Abstract
Introduction
There is significant variation in processes and outcomes of care for patients with inflammatory bowel disease (IBD), suggesting opportunities to improve quality of care. Recent efforts to define quality measures for IBD have identified emergency room (ER) visits, hospitalizations, corticosteroid use, and opioid use as indicators of care quality. We hypothesized that IBD care could be improved through a structured quality improvement (QI) program.
Methods
We utilized the Breakthrough Series Collaborative approach developed by the Institute for Healthcare Improvement to improve care for adults with IBD. We identified primary and secondary drivers of urgent care need for patients including those at high risk for ER use, and a multi-stakeholder panel developed 19 practice change ideas that could influence those drivers. Between January 2018 and May 2019, clinical sites participating in a QI collaborative across the United States tested and implemented various change ideas, shared ongoing results during coached monthly webinars, and participated in 3 in-person meetings to learn QI methods and share best practices. Patient-reported outcomes (PROs) were collected at clinical visits, including recent ER use and hospitalizations, use of steroids and narcotics, and measures of care utilization. Providers rated whether patients were at high risk for urgent care needs. Site performance on key measures were monitored using statistical control charts, with assessment for common cause (due to chance) variation and special cause (non-random) variation.
Results
We collected data prospectively from 20,382 discrete visits at twenty-six participating clinical practices (14 academic/university, 12 private/community). Disease type included Crohn’s disease (58%), ulcerative colitis (39%), and other (3%); 54% were female. During the 15-month project period, improvement with special cause variation was noted across multiple measures. Collaborative-wide decreases were seen in ER utilization (18% to 14%, relative reduction of 22%; Figure), hospitalization (14% to 11%, relative reduction of 21%), steroid use (14% to 10%, relative reduction of 29%), and narcotic utilization (8% to 4%, relative reduction of 50%). Successful change ideas tested by sites included proactive maintenance of a “high risk” patient list, reserved outpatient visits for urgent needs, “morning-after” contact with patients who went to the ER, patient education about how and when to get help, and proactively scheduling earlier follow-up for high risk patients.
Conclusions
Outcomes of IBD care were improved using a structured QI program that facilitates small changes in practice structure, sharing of best practices across sites, and ongoing feedback. Spread of successful change ideas may facilitate broad improvement in IBD care and significant cost savings when applied to a large population.
Changes in Key Measures Over Time
Statistical Process Control Chart Showing Monthly Proportion of Patients Reporting Recent ER Utilization
Trajectories of Stroke Care in Ontario: Which Path to Best Care?
