the first six months predict absenteeism and presenteeism at one year in patients commencing treatment with biologic therapy for RA.

Methods: Patients recruited to the British Society for Rheumatology Biologics Register for RA (BSRBR-RA) commencing biologics for RA and in full or part-time paid employment at inclusion were included in this study. DAS28, HAQ, and QoL (EQ-5D) were measured at baseline, six months, and 1 year. Patients were categorised into good/moderate vs. poor responders according to the EULAR response criteria at 6 months. Changes from baseline in physical ability and QoL at 6 months were calculated. Absenteeism in the last month, and presenteeism (0=no interference – 10=complete interference), were measured at all three time points using the RA specific Work Productivity Survey (WPS-RA). Due to excessive zero values in the WPS-RA, zero inflated negative binomial regression (ZINB) was used to assess whether treatment response, HAQ and EQ-5D changes at six months predicted absenteeism and presenteeism at one year.

Results: Since most patients were still in employment at 1 year (work disabled n=4), data on 263 patients with baseline and 1 year WPS-RA were used. The mean age was 51 years (SD 8.6), median disease duration was 6 years (IQR 3-13); 78% were female. At baseline, median [IQR] scores were: DAS28 5.8 (5.3-6.4), HAQ 1.3 (0.6-1.6) and EQ-5D 0.6 (0.5-0.7). At baseline, 21% reported ≥1 days absent from work, and the median presenteeism score was 3.0 (IQR 0-5). Presenteeism scores significantly improved over 1 year (median 1.0 [IQR 0-4], Kruskal-Wallis Test: H(2) = 18.70, p<0.001), as did absenteeism (17% reported ≥1 days absent) although a Chi-square test demonstrated this to be non-significant. Compared to non-responders, EULAR good/moderate responders were 2.8x more likely to be classified in the zero group for presenteeism at 1 year, (OR 2.89, [95% CI 0.99, 8.50], p=.0.05), and although non-significant there was a trend towards increases in HAQ at 6 months resulting in lower probability of being in the zero group presenteeism at 1 year (p<0.07) (table 1). Changes in EQ-5D scores at 6 months did not appear to predict presenteeism or absenteeism at one year.

Conclusion: The results suggest that if a moderate/good response to biologic therapy is observed at six months, a reduction in presenteeism over the following six months can be expected.  The results of this analysis are promising, particularly with regard to presenteeism which remains to be a major economic issue. Table 1. Six month predictors of presenteeism and absenteeism at one year follow up.

ZINB: Zero inflated negative binomial regression model

†Adjusted for age, gender, and disease duration.

*p<0.05.

¶ Relative change in count part of ZINB model, assuming non-zero score

§ Odds ratio for being in zero scoring group of ZINB model.

≠ A unit increase of 0.06 (a minimal clinically important difference in EQ-5D scores)