Collider Bias in Administrative Workers' Compensation Claims Data: A Challenge for Cross-Jurisdictional Research

Injured worker outcomes after compensation system overhaul: an interrupted time series study

OBJECTIVE

In 2015, South Australia replaced its workers' compensation system with the aim of improving return to work rates. We examined whether this was achieved by focusing on the duration of time off work, as well as claim processing times and claim volumes to understand how this may have been achieved.

METHODS

The primary outcome was mean weeks of compensated disability duration. Secondary outcomes tested alternative mechanisms of a change in disability duration: (1) mean employer report and insurer decision times to evaluate whether there had been changes in claim processing and (2) claim volumes to determine whether the new system altered the cohort under investigation. Outcomes were aggregated into monthly units and analysed with an interrupted time series design. Three condition subgroups-injury, disease and mental health-were compared in separate analyses.

RESULTS

While disability duration steadily declined before the RTW Act came into effect, afterwards it flatlined. A similar effect was observed in insurer decision time. Claim volumes gradually increased. Employer report time gradually decreased. Condition subgroups mostly followed a similar pattern to overall claims, though the increase in insurer decision time appears largely driven by changes in injury claims.

CONCLUSIONS

The increase in disability duration after the RTW Act took effect may be attributable to an increase in insurer decision time, which itself could be due to the disruption of overhauling a compensation system or the elimination of provisional liability entitlements that incentivised early decision making and provided early intervention.

Cross-Jurisdiction Comparative Research in Work Disability Prevention: Introduction to the Special Series

The systems that societies construct to support work disability prevention can have powerful effects on both the experiences of people with work disability as well as their health and employment outcomes. Comparative studies between jurisdictions provide an opportunity to gain insights into these system level impacts, by comparing system features, processes and experiences; and by determining if jurisdictional variation affects outcomes. In turn, this can prompt policy and practice reform. Reflecting the diversity of work disability systems globally, there is growing interest in cross-jurisdiction comparative research in the field. This special series presents seven articles addressing important methodological and conceptual aspects of comparative research in work disability prevention, and presents practical examples of how jurisdictions vary and the impact this can have on workers.

Exploring Differences in Work Disability Duration by Size of Firm in Canada and Australia

Purpose To identify whether there were differences in work disability duration between injured workers employed by small, medium, large, and self-insured firms and whether these differences varied between workers' compensation jurisdictions in Canada and Australia. Methods Workers' compensation data were used to identify comparable lost-time, work-related injury and musculoskeletal disorder claims in five Canadian and five Australian jurisdictions between 2011 and 2015. Work disability duration was measured using cumulative disability days paid up to one-year post-injury. Jurisdiction-specific quantile regression models were used to estimate differences in cumulative disability days paid to claims from small (< 20 full-time equivalents (FTEs)) medium (20-199 FTEs), large (200 + FTEs) and self-insured firms at the 25th, 50th, and 70th percentiles in the disability distribution, adjusting for confounders. Results Compared to large firms, workers in small firms generally had longer work disability duration at each percentile, particularly in Saskatchewan and Alberta (Canada), Victoria and Australian Capital Territory (Australia), where an additional 31.1, 18.4, 58.5 and 37.0 days were paid at the 75th percentiles, respectively. The disability duration of workers from self-insured firms was longer than large firms in all Canadian jurisdictions but was shorter or no different in Australian jurisdictions. Smaller differences were observed between claims from large and medium-sized firms. Conclusions Workers in small firms had longer work disability duration than those in large firms in all but one of the study jurisdictions. Claims management processes need to be sensitive to the challenges that small firms face in accommodating and returning injured workers back to work.