Original Research

Reducing Unneeded Clinical Variation in Sepsis and Heart Failure Care to Improve Outcomes and Reduce Cost: A Collaborative Engagement with Hospitalists in a MultiState System


OBJECTIVE: To (1) measure hospitalist care for sepsis and heart failure patients using online simulated patients, (2) improve quality and reduce cost through customized feedback, and (3) compare patient-level outcomes between project participants and nonparticipants.
METHODS: We conducted a prospective, quasi-controlled cohort study of hospitalists in eight hospitals matched with comparator hospitalists in six nonparticipating hospitals across the AdventHealth system. We provided measurement and feedback to participants using Clinical Performance and Value (CPV) vignettes to measure and track quality improvement. We then compared length of stay (LOS) and cost results between the two groups.
RESULTS: 107 providers participated in the study. Over two years, participants improved CPV scores by nearly 8% (P < .001), with improvements in utilization of the three-hour sepsis bundle (46.0% vs 57.7%; P = .034) and ordering essential medical treatment elements for heart failure (58.2% vs 72.1%; P = .038). In study year one, average LOS observed/expected (O/E) rates dropped by 8% for participants, compared to 2.5% in the comparator group, equating to an additional 570 hospital days saved among project participants. In study year two, cost O/E rates improved from 1.16 to 0.98 for participants versus 1.14 to 1.01 in the comparator group. Based on these improvements, we calculated total cost savings of $6.2 million among study participants, with $3.8 million linked to system-wide improvements and an additional $2.4 million in savings attributable to this project.
CONCLUSIONS: CPV case simulation-based measurement and feedback helped drive improvements in evidence-based care that translated into lower costs and LOS, above-and-beyond other improvements at AdventHealth.

   Comments ()

Related Articles from PubMed

BMJ Qual Saf. 2019 Mar 20. pii: bmjqs-2018-008829. doi: 10.1136/bmjqs-2018-008829. [Epub ahead of print]
Popul Health Manag. 2019 Aug;22(4):321-329. doi: 10.1089/pop.2018.0111. Epub 2018 Oct 17.