Impact of Provider Feedback on Quality Improvement in Rehabilitation Services: An Interrupted Time Series Analysis

Anne-Lene Sand-Svartrud ^1* Hanne Dagfinrud ¹ Johanne Fossen ² Heidi Framstad ³ Eirik Lind Irgens ⁴ Hild Kristin Morvik ⁵ Joseph Sexton ⁶ Rikke Helene Moe ¹ Ingvild Kjeken ¹

• ¹ Health Services Reserach and Innovation Unit, Centre for Treatment of Rheumatic and Musculoskeletal Diseases (REMEDY), Diakonhjemmet Hospital, Oslo, Norway
• ² HLF Rehabilitation AS, Oslo, Norway
• ³ Røysumtunet Rehabilitation Centre, Jaren, Norway
• ⁴ Helsepartner Rehabilitation Centre, Alta, Norway
• ⁵ Åstveit Health Centre, Åsane i Bergen, Norway
• ⁶ Centre for Treatment of Rheumatic and Musculoskeletal Diseases (REMEDY), Diakonhjemmet Hospital, Oslo, Norway

Introduction: Quality improvement in rehabilitation is needed due to unwarranted variations and suboptimal service coordination. Audit and feedback strategies are commonly used to improve healthcare quality, but evidence of their effectiveness in rehabilitation settings is limited. Objective: To evaluate the impact of an audit and feedback strategy on rehabilitation quality, as measured by a set of quality indicators (QIs) specifically designed for rehabilitation. Methods: Interrupted time series analysis was conducted across 16 Norwegian institutions delivering specialized rehabilitation for long-term diseases. Patient-reported rehabilitation quality data was collected continuously before and after a provider feedback intervention, while provider-reported quality was measured once before and after the intervention. We compared 11 pre-and 9 postintervention observations, each spanning 3 weeks, over a 15-months study period.The analyses included 2415 patients, with 1444 (59.8%) pre-intervention and 971 (40.2%) post-intervention. Mixed model analyses revealed that the mean differences in patient-reported QIs between the pre-and post-intervention phase were small and statistically non-significant. The expected impact model, including a gradually higher quality after the feedback to institution managers and clinical team members, was not confirmed. We observed variations in service quality among institutions, also post-intervention. The lowest pass rates were observed for indicators addressing the follow-up, involvement of external services and next of kin. Conclusions: In this multicentre study, the audit and feedback intervention did not lead to improvements in the quality of rehabilitation services, as measured by changes in QI pass rates covering health service structures, processes and patient outcomes. Trial registration number: ClinicalTrials.gov NCT03764982